Venous thromboembolism: Risk factors, comorbidities, and treatment-associated risk of bleeding Katarina Glise Sandblad 2023 1 To Staffan, Emma, and Gustav Venous thromboembolism: Risk factors, comorbidities and treatment-associated risk of bleeding © Katarina Glise Sandblad 2023 katarina.glise.sandblad@vgregion.se ISBN 978-91-8069-463-6 (PRINT) ISBN 978-91-8069-464-3 (PDF) http://hdl.handle.net/2077/78554 Cover illustration: Lungor med lunginfarkt by Emma Sandblad, 11 years old. Printed by Stema Specialtryck AB, Borås, Sweden ENM ANNENM Ä ÄRRKVA KEV E Trycksak 3T0r4y1c k0s2a3k4 3041 0234 2 3 SS TT To Staffan, Emma, and Gustav Venous thromboembolism: Risk factors, comorbidities and treatment-associated risk of bleeding © Katarina Glise Sandblad 2023 katarina.glise.sandblad@vgregion.se ISBN 978-91-8069-463-6 (PRINT) ISBN 978-91-8069-464-3 (PDF) http://hdl.handle.net/2077/78554 Cover illustration: Lungor med lunginfarkt by Emma Sandblad, 11 years old. Printed by Stema Specialtryck AB, Borås, Sweden 2 3 ABSTRACT Background: Venous thromboembolism (VTE) is the third most common cardiovas- cular disease, consisting mainly of deep vein thrombosis (DVT) and pulmonary em- bolism (PE). Since VTE often is a preventable disease, knowledge of risk factors is critical. Following a VTE, many patients are subjected to extended anticoagulant treatment. However, the bleeding risk during extended treatment is largely unknown. Aim: To study risk factors in patients with VTE and to determine the occurrence of major bleeding during VTE treatment. Methods: Paper I: 1.6 million men from The Swedish Military Service Conscription Register were grouped based on BMI and followed through nationwide registries to determine the risk of a  rst-time VTE. Papers II-IV: The National Patient Register, the National Cause of Death Register, the National Prescribed Drug Register, and the Total Population Register were used to identify almost 300,000 patients with  rst- time PE or DVT and 1,200,000 matched controls. PE and DVT patients and their respective controls were compared regarding comorbidities and temporary provoking factors (II), the prevalence of different cancers (III), and, between 2014–2020, the risk of bleeding during anticoagulant treatment (IV). Results: Paper I: Men who were overweight or obese at enlistment had a high risk of VTE later in life. Paper II: Patients with PE more often had underlying cardiovascular disease, while patients with DVT were more likely to have recent musculoskeletal surgery or fracture. Paper III: VTE had a strong association with pancreatic, brain, or liver cancer, while the association was weak with recent diagnoses of bladder/urinary tract cancer, kidney cancer, or uterine cancer. Paper IV: During initial treatment (0–6 months), patients treated with apixaban had a lower bleeding risk than patients treated with warfarin or rivaroxaban. During extended treatment (6 months–5 years), both apixaban and rivaroxaban had a low bleeding risk, lower than warfarin. Conclusion: The increasing prevalence of obesity might imply an increase in VTE in the coming decades. Patients with cardiopulmonary disease had a higher risk of PE than DVT. Risks of VTE differ widely for various cancers. Apixaban carried a lower risk of bleeding than rivaroxaban and warfarin in the initial treatment, while both apixaban and rivaroxaban had a low risk of bleeding, and lower than warfarin, in extended treatment. These  ndings are important for VTE prophylaxis and treatment in clinical praxis. Keywords: Venous thromboembolism, registries, overweight, body mass index, inci- dence, cardiovascular disease, anticoagulants, apixaban, rivaroxaban, warfarin, hem- orrhage. ISBN 978-91-8069-463-6 (PRINT) ISBN 978-91-8069-464-3 (PDF) http://hdl.handle.net/2077/78554 4 5 ABSTRACT Background: Venous thromboembolism (VTE) is the third most common cardiovas- cular disease, consisting mainly of deep vein thrombosis (DVT) and pulmonary em- bolism (PE). Since VTE often is a preventable disease, knowledge of risk factors is critical. Following a VTE, many patients are subjected to extended anticoagulant treatment. However, the bleeding risk during extended treatment is largely unknown. Aim: To study risk factors in patients with VTE and to determine the occurrence of major bleeding during VTE treatment. Methods: Paper I: 1.6 million men from The Swedish Military Service Conscription Register were grouped based on BMI and followed through nationwide registries to determine the risk of a  rst-time VTE. Papers II-IV: The National Patient Register, the National Cause of Death Register, the National Prescribed Drug Register, and the Total Population Register were used to identify almost 300,000 patients with  rst- time PE or DVT and 1,200,000 matched controls. PE and DVT patients and their respective controls were compared regarding comorbidities and temporary provoking factors (II), the prevalence of different cancers (III), and, between 2014–2020, the risk of bleeding during anticoagulant treatment (IV). Results: Paper I: Men who were overweight or obese at enlistment had a high risk of VTE later in life. Paper II: Patients with PE more often had underlying cardiovascular disease, while patients with DVT were more likely to have recent musculoskeletal surgery or fracture. Paper III: VTE had a strong association with pancreatic, brain, or liver cancer, while the association was weak with recent diagnoses of bladder/urinary tract cancer, kidney cancer, or uterine cancer. Paper IV: During initial treatment (0–6 months), patients treated with apixaban had a lower bleeding risk than patients treated with warfarin or rivaroxaban. During extended treatment (6 months–5 years), both apixaban and rivaroxaban had a low bleeding risk, lower than warfarin. Conclusion: The increasing prevalence of obesity might imply an increase in VTE in the coming decades. Patients with cardiopulmonary disease had a higher risk of PE than DVT. Risks of VTE differ widely for various cancers. Apixaban carried a lower risk of bleeding than rivaroxaban and warfarin in the initial treatment, while both apixaban and rivaroxaban had a low risk of bleeding, and lower than warfarin, in extended treatment. These  ndings are important for VTE prophylaxis and treatment in clinical praxis. Keywords: Venous thromboembolism, registries, overweight, body mass index, inci- dence, cardiovascular disease, anticoagulants, apixaban, rivaroxaban, warfarin, hem- orrhage. ISBN 978-91-8069-463-6 (PRINT) ISBN 978-91-8069-464-3 (PDF) http://hdl.handle.net/2077/78554 4 5 LIST OF PAPERS SAMMANFATTNING PÅ SVENSKA This thesis is based on the following studies, referred to in the text by their Roman Blodproppar i benens vener (DVT) och i lungans blodkärl (LE) kallas venös trombo- numerals. embolism (VTE). Det är den tredje vanligaste hjärt-kärlsjukdomen i världen, med hög dödlighet och lidande samt höga, ökande sjukdomsrelaterade kostnader. Syftet med I Glise Sandblad K, Jern S, Åberg M, Robertson J, Torén K, Lindgren M, denna avhandling var att 1) kartlägga bakomliggande sjukdomar vid venös trombo- Adiels M, Hansson PO, Rosengren A. Obesity in adolescent men increases embolism (VTE), främst övervikt och olika typer av cancer, 2) skillnader i riskfakto- the risk of venous thromboembolism in adult life. rer mellan blodproppar i ben och lungor samt 3) förekomst av allvarliga blödningar J Intern Med. 2020;287:734–745. vid behandling med blodförtunnande läkemedel efter VTE. II Glise Sandblad K, Rosengren A, Sörbo J, Jern S, Hansson PO. Pulmonary Metod embolism and deep vein thrombosis—comorbidities and temporary provok- ing factors in a register-based study of 1.48 million people. Delarbete I: 1.6 miljoner män från Värnpliktsregistret delades upp i grupper beroende Res Pract Thromb Haemost. 2022;6:e12714. på Body mass index (BMI) vid mönstring och följdes upp i landsomfattande register för att fastställa hur många som drabbades av VTE under uppföljningstiden. III Glise Sandblad K, Hansson PO, Philipson J, Mahmoud A, Karlsson P, Rosen- gren A, Sörbo J. Prevalence of Cancer in Patients with Venous Thromboem- Delarbete II: I data från Patientregistret, Dödsorsaksregistret, Läkemedelsregistret bolism: A Retrospective Nationwide Case-Control Study in Sweden. och Totalbefolkningsregistret identi erades alla VTE-patienter mellan 1987–2018 (ca Clinical and Applied Thrombosis/Hemostasis. 2023;29:1-10. 300,000 st) och deras matchade kontrollpersoner utan VTE (ca 1,200,000 st). Patien- ter och kontroller jämfördes för att avgöra förekomst av samsjuklighet mellan DVT IV Glise Sandblad K, Schulman S, Rosengren A, Sörbo J, Philipson J, Hans- och LE. son PO. Association of type of oral anticoagulation with risk of bleeding in 45,114 patients with venous thromboembolism during initial and extended Delarbete III: Samma registeruttag som i Delarbete II. Män och kvinnor med VTE treatment - A nationwide register-based study. jämfördes med sina respektive kontroller avseende förekomst av olika typer av cancer J Intern Med. 2023;00:1–18. inom ett år före VTE-diagnosen. Delarbete IV: Samma registeruttag som i Delarbeten II och III, men i denna studie följdes patienter med VTE 2014–2020 för att avgöra blödningsförekomst under initial behandling (0–6 månader) och förlängd behandling (6 månader upp till 5 år). Resultat Delarbete I: Vi fann att män med övervikt och fetma vid mönstring hade hög risk för VTE senare i livet. Delarbete II: Patienter med LE hade i större utsträckning bakomliggande hjärt-kärl- sjukdomar, medan patienter med DVT i högre utsträckning nyligen hade genomgått ortopedisk operation eller haft en nedre extremitetsfraktur. Delarbete III: Samsjukligheten mellan VTE och cancer varierade mycket mellan oli- ka cancertyper. Patienter med VTE hade höga odds för en nylig diagnos av cancer i bukspottskörteln, hjärnan eller levern. Delarbete IV: Patienter som behandlades med läkemedlet apixaban (Eliquis) hade lägre blödningsrisk än patienter med warfarin- (Waran) eller rivaroxaban- (Xarelto) behandling under de första sex månadernas behandling. Vid förlängd behandling sågs ingen säker skillnad mellan patienter som behandlades med apixaban jämfört med rivaroxaban, men båda hade en lägre risk än warfarin. 6 7 LIST OF PAPERS SAMMANFATTNING PÅ SVENSKA This thesis is based on the following studies, referred to in the text by their Roman Blodproppar i benens vener (DVT) och i lungans blodkärl (LE) kallas venös trombo- numerals. embolism (VTE). Det är den tredje vanligaste hjärt-kärlsjukdomen i världen, med hög dödlighet och lidande samt höga, ökande sjukdomsrelaterade kostnader. Syftet med I Glise Sandblad K, Jern S, Åberg M, Robertson J, Torén K, Lindgren M, denna avhandling var att 1) kartlägga bakomliggande sjukdomar vid venös trombo- Adiels M, Hansson PO, Rosengren A. Obesity in adolescent men increases embolism (VTE), främst övervikt och olika typer av cancer, 2) skillnader i riskfakto- the risk of venous thromboembolism in adult life. rer mellan blodproppar i ben och lungor samt 3) förekomst av allvarliga blödningar J Intern Med. 2020;287:734–745. vid behandling med blodförtunnande läkemedel efter VTE. II Glise Sandblad K, Rosengren A, Sörbo J, Jern S, Hansson PO. Pulmonary Metod embolism and deep vein thrombosis—comorbidities and temporary provok- ing factors in a register-based study of 1.48 million people. Delarbete I: 1.6 miljoner män från Värnpliktsregistret delades upp i grupper beroende Res Pract Thromb Haemost. 2022;6:e12714. på Body mass index (BMI) vid mönstring och följdes upp i landsomfattande register för att fastställa hur många som drabbades av VTE under uppföljningstiden. III Glise Sandblad K, Hansson PO, Philipson J, Mahmoud A, Karlsson P, Rosen- gren A, Sörbo J. Prevalence of Cancer in Patients with Venous Thromboem- Delarbete II: I data från Patientregistret, Dödsorsaksregistret, Läkemedelsregistret bolism: A Retrospective Nationwide Case-Control Study in Sweden. och Totalbefolkningsregistret identi erades alla VTE-patienter mellan 1987–2018 (ca Clinical and Applied Thrombosis/Hemostasis. 2023;29:1-10. 300,000 st) och deras matchade kontrollpersoner utan VTE (ca 1,200,000 st). Patien- ter och kontroller jämfördes för att avgöra förekomst av samsjuklighet mellan DVT IV Glise Sandblad K, Schulman S, Rosengren A, Sörbo J, Philipson J, Hans- och LE. son PO. Association of type of oral anticoagulation with risk of bleeding in 45,114 patients with venous thromboembolism during initial and extended Delarbete III: Samma registeruttag som i Delarbete II. Män och kvinnor med VTE treatment - A nationwide register-based study. jämfördes med sina respektive kontroller avseende förekomst av olika typer av cancer J Intern Med. 2023;00:1–18. inom ett år före VTE-diagnosen. Delarbete IV: Samma registeruttag som i Delarbeten II och III, men i denna studie följdes patienter med VTE 2014–2020 för att avgöra blödningsförekomst under initial behandling (0–6 månader) och förlängd behandling (6 månader upp till 5 år). Resultat Delarbete I: Vi fann att män med övervikt och fetma vid mönstring hade hög risk för VTE senare i livet. Delarbete II: Patienter med LE hade i större utsträckning bakomliggande hjärt-kärl- sjukdomar, medan patienter med DVT i högre utsträckning nyligen hade genomgått ortopedisk operation eller haft en nedre extremitetsfraktur. Delarbete III: Samsjukligheten mellan VTE och cancer varierade mycket mellan oli- ka cancertyper. Patienter med VTE hade höga odds för en nylig diagnos av cancer i bukspottskörteln, hjärnan eller levern. Delarbete IV: Patienter som behandlades med läkemedlet apixaban (Eliquis) hade lägre blödningsrisk än patienter med warfarin- (Waran) eller rivaroxaban- (Xarelto) behandling under de första sex månadernas behandling. Vid förlängd behandling sågs ingen säker skillnad mellan patienter som behandlades med apixaban jämfört med rivaroxaban, men båda hade en lägre risk än warfarin. 6 7 Betydelse CONTENTS Ökande förekomst av fetma i befolkningen kommer troligen att medföra en ökning av VTE de kommande decennierna. Att DVT och LE har olika samsjuklighet kan ha ABSTRACT 5 betydelse för vilka patienter som bör få förebyggande behandling (profylax) mot VTE eftersom LE är mer allvarligt än DVT. Risken för VTE vid olika cancerformer varie- LIST OF PAPERS 6 rar mycket och kräver individuell bedömning. Apixaban verkar medföra lägre risk för blödning än rivaroxaban och warfarin vid initial behandling. Vid förlängd behandling SAMMANFATTNING PÅ SVENSKA 7 är både apixaban och rivaroxaban förknippade med en låg blödningsrisk, lägre än för warfarin. ABBREVIATIONS 11 INTRODUCTION 13 Venous thromboembolism 13 Hemostasis 13 Blood clots 13 Development of VTE - Virchow´s triad 13 Historical remarks 14 Incidence, risk factors, and preventive measures 15 Incidence 15 Risk factors 15 Preventive measures 17 Clinical presentation and diagnosis 18 Diagnostics 18 Treatment 20 Thrombolytic treatment 20 Additional advanced treatment options in the acute setting 21 Anticoagulant treatment 22 Duration of anticoagulant treatment 23 Risk of bleeding on anticoagulant treatment 24 Prognosis 25 Mortality 25 Long-term complications 25 AIM 27 PATIENTS AND METHODS 28 Study populations 28 Paper I 28 Papers II and III 28 Papers IV 28 Registers 28 The Swedish Military Service Conscription Register 28 The Longitudinal Integration Database for Health Insurance and 28 Labor Market Services (LISA) register 8 9 Betydelse CONTENTS Ökande förekomst av fetma i befolkningen kommer troligen att medföra en ökning av VTE de kommande decennierna. Att DVT och LE har olika samsjuklighet kan ha ABSTRACT 5 betydelse för vilka patienter som bör få förebyggande behandling (profylax) mot VTE eftersom LE är mer allvarligt än DVT. Risken för VTE vid olika cancerformer varie- LIST OF PAPERS 6 rar mycket och kräver individuell bedömning. Apixaban verkar medföra lägre risk för blödning än rivaroxaban och warfarin vid initial behandling. Vid förlängd behandling SAMMANFATTNING PÅ SVENSKA 7 är både apixaban och rivaroxaban förknippade med en låg blödningsrisk, lägre än för warfarin. ABBREVIATIONS 11 INTRODUCTION 13 Venous thromboembolism 13 Hemostasis 13 Blood clots 13 Development of VTE - Virchow´s triad 13 Historical remarks 14 Incidence, risk factors, and preventive measures 15 Incidence 15 Risk factors 15 Preventive measures 17 Clinical presentation and diagnosis 18 Diagnostics 18 Treatment 20 Thrombolytic treatment 20 Additional advanced treatment options in the acute setting 21 Anticoagulant treatment 22 Duration of anticoagulant treatment 23 Risk of bleeding on anticoagulant treatment 24 Prognosis 25 Mortality 25 Long-term complications 25 AIM 27 PATIENTS AND METHODS 28 Study populations 28 Paper I 28 Papers II and III 28 Papers IV 28 Registers 28 The Swedish Military Service Conscription Register 28 The Longitudinal Integration Database for Health Insurance and 28 Labor Market Services (LISA) register 8 9 The National Patient Register 29 ABBREVIATIONS The National Cause of Death Register 29 The National Prescribed Drug Register 29 The Total Population Register 29 BMI Body Mass Index De nitions 30 Paper I 30 CDT Catheter-directed treatment Paper II 31 CI Con dence interval Paper III 32 Paper IV 33 COPD Chronic obstructive pulmonary disease Statistical analysis 34 CT Computed tomography RESULTS 36 CTEPH Chronic thromboembolic pulmonary hypertension Paper I 36 CTPA Computed tomography pulmonary angiography Baseline characteristics 36 VTE events 36 DVT Deep vein thrombosis, djup ventrombos Hazard ratios 36 ECMO Extracorporeal membrane oxygenation Paper II 36 Study population 36 HR Hazard ratio Incidence rates 36 ICD International Classi cation of Diseases Odds ratios and PAR for comorbidities and temporary provoking factors 37 Paper III 38 IQR Interquartile range Study population 38 LE Lungemboli Incidence rate 39 Multivariable adjusted odds ratios of various types of cancer 39 DOAC Direct oral anticoagulant Paper IV 39 OR Odds ratio Study population 39 PAR Population attributable risk Risk of major bleeding 40 Risk factors 40 PE Pulmonary embolism DISCUSSION 41 PPV Positive predictive value – the proportion of subjects with a positive Clinical implications 43 test result who have the disease of interest Limitations 44 PTS Post thrombotic syndrome CONCLUSIONS 45 RCT Randomized controlled trial VTE Venous thromboembolism, venös tromboembolism FUTURE PERSPECTIVES 46 VKA Vitamin-K Antagonist ACKNOWLEDGEMENTS 47 REFERENCES 49 PAPER I-IV 10 11 The National Patient Register 29 ABBREVIATIONS The National Cause of Death Register 29 The National Prescribed Drug Register 29 The Total Population Register 29 BMI Body Mass Index De nitions 30 Paper I 30 CDT Catheter-directed treatment Paper II 31 CI Con dence interval Paper III 32 Paper IV 33 COPD Chronic obstructive pulmonary disease Statistical analysis 34 CT Computed tomography RESULTS 36 CTEPH Chronic thromboembolic pulmonary hypertension Paper I 36 CTPA Computed tomography pulmonary angiography Baseline characteristics 36 VTE events 36 DVT Deep vein thrombosis, djup ventrombos Hazard ratios 36 ECMO Extracorporeal membrane oxygenation Paper II 36 Study population 36 HR Hazard ratio Incidence rates 36 ICD International Classi cation of Diseases Odds ratios and PAR for comorbidities and temporary provoking factors 37 Paper III 38 IQR Interquartile range Study population 38 LE Lungemboli Incidence rate 39 Multivariable adjusted odds ratios of various types of cancer 39 DOAC Direct oral anticoagulant Paper IV 39 OR Odds ratio Study population 39 PAR Population attributable risk Risk of major bleeding 40 Risk factors 40 PE Pulmonary embolism DISCUSSION 41 PPV Positive predictive value – the proportion of subjects with a positive Clinical implications 43 test result who have the disease of interest Limitations 44 PTS Post thrombotic syndrome CONCLUSIONS 45 RCT Randomized controlled trial VTE Venous thromboembolism, venös tromboembolism FUTURE PERSPECTIVES 46 VKA Vitamin-K Antagonist ACKNOWLEDGEMENTS 47 REFERENCES 49 PAPER I-IV 10 11 INTRODUCTION Venous thromboembolism Venous thromboembolism (VTE) occurs when a blood clot forms in a vein, most commonly as deep vein thrombosis (DVT) in the lower extremities or pulmonary embolism (PE). Most PEs embolize from a DVT in a lower extremity and travel to the lung arteries. However, they can also develop in the right atrium1 of the heart or in situ in the lung vessels, i.e., as pulmonary thrombosis2. Venous thrombosis can also occur in other veins, such as upper extremity veins, splanchnic veins, and the cerebral venous sinuses. Hemostasis Hemostasis is a tightly balanced, local process that produces a blood clot in areas of blood vessel injury but also limits the clot’s extension and lyses the clot when vascu- lar integrity is restored3. It constitutes a highly complex system of many interrelated components, including the vessel wall, platelets, coagulation factors, anticoagulant proteins, and the  brinolytic system. If the balance in the hemostatic system is tilted, the result is either bleeding or a blood clot. Blood clots Blood clot formation depends on the type of vessel in which the clot develops. Arterial blood clots are formed when an atherosclerotic plaque ruptures and plate- lets are recruited to the site4. The platelets are activated, and additional platelets are recruited. Following the exposure of tissue factor from the atherosclerotic plaque, coagulation factors are activated, generating  brin, and the clot develops. By contrast, venous thromboembolism is often initiated in venous valve pockets, where the  ow is irregular, and the oxygen tension may be low, promoting thrombi5. The clots are rich in  brin and are often called red clots due to trapped red blood cells. Development of VTE – Virchow’s triad The development of VTEs is traditionally looked upon as the result of alterations in the so-called Virchow’s triad; stasis/changes in blood  ow, endothelial damage or dysfunction, and hypercoagulability6. Even though this simpli ed explanatory model has been questioned, both concerning its origin and accuracy7, many risk factors of VTE can be derived from this triad hypothesis. Stasis: Reduced blood  ow and stasis can be caused by surgery, obesity, immobiliza- tion, and paralysis5. It can also explain the increased risk of left-sided DVT in preg- nancy (compression of the left common iliac vein) and patients with May-Thurner 12 13 INTRODUCTION Venous thromboembolism Venous thromboembolism (VTE) occurs when a blood clot forms in a vein, most commonly as deep vein thrombosis (DVT) in the lower extremities or pulmonary embolism (PE). Most PEs embolize from a DVT in a lower extremity and travel to the lung arteries. However, they can also develop in the right atrium1 of the heart or in situ in the lung vessels, i.e., as pulmonary thrombosis2. Venous thrombosis can also occur in other veins, such as upper extremity veins, splanchnic veins, and the cerebral venous sinuses. Hemostasis Hemostasis is a tightly balanced, local process that produces a blood clot in areas of blood vessel injury but also limits the clot’s extension and lyses the clot when vascu- lar integrity is restored3. It constitutes a highly complex system of many interrelated components, including the vessel wall, platelets, coagulation factors, anticoagulant proteins, and the  brinolytic system. If the balance in the hemostatic system is tilted, the result is either bleeding or a blood clot. Blood clots Blood clot formation depends on the type of vessel in which the clot develops. Arterial blood clots are formed when an atherosclerotic plaque ruptures and plate- lets are recruited to the site4. The platelets are activated, and additional platelets are recruited. Following the exposure of tissue factor from the atherosclerotic plaque, coagulation factors are activated, generating  brin, and the clot develops. By contrast, venous thromboembolism is often initiated in venous valve pockets, where the  ow is irregular, and the oxygen tension may be low, promoting thrombi5. The clots are rich in  brin and are often called red clots due to trapped red blood cells. Development of VTE – Virchow’s triad The development of VTEs is traditionally looked upon as the result of alterations in the so-called Virchow’s triad; stasis/changes in blood  ow, endothelial damage or dysfunction, and hypercoagulability6. Even though this simpli ed explanatory model has been questioned, both concerning its origin and accuracy7, many risk factors of VTE can be derived from this triad hypothesis. Stasis: Reduced blood  ow and stasis can be caused by surgery, obesity, immobiliza- tion, and paralysis5. It can also explain the increased risk of left-sided DVT in preg- nancy (compression of the left common iliac vein) and patients with May-Thurner 12 13 syndrome (compression of the left common iliac vein by the left common iliac artery Incidence, risk factors, and preventive measures leading to scarring and reduced blood  ow in the vein) 8. Mechanistically, the reduced blood  ow is thought to lead to activation of the endothelium with expression of adhe- Incidence sion proteins such as von Willebrand factor and p-selectin9. Leucocytes and platelets The annual incidence of VTE has been reported to be 75–269 per 100,000 persons are accumulated, and  brin is formed. in the Western world and Southern Latin America23. Studying PE separately, inci- Endothelial damage: Under physiological conditions, the endothelial surface prevents dences of 23–115 per 100,000 patient-years have been reported in North America and attachment of proteins required for clotting and suppresses coagulation by release of Europe. For DVT, the corresponding numbers are 48–162 per 100,000. By contrast, anticoagulant and pro brinolytic factors5. However, endothelial damage/dysfunction yearly incidence numbers from Southeast Asia show a much lower rate of 8–17 per 23 can lead to downregulation of their expression and upregulation of the tissue factor 100,000 population . These ethnic differences have been debated since the incidence expression, which has a strong procoagulant action. numbers may have been skewed by a probable under-diagnosis in Asian countries 24. However, in studies on populations with different ethnicities living in the same geo- 24 Hypercoagulability: Genetic and acquired forms of thrombophilia, advanced age, can- graphical area, the differences remain, suggesting that the differences are real . cer, obesity, hormone contraceptives, and pregnancy are all hypercoagulable states5. Regardless of nationality, the incidence increases sharply with increasing age23,25Hypercoagulability can, as in pregnancy, be due to both an up-regulation of proco- . Pre- agulant factors such as factor VII, VIII, X,  brinogen, and von Willebrand factor and vious reports have indicated a ten times higher risk in patients above the age of 80 23 a decrease of anticoagulants such as Protein S10. Hypercoagulability can also, as in compared to patients between 40 and 50 years of age . obesity, result from lower  brinolytic capacity due to decreased levels of plasminogen activator inhibitor (PAI-1) in combination with increased procoagulant tissue factor The incidence of VTE is increasing over time, mainly driven by the incidence of 20,26 and increased platelet activation11,12. PE . This development has been parallel to better accessibility of computed to-mographic pulmonary angiography (CTPA) 26, and a part of the increase is probably Historical remarks due to the diagnosis of PEs that might be clinically insigni cant, for example, small emboli in asymptomatic patients undergoing CT scan for cancer follow-up27. Other One of the  rst descriptions of well-documented cases of DVT originates from the possible contributing reasons for the increased incidence are improved cancer treat- 13th century13 28 29,30 . Pregnancy and the post-partum period were early recognized as risk ment and survival and the global burden of obesity , increasing the number of factors, leading to various theories on origin, including breast milk accumulating in individuals at increased risk of VTE. the legs14. In the 19th and early 20th centuries, DVT treatment included ligating large veins such as the inferior vena cava (to prevent PE), strict bed rest, and bloodletting13. Risk factors VTE is, in many cases, a preventable disease. The development of VTE is often a con- Heparin was discovered in the 1910s, and following its puri cation in the 1930s, hu- sequence of the combination of underlying patient-related and situation-related risk man treatment was made possible13. Heparin was reported to represent a breakthrough factors20. Identi cation of high-risk individuals and situations is therefore of interest. in treatment with dramatically improved survival in patients with DVT15. Around the Among the most potent patient-related risk factors are thrombophilias (notably an- same time, the  rst discoveries leading to Vitamin K- antagonist (VKA) treatment tiphospholipid syndrome) and metastatic cancer, whereas high-risk situations include were made13. A hemorrhagic disease was spread in cattle in the US in the early 20th major trauma and fractures, as well as hospitalization per se31. century and was later found to be caused by spoiled clover. About 20 years later, di- cumarol was found in moldy hay, and it was discovered that the effect of both spoiled Obesity clover and dicumarol could be reversed by Vitamin K. Dicumarol was thereafter in- troduced to treat VTE. Obesity is a growing problem in large parts of the world but with a widely varying magnitude of the problem. The mean BMI in 2019 among 19-year-olds in Paci c 2 A bene t of a combination of heparin and VKA treatment was reported in the late Island countries in Oceania was above 28 kg/m , closely followed by the USA, New 1940s16. With the introduction of LMWH in the 1980s17 and new insights into the Zealand, Middle Eastern countries, North African countries, and the Caribbean Is- bene t of early ambulation and compression stockings18, the treatment of DVT was lands. By contrast, countries in east and central Africa, central European countries 2 32 changed from a bedridden in-hospital treatment to ambulatory outpatient treatment. such as Romania and Bosnia, and Japan had average BMIs 9–10 kg/m lower . Eu- Over the past decade, outpatient treatment has also been included in guidelines as a ropean guideline estimations on the effect of overweight and obesity on the risk of 20 33 treatment option for patients with low-risk PE19-22, facilitated by the introduction of VTE are based on studies performed decades ago and might not represent the cur- direct oral anticoagulants, DOACs. rent situation, when more severe forms of obesity are increasing 34. Also, the effect 14 15 syndrome (compression of the left common iliac vein by the left common iliac artery Incidence, risk factors, and preventive measures leading to scarring and reduced blood  ow in the vein) 8. Mechanistically, the reduced blood  ow is thought to lead to activation of the endothelium with expression of adhe- Incidence sion proteins such as von Willebrand factor and p-selectin9. Leucocytes and platelets The annual incidence of VTE has been reported to be 75–269 per 100,000 persons are accumulated, and  brin is formed. in the Western world and Southern Latin America23. Studying PE separately, inci- Endothelial damage: Under physiological conditions, the endothelial surface prevents dences of 23–115 per 100,000 patient-years have been reported in North America and attachment of proteins required for clotting and suppresses coagulation by release of Europe. For DVT, the corresponding numbers are 48–162 per 100,000. By contrast, anticoagulant and pro brinolytic factors5. However, endothelial damage/dysfunction yearly incidence numbers from Southeast Asia show a much lower rate of 8–17 per 23 can lead to downregulation of their expression and upregulation of the tissue factor 100,000 population . These ethnic differences have been debated since the incidence expression, which has a strong procoagulant action. numbers may have been skewed by a probable under-diagnosis in Asian countries 24. However, in studies on populations with different ethnicities living in the same geo- 24 Hypercoagulability: Genetic and acquired forms of thrombophilia, advanced age, can- graphical area, the differences remain, suggesting that the differences are real . cer, obesity, hormone contraceptives, and pregnancy are all hypercoagulable states5. Regardless of nationality, the incidence increases sharply with increasing age23,25Hypercoagulability can, as in pregnancy, be due to both an up-regulation of proco- . Pre- agulant factors such as factor VII, VIII, X,  brinogen, and von Willebrand factor and vious reports have indicated a ten times higher risk in patients above the age of 80 23 a decrease of anticoagulants such as Protein S10. Hypercoagulability can also, as in compared to patients between 40 and 50 years of age . obesity, result from lower  brinolytic capacity due to decreased levels of plasminogen activator inhibitor (PAI-1) in combination with increased procoagulant tissue factor The incidence of VTE is increasing over time, mainly driven by the incidence of 20,26 and increased platelet activation11,12. PE . This development has been parallel to better accessibility of computed to-mographic pulmonary angiography (CTPA) 26, and a part of the increase is probably Historical remarks due to the diagnosis of PEs that might be clinically insigni cant, for example, small emboli in asymptomatic patients undergoing CT scan for cancer follow-up27. Other One of the  rst descriptions of well-documented cases of DVT originates from the possible contributing reasons for the increased incidence are improved cancer treat- 13th century13 28 29,30 . Pregnancy and the post-partum period were early recognized as risk ment and survival and the global burden of obesity , increasing the number of factors, leading to various theories on origin, including breast milk accumulating in individuals at increased risk of VTE. the legs14. In the 19th and early 20th centuries, DVT treatment included ligating large veins such as the inferior vena cava (to prevent PE), strict bed rest, and bloodletting13. Risk factors VTE is, in many cases, a preventable disease. The development of VTE is often a con- Heparin was discovered in the 1910s, and following its puri cation in the 1930s, hu- sequence of the combination of underlying patient-related and situation-related risk man treatment was made possible13. Heparin was reported to represent a breakthrough factors20. Identi cation of high-risk individuals and situations is therefore of interest. in treatment with dramatically improved survival in patients with DVT15. Around the Among the most potent patient-related risk factors are thrombophilias (notably an- same time, the  rst discoveries leading to Vitamin K- antagonist (VKA) treatment tiphospholipid syndrome) and metastatic cancer, whereas high-risk situations include were made13. A hemorrhagic disease was spread in cattle in the US in the early 20th major trauma and fractures, as well as hospitalization per se31. century and was later found to be caused by spoiled clover. About 20 years later, di- cumarol was found in moldy hay, and it was discovered that the effect of both spoiled Obesity clover and dicumarol could be reversed by Vitamin K. Dicumarol was thereafter in- troduced to treat VTE. Obesity is a growing problem in large parts of the world but with a widely varying magnitude of the problem. The mean BMI in 2019 among 19-year-olds in Paci c 2 A bene t of a combination of heparin and VKA treatment was reported in the late Island countries in Oceania was above 28 kg/m , closely followed by the USA, New 1940s16. With the introduction of LMWH in the 1980s17 and new insights into the Zealand, Middle Eastern countries, North African countries, and the Caribbean Is- bene t of early ambulation and compression stockings18, the treatment of DVT was lands. By contrast, countries in east and central Africa, central European countries 2 32 changed from a bedridden in-hospital treatment to ambulatory outpatient treatment. such as Romania and Bosnia, and Japan had average BMIs 9–10 kg/m lower . Eu- Over the past decade, outpatient treatment has also been included in guidelines as a ropean guideline estimations on the effect of overweight and obesity on the risk of 20 33 treatment option for patients with low-risk PE19-22, facilitated by the introduction of VTE are based on studies performed decades ago and might not represent the cur- direct oral anticoagulants, DOACs. rent situation, when more severe forms of obesity are increasing 34. Also, the effect 14 15 of being overweight or obese early in life on the future risk of VTE, important for zygous mutation. Prothrombin G20210A mutation exists almost exclusively in Cau- prognostication of future health care needs, has been scarcely studied. casians, among whom it has a prevalence of 2%, giving a risk increase of VTE of 2-3 times in heterozygotes and 30 times in homozygotes. Protein C, S, and antithrombin Cancer de ciencies are rare but strong risk factors with several different subtypes50. Antiphos- Cancer is an important risk factor for VTE, with around 20% of all VTE cases oc- pholipid syndrome is the most common acquired thrombophilia. It is an autoimmune curring in patients with cancer35. VTE can also be the  rst manifestation of occult disorder, which requires persisting positive antiphospholipid antibodies (lupus antico- cancer, detected in 5% of patients within a year of a seemingly unprovoked VTE36. agulant, anticardiolipin antibodies, or anti-2glykoprotein antibodies) in combination The association between VTE and cancer varies with the type, stage, and treatment of with clinical symptoms (thrombosis or obstetric complications) for diagnosis51. cancer28. Reports on high thrombogenicity of speci c cancer types, such as pancreatic and brain cancer, have been relatively consistent28. However, in other types of cancer, Immobilization such as kidney and bladder cancer, data is less clear37-39. Short-term immobilization is a strong risk factor for VTE, contributing to the increased risk of hospitalization, trauma, surgery, and orthopedic casting. However, patients Cardiovascular comorbidities with long-term immobilization, for example, due to spinal cord injury, do not have The association between VTE and atherosclerotic diseases such as myocardial in- an increased risk of VTE52. This has recently, following studies on hibernating bears, farction has led to discussions concerning whether risk factors for arterial diseases been proposed to be mediated by down-regulating the pro-thomboin ammatory state also predispose to VTE20. However, although some risk factors are common to both that immobilization normally induces. conditions, such as obesity and cigarette smoking, the most important association seems to be increased risk of VTE, mainly PE, after hospitalization for acute cardiac Sex differences disease40-42. The increased risk associated with cigarette smoking could be mediated There are important sex differences in VTE. Women have an increased risk of VTE by increased risk of myocardial infarction and cancer43. Stroke is also a well-recog- during fertile years, whereas men have a higher risk in middle age. The excess risk in nized risk factor of VTE, with a high occurrence of VTE in particular within the  rst fertile women has been reported to be associated with endogenous estrogen exposure months44 and with previous data indicating higher VTE rates for hemorrhagic than with higher VTE risk in women with late menopause and higher parity compared to ischemic stroke45. early or normal menopause and lower parity53. Pregnancy entails an increased risk of VTE, particularly in the third trimester, with a reported 6-fold risk increase compared Chronic obstructive pulmonary disease to time outside of pregnancy54. The risk was considerably lower during the  rst and PE is an important differential diagnosis in acute exacerbations of chronic obstructive second trimesters, while the postpartum period conferred an even higher risk increase. pulmonary disease (COPD), with a reported prevalence of 14.4% in a meta-analysis It has been proposed that men have a higher baseline risk of VTE than women but that of studies with a standardized protocol for examination of PE46. Previous data indicate female exposure to reproductive risk factors such as oral contraceptives, pregnancy/ that patients with COPD are at a higher risk of PE than DVT47. puerperium, and postmenopausal hormone therapy, out-balance this difference55. This aligns with the higher recurrence risk of VTE in men than women55. Surgery, trauma Surgery is among the most well-recognized risk factors for VTE, and the use of phar- Differences between DVT and PE macologic thromboprophylaxis is widely spread. This, together with re ned surgical Previous reports suggest that some risk factors are more strongly associated with DVT 41,56 techniques, earlier mobilization, and shorter hospital stays, has reduced the incidence than with PE, or contrarily, with PE than with DVT. For example, Factor V Leiden of postoperative VTE. For instance, patients undergoing major orthopedic surgery mutation is more often associated with DVT than with PE, which is sometimes called 56 before 1980 had a reported incidence of symptomatic VTE of around 30%, including the Factor V Leiden paradox . Although many patients with DVT have a concur- cases of fatal PE48, while the current untreated 35-day baseline risk has been calcu- rent asymptomatic PE and many patients with PE have a simultaneous asymptomatic 20 lated at 4.3% for major orthopedic surgery49. The usage of pharmacologic thrombo- DVT , there are differences in mortality between the two VTE manifestations. The prophylaxis reduces the risk by approximately 50%49. overall mortality during the  rst year following a  rst-time PE has been reported to be higher than after a  rst-time DVT, even when excluding the  rst month after diagno- 57 Thrombophilias sis . This suggests that it might be of greater importance to prevent a PE than a DVT Thrombophilias can be both inherited and acquired. Among widely recognized inher- and thereby to identify which patients are at the highest risk of a PE. ited thrombophilias are factor V Leiden (the most common mutation leading to acti- vated protein C-resistance), prothrombin G20210A mutation, de ciencies of protein Preventive measures C, protein S, and antithrombin50. Of these, factor V Leiden is the most common in Pharmacological thromboprophylaxis with low-molecular-weight heparin (or, in the Caucasians (5–10% are carriers), with a 5-fold risk increase for thrombosis in indi- case of orthopedic prosthetic surgery, DOACs) is the standard care in high-risk situ- viduals with a heterozygous mutation and a 50-fold increase in patients with a homo- ations. Such situations include major orthopedic and non-orthopedic surgery49,58. In 16 17 of being overweight or obese early in life on the future risk of VTE, important for zygous mutation. Prothrombin G20210A mutation exists almost exclusively in Cau- prognostication of future health care needs, has been scarcely studied. casians, among whom it has a prevalence of 2%, giving a risk increase of VTE of 2-3 times in heterozygotes and 30 times in homozygotes. Protein C, S, and antithrombin Cancer de ciencies are rare but strong risk factors with several different subtypes50. Antiphos- Cancer is an important risk factor for VTE, with around 20% of all VTE cases oc- pholipid syndrome is the most common acquired thrombophilia. It is an autoimmune curring in patients with cancer35. VTE can also be the  rst manifestation of occult disorder, which requires persisting positive antiphospholipid antibodies (lupus antico- cancer, detected in 5% of patients within a year of a seemingly unprovoked VTE36. agulant, anticardiolipin antibodies, or anti-2glykoprotein antibodies) in combination The association between VTE and cancer varies with the type, stage, and treatment of with clinical symptoms (thrombosis or obstetric complications) for diagnosis51. cancer28. Reports on high thrombogenicity of speci c cancer types, such as pancreatic and brain cancer, have been relatively consistent28. However, in other types of cancer, Immobilization such as kidney and bladder cancer, data is less clear37-39. Short-term immobilization is a strong risk factor for VTE, contributing to the increased risk of hospitalization, trauma, surgery, and orthopedic casting. However, patients Cardiovascular comorbidities with long-term immobilization, for example, due to spinal cord injury, do not have The association between VTE and atherosclerotic diseases such as myocardial in- an increased risk of VTE52. This has recently, following studies on hibernating bears, farction has led to discussions concerning whether risk factors for arterial diseases been proposed to be mediated by down-regulating the pro-thomboin ammatory state also predispose to VTE20. However, although some risk factors are common to both that immobilization normally induces. conditions, such as obesity and cigarette smoking, the most important association seems to be increased risk of VTE, mainly PE, after hospitalization for acute cardiac Sex differences disease40-42. The increased risk associated with cigarette smoking could be mediated There are important sex differences in VTE. Women have an increased risk of VTE by increased risk of myocardial infarction and cancer43. Stroke is also a well-recog- during fertile years, whereas men have a higher risk in middle age. The excess risk in nized risk factor of VTE, with a high occurrence of VTE in particular within the  rst fertile women has been reported to be associated with endogenous estrogen exposure months44 and with previous data indicating higher VTE rates for hemorrhagic than with higher VTE risk in women with late menopause and higher parity compared to ischemic stroke45. early or normal menopause and lower parity53. Pregnancy entails an increased risk of VTE, particularly in the third trimester, with a reported 6-fold risk increase compared Chronic obstructive pulmonary disease to time outside of pregnancy54. The risk was considerably lower during the  rst and PE is an important differential diagnosis in acute exacerbations of chronic obstructive second trimesters, while the postpartum period conferred an even higher risk increase. pulmonary disease (COPD), with a reported prevalence of 14.4% in a meta-analysis It has been proposed that men have a higher baseline risk of VTE than women but that of studies with a standardized protocol for examination of PE46. Previous data indicate female exposure to reproductive risk factors such as oral contraceptives, pregnancy/ that patients with COPD are at a higher risk of PE than DVT47. puerperium, and postmenopausal hormone therapy, out-balance this difference55. This aligns with the higher recurrence risk of VTE in men than women55. Surgery, trauma Surgery is among the most well-recognized risk factors for VTE, and the use of phar- Differences between DVT and PE macologic thromboprophylaxis is widely spread. This, together with re ned surgical Previous reports suggest that some risk factors are more strongly associated with DVT 41,56 techniques, earlier mobilization, and shorter hospital stays, has reduced the incidence than with PE, or contrarily, with PE than with DVT. For example, Factor V Leiden of postoperative VTE. For instance, patients undergoing major orthopedic surgery mutation is more often associated with DVT than with PE, which is sometimes called 56 before 1980 had a reported incidence of symptomatic VTE of around 30%, including the Factor V Leiden paradox . Although many patients with DVT have a concur- cases of fatal PE48, while the current untreated 35-day baseline risk has been calcu- rent asymptomatic PE and many patients with PE have a simultaneous asymptomatic 20 lated at 4.3% for major orthopedic surgery49. The usage of pharmacologic thrombo- DVT , there are differences in mortality between the two VTE manifestations. The prophylaxis reduces the risk by approximately 50%49. overall mortality during the  rst year following a  rst-time PE has been reported to be higher than after a  rst-time DVT, even when excluding the  rst month after diagno- 57 Thrombophilias sis . This suggests that it might be of greater importance to prevent a PE than a DVT Thrombophilias can be both inherited and acquired. Among widely recognized inher- and thereby to identify which patients are at the highest risk of a PE. ited thrombophilias are factor V Leiden (the most common mutation leading to acti- vated protein C-resistance), prothrombin G20210A mutation, de ciencies of protein Preventive measures C, protein S, and antithrombin50. Of these, factor V Leiden is the most common in Pharmacological thromboprophylaxis with low-molecular-weight heparin (or, in the Caucasians (5–10% are carriers), with a 5-fold risk increase for thrombosis in indi- case of orthopedic prosthetic surgery, DOACs) is the standard care in high-risk situ- viduals with a heterozygous mutation and a 50-fold increase in patients with a homo- ations. Such situations include major orthopedic and non-orthopedic surgery49,58. In 16 17 some situations, such as certain major orthopedic surgery, major abdominal or pelvic Table 1. Wells Clinical Prediction Rule for DVT and PE surgery in cancer patients, postoperative thromboprophylaxis is often extended to 28- 35 days49,58,59. For medical inpatients, there are large differences in the usage of throm- The Wells Clinical Prediction Rule for DVT boprophylaxis60,61. Landmark studies on the bene t of thromboprophylaxis in medical Items Points inpatients62,63 are two decades old, and the main bene t was seen in the prevention of Active cancer (treatment or palliation within 6 months) 1 asymptomatic VTEs. Unfortunately, however, reports on bene ts in clinical practice Paralysis, paresis or recent plaster immobilization of the lower are not consistent64-67, and it is unclear which medical inpatients bene t from throm- extremities 1 boprophylaxis. Consistent with this, there is a wide array of different risk assessment Recently bedridden for 3 days or more, or major surgery within tools that categorize a widely varying proportion of hospitalized patients as suitable the previous 12 weeks requiring general or regional anesthesia 1 for thromboprophylaxis68. Localized tenderness along the distribution of the deep venoussystem 1 Entire leg swollen 1 Cancer is one of the most important risk factors for VTE, and consequently, primary Calf swelling at least 3 cm larger than that on the asymptomatic prophylaxis has been studied in selected out-patient groups of cancer patients69,70. leg (measured 10 cm below the tibial tuberosity) 1 However, identifying patients who bene t most from this prophylaxis appears chal- Pitting oedema confined to the symptomatic leg 1 lenging71. Cancer patients have a higher risk of bleeding, a risk that may be increased Collateral superficial veins (nonvaricose) 1 with the administration of thromboprophylaxis69. In this context, it is important not Previously documented DVT 1 only to consider which cancer patients have a high risk of VTE but also which patients Alternative diagnosis at least as likely as DVT 2 do not. Clinical probability Two level score Graduated compression stockings are mainly used for thromboprophylaxis in acutely DVT unlikely 1 ill patients who are perceived to be at high risk of VTE but deemed to be unsuitable DVT likely 2 for pharmacological thromboprophylaxis due to a high risk of bleeding. In theory, graduated compression stockings reduce venous stasis and could thereby reduce the The Wells Clinical Prediction Rule for PE risk of VTE. However, in practice, the bene t of mechanical thromboprophylaxis re- Original Simplified mains unclear72. Items version version Previous PE or DVT 1.5 1 Clinical presentation and diagnosis Heart rate >100 beats per minute 1.5 1 Immobilization at least 3 days OR surgery in the The clinical presentation of PE is non-speci c and highly variable among patients. previous 4 weeks 1.5 1 Nevertheless, the most common signs and symptoms are dyspnea, hypoxemia, tachy- Hemoptysis 1 1 cardia, and pleuritic chest pain31. Hemoptysis is only present in a small proportion of Active cancer (treatment or palliation within 6 patients but should raise the suspicion of PE, whereas hemodynamic compromise, months) 1 1 confusion, and sudden death are all potential but non-speci c manifestations of a Clinical signs of DVT 3 1 more extensive clot burden31. Alternative diagnosis less likely than PE 3 1 Clinical probability Patients with DVT frequently present with leg swelling, pain, localized tenderness Two level score over deep veins, redness, and prominent super cial veins31. In rare cases, the venous PE unlikely 0–4 0–1 occlusion and swelling of the extremity become severe with arterial ischemia, called PE likely 5 2 phlegmasia cerulea dolens (cyanotic leg) or phlegmasia alba dolens (white leg). Both these conditions are associated with high amputation and death rates73,74. In patients strati ed to the DVT or PE-unlikely group, D-dimer measurement is rec- Diagnostics ommended. The negative predictive value of a low clinical probability according to Wells score in combination with a negative D-dimer is high, 99.5% (95% CI 98.4– Upon suspicion of PE and DVT, the  rst step is to assess clinical probability, either by 99.9%)for PE76. However, the positive predictive value of D-dimer is low since D- clinical judgment based on experience or by using a prediction algorithm20,75. Among dimer can be elevated for many other reasons such as infection, cancer, pregnancy, or the most frequently used prediction models for DVT and PE is the Wells score, see simply high age20. Therefore, D-dimer cannot be used to con rm VTE, and if D-dimer Table 1. is positive, diagnostic imaging is warranted20,75. 18 19 some situations, such as certain major orthopedic surgery, major abdominal or pelvic Table 1. Wells Clinical Prediction Rule for DVT and PE surgery in cancer patients, postoperative thromboprophylaxis is often extended to 28- 35 days49,58,59. For medical inpatients, there are large differences in the usage of throm- The Wells Clinical Prediction Rule for DVT boprophylaxis60,61. Landmark studies on the bene t of thromboprophylaxis in medical Items Points inpatients62,63 are two decades old, and the main bene t was seen in the prevention of Active cancer (treatment or palliation within 6 months) 1 asymptomatic VTEs. Unfortunately, however, reports on bene ts in clinical practice Paralysis, paresis or recent plaster immobilization of the lower are not consistent64-67, and it is unclear which medical inpatients bene t from throm- extremities 1 boprophylaxis. Consistent with this, there is a wide array of different risk assessment Recently bedridden for 3 days or more, or major surgery within tools that categorize a widely varying proportion of hospitalized patients as suitable the previous 12 weeks requiring general or regional anesthesia 1 for thromboprophylaxis68. Localized tenderness along the distribution of the deep venoussystem 1 Entire leg swollen 1 Cancer is one of the most important risk factors for VTE, and consequently, primary Calf swelling at least 3 cm larger than that on the asymptomatic prophylaxis has been studied in selected out-patient groups of cancer patients69,70. leg (measured 10 cm below the tibial tuberosity) 1 However, identifying patients who bene t most from this prophylaxis appears chal- Pitting oedema confined to the symptomatic leg 1 lenging71. Cancer patients have a higher risk of bleeding, a risk that may be increased Collateral superficial veins (nonvaricose) 1 with the administration of thromboprophylaxis69. In this context, it is important not Previously documented DVT 1 only to consider which cancer patients have a high risk of VTE but also which patients Alternative diagnosis at least as likely as DVT 2 do not. Clinical probability Two level score Graduated compression stockings are mainly used for thromboprophylaxis in acutely DVT unlikely 1 ill patients who are perceived to be at high risk of VTE but deemed to be unsuitable DVT likely 2 for pharmacological thromboprophylaxis due to a high risk of bleeding. In theory, graduated compression stockings reduce venous stasis and could thereby reduce the The Wells Clinical Prediction Rule for PE risk of VTE. However, in practice, the bene t of mechanical thromboprophylaxis re- Original Simplified mains unclear72. Items version version Previous PE or DVT 1.5 1 Clinical presentation and diagnosis Heart rate >100 beats per minute 1.5 1 Immobilization at least 3 days OR surgery in the The clinical presentation of PE is non-speci c and highly variable among patients. previous 4 weeks 1.5 1 Nevertheless, the most common signs and symptoms are dyspnea, hypoxemia, tachy- Hemoptysis 1 1 cardia, and pleuritic chest pain31. Hemoptysis is only present in a small proportion of Active cancer (treatment or palliation within 6 patients but should raise the suspicion of PE, whereas hemodynamic compromise, months) 1 1 confusion, and sudden death are all potential but non-speci c manifestations of a Clinical signs of DVT 3 1 more extensive clot burden31. Alternative diagnosis less likely than PE 3 1 Clinical probability Patients with DVT frequently present with leg swelling, pain, localized tenderness Two level score over deep veins, redness, and prominent super cial veins31. In rare cases, the venous PE unlikely 0–4 0–1 occlusion and swelling of the extremity become severe with arterial ischemia, called PE likely 5 2 phlegmasia cerulea dolens (cyanotic leg) or phlegmasia alba dolens (white leg). Both these conditions are associated with high amputation and death rates73,74. In patients strati ed to the DVT or PE-unlikely group, D-dimer measurement is rec- Diagnostics ommended. The negative predictive value of a low clinical probability according to Wells score in combination with a negative D-dimer is high, 99.5% (95% CI 98.4– Upon suspicion of PE and DVT, the  rst step is to assess clinical probability, either by 99.9%)for PE76. However, the positive predictive value of D-dimer is low since D- clinical judgment based on experience or by using a prediction algorithm20,75. Among dimer can be elevated for many other reasons such as infection, cancer, pregnancy, or the most frequently used prediction models for DVT and PE is the Wells score, see simply high age20. Therefore, D-dimer cannot be used to con rm VTE, and if D-dimer Table 1. is positive, diagnostic imaging is warranted20,75. 18 19 In patients strati ed to the DVT or PE-likely group, there is no use for D-dimer test- reduced, whereas the risks of bleeding were tripled with thrombolytic therapy78. In ing since the negative predictive value of a negative D-dimer in this group is not clinical practice, thrombolysis is used in patients with high-risk PE despite the lack suf ciently high76. Instead, diagnostic imaging is required for all patients. The most of strong scienti c evidence due to the imminent risk of death if thrombolysis is not frequently used imaging modality for PE is CTPA, and the most frequently used mo- administered22. dality for DVT diagnosis in the upper or lower extremities is ultrasonography20,75. Systemic thrombolysis for patients with intermediate high-risk PE was assessed in the The introduction of CTPA as the  rst-line diagnostic approach to PE instead of ven- randomized PEITHO trial79. This trial showed that thrombolytic treatment plus hepa- tilation-perfusion scans and invasive pulmonary angiography has revolutionized the rin versus placebo plus heparin led to a lower risk of hemodynamic decompensation diagnostic work-up of PE. CTPA is currently widely available in hospitals with an but at the cost of a higher bleeding rate and stroke (mainly hemorrhagic stroke). The emergency department, usually on a 24/7 basis. It is sensitive and capable of detecting conclusion of the trial was that great caution is warranted when considering thrombo- emboli down to the subsegmental level as well as being able to give information on lytic treatment for this group of patients. differential diagnosis20. The diagnostic imaging of DVT has also been simpli ed fol- lowing the introduction of ultrasound instead of phlebography (also called venogra- In DVT, local transcatheter thrombolytic treatment is used in severe cases such as phy) due to its non-invasive nature and the increasing use of point-of-care ultrasound phlegmasia cerulea dolens and in carefully selected young patients with extensive in emergency rooms77. clots in the iliac and common femoral veins associated with severe symptoms22. This is a lengthy and, for many patients, demanding procedure with largely unclear ben- Treatment e ts. In the largest randomized trial to date, pharmacomechanical catheter-directed thrombolysis in combination with anticoagulation compared to anticoagulation alone Treatment of VTE varies depending on the urgency of the VTE and the patient’s risk did not lead to a lower risk of post-thrombotic syndrome (PTS). Still, on average, the of bleeding. Anticoagulant treatment, which is the sole treatment for the majority of PTS was less severe. However, major bleeding was signi cantly more common in the patients with VTE, prevents thrombus growth and embolization in the acute phase as intervention group80. well as recurrence in the post-acute phase. In this situation, it is the patients’ endog- enous  brinolysis that dissolves the clot. However, in urgent situations, the thrombus must be dissolved more rapidly, and thrombolytic treatment is used. In addition, more Additional advanced treatment options in the acute setting advanced invasive treatment options can be used in select cases if clinical expertise Catheter-directed treatment (CDT) for PE, extracorporeal membrane oxygenation is available. (ECMO), surgical embolectomy, and vena cava  lters can be used in selected patients with VTE if the expertise is available. In PE, treatment selection is based on risk strati cation of short-term mortality due to PE. Patients with high-risk PE20 (massive PE according to North American classi ca- CDT for PE is an area of high interest with several ongoing trials81. The indication for tion22) have hemodynamic instability. Patients with intermediate-high risk (submas- CDT is patients with PE who require thrombolysis but have a contraindication to this sive according to North American classi cation) PE have right ventricle strain on treatment or patients for which thrombolysis has failed81. In patients with PE associ- imaging and myocardial injury with elevated biomarkers (often troponins). Patients ated with cardiac arrest or refractory circulatory collapse, using ECMO to maintain with intermediate-low risk have only one or none of the two, whereas the distinction circulation and oxygenate critical organs can be helpful. In these cases, ECMO is of- between intermediate-low and low-risk PE is based on signs of clinical severity and ten used to buy time to prepare for clot removal with CDT or surgical embolectomy20. comorbidities, which can be assessed by pulmonary embolism severity index (PESI/ Surgical embolectomy is a treatment option used almost exclusively in high-risk PE sPESI) or the Hestia criteria20. when other treatment options have failed. With this procedure, a sternotomy is per- formed, and the patient is on cardiopulmonary bypass during the procedure, in which Thrombolytic treatment a pulmonary arteriotomy is performed, and the clot is extracted82. Systemic thrombolysis is recommended for patients with high-risk PE20-22. In a meta- analysis of randomized controlled trials including 2,057 patients, the odds ratio of Surgical embolectomy or percutaneous thrombectomy for DVT are treatment options death in patients receiving systemic thrombolysis compared to heparin alone was re- when local thrombolysis has failed or in patients with a contraindication to throm-83,84 ported to be 0.59 (95% CI 0.36–0.96)78. The reduction was not signi cant when ex- bolysis. However, the evidence for these treatment options is weak . In patients cluding patients with high-risk PE. However, the results on high-risk PE were weak with PE or proximal DVT who have a contraindication to anticoagulant treatment or since the included studies were old (three studies from the 1970s and one from 1995) who have PE recurrence despite therapeutic anticoagulation, vena cava  lters can be with few patients (115 patients in total who received thrombolysis). Additionally, the an option. However, the bene t of vena cava  lters in terms of a lower incidence of reported bene t depended on a study of only four patients. In the PE group as a whole, PE is out-balanced by an increased risk of DVT, high complication rates, and lack of 20 the risks of death or treatment escalation, PE-related death, and PE recurrence were evidence of a mortality bene t . 20 21 In patients strati ed to the DVT or PE-likely group, there is no use for D-dimer test- reduced, whereas the risks of bleeding were tripled with thrombolytic therapy78. In ing since the negative predictive value of a negative D-dimer in this group is not clinical practice, thrombolysis is used in patients with high-risk PE despite the lack suf ciently high76. Instead, diagnostic imaging is required for all patients. The most of strong scienti c evidence due to the imminent risk of death if thrombolysis is not frequently used imaging modality for PE is CTPA, and the most frequently used mo- administered22. dality for DVT diagnosis in the upper or lower extremities is ultrasonography20,75. Systemic thrombolysis for patients with intermediate high-risk PE was assessed in the The introduction of CTPA as the  rst-line diagnostic approach to PE instead of ven- randomized PEITHO trial79. This trial showed that thrombolytic treatment plus hepa- tilation-perfusion scans and invasive pulmonary angiography has revolutionized the rin versus placebo plus heparin led to a lower risk of hemodynamic decompensation diagnostic work-up of PE. CTPA is currently widely available in hospitals with an but at the cost of a higher bleeding rate and stroke (mainly hemorrhagic stroke). The emergency department, usually on a 24/7 basis. It is sensitive and capable of detecting conclusion of the trial was that great caution is warranted when considering thrombo- emboli down to the subsegmental level as well as being able to give information on lytic treatment for this group of patients. differential diagnosis20. The diagnostic imaging of DVT has also been simpli ed fol- lowing the introduction of ultrasound instead of phlebography (also called venogra- In DVT, local transcatheter thrombolytic treatment is used in severe cases such as phy) due to its non-invasive nature and the increasing use of point-of-care ultrasound phlegmasia cerulea dolens and in carefully selected young patients with extensive in emergency rooms77. clots in the iliac and common femoral veins associated with severe symptoms22. This is a lengthy and, for many patients, demanding procedure with largely unclear ben- Treatment e ts. In the largest randomized trial to date, pharmacomechanical catheter-directed thrombolysis in combination with anticoagulation compared to anticoagulation alone Treatment of VTE varies depending on the urgency of the VTE and the patient’s risk did not lead to a lower risk of post-thrombotic syndrome (PTS). Still, on average, the of bleeding. Anticoagulant treatment, which is the sole treatment for the majority of PTS was less severe. However, major bleeding was signi cantly more common in the patients with VTE, prevents thrombus growth and embolization in the acute phase as intervention group80. well as recurrence in the post-acute phase. In this situation, it is the patients’ endog- enous  brinolysis that dissolves the clot. However, in urgent situations, the thrombus must be dissolved more rapidly, and thrombolytic treatment is used. In addition, more Additional advanced treatment options in the acute setting advanced invasive treatment options can be used in select cases if clinical expertise Catheter-directed treatment (CDT) for PE, extracorporeal membrane oxygenation is available. (ECMO), surgical embolectomy, and vena cava  lters can be used in selected patients with VTE if the expertise is available. In PE, treatment selection is based on risk strati cation of short-term mortality due to PE. Patients with high-risk PE20 (massive PE according to North American classi ca- CDT for PE is an area of high interest with several ongoing trials81. The indication for tion22) have hemodynamic instability. Patients with intermediate-high risk (submas- CDT is patients with PE who require thrombolysis but have a contraindication to this sive according to North American classi cation) PE have right ventricle strain on treatment or patients for which thrombolysis has failed81. In patients with PE associ- imaging and myocardial injury with elevated biomarkers (often troponins). Patients ated with cardiac arrest or refractory circulatory collapse, using ECMO to maintain with intermediate-low risk have only one or none of the two, whereas the distinction circulation and oxygenate critical organs can be helpful. In these cases, ECMO is of- between intermediate-low and low-risk PE is based on signs of clinical severity and ten used to buy time to prepare for clot removal with CDT or surgical embolectomy20. comorbidities, which can be assessed by pulmonary embolism severity index (PESI/ Surgical embolectomy is a treatment option used almost exclusively in high-risk PE sPESI) or the Hestia criteria20. when other treatment options have failed. With this procedure, a sternotomy is per- formed, and the patient is on cardiopulmonary bypass during the procedure, in which Thrombolytic treatment a pulmonary arteriotomy is performed, and the clot is extracted82. Systemic thrombolysis is recommended for patients with high-risk PE20-22. In a meta- analysis of randomized controlled trials including 2,057 patients, the odds ratio of Surgical embolectomy or percutaneous thrombectomy for DVT are treatment options death in patients receiving systemic thrombolysis compared to heparin alone was re- when local thrombolysis has failed or in patients with a contraindication to throm-83,84 ported to be 0.59 (95% CI 0.36–0.96)78. The reduction was not signi cant when ex- bolysis. However, the evidence for these treatment options is weak . In patients cluding patients with high-risk PE. However, the results on high-risk PE were weak with PE or proximal DVT who have a contraindication to anticoagulant treatment or since the included studies were old (three studies from the 1970s and one from 1995) who have PE recurrence despite therapeutic anticoagulation, vena cava  lters can be with few patients (115 patients in total who received thrombolysis). Additionally, the an option. However, the bene t of vena cava  lters in terms of a lower incidence of reported bene t depended on a study of only four patients. In the PE group as a whole, PE is out-balanced by an increased risk of DVT, high complication rates, and lack of 20 the risks of death or treatment escalation, PE-related death, and PE recurrence were evidence of a mortality bene t . 20 21 Anticoagulant treatment Duration of anticoagulant treatment As previously described,  rst-line treatment for VTE was for many decades LMWH/ Initial treatment heparin and VKA. LMWH activates antithrombin, leading to accelerated interac- Initial treatment is the minimal treatment duration needed for VTE before deciding tion with factor Xa and, to some extent, thrombin (factor II). Unfractionated heparin on extended treatment or termination of treatment. Numerous trials have examined (UFH) has a similar mechanism but inhibits thrombin to a greater extent than LMWH the optimal length of initial treatment after a VTE. A pooled analysis of individual due to its longer heparin chain, enabling it to bind both antithrombin and thrombin, participants’ data from randomized trials concluded that patients with 1–1.5 months which is needed for the thrombin-related effect85. VKA inhibits vitamin K, leading of treatment after a VTE had a higher risk of recurrence after stopping treatment than to the K-vitamin dependent coagulation factors II, VII, IX, and X and anticoagulant patients who completed three months of treatment87. There was no difference in recur- proteins C, S, and Z being functionally incompentent86. rence after cessation of treatment in patients who completed three or six months or longer except for a borderline signi cance in patients with unprovoked VTE, where In 2014, the Swedish Medical Product Agency approved the  rst DOAC, rivaroxa- six months of treatment seemed more favorable than three months. The recurrence ban (Xarelto), for VTE treatment. Rivaroxaban was followed by apixaban (Eliquis), rate after treatment termination was highest during the  rst six months after cessation dabigatran (Pradaxa) and edoxaban (Lixiana). Rivaroxaban, apixaban, and edoxaban of treatment. Hence, in patients who do not have a high enough risk for recurrence are direct factor Xa-inhibitors, preventing the conversion of prothrombin to throm- to merit inde nite treatment, three months of treatment seems to suf ce for many bin, whereas dabigatran is a direct thrombin inhibitor, preventing the conversion of patients. However, in clinical practice in Sweden, many patients, particularly those  brinogen to  brin. DOACs are not only associated with lower bleeding risks but also who suffer from PE, receive six months of initial treatment before assessing residual are easier for both patients and caregivers since they do not require monitoring with symptoms, and a decision on treatment duration is made. blood samples. Therefore, their introduction has led to a drastic change in the treat- ment patterns, see Figure 1. DOACs are now  rst-line treatment in most patients with Extended treatment VTE except in pregnant patients (requiring LMWH), breastfeeding patients (LMWH or warfarin) patients with antiphospholipid syndrome (in particular triple positive, Anticoagulant treatment effectively reduces the risk of recurrent thromboembolic 88 requiring warfarin)20, patients with certain cancers (LMWH)59, or patients with other events during treatment . However, the risk of recurrence after cessation of treatment 89 speci c indications for warfarin treatment such as mechanical heart valves. is highly dependent on the circumstances of the initial VTE , see Table 2. In cases Table 2. Risk of recurrence after termination of anticoagulant treatment based on risk factors present at the initial VTE event. From 2019 ESC guidelines on acute pulmonary embolism20, Reprinted with permission from Oxford University Press Figure 1. Temporal changes in expeditions of anticoagulant treatment for VTE in Swe- den. Based data on fi lled prescriptions from the Prescribed Drug Register, Study IV. 22 23 Anticoagulant treatment Duration of anticoagulant treatment As previously described,  rst-line treatment for VTE was for many decades LMWH/ Initial treatment heparin and VKA. LMWH activates antithrombin, leading to accelerated interac- Initial treatment is the minimal treatment duration needed for VTE before deciding tion with factor Xa and, to some extent, thrombin (factor II). Unfractionated heparin on extended treatment or termination of treatment. Numerous trials have examined (UFH) has a similar mechanism but inhibits thrombin to a greater extent than LMWH the optimal length of initial treatment after a VTE. A pooled analysis of individual due to its longer heparin chain, enabling it to bind both antithrombin and thrombin, participants’ data from randomized trials concluded that patients with 1–1.5 months which is needed for the thrombin-related effect85. VKA inhibits vitamin K, leading of treatment after a VTE had a higher risk of recurrence after stopping treatment than to the K-vitamin dependent coagulation factors II, VII, IX, and X and anticoagulant patients who completed three months of treatment87. There was no difference in recur- proteins C, S, and Z being functionally incompentent86. rence after cessation of treatment in patients who completed three or six months or longer except for a borderline signi cance in patients with unprovoked VTE, where In 2014, the Swedish Medical Product Agency approved the  rst DOAC, rivaroxa- six months of treatment seemed more favorable than three months. The recurrence ban (Xarelto), for VTE treatment. Rivaroxaban was followed by apixaban (Eliquis), rate after treatment termination was highest during the  rst six months after cessation dabigatran (Pradaxa) and edoxaban (Lixiana). Rivaroxaban, apixaban, and edoxaban of treatment. Hence, in patients who do not have a high enough risk for recurrence are direct factor Xa-inhibitors, preventing the conversion of prothrombin to throm- to merit inde nite treatment, three months of treatment seems to suf ce for many bin, whereas dabigatran is a direct thrombin inhibitor, preventing the conversion of patients. However, in clinical practice in Sweden, many patients, particularly those  brinogen to  brin. DOACs are not only associated with lower bleeding risks but also who suffer from PE, receive six months of initial treatment before assessing residual are easier for both patients and caregivers since they do not require monitoring with symptoms, and a decision on treatment duration is made. blood samples. Therefore, their introduction has led to a drastic change in the treat- ment patterns, see Figure 1. DOACs are now  rst-line treatment in most patients with Extended treatment VTE except in pregnant patients (requiring LMWH), breastfeeding patients (LMWH or warfarin) patients with antiphospholipid syndrome (in particular triple positive, Anticoagulant treatment effectively reduces the risk of recurrent thromboembolic 88 requiring warfarin)20, patients with certain cancers (LMWH)59, or patients with other events during treatment . However, the risk of recurrence after cessation of treatment 89 speci c indications for warfarin treatment such as mechanical heart valves. is highly dependent on the circumstances of the initial VTE , see Table 2. In cases Table 2. Risk of recurrence after termination of anticoagulant treatment based on risk factors present at the initial VTE event. From 2019 ESC guidelines on acute pulmonary embolism20, Reprinted with permission from Oxford University Press Figure 1. Temporal changes in expeditions of anticoagulant treatment for VTE in Swe- den. Based data on fi lled prescriptions from the Prescribed Drug Register, Study IV. 22 23 where the initial VTE was provoked by a temporary major risk factor, such as major Prognosis surgery or major trauma with fractures, the recurrence risk is low after cessation of treatment, and most patients can stop treatment after 3–6 months87. However, when Mortality no provoking factor can be identi ed (also called unprovoked or idiopathic VTE), the The 30-day adjusted mortality rate ratios (MRR) compared to population controls recurrence rate is high, and guidelines recommend considering extended therapy in has been reported to be 33.0 (95% CI 31.6–34.5) for VTE, with a signi cant dif- most of these patients who do not have a high risk of bleeding20-22. ference between DVT (MRR 5.4 [95% CI 5.0–5.8]) and PE (MRR 80.9 [95% CI 76.0–86.0])57. From 2000 to 2015, the age-standardized annual mortality rate from The risk of recurrence in patients without provoking factors has been estimated to be PE (registered as the primary cause of death in the WHO European Region Mortality 10% during the  rst year after cessation of treatment, 25% after  ve years, and 36% database) has decreased linearly from 12.8 (95% CI 11.4–14.2) deaths per 100,000 after ten years90. The overall risk of recurrence was 1.4 (95% CI 1.3–1.6) times higher population in 2000 to 6.5 (95% CI 5.3–7.7) deaths per 100,000 population in 2015100. in men compared to women. For patients with non-surgical provoking factors, the re- Among the proposed reasons for this change are improved disease management and a currence rate during the  rst year after cessation of anticoagulation has been reported lower autopsy rate, the latter of which could lead to a lower detection rate. However, to be 5.8% (95% CI 3.2–8.3%)91. The lowest recurrence rate after cessation of treat- data on mortality trends are not unanimous. In a similar study on data from 2000 to ment has been reported for patients with an initial VTE provoked by a surgical risk 2017 from the WHO Mortality Database from North America, an increase in PE- factor, 1% (95% CI 0–2.3%) during the  rst year91. However, the risk of recurrence related mortality in the USA among young and middle-aged adults was seen after during extended treatment of an unprovoked VTE event (after initial treatment of 3 2006101. Proposed mechanisms for this are increasing inequities in risk factors and months) is not negligible. In a meta-analysis of prospective cohort studies and ran- availability of health care. The age-standardized annual mortality rate in the USA in domized controlled trials (RCTs), it was estimated to be 3.3% (95% CI 2.0%–5.1%) 2017 was 4.1 [95% CI 4.0-4.2] per 100,000 inhabitants for women and 4.5 [95% CI during the  rst two years and 7.1% (95% CI 3.0%–13.2%) after  ve years92. 4.4–4.7] per 100,000 for men. Risk of bleeding on anticoagulant treatment Long-term complications In decisions on treatment duration, the risk of recurrence needs to be balanced against Complications of PE the risk of bleeding for each patient. Given the high recurrence rates in many VTE patients and the reduced risk of bleeding of DOACs compared to VKA, a large num- Persisting dyspnea is common after PE, with a reported prevalence of around 50% of 102,103 ber of patients bene t from extended treatment. However, there are still uncertainties PE survivors after six months to three years of anticoagulation . In 0.5–4% of pa- concerning the long-term risk of bleeding in the extended treatment of VTE93. tients suffering from PE, chronic thromboembolic pulmonary hypertension (CTEPH) develops102. In patients with this condition, thrombi in the pulmonary vasculature are The risk of major bleeding (as de ned by the ISTH)94 for patients on initial anticoagu- not dissolved but organized into  brous obstructions of the arteries. Distal arteries lant treatment for VTE in phase III RCTs has been reported to be 1.1% for DOACs in both affected and unaffected areas of the lungs are remodeled, leading to an in-102 and 1.7% for Warfarin95. In real-world data, the risk of major bleeding (bleeding re- crease in pulmonary artery pressure . Patients develop progressive right heart failure quiring hospitalization, a wider de nition compared to the phase III trials) during the with high mortality unless treated. Patients with CTEPH are recommended treatment  rst six months of treatment for VTE patients without cancer has been reported to be with life-long anticoagulation to prevent additional thrombosis. VKA as a  rst-hand 2.496– 4.097 per 100 patient-years for rivaroxaban, 1.997–4.298 per 100 patient-years for treatment choice is increasingly being replaced by DOACs, except in patients with 104 apixaban and 2.096–5.598 per 100 patient-years for warfarin. However, available data antiphospholipid antibodies (10% of CTEPH patients) . In cases with surgically ac- are sparse for extended treatment. In a meta-analysis of RCTs and prospective cohort cessible  brotic lesions, pulmonary endarterectomy is recommended if the patient studies, the  ve-year cumulative incidence of major bleeding with VKA was 6.3% is considered operable. Pulmonary endarterectomy, in which  brotic lesions are re- (95% CI 3.6%–10.0%), whereas data were insuf cient to estimate bleeding incidence moved down to subsegmental arteries, is associated with improved quality of life and 104 for patients on extended DOAC treatment beyond one year93. improved 5-year survival of 83% compared to 53% without surgery . One study on real-world data on extended treatment beyond 90 days of initial treat- Complications of DVT ment showed a rate of major bleeding (de ned as bleeding requiring hospitalization) Post-thrombotic syndrome (PTS) is a common complication of DVT, affecting about 105 of 44.5 per 1000 person-years for apixaban, 50.0 for rivaroxaban, and 47.1 for war- half of all DVT patients 1–2 years after the initial event . The risk of PTS is higher farin99. However, in this study, more than one-fourth of patients had cancer, making in patients with recurrent ipsilateral DVT and proximal DVT (in particular iliofemoral the results dif cult to generalize to non-cancer patients. Studies on the risk of major DVT). Following an acute DVT, an in ammatory response and recanalization of the bleeding during extended treatment in different patient groups in a real-world setting vein starts. Both these events have been described to damage venous valves, result- are needed for better risk-bene t analysis in treatment decisions. ing in valvular re ux. The re ux, in combination with residual obstruction, increases 24 25 where the initial VTE was provoked by a temporary major risk factor, such as major Prognosis surgery or major trauma with fractures, the recurrence risk is low after cessation of treatment, and most patients can stop treatment after 3–6 months87. However, when Mortality no provoking factor can be identi ed (also called unprovoked or idiopathic VTE), the The 30-day adjusted mortality rate ratios (MRR) compared to population controls recurrence rate is high, and guidelines recommend considering extended therapy in has been reported to be 33.0 (95% CI 31.6–34.5) for VTE, with a signi cant dif- most of these patients who do not have a high risk of bleeding20-22. ference between DVT (MRR 5.4 [95% CI 5.0–5.8]) and PE (MRR 80.9 [95% CI 76.0–86.0])57. From 2000 to 2015, the age-standardized annual mortality rate from The risk of recurrence in patients without provoking factors has been estimated to be PE (registered as the primary cause of death in the WHO European Region Mortality 10% during the  rst year after cessation of treatment, 25% after  ve years, and 36% database) has decreased linearly from 12.8 (95% CI 11.4–14.2) deaths per 100,000 after ten years90. The overall risk of recurrence was 1.4 (95% CI 1.3–1.6) times higher population in 2000 to 6.5 (95% CI 5.3–7.7) deaths per 100,000 population in 2015100. in men compared to women. For patients with non-surgical provoking factors, the re- Among the proposed reasons for this change are improved disease management and a currence rate during the  rst year after cessation of anticoagulation has been reported lower autopsy rate, the latter of which could lead to a lower detection rate. However, to be 5.8% (95% CI 3.2–8.3%)91. The lowest recurrence rate after cessation of treat- data on mortality trends are not unanimous. In a similar study on data from 2000 to ment has been reported for patients with an initial VTE provoked by a surgical risk 2017 from the WHO Mortality Database from North America, an increase in PE- factor, 1% (95% CI 0–2.3%) during the  rst year91. However, the risk of recurrence related mortality in the USA among young and middle-aged adults was seen after during extended treatment of an unprovoked VTE event (after initial treatment of 3 2006101. Proposed mechanisms for this are increasing inequities in risk factors and months) is not negligible. In a meta-analysis of prospective cohort studies and ran- availability of health care. The age-standardized annual mortality rate in the USA in domized controlled trials (RCTs), it was estimated to be 3.3% (95% CI 2.0%–5.1%) 2017 was 4.1 [95% CI 4.0-4.2] per 100,000 inhabitants for women and 4.5 [95% CI during the  rst two years and 7.1% (95% CI 3.0%–13.2%) after  ve years92. 4.4–4.7] per 100,000 for men. Risk of bleeding on anticoagulant treatment Long-term complications In decisions on treatment duration, the risk of recurrence needs to be balanced against Complications of PE the risk of bleeding for each patient. Given the high recurrence rates in many VTE patients and the reduced risk of bleeding of DOACs compared to VKA, a large num- Persisting dyspnea is common after PE, with a reported prevalence of around 50% of 102,103 ber of patients bene t from extended treatment. However, there are still uncertainties PE survivors after six months to three years of anticoagulation . In 0.5–4% of pa- concerning the long-term risk of bleeding in the extended treatment of VTE93. tients suffering from PE, chronic thromboembolic pulmonary hypertension (CTEPH) develops102. In patients with this condition, thrombi in the pulmonary vasculature are The risk of major bleeding (as de ned by the ISTH)94 for patients on initial anticoagu- not dissolved but organized into  brous obstructions of the arteries. Distal arteries lant treatment for VTE in phase III RCTs has been reported to be 1.1% for DOACs in both affected and unaffected areas of the lungs are remodeled, leading to an in-102 and 1.7% for Warfarin95. In real-world data, the risk of major bleeding (bleeding re- crease in pulmonary artery pressure . Patients develop progressive right heart failure quiring hospitalization, a wider de nition compared to the phase III trials) during the with high mortality unless treated. Patients with CTEPH are recommended treatment  rst six months of treatment for VTE patients without cancer has been reported to be with life-long anticoagulation to prevent additional thrombosis. VKA as a  rst-hand 2.496– 4.097 per 100 patient-years for rivaroxaban, 1.997–4.298 per 100 patient-years for treatment choice is increasingly being replaced by DOACs, except in patients with 104 apixaban and 2.096–5.598 per 100 patient-years for warfarin. However, available data antiphospholipid antibodies (10% of CTEPH patients) . In cases with surgically ac- are sparse for extended treatment. In a meta-analysis of RCTs and prospective cohort cessible  brotic lesions, pulmonary endarterectomy is recommended if the patient studies, the  ve-year cumulative incidence of major bleeding with VKA was 6.3% is considered operable. Pulmonary endarterectomy, in which  brotic lesions are re- (95% CI 3.6%–10.0%), whereas data were insuf cient to estimate bleeding incidence moved down to subsegmental arteries, is associated with improved quality of life and 104 for patients on extended DOAC treatment beyond one year93. improved 5-year survival of 83% compared to 53% without surgery . One study on real-world data on extended treatment beyond 90 days of initial treat- Complications of DVT ment showed a rate of major bleeding (de ned as bleeding requiring hospitalization) Post-thrombotic syndrome (PTS) is a common complication of DVT, affecting about 105 of 44.5 per 1000 person-years for apixaban, 50.0 for rivaroxaban, and 47.1 for war- half of all DVT patients 1–2 years after the initial event . The risk of PTS is higher farin99. However, in this study, more than one-fourth of patients had cancer, making in patients with recurrent ipsilateral DVT and proximal DVT (in particular iliofemoral the results dif cult to generalize to non-cancer patients. Studies on the risk of major DVT). Following an acute DVT, an in ammatory response and recanalization of the bleeding during extended treatment in different patient groups in a real-world setting vein starts. Both these events have been described to damage venous valves, result- are needed for better risk-bene t analysis in treatment decisions. ing in valvular re ux. The re ux, in combination with residual obstruction, increases 24 25 venous pressure, leading to edema and, in more severe cases,  brosis, hypoxia, and AIM ulceration of the leg. Elastic compression stockings have been used for patients with DVT to reduce the This thesis explored risk factors for  rst-time VTE and bleeding risk during antico- risk of PTS. However, in the randomized SOX trial106, no bene t was shown in pre- agulant treatment in patients with a  rst-time VTE. venting PTS using elastic compression stockings compared to placebo stockings in patients with a  rst proximal DVT. Hence, guidelines do not routinely recommend The speci c aims of the separate papers were: compression stockings to all DVT patients22. However, in the event of established PTS, compression stockings are the  rst-line treatment. - P aper I: To explore the relationship between BMI levels in young adulthood and later development of VTE using a long-term follow-up of a large population of men from the Swedish Military Service Conscription Register. - Paper II: To determine age-speci c incidence rates of PE and DVT and the preva- lence of comorbidities and temporary provoking factors at the time of a  rst-time PE or DVT compared with a matched control population. - P aper III: To determine the incidence of VTE in association with a recent diag- nosis of cancer and provide information on the sex and age distributions for both all cancers and separate cancer types in patients with VTE. We also aimed to estimate sex-speci c odds ratios (ORs) for various cancers in patients with VTE compared to matched population controls. - P aper IV: To describe the risk of major bleeding depending on the choice of anticoagulant treatment, both during initial VTE treatment (0–6 months) and ex- tended treatment (6 months up to 5 years) in patients with a  rst-time VTE. We also aimed to identify risk factors for predicting an increased risk of bleeding during initial and extended anticoagulant therapy. 26 27 venous pressure, leading to edema and, in more severe cases,  brosis, hypoxia, and AIM ulceration of the leg. Elastic compression stockings have been used for patients with DVT to reduce the This thesis explored risk factors for  rst-time VTE and bleeding risk during antico- risk of PTS. However, in the randomized SOX trial106, no bene t was shown in pre- agulant treatment in patients with a  rst-time VTE. venting PTS using elastic compression stockings compared to placebo stockings in patients with a  rst proximal DVT. Hence, guidelines do not routinely recommend The speci c aims of the separate papers were: compression stockings to all DVT patients22. However, in the event of established PTS, compression stockings are the  rst-line treatment. - P aper I: To explore the relationship between BMI levels in young adulthood and later development of VTE using a long-term follow-up of a large population of men from the Swedish Military Service Conscription Register. - Paper II: To determine age-speci c incidence rates of PE and DVT and the preva- lence of comorbidities and temporary provoking factors at the time of a  rst-time PE or DVT compared with a matched control population. - P aper III: To determine the incidence of VTE in association with a recent diag- nosis of cancer and provide information on the sex and age distributions for both all cancers and separate cancer types in patients with VTE. We also aimed to estimate sex-speci c odds ratios (ORs) for various cancers in patients with VTE compared to matched population controls. - P aper IV: To describe the risk of major bleeding depending on the choice of anticoagulant treatment, both during initial VTE treatment (0–6 months) and ex- tended treatment (6 months up to 5 years) in patients with a  rst-time VTE. We also aimed to identify risk factors for predicting an increased risk of bleeding during initial and extended anticoagulant therapy. 26 27 PATIENTS AND METHODS years of age (from 2010, all registered inhabitants from the age of 15). It includes variables such as education, migration, employment status, income, and sick leave. Study populations The completeness of education data has been reported to be >98% with an estimated accuracy for the highest achieved level of education of 85%109. Paper I The National Patient Register Paper I was based on a cohort of all men enlisting for military service between 1969– 2005, registered in the Swedish Military Service Conscription Register. Sweden has a universal healthcare system that provides low-cost out-patient and hos- pital care to all citizens. The National Patient Register includes the National Inpatient Papers II and III Register, which has a coverage that increased gradually from 1964 and has been com- plete since 1987, and the National Outpatient Register, which has existed since 2001. Papers II and III were based on a cohort of all VTE patients with a  rst-time DVT or The outpatient register has had lower coverage than the inpatient register, approxi- PE in Sweden registered in the National Patient Register or National Cause of Death mately 80% up until 2007, mainly due to missing data from private caregivers. The Register between 1987–2018 and their matched population controls from the Total proportion of outpatient visits without a valid principal diagnosis has declined since Population Register. then and was at only 2% in 2021110. The National Patient Register covers data from all hospital care and specialized outpatient clinics in Sweden, but not primary care111. Paper IV Paper IV was based on all patients with a  rst-time DVT or PE in the National Patient The National Cause of Death Register Register 2014–2020. Registration of death has two parts in Sweden; Noti cation of death, which is sent by a physician to the Swedish Tax Agency immediately following the con rmation Registers of death, and the Medical death certi cate (with information on the cause of death), The Swedish Military Service Conscription Register which is sent to the National Board of Health and Welfare within three weeks of the death. The register includes all men enlisting for compulsory military service in Sweden between 1969 and 2006. Exemption from conscription was granted only for 1) Swed- The National Cause of Death Register includes the date of death and cause of death ish citizens living abroad, 2) men with certain psychiatric disorders, 3) men receiving for all citizens of Sweden111. A vast majority, around 96%, of all individuals in Sweden assistance allowance (or with parents receiving care allowance for the youth), 4) men are assigned a speci c cause of death code. A minor proportion of persons who die, receiving support for physical impairments, and 5) men who were members of Jeho- mainly older individuals with multiple diseases, receive unspeci c codes since the vah’s Witness congregations. During the period 1969–2006, about 90% of all men of speci c cause of death is dif cult to determine. Individuals with obvious or suspected eligible age were tested107. unnatural deaths, such as suicides or homicides, or obscure cases, such as previously healthy individuals who die suddenly, are reported to the police authorities by the During conscription, height and weight were measured and used to calculate Body physician who con rms the death. Of these, 95% undergo forensic autopsy112, and the mass index (BMI): weight / height2. Blood pressure was measured after 5–10 min cause of death is reported to the National Cause of Death Register. of supine rest. Cognitive testing protocols changed over time but included several domains, such as spatial capacity and technical understanding. Muscle strength was The National Prescribed Drug Register previously assessed with strength in knee extension, elbow  exion, and handgrip, but since 1994, only with one vertical lifting procedure with resistance depending on The National Prescribed Drug Register includes virtually all prescription medications 113 the strength of the person performing the lift107,108. Cognitive evaluation and muscle Swedish pharmacies have dispensed since July 2005 . The register does not include strength test results were standardized and transformed into STAndard NINE (stan- over-the-counter medications or drugs such as antitumoral agents administered in ine) scores (0–9, normally distributed). Both were divided into low (0–3), medium hospital daycare facilities. Overall, the register has been reported to include 84% of (4–6) and high (7–9). Maximum work capacity was tested with a bicycle ergometer, all drugs sold in the country. divided by weight, and transformed into scores 0–9, where 0–4 was considered low, 5–6 medium, and 7–9 high. The Total Population Register This register consists of all Swedish inhabitants since 1968 and includes information The Longitudinal Integration Database for Health Insurance and Labor Mar- on birthdate, date of death, emigration, marital status, and area of residence114. Data ket Services (LISA) register quality is regarded as high, particularly for births, deaths, and civil status, which pro- This register was established in 1990 and includes all registered inhabitants from 16 fessionals usually report. However, data on residence or migration might be of lower 28 29 PATIENTS AND METHODS years of age (from 2010, all registered inhabitants from the age of 15). It includes variables such as education, migration, employment status, income, and sick leave. Study populations The completeness of education data has been reported to be >98% with an estimated accuracy for the highest achieved level of education of 85%109. Paper I The National Patient Register Paper I was based on a cohort of all men enlisting for military service between 1969– 2005, registered in the Swedish Military Service Conscription Register. Sweden has a universal healthcare system that provides low-cost out-patient and hos- pital care to all citizens. The National Patient Register includes the National Inpatient Papers II and III Register, which has a coverage that increased gradually from 1964 and has been com- plete since 1987, and the National Outpatient Register, which has existed since 2001. Papers II and III were based on a cohort of all VTE patients with a  rst-time DVT or The outpatient register has had lower coverage than the inpatient register, approxi- PE in Sweden registered in the National Patient Register or National Cause of Death mately 80% up until 2007, mainly due to missing data from private caregivers. The Register between 1987–2018 and their matched population controls from the Total proportion of outpatient visits without a valid principal diagnosis has declined since Population Register. then and was at only 2% in 2021110. The National Patient Register covers data from all hospital care and specialized outpatient clinics in Sweden, but not primary care111. Paper IV Paper IV was based on all patients with a  rst-time DVT or PE in the National Patient The National Cause of Death Register Register 2014–2020. Registration of death has two parts in Sweden; Noti cation of death, which is sent by a physician to the Swedish Tax Agency immediately following the con rmation Registers of death, and the Medical death certi cate (with information on the cause of death), The Swedish Military Service Conscription Register which is sent to the National Board of Health and Welfare within three weeks of the death. The register includes all men enlisting for compulsory military service in Sweden between 1969 and 2006. Exemption from conscription was granted only for 1) Swed- The National Cause of Death Register includes the date of death and cause of death ish citizens living abroad, 2) men with certain psychiatric disorders, 3) men receiving for all citizens of Sweden111. A vast majority, around 96%, of all individuals in Sweden assistance allowance (or with parents receiving care allowance for the youth), 4) men are assigned a speci c cause of death code. A minor proportion of persons who die, receiving support for physical impairments, and 5) men who were members of Jeho- mainly older individuals with multiple diseases, receive unspeci c codes since the vah’s Witness congregations. During the period 1969–2006, about 90% of all men of speci c cause of death is dif cult to determine. Individuals with obvious or suspected eligible age were tested107. unnatural deaths, such as suicides or homicides, or obscure cases, such as previously healthy individuals who die suddenly, are reported to the police authorities by the During conscription, height and weight were measured and used to calculate Body physician who con rms the death. Of these, 95% undergo forensic autopsy112, and the mass index (BMI): weight / height2. Blood pressure was measured after 5–10 min cause of death is reported to the National Cause of Death Register. of supine rest. Cognitive testing protocols changed over time but included several domains, such as spatial capacity and technical understanding. Muscle strength was The National Prescribed Drug Register previously assessed with strength in knee extension, elbow  exion, and handgrip, but since 1994, only with one vertical lifting procedure with resistance depending on The National Prescribed Drug Register includes virtually all prescription medications 113 the strength of the person performing the lift107,108. Cognitive evaluation and muscle Swedish pharmacies have dispensed since July 2005 . The register does not include strength test results were standardized and transformed into STAndard NINE (stan- over-the-counter medications or drugs such as antitumoral agents administered in ine) scores (0–9, normally distributed). Both were divided into low (0–3), medium hospital daycare facilities. Overall, the register has been reported to include 84% of (4–6) and high (7–9). Maximum work capacity was tested with a bicycle ergometer, all drugs sold in the country. divided by weight, and transformed into scores 0–9, where 0–4 was considered low, 5–6 medium, and 7–9 high. The Total Population Register This register consists of all Swedish inhabitants since 1968 and includes information The Longitudinal Integration Database for Health Insurance and Labor Mar- on birthdate, date of death, emigration, marital status, and area of residence114. Data ket Services (LISA) register quality is regarded as high, particularly for births, deaths, and civil status, which pro- This register was established in 1990 and includes all registered inhabitants from 16 fessionals usually report. However, data on residence or migration might be of lower 28 29 quality since it is reported by the individual, who for various reasons might fail to Paper II report a change of residence. DVT and PE diagnostic codes were identical as in Paper I. Only  rst-time diagnoses Defi nitions were included. For diagnoses registered before July 1, 2005, a VTE diagnosis was de ned as either a VTE diagnosis in the inpatient or National Cause of Death Register Paper I or one outpatient diagnosis of VTE and a subsequent, identical diagnosis within three months. For patients with VTE after July 1, 2005, the de nition was the following: Data from the Swedish Military Service Conscription Register 1969–2005 were used 1) one inpatient diagnosis and a  lling of a prescription of anticoagulants within six for baseline data. Information about the highest achieved parental education (a proxy months or 2) one outpatient VTE diagnosis and a  lling of a prescription of antico- for socio-economic status) was found in the LISA register. For this study, education agulants within three months, or 3) one VTE diagnosis in the National Cause of Death level was categorized as low, medium, or high. Register. All patients with a previous VTE diagnosis from January 1, 1980, to Decem- ber 31, 1986, were excluded. VTE outcomes were de ned as follows: A primary or secondary diagnostic code in the National Patient Register or National Cause of Death Register; International Clas- Comorbidities were registered within seven years or on the same date as the VTE si cation of Diseases (ICD) 8: PE: 450, DVT: 451. ICD 9: PE: 415B, 416W, DVT: event, recent comorbidities within six months or on the same date, and temporary 451 except 451A, ICD 10: PE: I26, DVT: I80 except I80.0. Before January 2006, only provoking factors within three months or on the same date, see Table 3 for included inpatient diagnoses were included. From January 2006, both inpatient and outpatient factors and comorbidities. diagnoses were included if followed by a  lling of a prescription of anticoagulant medication. Patients with a VTE diagnosis in the National Cause of Death Register Table 3. Demographic factors (at the time of the did not need a prescription  lling. Only the  rst registered diagnosis was included. VTE event), comorbidities (within seven years or on the same date as the VTE event), and tempo- rary provoking factors (within three months or on Exclusion criteria were: 1) enlistment at an early or late age (≤16 years or ≥25 years), the same date as the VTE event). 2) female sex, 3) individuals whose Swedish personal identi cation number was re- used after their death or emigration, 4) missing BMI values, 5) diagnosis of VTE or Demographic factors stroke before enlistment, or 6) lower extremity fracture within one year prior to enlist- Age ment, see Figure 2. Female sex Comorbidities Men were followed until 1) a diagnosis of PE or DVT, 2) death, 3) emigration, or 4) Heart failure end of follow-up (December 31, 2014). Ischemic heart disease Atrial fibrillation Ischemic stroke Hemorrhagic stroke Chronic obstructive pulmonary disease (COPD) Cancer Systemic connective tissue disorders Inflammatory bowel syndrome Liver disease Kidney failure Depression Psychosis Alcohol abuse Temporary provoking factors Gastrointestinal surgery Obstetrical surgery Surgery of the musculoskeletal system Surgery, other major Trauma Figure 2. Overview of included patients and reasons for exclusion from the study. BMI= Body Lower extremity fracture Mass Index. ID= Identity. VTE= Venous Thromboembolism. Reprint from Paper 1115. 30 31 quality since it is reported by the individual, who for various reasons might fail to Paper II report a change of residence. DVT and PE diagnostic codes were identical as in Paper I. Only  rst-time diagnoses Defi nitions were included. For diagnoses registered before July 1, 2005, a VTE diagnosis was de ned as either a VTE diagnosis in the inpatient or National Cause of Death Register Paper I or one outpatient diagnosis of VTE and a subsequent, identical diagnosis within three months. For patients with VTE after July 1, 2005, the de nition was the following: Data from the Swedish Military Service Conscription Register 1969–2005 were used 1) one inpatient diagnosis and a  lling of a prescription of anticoagulants within six for baseline data. Information about the highest achieved parental education (a proxy months or 2) one outpatient VTE diagnosis and a  lling of a prescription of antico- for socio-economic status) was found in the LISA register. For this study, education agulants within three months, or 3) one VTE diagnosis in the National Cause of Death level was categorized as low, medium, or high. Register. All patients with a previous VTE diagnosis from January 1, 1980, to Decem- ber 31, 1986, were excluded. VTE outcomes were de ned as follows: A primary or secondary diagnostic code in the National Patient Register or National Cause of Death Register; International Clas- Comorbidities were registered within seven years or on the same date as the VTE si cation of Diseases (ICD) 8: PE: 450, DVT: 451. ICD 9: PE: 415B, 416W, DVT: event, recent comorbidities within six months or on the same date, and temporary 451 except 451A, ICD 10: PE: I26, DVT: I80 except I80.0. Before January 2006, only provoking factors within three months or on the same date, see Table 3 for included inpatient diagnoses were included. From January 2006, both inpatient and outpatient factors and comorbidities. diagnoses were included if followed by a  lling of a prescription of anticoagulant medication. Patients with a VTE diagnosis in the National Cause of Death Register Table 3. Demographic factors (at the time of the did not need a prescription  lling. Only the  rst registered diagnosis was included. VTE event), comorbidities (within seven years or on the same date as the VTE event), and tempo- rary provoking factors (within three months or on Exclusion criteria were: 1) enlistment at an early or late age (≤16 years or ≥25 years), the same date as the VTE event). 2) female sex, 3) individuals whose Swedish personal identi cation number was re- used after their death or emigration, 4) missing BMI values, 5) diagnosis of VTE or Demographic factors stroke before enlistment, or 6) lower extremity fracture within one year prior to enlist- Age ment, see Figure 2. Female sex Comorbidities Men were followed until 1) a diagnosis of PE or DVT, 2) death, 3) emigration, or 4) Heart failure end of follow-up (December 31, 2014). Ischemic heart disease Atrial fibrillation Ischemic stroke Hemorrhagic stroke Chronic obstructive pulmonary disease (COPD) Cancer Systemic connective tissue disorders Inflammatory bowel syndrome Liver disease Kidney failure Depression Psychosis Alcohol abuse Temporary provoking factors Gastrointestinal surgery Obstetrical surgery Surgery of the musculoskeletal system Surgery, other major Trauma Figure 2. Overview of included patients and reasons for exclusion from the study. BMI= Body Lower extremity fracture Mass Index. ID= Identity. VTE= Venous Thromboembolism. Reprint from Paper 1115. 30 31 Ideally,  ve controls matched for sex, year of birth, and county of residence, were The following comorbidities were used for multivariable adjustment: heart failure, selected from the Total Population Register. However, in some cases fewer controls ischemic heart disease, atrial  brillation, ischemic stroke, hemorrhagic stroke, COPD, were used when it was not possible to  nd  ve unique controls for every case. The and in ammatory bowel disease. The following temporary provoking factors were same person could  rst be included as a control and later, after suffering a VTE, be used: major surgery, lower extremity fracture, and trauma. included as a case. Paper IV Paper III Included patients came from the same dataset as in Papers II and III, with a diagnosis This study included the same VTE population as in Paper II. Patients with an in- or of VTE and a subsequent  lling of a prescription of anticoagulant medication within outpatient diagnostic code of cancer within a year or on the same date as the VTE 30 days from January 2014 to December 2020. We excluded: 1) patients with a prior diagnosis were identi ed; see Table 4 for included cancer types. Among controls, VTE. 2) patients with atrial  brillation at any time point until the day of censoring 3) individuals with a diagnostic code of cancer within or on the same date as the VTE of patients who were pregnant at inclusion, 4) patients with a diagnosis of cancer within the corresponding case were identi ed. 1 year before the VTE, and 5) patients who  lled a prescription of anticoagulant medi- cation within 6 months prior to the VTE, see Figure 3. Table 4. Types of cancer (registered in the National Patient Register within 1 year be- fore, or on the same date as the VTE) in- Patients with a first VTE after January 2014 cluded in the study. n=94,485 Cancer groups All cancers except non-melanoma Atrial fibrillation before end of study, n= 19,530 skin cancer Pregnancy at time of VTE, n=4,001 Cancer <1 year prior to VTE, n= 11,038 Individual cancer types Anticoagulant <6 months pre VTE, n= 10,778 Death registered in Cause of death register but Esophageal, stomach month or year unknown, n=11 No retrieval of oral anticoagulant within 30 days of Small intestinal VTE diagnosis, n=4,013 Colon Rectal, anal Included in study population Pancreatic n= 45,114 Liver Biliary Lung Brain Warfarin Apixaban Rivaroxaban Dabigatran Edoxabann= 6,558 n= 19,498 n=18,196 n= 642 n= 220 Malignant melanoma Kidney Figure 3. Flow chart for study inclusion and number of patients with a fi rst venous thromboembolism (VTE) in each study group. Reprint from Paper IV116. Bladder and urothelial cancer Uterine Ovarian Cervix Patients were considered to be on treatment if they  lled a prescription within 30 Leukemia days of VTE diagnosis. They were considered to have ongoing treatment if they  lled Lymphoma at least two prescriptions of the same medication (dose adjustment allowed) per 12 Multiple myeloma months. Prostate Major bleeding was de ned as a primary inpatient diagnosis of bleeding. Fatal bleed- Testicular ing was de ned as 1) a major bleeding in combination with death within 30 days or 2) Breast a diagnosis of bleeding in any position on a death certi cate. 32 33 Ideally,  ve controls matched for sex, year of birth, and county of residence, were The following comorbidities were used for multivariable adjustment: heart failure, selected from the Total Population Register. However, in some cases fewer controls ischemic heart disease, atrial  brillation, ischemic stroke, hemorrhagic stroke, COPD, were used when it was not possible to  nd  ve unique controls for every case. The and in ammatory bowel disease. The following temporary provoking factors were same person could  rst be included as a control and later, after suffering a VTE, be used: major surgery, lower extremity fracture, and trauma. included as a case. Paper IV Paper III Included patients came from the same dataset as in Papers II and III, with a diagnosis This study included the same VTE population as in Paper II. Patients with an in- or of VTE and a subsequent  lling of a prescription of anticoagulant medication within outpatient diagnostic code of cancer within a year or on the same date as the VTE 30 days from January 2014 to December 2020. We excluded: 1) patients with a prior diagnosis were identi ed; see Table 4 for included cancer types. Among controls, VTE. 2) patients with atrial  brillation at any time point until the day of censoring 3) individuals with a diagnostic code of cancer within or on the same date as the VTE of patients who were pregnant at inclusion, 4) patients with a diagnosis of cancer within the corresponding case were identi ed. 1 year before the VTE, and 5) patients who  lled a prescription of anticoagulant medi- cation within 6 months prior to the VTE, see Figure 3. Table 4. Types of cancer (registered in the National Patient Register within 1 year be- fore, or on the same date as the VTE) in- Patients with a first VTE after January 2014 cluded in the study. n=94,485 Cancer groups All cancers except non-melanoma Atrial fibrillation before end of study, n= 19,530 skin cancer Pregnancy at time of VTE, n=4,001 Cancer <1 year prior to VTE, n= 11,038 Individual cancer types Anticoagulant <6 months pre VTE, n= 10,778 Death registered in Cause of death register but Esophageal, stomach month or year unknown, n=11 No retrieval of oral anticoagulant within 30 days of Small intestinal VTE diagnosis, n=4,013 Colon Rectal, anal Included in study population Pancreatic n= 45,114 Liver Biliary Lung Brain Warfarin Apixaban Rivaroxaban Dabigatran Edoxabann= 6,558 n= 19,498 n=18,196 n= 642 n= 220 Malignant melanoma Kidney Figure 3. Flow chart for study inclusion and number of patients with a fi rst venous thromboembolism (VTE) in each study group. Reprint from Paper IV116. Bladder and urothelial cancer Uterine Ovarian Cervix Patients were considered to be on treatment if they  lled a prescription within 30 Leukemia days of VTE diagnosis. They were considered to have ongoing treatment if they  lled Lymphoma at least two prescriptions of the same medication (dose adjustment allowed) per 12 Multiple myeloma months. Prostate Major bleeding was de ned as a primary inpatient diagnosis of bleeding. Fatal bleed- Testicular ing was de ned as 1) a major bleeding in combination with death within 30 days or 2) Breast a diagnosis of bleeding in any position on a death certi cate. 32 33 Concomitant pharmacological treatment and comorbidities were registered, both for Paper III: For total incidence rates, the total number of patients with VTE and can- assessment of risk factors and for multivariable adjustment in calculation of hazard cer was divided by the total number of Swedish inhabitants each year (equaling the ratios, see Table 5. patient years at risk), using data from Statistics Sweden. For incidence rates with sex and age strati cation, VTE cases were divided by the number of Swedish inhabitants Table 5. Comorbidities and concomitant pharmacologic treatments included in Paper IV in the corresponding age and sex category in the same calendar year. Adjusted odds ratios were calculated using conditional logistic regression for case-control matching Concomitant with 99% CI, adjusting for comorbidities and temporary provoking factors. Comorbidities pharmacological treatment Paper IV: Cumulative incidence functions were calculated with other bleedings, death, and cancer as competing risks. Event rates for different bleedings depending Previous bleeding Renal failure Stroke (hemorrhagic) Antiplatelet on treatment choice and time period were calculated using Poisson regression. Cox (inpatient treatment proportional hazard regression was used to estimate adjusted hazard ratios for bleed- diagnosis) ing depending on the choice of anticoagulant medication and depending on different Hypertension Liver disease Systemic connective Proton pump risk factors. Adjustment of hazard ratios was made for bleeding risk factors. In sub tissue disease inhibitor analyses with few events, we reduced the number of adjustment variables. All signi - Heart failure Ischemic heart IBD (inflammatory SSRI cance tests were two-sided and performed with a signi cance level of 5%. disease bowel disease) COPD Peripheral arterial Dementia Statins disease Diabetes Stroke (ischemic) Patients were followed until one of the following events occurred: 1) major bleeding, 2) death, 3) diagnosis of cancer, 4)  ve years after the diagnosis of VTE, 5) end of treatment, or 6) end of follow-up (December 31, 2020). Statistical analysis In all papers, continuous variables were presented as mean and standard deviation or median with  rst and third quartiles, interquartile range (IQR). Categorical data was presented as numbers with percentages. Paper I: We used Poisson regression to calculate incidence rates and corresponding 95% con dence intervals (CI). Cox proportional hazard regression was used to calcu- late the risk of VTE during follow-up based on BMI at conscription. Covariates from baseline data were grouped in different models, and adjustment was made by combin- ing these models in different ways. Paper II: When comparing cases and controls, chi-square tests were used for dichoto- mous variables and t-tests for continuous variables. Tests were two-tailed, and 1% sig- ni cance levels were chosen due to a large number of analyzed risk factors. Adjusted odds ratios (ORs) were calculated with conditional logistic regression according to case-control matching, with adjustment for comorbidities and temporary provoking factors. Population attributable risk (PAR) was calculated according to Bruzzi et al117, to estimate the possible reduction of PE or DVT in the population if a risk factor is eliminated. 34 35 Concomitant pharmacological treatment and comorbidities were registered, both for Paper III: For total incidence rates, the total number of patients with VTE and can- assessment of risk factors and for multivariable adjustment in calculation of hazard cer was divided by the total number of Swedish inhabitants each year (equaling the ratios, see Table 5. patient years at risk), using data from Statistics Sweden. For incidence rates with sex and age strati cation, VTE cases were divided by the number of Swedish inhabitants Table 5. Comorbidities and concomitant pharmacologic treatments included in Paper IV in the corresponding age and sex category in the same calendar year. Adjusted odds ratios were calculated using conditional logistic regression for case-control matching Concomitant with 99% CI, adjusting for comorbidities and temporary provoking factors. Comorbidities pharmacological treatment Paper IV: Cumulative incidence functions were calculated with other bleedings, death, and cancer as competing risks. Event rates for different bleedings depending Previous bleeding Renal failure Stroke (hemorrhagic) Antiplatelet on treatment choice and time period were calculated using Poisson regression. Cox (inpatient treatment proportional hazard regression was used to estimate adjusted hazard ratios for bleed- diagnosis) ing depending on the choice of anticoagulant medication and depending on different Hypertension Liver disease Systemic connective Proton pump risk factors. Adjustment of hazard ratios was made for bleeding risk factors. In sub tissue disease inhibitor analyses with few events, we reduced the number of adjustment variables. All signi - Heart failure Ischemic heart IBD (inflammatory SSRI cance tests were two-sided and performed with a signi cance level of 5%. disease bowel disease) COPD Peripheral arterial Dementia Statins disease Diabetes Stroke (ischemic) Patients were followed until one of the following events occurred: 1) major bleeding, 2) death, 3) diagnosis of cancer, 4)  ve years after the diagnosis of VTE, 5) end of treatment, or 6) end of follow-up (December 31, 2020). Statistical analysis In all papers, continuous variables were presented as mean and standard deviation or median with  rst and third quartiles, interquartile range (IQR). Categorical data was presented as numbers with percentages. Paper I: We used Poisson regression to calculate incidence rates and corresponding 95% con dence intervals (CI). Cox proportional hazard regression was used to calcu- late the risk of VTE during follow-up based on BMI at conscription. Covariates from baseline data were grouped in different models, and adjustment was made by combin- ing these models in different ways. Paper II: When comparing cases and controls, chi-square tests were used for dichoto- mous variables and t-tests for continuous variables. Tests were two-tailed, and 1% sig- ni cance levels were chosen due to a large number of analyzed risk factors. Adjusted odds ratios (ORs) were calculated with conditional logistic regression according to case-control matching, with adjustment for comorbidities and temporary provoking factors. Population attributable risk (PAR) was calculated according to Bruzzi et al117, to estimate the possible reduction of PE or DVT in the population if a risk factor is eliminated. 34 35 RESULTS Cancer was the most prevalent comorbidity (PE: 21.3%, DVT: 19.3%). Heart failure (PE: 18.9%, DVT: 10.3%) and ischemic heart disease (PE: 18.6%, DVT: 11.0%) were Paper I among the conditions that were more common in patients with PE than DVT. Baseline characteristics Incidence rates The study included 1,639,838 men. Of these, 79.8% had normal weight (BMI 18.5– The annual incidence rate for a  rst-time VTE was 105.2 per 100,000 inhabitants, <25 kg/m2), 9.9% were overweight (BMI 25–<30 kg/m2), and 2.2% were obese (BMI 49.8 for PE, and 55.4 for DVT. The incidence rate increased markedly with age, see >30 kg/m2). Men were followed for a median of 28 years (interquartile range 20–36 Figure 4. Women had a higher incidence rate of VTE before the age of 40 and 80 years years). and older, whereas men had higher incidence rates in ages 40–79 years. When study subjects were strati ed according to BMI, men with a high BMI had a shorter follow-up. Men with BMI 15–<18.8 kg/m2 had a median follow-up of 31 (IQR 22–40) years, decreasing gradually to 19 (IQR 13–27) years in men with BMI 35–<60 kg/m2. Men with normal BMIs had better cardiorespiratory  tness than thin and obese men. Muscle strength was lower in men with a low BMI than in normal or obese men. Men who were obese more often had a low intelligence quotient, hyper- tension, and diabetes than other study subjects. VTE events In total, 17,805  rst VTE events were registered. There was a clear increase in in- cidence rates with higher BMI class. In lean men with a BMI of 18.5 –<20 kg/m2, the event rate was 32.8 per 100,000 patient-years. In obese men (class I obesity, BMI 30–<35), the event rate was 73.7 per 100,000 patient-years. In men with severe obe- sity (class II–III obesity, BMI 35–<60 kg/m2), corresponding numbers were 112.1 per 100,000. Hazard ratios After multivariable adjustment for baseline characteristics, obese men (BMI 30–<35) had an aHR of VTE of 2.9 (95% CI 2.7–3.2) and severely obese men (BMI 35–<60 kg/m2) had an aHR of 5.0 (95% CI 4.2–5.9), compared to lean men with (BMI 18.5 Figure 4. VTE incidence per 100,000 inhabitants divided into men –<20 kg/m2). In a sub-analysis, men who died within two years after a VTE or who and women aged 0-100 years. Reprint from Paper II 118. had a lower extremity fracture within three months before the VTE were excluded. This only affected aHR marginally. Paper II Odds ratios and PAR for comorbidities and temporary provoking factors When comparing VTE patients to their respective matched controls, patients with PE Study population had higher adjusted odds ratios (aOR) for cardiopulmonary diseases such as heart We included 298,172 patients with a  rst-time VTE in the National Patient Regis- failure, ischemic heart disease, atrial  brillation, and COPD, see Figure 5. Depression ter or National Cause of Death Register. The matched control population included also had higher aORs in patients with PE than patients with DVT. Patients with DVT 1,185,079 individuals. had higher aORs for temporary provoking factors such as musculoskeletal surgery and lower-limb fracture. Patients with PE were more often women than patients with DVT (PE: 53.4% women, DVT 52.1% women) and, on average, three years older (PE: 69 years vs DVT: 66 Risk factors with the highest PAR for patients with PE were cancer (13.0%), heart years). Comorbidities registered within seven years and temporary provoking factors failure (11.7%), and ischemic heart disease (6.3%). The highest PARs for DVT were present within three months of the VTE were compared between DVT and PE patients. seen for cancer (11.9%) and musculoskeletal surgery (8.6%). 36 37 RESULTS Cancer was the most prevalent comorbidity (PE: 21.3%, DVT: 19.3%). Heart failure (PE: 18.9%, DVT: 10.3%) and ischemic heart disease (PE: 18.6%, DVT: 11.0%) were Paper I among the conditions that were more common in patients with PE than DVT. Baseline characteristics Incidence rates The study included 1,639,838 men. Of these, 79.8% had normal weight (BMI 18.5– The annual incidence rate for a  rst-time VTE was 105.2 per 100,000 inhabitants, <25 kg/m2), 9.9% were overweight (BMI 25–<30 kg/m2), and 2.2% were obese (BMI 49.8 for PE, and 55.4 for DVT. The incidence rate increased markedly with age, see >30 kg/m2). Men were followed for a median of 28 years (interquartile range 20–36 Figure 4. Women had a higher incidence rate of VTE before the age of 40 and 80 years years). and older, whereas men had higher incidence rates in ages 40–79 years. When study subjects were strati ed according to BMI, men with a high BMI had a shorter follow-up. Men with BMI 15–<18.8 kg/m2 had a median follow-up of 31 (IQR 22–40) years, decreasing gradually to 19 (IQR 13–27) years in men with BMI 35–<60 kg/m2. Men with normal BMIs had better cardiorespiratory  tness than thin and obese men. Muscle strength was lower in men with a low BMI than in normal or obese men. Men who were obese more often had a low intelligence quotient, hyper- tension, and diabetes than other study subjects. VTE events In total, 17,805  rst VTE events were registered. There was a clear increase in in- cidence rates with higher BMI class. In lean men with a BMI of 18.5 –<20 kg/m2, the event rate was 32.8 per 100,000 patient-years. In obese men (class I obesity, BMI 30–<35), the event rate was 73.7 per 100,000 patient-years. In men with severe obe- sity (class II–III obesity, BMI 35–<60 kg/m2), corresponding numbers were 112.1 per 100,000. Hazard ratios After multivariable adjustment for baseline characteristics, obese men (BMI 30–<35) had an aHR of VTE of 2.9 (95% CI 2.7–3.2) and severely obese men (BMI 35–<60 kg/m2) had an aHR of 5.0 (95% CI 4.2–5.9), compared to lean men with (BMI 18.5 Figure 4. VTE incidence per 100,000 inhabitants divided into men –<20 kg/m2). In a sub-analysis, men who died within two years after a VTE or who and women aged 0-100 years. Reprint from Paper II 118. had a lower extremity fracture within three months before the VTE were excluded. This only affected aHR marginally. Paper II Odds ratios and PAR for comorbidities and temporary provoking factors When comparing VTE patients to their respective matched controls, patients with PE Study population had higher adjusted odds ratios (aOR) for cardiopulmonary diseases such as heart We included 298,172 patients with a  rst-time VTE in the National Patient Regis- failure, ischemic heart disease, atrial  brillation, and COPD, see Figure 5. Depression ter or National Cause of Death Register. The matched control population included also had higher aORs in patients with PE than patients with DVT. Patients with DVT 1,185,079 individuals. had higher aORs for temporary provoking factors such as musculoskeletal surgery and lower-limb fracture. Patients with PE were more often women than patients with DVT (PE: 53.4% women, DVT 52.1% women) and, on average, three years older (PE: 69 years vs DVT: 66 Risk factors with the highest PAR for patients with PE were cancer (13.0%), heart years). Comorbidities registered within seven years and temporary provoking factors failure (11.7%), and ischemic heart disease (6.3%). The highest PARs for DVT were present within three months of the VTE were compared between DVT and PE patients. seen for cancer (11.9%) and musculoskeletal surgery (8.6%). 36 37 Incidence rate The annual incidence rate of VTE with a registered cancer diagnosis within the pre- ceding year was 15.9 per 100,000 patient years for men and 15.7 for women. The incidence rate increased markedly after the age of 60. In women, the most common cancer types among VTE patients were breast cancer (23.3%), lung cancer (10.3%) and colon cancer (9.2%). In men, the most prevalent cancer types among patients with VTE were prostate cancer (32.1%), lung cancer (10.5%), and bladder/urothelial cancer (8.6%). Multivariable adjusted odds ratios of various types of cancer The adjusted odds ratios of any cancer were higher in females (aOR 5.5 [99% CI: 5.4–5.7]) than in male VTE patients (aOR 3.9 [99% CI: 3.8–4.0]) compared to their respective population controls. However, when excluding sex-speci c cancers, the difference was small. Among cancers with the highest aOR in VTE patients compared to matched controls were brain cancer (women: aOR 17.4 [99% CI 12.9–23.4], men: aOR 17.5 [99% CI 13.8-–22.2]), pancreatic cancer (women aOR: 19.6 [99% CI: 15.8–24.4], men aOR 17.2 [99% CI 13.7–21.6]) and biliary cancer (women: aOR 16.7 [99% CI 12.3–22.8], men: aOR 10.4 [99% CI 7.1–15.2]). Among cancers with the lowest aOR, we found bladder/urothelial cancer (women: aOR 1.3 [99% CI 1.1–1.5], men: aOR 1.3 [99% CI 1.2–1.5]), malignant melanoma (women: aOR 2.5 [99% CI 2.1–3.1], men: aOR 2.7 [99% CI 2.2–3.2]). Female VTE patients also had low OR for uterine cancer (aOR 3.4 [99% CI 3.1–3.9]), whereas men had a low OR for prostate cancer (aOR 2.2 [99% CI 2.1–2.3]) compared to matched population controls. Paper IV Study population Figure 5. Multivariable adjusted ORs and Population attributable risk (PARs) with correspond- ing 99% CI for various comorbidities (registered within seven years) and temporary provoking The study included 45,114 patients with a  rst-time VTE between 2014–2020. The factors (surgery, trauma, lower extremity fracture registered within three months). Reprint from median age was 68 years, and 54.7% were male. The median follow-up after initiation Paper II118. of treatment was 0.6 years (mean 1.0 years). Among the patients, 6,558 were treated with warfarin, 18,196 with rivaroxaban, and 19,498 with apixaban. Dabigatran and Paper III edoxaban were initiated in too few patients to permit the calculation of reliable out- come data. Study population Our register data comprised 298,172 patients diagnosed with a  rst-time VTE be- Patients treated with warfarin were slightly older and had more comorbidities such tween 1987–2018, and their 1,185,079 matched population controls. In the VTE pop- as renal failure, heart failure, diabetes, peripheral arterial disease, and systemic con- ulation, we identi ed 44,685 (15%) patients with a registered cancer diagnosis within nective tissue disorder than patients on other anticoagulants. Patients on rivaroxaban one year before the VTE. Of these, 22,000 had DVT, and 22,685 had PE. The median were generally younger and healthier than patients on other anticoagulants. age was 71 years. 38 39 Incidence rate The annual incidence rate of VTE with a registered cancer diagnosis within the pre- ceding year was 15.9 per 100,000 patient years for men and 15.7 for women. The incidence rate increased markedly after the age of 60. In women, the most common cancer types among VTE patients were breast cancer (23.3%), lung cancer (10.3%) and colon cancer (9.2%). In men, the most prevalent cancer types among patients with VTE were prostate cancer (32.1%), lung cancer (10.5%), and bladder/urothelial cancer (8.6%). Multivariable adjusted odds ratios of various types of cancer The adjusted odds ratios of any cancer were higher in females (aOR 5.5 [99% CI: 5.4–5.7]) than in male VTE patients (aOR 3.9 [99% CI: 3.8–4.0]) compared to their respective population controls. However, when excluding sex-speci c cancers, the difference was small. Among cancers with the highest aOR in VTE patients compared to matched controls were brain cancer (women: aOR 17.4 [99% CI 12.9–23.4], men: aOR 17.5 [99% CI 13.8-–22.2]), pancreatic cancer (women aOR: 19.6 [99% CI: 15.8–24.4], men aOR 17.2 [99% CI 13.7–21.6]) and biliary cancer (women: aOR 16.7 [99% CI 12.3–22.8], men: aOR 10.4 [99% CI 7.1–15.2]). Among cancers with the lowest aOR, we found bladder/urothelial cancer (women: aOR 1.3 [99% CI 1.1–1.5], men: aOR 1.3 [99% CI 1.2–1.5]), malignant melanoma (women: aOR 2.5 [99% CI 2.1–3.1], men: aOR 2.7 [99% CI 2.2–3.2]). Female VTE patients also had low OR for uterine cancer (aOR 3.4 [99% CI 3.1–3.9]), whereas men had a low OR for prostate cancer (aOR 2.2 [99% CI 2.1–2.3]) compared to matched population controls. Paper IV Study population Figure 5. Multivariable adjusted ORs and Population attributable risk (PARs) with correspond- ing 99% CI for various comorbidities (registered within seven years) and temporary provoking The study included 45,114 patients with a  rst-time VTE between 2014–2020. The factors (surgery, trauma, lower extremity fracture registered within three months). Reprint from median age was 68 years, and 54.7% were male. The median follow-up after initiation Paper II118. of treatment was 0.6 years (mean 1.0 years). Among the patients, 6,558 were treated with warfarin, 18,196 with rivaroxaban, and 19,498 with apixaban. Dabigatran and Paper III edoxaban were initiated in too few patients to permit the calculation of reliable out- come data. Study population Our register data comprised 298,172 patients diagnosed with a  rst-time VTE be- Patients treated with warfarin were slightly older and had more comorbidities such tween 1987–2018, and their 1,185,079 matched population controls. In the VTE pop- as renal failure, heart failure, diabetes, peripheral arterial disease, and systemic con- ulation, we identi ed 44,685 (15%) patients with a registered cancer diagnosis within nective tissue disorder than patients on other anticoagulants. Patients on rivaroxaban one year before the VTE. Of these, 22,000 had DVT, and 22,685 had PE. The median were generally younger and healthier than patients on other anticoagulants. age was 71 years. 38 39 Risk of major bleeding DISCUSSION Initial treatment During the  rst six months of treatment, 494 patients (1.1% of patients who started This thesis explored risk factors for  rst-time VTE and bleeding risk during antico- anticoagulant treatment) suffered major bleeding. The event rate was 3.9 (95% CI agulant treatment in patients with a  rst-time VTE. 3.1–4.6) for warfarin, 2.9 (95% CI 2.6–3.3) for rivaroxaban, and 2.0 (95% CI 1.7–2.3) for apixaban, per 100 patient-years. The thesis has contributed with the following knowledge: Paper I: Prior to our re- port, overweight and obesity at a young age and its impact on future risk of VTE was After multivariable adjustment for comorbidities and concomitant medications, pa- scarcely studied. We showed that even mildly elevated BMI in young men was a risk tients on apixaban had a lower risk of major bleeding compared to patients on warfa- factor for VTE in adult life, and the risk increased incrementally with higher BMI. rin (aHR 0.56 [95% CI 0.4–0.7]) and rivaroxaban (aHR 0.6; [95% CI 0.5–0.8]). There Since overweight and obesity in young adulthood are rapidly increasing in many was no signi cant difference between rivaroxaban and warfarin (aHR 0.9 [95% CI countries32, this association may have large effects on the future VTE incidence and 0.7–1.1]). may be important for prognostication of the need for future health care. Paper II: Our study contributed a comprehensive picture of the comorbidities associated with PE In a sub analysis of the  rst month of treatment, 44 patients on apixaban and 98 on and DVT, respectively, where cardiopulmonary diseases are more strongly associ- rivaroxaban experienced major bleeding, yielding an aHR of 0.4 (95% CI 0.3–0.6). ated with PE and recent musculoskeletal surgery, and lower extremity fractures are During the subsequent  ve months, 116 patients on apixaban and 126 patients on ri- more strongly associated with DVT. Paper III: Our results con rmed the previously varoxaban experienced major bleeding, resulting in a non-signi cant risk difference known, strong association between VTE and pancreatic, brain, liver, and biliary can- of aHR 0.8 (95% CI 0.6–1.0). cer. However, the risk was not as high in bladder/urothelial cancer, kidney cancer, and some gynecological cancers, which are commonly considered to be associated with a Extended treatment high VTE risk35,37. Previous risk estimates were mainly based on studies on patients 37,38,119-121 During extended treatment (from 6 months of treatment up to 5 years of treatment), on chemotherapy . Our results indicate that at least a proportion of patients a total of 267 patients (1.0% of patients who remained on anticoagulant treatment at with these cancer types are at low risk of VTE. Paper IV: We reported a lower risk six months) suffered major bleeding. The event rate was 1.6 (95% CI 1.2–1.9) per of major bleeding in VTE patients without known cancer for apixaban compared to 100 patient-years for warfarin, 1.1 (95% CI 0.9–1.3) for rivaroxaban and 1.0 (95% CI rivaroxaban or warfarin in patients during initial treatment (0–6 months). This was in 0.8–1.2) apixaban. line with previous real-life data on warfarin compared to the respective DOACs 96-98, but no previous study comparing apixaban to rivaroxaban was available in this pa- After multivariable adjustment for comorbidities and concomitant medications, the tient group. During extended treatment (6 months up to 5 years), the risk was low for aHR was 0.7 (95% CI 0.5–1.0) for rivaroxaban and aHR 0.6 (95% CI 0.4–0.8) for both apixaban and rivaroxaban, while warfarin had a higher risk of major bleeding. apixaban vs warfarin. The difference between apixaban and rivaroxaban was non- We found no earlier real-life data study addressing this question in a VTE population signi cant, aHR of 0.9 (0.6–1.1). without known cancer. Risk factors We used national registers to try to answer our scienti c questions. In Paper I, a co- hort of men was followed until the  rst VTE diagnosis. Papers II–III, a cohort of The most important risk factors associated with major bleeding during the  rst six patients with a  rst-time VTE (cases) were compared to matched population controls months of treatment were age (per additional year), previous bleeding, liver disease, regarding prior or concomitant comorbidities. In Paper IV, the same VTE cohort as renal failure, and antiplatelet treatment. PPIs were also associated with an increased in Papers II and III was followed prospectively after diagnosis. These methods have risk. both advantages and disadvantages. Among the advantages of register-based studies are the following: After the  rst six months of treatment, age (per additional year), previous bleeding, COPD, antiplatelet treatment, and selective serotonin reuptake inhibitors were the Possibility to study large populations over long periods of time. This was valuable in most important factors associated with increased bleeding risk. Paper I, since VTE incidence increases markedly with age. Included individuals were adolescents, and the registries gave a possibility of a median follow-up of 28 years, which would have been dif cult to achieve for a large population with other study designs. Most previous studies were done on older populations122,123, and it is valuable to have access to data from early life when studying risk factors for disease. When studying BMI at higher ages, diseases such as various cancer types124,125 may have developed. These diseases are more frequent in obese patients and may contribute to 40 41 Risk of major bleeding DISCUSSION Initial treatment During the  rst six months of treatment, 494 patients (1.1% of patients who started This thesis explored risk factors for  rst-time VTE and bleeding risk during antico- anticoagulant treatment) suffered major bleeding. The event rate was 3.9 (95% CI agulant treatment in patients with a  rst-time VTE. 3.1–4.6) for warfarin, 2.9 (95% CI 2.6–3.3) for rivaroxaban, and 2.0 (95% CI 1.7–2.3) for apixaban, per 100 patient-years. The thesis has contributed with the following knowledge: Paper I: Prior to our re- port, overweight and obesity at a young age and its impact on future risk of VTE was After multivariable adjustment for comorbidities and concomitant medications, pa- scarcely studied. We showed that even mildly elevated BMI in young men was a risk tients on apixaban had a lower risk of major bleeding compared to patients on warfa- factor for VTE in adult life, and the risk increased incrementally with higher BMI. rin (aHR 0.56 [95% CI 0.4–0.7]) and rivaroxaban (aHR 0.6; [95% CI 0.5–0.8]). There Since overweight and obesity in young adulthood are rapidly increasing in many was no signi cant difference between rivaroxaban and warfarin (aHR 0.9 [95% CI countries32, this association may have large effects on the future VTE incidence and 0.7–1.1]). may be important for prognostication of the need for future health care. Paper II: Our study contributed a comprehensive picture of the comorbidities associated with PE In a sub analysis of the  rst month of treatment, 44 patients on apixaban and 98 on and DVT, respectively, where cardiopulmonary diseases are more strongly associ- rivaroxaban experienced major bleeding, yielding an aHR of 0.4 (95% CI 0.3–0.6). ated with PE and recent musculoskeletal surgery, and lower extremity fractures are During the subsequent  ve months, 116 patients on apixaban and 126 patients on ri- more strongly associated with DVT. Paper III: Our results con rmed the previously varoxaban experienced major bleeding, resulting in a non-signi cant risk difference known, strong association between VTE and pancreatic, brain, liver, and biliary can- of aHR 0.8 (95% CI 0.6–1.0). cer. However, the risk was not as high in bladder/urothelial cancer, kidney cancer, and some gynecological cancers, which are commonly considered to be associated with a Extended treatment high VTE risk35,37. Previous risk estimates were mainly based on studies on patients 37,38,119-121 During extended treatment (from 6 months of treatment up to 5 years of treatment), on chemotherapy . Our results indicate that at least a proportion of patients a total of 267 patients (1.0% of patients who remained on anticoagulant treatment at with these cancer types are at low risk of VTE. Paper IV: We reported a lower risk six months) suffered major bleeding. The event rate was 1.6 (95% CI 1.2–1.9) per of major bleeding in VTE patients without known cancer for apixaban compared to 100 patient-years for warfarin, 1.1 (95% CI 0.9–1.3) for rivaroxaban and 1.0 (95% CI rivaroxaban or warfarin in patients during initial treatment (0–6 months). This was in 0.8–1.2) apixaban. line with previous real-life data on warfarin compared to the respective DOACs 96-98, but no previous study comparing apixaban to rivaroxaban was available in this pa- After multivariable adjustment for comorbidities and concomitant medications, the tient group. During extended treatment (6 months up to 5 years), the risk was low for aHR was 0.7 (95% CI 0.5–1.0) for rivaroxaban and aHR 0.6 (95% CI 0.4–0.8) for both apixaban and rivaroxaban, while warfarin had a higher risk of major bleeding. apixaban vs warfarin. The difference between apixaban and rivaroxaban was non- We found no earlier real-life data study addressing this question in a VTE population signi cant, aHR of 0.9 (0.6–1.1). without known cancer. Risk factors We used national registers to try to answer our scienti c questions. In Paper I, a co- hort of men was followed until the  rst VTE diagnosis. Papers II–III, a cohort of The most important risk factors associated with major bleeding during the  rst six patients with a  rst-time VTE (cases) were compared to matched population controls months of treatment were age (per additional year), previous bleeding, liver disease, regarding prior or concomitant comorbidities. In Paper IV, the same VTE cohort as renal failure, and antiplatelet treatment. PPIs were also associated with an increased in Papers II and III was followed prospectively after diagnosis. These methods have risk. both advantages and disadvantages. Among the advantages of register-based studies are the following: After the  rst six months of treatment, age (per additional year), previous bleeding, COPD, antiplatelet treatment, and selective serotonin reuptake inhibitors were the Possibility to study large populations over long periods of time. This was valuable in most important factors associated with increased bleeding risk. Paper I, since VTE incidence increases markedly with age. Included individuals were adolescents, and the registries gave a possibility of a median follow-up of 28 years, which would have been dif cult to achieve for a large population with other study designs. Most previous studies were done on older populations122,123, and it is valuable to have access to data from early life when studying risk factors for disease. When studying BMI at higher ages, diseases such as various cancer types124,125 may have developed. These diseases are more frequent in obese patients and may contribute to 40 41 the increased risk of VTE, meaning that they are effect modi ers. In studies on adults, Paper II was performed as a case-control study where we compared patients with PE however, results on VTE risk in overweight individuals might be adjusted for cancer, and patients with DVT, respectively, to their matched population controls. We chose and the true effect of BMI might be underestimated. When measuring weight in more this design instead of comparing PE to DVT patients since patients with PE gener- advanced ages, it is also possible that diseases such as cancer might have led to weight ally are older than DVT patients, which could confound the results on risk factors. loss, and the relationship between weight and VTE might be missed. The possibility However, the chosen design might be somewhat less intuitive and might be dif cult to study a large population longitudinally was also a strength in Paper IV since stud- to understand. ies of treatment effects with anticoagulants for VTE (in randomized controlled trials) commonly have a short follow-up and a limited sample size. Paper III would have had more clinical value if we, instead of the case-control de- sign, had studied a cohort of cancer patients with information on cancer stage from Generalizability. All Swedish citizens have access to publicly  nanced health care, the National Cancer Register with VTE as the outcome. This would not only have and the national registers have high coverage. Hence, the studies include patients who rendered more easily understood results, but the National Cancer Register could also would not have been included in randomized controlled trials or population studies have provided information on cancer stage at diagnosis. Our results on uterine, kid- due to severe comorbidities or incapacity to complete follow-up. However, the results ney, and bladder/urothelial cancer are likely to be highly in uenced by the cancer may not be equally valid for populations who have been shown to have a lower VTE stage, as proposed by earlier studies in the California Cancer Register39,128. In the incidence than the Swedish population, such as Southeast Asian populations23. Our previous studies, these cancers were often localized at diagnosis and were associated results may also differ from other countries due to differences in access to healthcare, with a low risk of VTE, whereas the same cancers were associated with a high VTE leading to differences in healthcare-seeking patterns and treatments. risk in more advanced stages. It would also have been interesting to have data on anti- tumoral treatment since this in uences the risk of VTE. Inpatient medical treatment National registries also have some disadvantages: (such as chemotherapy) is mainly reported in other registers that do not have national Accuracy of diagnosis. According to previous studies on included registries, PE coverage. Therefore, it would also be relevant to perform separate studies on differ- has a better positive predictive value (PPV) of 80.7% than DVT, with a low PPV of ent cancer types, including data from national quality registers such as the National 59.2%126, and inpatient diagnoses are more reliable than outpatient diagnoses127. For Quality Registry for Hematology or the National Quality Registry for Gynecological measures to increase PPV, please see the Limitations section. Oncology. Missing data/ unmeasured exposures. Swedish national healthcare registries lack in- Paper IV could have been conducted as a large, randomized, controlled trial with formation on data such as laboratory values, BMI, smoking, and diagnostic codes a long follow-up. This would not have given a true picture of clinical practice and from primary care. In Paper I, the result is a lack of information on subsequent weight would have required large resources, but it had reduced the risk of residual confound- development after enlistment for military service. In Papers II and IV, the consequence ing. The study could also have been performed as a prospective cohort study in a qual- is that we do not have complete data on comorbidities such as depression or hyperten- ity or research register in which parameters of speci c interest would be more care- sion, which is often cared for in primary care. We also lack granularity of data on for fully documented. These parameters could include the reason for treatment choice, example, kidney failure, which is likely to be correctly registered when the patient is data on bleeding that do not require inpatient care, details on kidney function, and followed by a nephrologist, but not in elderly patients with severe comorbidities. In use of over-the-counter medications such as NSAIDs. This approach would, however, Paper III, we lack information on antitumoral treatment and the stage of cancer, which require a lot of time and resources since there to date is no Swedish quality/ research are both known to in uence the risk of VTE28,128. register for VTE. Residual confounding. In Paper IV, the baseline data of patients treated with the dif- Clinical implications ferent anticoagulants differ. We have adjusted for all comorbidities we have judged as potential confounders, but it is possible that residual confounding remains. One indi- Our results on obesity in adolescence leading to an increased risk of future VTE may cation of this is the increased risk of bleeding in patients with concomitant treatment hopefully be an incentive for weight loss in overweight patients. It is likely that the with PPI during initial treatment with anticoagulation. The cause of this is probably incidence of VTE will increase with the growing prevalence of obesity. This calls for that PPI is prescribed to patients who are assessed to have a high bleeding risk by their more knowledge within all areas of VTE in relation to obesity, from concomitant risk doctor. factors to treatment and prognosis. Could our research questions have been better addressed with another study design? The strong correlation between PE and cardiopulmonary diseases suggests that when patients with these comorbidities present with acute respiratory or chest symptoms, With large economic resources, Paper I could have been performed as a prospective the suspicion of PE should be high. Our results could also be of importance for throm- cohort study with regular visits for follow-up of parameters such as weight to enable boprophylaxis. PE is a disease with higher mortality than DVT, indicating that throm- studies on time-updated data. However, it would not be possible with this study size. boprophylaxis is of higher importance in conditions with a strong correlation with PE. 42 43 the increased risk of VTE, meaning that they are effect modi ers. In studies on adults, Paper II was performed as a case-control study where we compared patients with PE however, results on VTE risk in overweight individuals might be adjusted for cancer, and patients with DVT, respectively, to their matched population controls. We chose and the true effect of BMI might be underestimated. When measuring weight in more this design instead of comparing PE to DVT patients since patients with PE gener- advanced ages, it is also possible that diseases such as cancer might have led to weight ally are older than DVT patients, which could confound the results on risk factors. loss, and the relationship between weight and VTE might be missed. The possibility However, the chosen design might be somewhat less intuitive and might be dif cult to study a large population longitudinally was also a strength in Paper IV since stud- to understand. ies of treatment effects with anticoagulants for VTE (in randomized controlled trials) commonly have a short follow-up and a limited sample size. Paper III would have had more clinical value if we, instead of the case-control de- sign, had studied a cohort of cancer patients with information on cancer stage from Generalizability. All Swedish citizens have access to publicly  nanced health care, the National Cancer Register with VTE as the outcome. This would not only have and the national registers have high coverage. Hence, the studies include patients who rendered more easily understood results, but the National Cancer Register could also would not have been included in randomized controlled trials or population studies have provided information on cancer stage at diagnosis. Our results on uterine, kid- due to severe comorbidities or incapacity to complete follow-up. However, the results ney, and bladder/urothelial cancer are likely to be highly in uenced by the cancer may not be equally valid for populations who have been shown to have a lower VTE stage, as proposed by earlier studies in the California Cancer Register39,128. In the incidence than the Swedish population, such as Southeast Asian populations23. Our previous studies, these cancers were often localized at diagnosis and were associated results may also differ from other countries due to differences in access to healthcare, with a low risk of VTE, whereas the same cancers were associated with a high VTE leading to differences in healthcare-seeking patterns and treatments. risk in more advanced stages. It would also have been interesting to have data on anti- tumoral treatment since this in uences the risk of VTE. Inpatient medical treatment National registries also have some disadvantages: (such as chemotherapy) is mainly reported in other registers that do not have national Accuracy of diagnosis. According to previous studies on included registries, PE coverage. Therefore, it would also be relevant to perform separate studies on differ- has a better positive predictive value (PPV) of 80.7% than DVT, with a low PPV of ent cancer types, including data from national quality registers such as the National 59.2%126, and inpatient diagnoses are more reliable than outpatient diagnoses127. For Quality Registry for Hematology or the National Quality Registry for Gynecological measures to increase PPV, please see the Limitations section. Oncology. Missing data/ unmeasured exposures. Swedish national healthcare registries lack in- Paper IV could have been conducted as a large, randomized, controlled trial with formation on data such as laboratory values, BMI, smoking, and diagnostic codes a long follow-up. This would not have given a true picture of clinical practice and from primary care. In Paper I, the result is a lack of information on subsequent weight would have required large resources, but it had reduced the risk of residual confound- development after enlistment for military service. In Papers II and IV, the consequence ing. The study could also have been performed as a prospective cohort study in a qual- is that we do not have complete data on comorbidities such as depression or hyperten- ity or research register in which parameters of speci c interest would be more care- sion, which is often cared for in primary care. We also lack granularity of data on for fully documented. These parameters could include the reason for treatment choice, example, kidney failure, which is likely to be correctly registered when the patient is data on bleeding that do not require inpatient care, details on kidney function, and followed by a nephrologist, but not in elderly patients with severe comorbidities. In use of over-the-counter medications such as NSAIDs. This approach would, however, Paper III, we lack information on antitumoral treatment and the stage of cancer, which require a lot of time and resources since there to date is no Swedish quality/ research are both known to in uence the risk of VTE28,128. register for VTE. Residual confounding. In Paper IV, the baseline data of patients treated with the dif- Clinical implications ferent anticoagulants differ. We have adjusted for all comorbidities we have judged as potential confounders, but it is possible that residual confounding remains. One indi- Our results on obesity in adolescence leading to an increased risk of future VTE may cation of this is the increased risk of bleeding in patients with concomitant treatment hopefully be an incentive for weight loss in overweight patients. It is likely that the with PPI during initial treatment with anticoagulation. The cause of this is probably incidence of VTE will increase with the growing prevalence of obesity. This calls for that PPI is prescribed to patients who are assessed to have a high bleeding risk by their more knowledge within all areas of VTE in relation to obesity, from concomitant risk doctor. factors to treatment and prognosis. Could our research questions have been better addressed with another study design? The strong correlation between PE and cardiopulmonary diseases suggests that when patients with these comorbidities present with acute respiratory or chest symptoms, With large economic resources, Paper I could have been performed as a prospective the suspicion of PE should be high. Our results could also be of importance for throm- cohort study with regular visits for follow-up of parameters such as weight to enable boprophylaxis. PE is a disease with higher mortality than DVT, indicating that throm- studies on time-updated data. However, it would not be possible with this study size. boprophylaxis is of higher importance in conditions with a strong correlation with PE. 42 43 A clinical dilemma is that patients with cancer not only have an increased risk of CONCLUSION thrombosis but also that their bleeding risk is higher. Our results on cancers with lower thrombotic risk highlight the need for an individualized approach to thrombo- prophylaxis and treatment of cancer-associated thrombosis based on more robust data This thesis concludes that the healthcare burden of VTE is likely to increase consid- on thrombotic risk and bleeding risk in various stages of different cancers. erably with an increasing prevalence of obesity in society. Women have higher VTE incidence in fertile years and ages over 80, whereas men have a higher incidence in With the introduction of DOACs, treatment practice has changed dramatically, largely middle age. Cardiopulmonary diseases and PE are closely linked, whereas DVT is due to the lower bleeding risk. However, both previous data and our results call for more closely related to musculoskeletal surgery and lower extremity fractures. Large caution in treating DOACs as a group when assessing bleeding risk. This is also ad- groups of cancer patients seem to have a low risk for VTE, even in some cancer types dressed in a randomized trial comparing bleeding risk in apixaban and rivaroxaban generally considered high-risk conditions, such as kidney and bladder cancer. Apixa- during initial treatment with an estimated completion date in December 2024129. In ban is associated with the lowest bleeding risk during initial treatment compared to current clinical practice, many patients with VTE without temporary provoking fac- warfarin and rivaroxaban, but in extended treatment, this difference is only statisti- tors are treated inde nitely. Our results support the view that this is a safe approach in cally signi cant compared to warfarin. Rivaroxaban is associated with a lower bleed- many patients, with a low incidence of treatment-associated bleeding during extended ing risk than warfarin in extended but not in initial treatment. treatment. At the same time, however, our results also indicate a need for careful consideration of the risks of anticoagulant treatment in older patients, patients with previous bleeding, and patients with liver disease. Limitations As stated earlier in the discussion, the present studies have some limitations. We lack external validation of the PE and DVT diagnoses. To increase the accuracy of diagnosis, we only included  rst-time VTEs. We also con rmed VTE diagnoses after July 2005 (January 2006 in Paper I), when the National Prescribed Drug Register was introduced, with the dispense of anticoagulant medication127. For diagnoses before the introduction of the National Prescribed Drug Register, only patients with inpatient diagnoses or diagnoses in the National Cause of Death Register were included in Pa- per I. In Papers II and III, we also included  rst-time outpatient diagnoses that were con rmed with the same diagnosis within three months. However, we do not have a validation study verifying the accuracy of these approaches. One limitation of Paper I is the lack of information on the subsequent weight develop- ment of the studied men after enlistment. However, it is well known that obesity in early adulthood is strongly predictive of obesity later in life130. The results in Paper II might have been in uenced by ascertainment bias. For exam- ple, heart failure patients are likely to seek medical care for shortness of breath more often than healthy persons. This might lead to a more frequent use of imaging and a higher likelihood of  nding PEs. Likewise, a patient who recently had orthopedic surgery in a hip or leg is more likely to have a swollen extremity and might be referred for an ultrasound to discover a DVT. However, the opposite is also true – patients with a plausible explanation for shortness of breath or a swollen leg might not be investi- gated for VTE1 31,132. 44 45 A clinical dilemma is that patients with cancer not only have an increased risk of CONCLUSION thrombosis but also that their bleeding risk is higher. Our results on cancers with lower thrombotic risk highlight the need for an individualized approach to thrombo- prophylaxis and treatment of cancer-associated thrombosis based on more robust data This thesis concludes that the healthcare burden of VTE is likely to increase consid- on thrombotic risk and bleeding risk in various stages of different cancers. erably with an increasing prevalence of obesity in society. Women have higher VTE incidence in fertile years and ages over 80, whereas men have a higher incidence in With the introduction of DOACs, treatment practice has changed dramatically, largely middle age. Cardiopulmonary diseases and PE are closely linked, whereas DVT is due to the lower bleeding risk. However, both previous data and our results call for more closely related to musculoskeletal surgery and lower extremity fractures. Large caution in treating DOACs as a group when assessing bleeding risk. This is also ad- groups of cancer patients seem to have a low risk for VTE, even in some cancer types dressed in a randomized trial comparing bleeding risk in apixaban and rivaroxaban generally considered high-risk conditions, such as kidney and bladder cancer. Apixa- during initial treatment with an estimated completion date in December 2024129. In ban is associated with the lowest bleeding risk during initial treatment compared to current clinical practice, many patients with VTE without temporary provoking fac- warfarin and rivaroxaban, but in extended treatment, this difference is only statisti- tors are treated inde nitely. Our results support the view that this is a safe approach in cally signi cant compared to warfarin. Rivaroxaban is associated with a lower bleed- many patients, with a low incidence of treatment-associated bleeding during extended ing risk than warfarin in extended but not in initial treatment. treatment. At the same time, however, our results also indicate a need for careful consideration of the risks of anticoagulant treatment in older patients, patients with previous bleeding, and patients with liver disease. Limitations As stated earlier in the discussion, the present studies have some limitations. We lack external validation of the PE and DVT diagnoses. To increase the accuracy of diagnosis, we only included  rst-time VTEs. We also con rmed VTE diagnoses after July 2005 (January 2006 in Paper I), when the National Prescribed Drug Register was introduced, with the dispense of anticoagulant medication127. For diagnoses before the introduction of the National Prescribed Drug Register, only patients with inpatient diagnoses or diagnoses in the National Cause of Death Register were included in Pa- per I. In Papers II and III, we also included  rst-time outpatient diagnoses that were con rmed with the same diagnosis within three months. However, we do not have a validation study verifying the accuracy of these approaches. One limitation of Paper I is the lack of information on the subsequent weight develop- ment of the studied men after enlistment. However, it is well known that obesity in early adulthood is strongly predictive of obesity later in life130. The results in Paper II might have been in uenced by ascertainment bias. For exam- ple, heart failure patients are likely to seek medical care for shortness of breath more often than healthy persons. This might lead to a more frequent use of imaging and a higher likelihood of  nding PEs. Likewise, a patient who recently had orthopedic surgery in a hip or leg is more likely to have a swollen extremity and might be referred for an ultrasound to discover a DVT. However, the opposite is also true – patients with a plausible explanation for shortness of breath or a swollen leg might not be investi- gated for VTE1 31,132. 44 45 FUTURE PERSPECTIVES ACKNOWLEDGEMENT With a growing number of obese patients with VTE, the underlying mechanisms of I want to thank my main supervisor, Per-Olof Hansson. Always encouraging, enthusias- the increased VTE risk need to be clari ed, as well as the effect of different concomi- tic, and reliable. Always up for new things, whether it’s going to thrombosis conferences tant risk factors and the effect of preventive measures such as thromboprophylaxis. or starting a VTE research group. You are dedicated and get the job done no matter how We need to establish the optimal dosage of anticoagulation for prophylaxis and treat- much other work you have. But also very clear about priorities between family and ment for this speci c patient group as well as the long-term prognosis after VTE. work. You are a truly good person, and I am honored to have you as my main supervisor. There are also uncertainties regarding the consequences of obesity treatments, such as bariatric surgery, on the uptake of peroral anticoagulant treatment and the possible Sverker Jern, my main supervisor for the original Ph.D. project on diagnostics on PE in risk of treatment failure, which need to be addressed. exhaled air, which was paused due to Covid-19. We met when you taught ECG when I was in medical school, and you invited me to do my master’s thesis in your lab. When Many clinical decisions regarding patients with risk of VTE or established VTE I moved back from Stockholm, I ran into you at the main entrance of Östra, and we de- require careful assessment of risk versus bene ts with anticoagulation. Due to the cided to continue where we left off six years earlier. Our Wednesday luncheons, discuss- lack of evidence of the bene t of clinical decision rules (in thromboprophylaxis)68 ing clinical work, science, and life outside of medicine, were weekly highlights. Thank or assessment of the risk of thrombosis versus the risk of bleeding (in decisions on you also for all the time you put into reviewing my thesis. treatment duration)31, the responsible clinician is often left without much guidance. Annika Rosengren, we ended up sitting next to each other at a dinner at Göteborgs Läk- However, arti cial intelligence, with its subset of machine learning, is evolving in aresällskap, and you offered to introduce me to register-based research by letting me everyday medicine. Today, it is an integrated part of areas such as the interpretation write the  rst paper of this thesis. That introduction was crucial to the change of the of electrocardiograms and diagnostic images133. The algorithms have also shown a Ph.D. project when Covid-19 made it impossible to continue with the studies on exhaled potential to aid decisions on thromboprophylaxis134 and VKA-dosing135. Hopefully, it air. You have exceptional writing skills, and your input in the papers of this thesis has can also be helpful in treatment decisions in patients with high bleeding risk, cancer taught me very much. patients, patients with previous bleeding on anticoagulation, or in decisions on ex- tended treatment in patients where the bene t is unclear. Sofi a Ekdahl, the head of the Internal Medicine, Geriatric, and Emergency Medicine De- partment at Sahlgrenska University Hospital/Östra, for creating a stimulating environ- Although the treatment of VTE has changed considerably over the last decade with ment that allows combining clinical and research work. And being a good human being. the introduction of DOACs, all currently available agents affect central factors in the coagulation cascade. Hence, the antithrombotic effect comes at a price of decreased Sara Lann and Henrik Norrsell, heads of the Hematology/VTE unit (353) and Acute hemostasis, leading to an increased risk of bleeding. New strategies for preventing Medicine unit (MAVA) at the Internal Medicine, Geriatric, and Emergency Medicine and treating VTE with inhibition of Factor XI/XIa and FXII are being explored, which Department at Sahlgrenska University Hospital/Östra, for providing time to work on seems to decrease thrombosis without signi cantly affecting hemostasis136,137. This is this project. Also for being sound colleagues and clinicians, patient and understanding particularly interesting in patients with a high bleeding risk, for whom we often need leaders and good friends. to terminate or dose-reduce anticoagulant treatment. Concomitantly, patent expira- tions are commencing for DOACs. With lower pricing of generic medications, DO- Sam Schulman, for helping me with articles and future research ideas even though you ACs will be affordable for more VTE patients worldwide, likely in uencing treatment have no personal gain in doing so. For being warm, helpful, and inclusive. I hope that patterns. one day, I have learned enough to have the possibility to pay it forward. Jan Sörbo, for sound suggestions on research work and complicated clinical cases. You are good at not complicating things; you identify what’s important and what’s not, which I highly appreciate. Per Karlsson, for sharing your knowledge on oncology, for encouragement and valuable advice. Martin Adiels and Bengt Bengtsson, for help with statistical analysis. Ulrica Forslund-Granheden, for helping me navigate the administration of being a Ph.D. student, and Eva Thydén, for traveling to Gothenburg just to help me with the layout of this thesis. 46 47 FUTURE PERSPECTIVES ACKNOWLEDGEMENT With a growing number of obese patients with VTE, the underlying mechanisms of I want to thank my main supervisor, Per-Olof Hansson. Always encouraging, enthusias- the increased VTE risk need to be clari ed, as well as the effect of different concomi- tic, and reliable. Always up for new things, whether it’s going to thrombosis conferences tant risk factors and the effect of preventive measures such as thromboprophylaxis. or starting a VTE research group. You are dedicated and get the job done no matter how We need to establish the optimal dosage of anticoagulation for prophylaxis and treat- much other work you have. But also very clear about priorities between family and ment for this speci c patient group as well as the long-term prognosis after VTE. work. You are a truly good person, and I am honored to have you as my main supervisor. There are also uncertainties regarding the consequences of obesity treatments, such as bariatric surgery, on the uptake of peroral anticoagulant treatment and the possible Sverker Jern, my main supervisor for the original Ph.D. project on diagnostics on PE in risk of treatment failure, which need to be addressed. exhaled air, which was paused due to Covid-19. We met when you taught ECG when I was in medical school, and you invited me to do my master’s thesis in your lab. When Many clinical decisions regarding patients with risk of VTE or established VTE I moved back from Stockholm, I ran into you at the main entrance of Östra, and we de- require careful assessment of risk versus bene ts with anticoagulation. Due to the cided to continue where we left off six years earlier. Our Wednesday luncheons, discuss- lack of evidence of the bene t of clinical decision rules (in thromboprophylaxis)68 ing clinical work, science, and life outside of medicine, were weekly highlights. Thank or assessment of the risk of thrombosis versus the risk of bleeding (in decisions on you also for all the time you put into reviewing my thesis. treatment duration)31, the responsible clinician is often left without much guidance. Annika Rosengren, we ended up sitting next to each other at a dinner at Göteborgs Läk- However, arti cial intelligence, with its subset of machine learning, is evolving in aresällskap, and you offered to introduce me to register-based research by letting me everyday medicine. Today, it is an integrated part of areas such as the interpretation write the  rst paper of this thesis. That introduction was crucial to the change of the of electrocardiograms and diagnostic images133. The algorithms have also shown a Ph.D. project when Covid-19 made it impossible to continue with the studies on exhaled potential to aid decisions on thromboprophylaxis134 and VKA-dosing135. Hopefully, it air. You have exceptional writing skills, and your input in the papers of this thesis has can also be helpful in treatment decisions in patients with high bleeding risk, cancer taught me very much. patients, patients with previous bleeding on anticoagulation, or in decisions on ex- tended treatment in patients where the bene t is unclear. Sofi a Ekdahl, the head of the Internal Medicine, Geriatric, and Emergency Medicine De- partment at Sahlgrenska University Hospital/Östra, for creating a stimulating environ- Although the treatment of VTE has changed considerably over the last decade with ment that allows combining clinical and research work. And being a good human being. the introduction of DOACs, all currently available agents affect central factors in the coagulation cascade. Hence, the antithrombotic effect comes at a price of decreased Sara Lann and Henrik Norrsell, heads of the Hematology/VTE unit (353) and Acute hemostasis, leading to an increased risk of bleeding. New strategies for preventing Medicine unit (MAVA) at the Internal Medicine, Geriatric, and Emergency Medicine and treating VTE with inhibition of Factor XI/XIa and FXII are being explored, which Department at Sahlgrenska University Hospital/Östra, for providing time to work on seems to decrease thrombosis without signi cantly affecting hemostasis136,137. This is this project. Also for being sound colleagues and clinicians, patient and understanding particularly interesting in patients with a high bleeding risk, for whom we often need leaders and good friends. to terminate or dose-reduce anticoagulant treatment. Concomitantly, patent expira- tions are commencing for DOACs. With lower pricing of generic medications, DO- Sam Schulman, for helping me with articles and future research ideas even though you ACs will be affordable for more VTE patients worldwide, likely in uencing treatment have no personal gain in doing so. For being warm, helpful, and inclusive. I hope that patterns. one day, I have learned enough to have the possibility to pay it forward. Jan Sörbo, for sound suggestions on research work and complicated clinical cases. You are good at not complicating things; you identify what’s important and what’s not, which I highly appreciate. Per Karlsson, for sharing your knowledge on oncology, for encouragement and valuable advice. Martin Adiels and Bengt Bengtsson, for help with statistical analysis. Ulrica Forslund-Granheden, for helping me navigate the administration of being a Ph.D. student, and Eva Thydén, for traveling to Gothenburg just to help me with the layout of this thesis. 46 47 Vladimir Radulovic, Fariba Baghaei, and Anna Olsson, for your warm welcome to REFERENCES the world of coagulation and thrombosis and never-ending patience with my numer- ous questions. Nina Jurander and Linda Myrin Westesson for all practical help at the Coagulation Unit. 1. Otoupalova E, Dalal B, Renard B. Right heart thrombus in transit: a series of two cases. Critical Ultrasound Journal. 2017/06/15 2017;9(1):14. doi:10.1186/s13089-017-0069- 9 Mazdak Tavoly and Kristina Tempelman Svennerholm, for great discussions on compli- cated clinical cases, research ideas, and collaboration on VTE guidelines and lectures. I 2. Joan L, Daniella AS, Michael N. COVID-19, immunothrombosis and venous throm- love being challenged with your knowledge. I hope we have many, many years together boembolism: biological mechanisms. Thorax. 2021;76(4):412. doi:10.1136/tho- ahead. raxjnl-2020-216243 Maria Roupe, for being my companion in VTE teaching and in being new at the coagu- 3. Paul Scott J, Flood V, Raf ni L. Hemostasis. Nelson Textbook of Pediatrics. Elsevier; lation unit. Who would have thought when we worked together in Alingsås 15 years 2020:2589-2594:chap 502. ago? Reliable, warm, and wise. Looking forward to working with you for the rest of my clinical life. 4. Mackman N. Triggers, targets and treatments for thrombosis. Nature. Feb 21 2008;451(7181):914-8. doi:10.1038/nature06797 Beatrice Aldenborg, Jacob Philipson, and Sara Hallström, for countless lunches, walks, 5. Mackman N. New insights into the mechanisms of venous thrombosis. J Clin Invest. texts, phone calls, runs (Jacob), and dinners, for making the research weeks less lonely Jul 2012;122(7):2331-6. doi:10.1172/jci60229 and life in general much more fun. I am truly grateful to have the three of you as my friends and colleagues. And Elin Axelsson Andrén, for lunch company, walks, laughs, 6. Virchow R. Gesammelte Abhandlungen zur wis senschaftlichen Medicin. Von and for being straight with me when I’m wrong. Meidinger & Sohn 1856. All colleagues at Medicin, Geriatrik och Akutmottagning, Östra sjukhuset, in particular 7. Introduction to the Study of Deep Venous Thrombosis. In: Malone PC, Agutter PS, eds. The Aetiology of Deep Venous Thrombosis: A Critical, Historical and Epistemological Maktgruppen, for collegiality. Survey. Springer Netherlands; 2008:1-9. My mom, Kristina, for always being there for me and my family with never-ending 8. Cockett FB, Thomas ML. The iliac compression syndrome. Br J Surg. Oct enthusiasm. In good times and bad. 1965;52(10):816-21. doi:10.1002/bjs.1800521028 My mother-in-law, Helena, for your help in our everyday life and for your numerous 9. Yamashita A, Asada Y. Underlying mechanisms of thrombus formation/growth in efforts to make me culturally educated. The whole Sandblad-family for giving a sense atherothrombosis and deep vein thrombosis. Pathology International. 2023/02/01 of belonging. 2023;73(2):65-80. doi:https://doi.org/10.1111/pin.13305 10. Brenner B. Haemostatic changes in pregnancy. Thromb Res. 2004;114(5-6):409-14. My brother, Lars, for always being there. For all great bike rides and runs. Great discus- doi:10.1016/j.thromres.2004.08.004 sions on life, research, and clinical work. And my sister-in-law, Josefi na, because you make things better for people around you. And all three of you (including Hedvig) for 11. Samad F, Ruf W. In ammation, obesity, and thrombosis. Blood. Nov 14 being there for my family when we need you. 2013;122(20):3415-22. doi:10.1182/blood-2013-05-427708 Staffan, my love. You are may rock. You remind me of what’s important in life and put 12. Frischmuth T, Hindberg K, Aukrust P, et al. Elevated plasma levels of plasminogen ac- things into perspective. You make me dare to make decisions that I would never have tivator inhibitor-1 are associated with risk of future incident venous thromboembolism. dared without you. J Thromb Haemost. Jul 2022;20(7):1618-1626. doi:10.1111/jth.15701 13. Galanaud JP, Laroche JP, Righini M. The history and historical treatments of deep vein Emma and Gustav, because the two of you are most important persons in my life, and I thrombosis. J Thromb Haemost. Mar 2013;11(3):402-11. doi:10.1111/jth.12127 love you both very, very much. 14. White C. An Inquiry Into the Nature and Cause of that Swelling, in One Or Both of the Lower Extremities, which Sometimes Happens to Lying-in Women: Together with an Examination Into the Propriety of Drawing the Breasts, of Those who Do, and Also of The work was supported by grants from the Swedish state under an agreement concerning Those who Do Not Give Suck. London: Warrington1784. research and education of doctors (ALF), Sahlgrenska University hospital funds, The Go- thenburg Medical Society, Elsa och Gustaf Lindhs stiftelse, Emelle fond, and the Swedish 15. Bauer G. Nine years’ Experience with Heparin in Acute Venous Thrombosis. Angiol-ogy. 1950;1(2):161-9. Heart and Lung association. 48 49 Vladimir Radulovic, Fariba Baghaei, and Anna Olsson, for your warm welcome to REFERENCES the world of coagulation and thrombosis and never-ending patience with my numer- ous questions. Nina Jurander and Linda Myrin Westesson for all practical help at the Coagulation Unit. 1. Otoupalova E, Dalal B, Renard B. Right heart thrombus in transit: a series of two cases. Critical Ultrasound Journal. 2017/06/15 2017;9(1):14. doi:10.1186/s13089-017-0069- 9 Mazdak Tavoly and Kristina Tempelman Svennerholm, for great discussions on compli- cated clinical cases, research ideas, and collaboration on VTE guidelines and lectures. I 2. Joan L, Daniella AS, Michael N. COVID-19, immunothrombosis and venous throm- love being challenged with your knowledge. I hope we have many, many years together boembolism: biological mechanisms. Thorax. 2021;76(4):412. doi:10.1136/tho- ahead. raxjnl-2020-216243 Maria Roupe, for being my companion in VTE teaching and in being new at the coagu- 3. Paul Scott J, Flood V, Raf ni L. Hemostasis. Nelson Textbook of Pediatrics. Elsevier; lation unit. Who would have thought when we worked together in Alingsås 15 years 2020:2589-2594:chap 502. ago? Reliable, warm, and wise. Looking forward to working with you for the rest of my clinical life. 4. Mackman N. Triggers, targets and treatments for thrombosis. Nature. Feb 21 2008;451(7181):914-8. doi:10.1038/nature06797 Beatrice Aldenborg, Jacob Philipson, and Sara Hallström, for countless lunches, walks, 5. Mackman N. New insights into the mechanisms of venous thrombosis. J Clin Invest. texts, phone calls, runs (Jacob), and dinners, for making the research weeks less lonely Jul 2012;122(7):2331-6. doi:10.1172/jci60229 and life in general much more fun. I am truly grateful to have the three of you as my friends and colleagues. And Elin Axelsson Andrén, for lunch company, walks, laughs, 6. Virchow R. Gesammelte Abhandlungen zur wis senschaftlichen Medicin. Von and for being straight with me when I’m wrong. Meidinger & Sohn 1856. All colleagues at Medicin, Geriatrik och Akutmottagning, Östra sjukhuset, in particular 7. Introduction to the Study of Deep Venous Thrombosis. In: Malone PC, Agutter PS, eds. The Aetiology of Deep Venous Thrombosis: A Critical, Historical and Epistemological Maktgruppen, for collegiality. Survey. Springer Netherlands; 2008:1-9. My mom, Kristina, for always being there for me and my family with never-ending 8. Cockett FB, Thomas ML. The iliac compression syndrome. Br J Surg. Oct enthusiasm. In good times and bad. 1965;52(10):816-21. doi:10.1002/bjs.1800521028 My mother-in-law, Helena, for your help in our everyday life and for your numerous 9. Yamashita A, Asada Y. Underlying mechanisms of thrombus formation/growth in efforts to make me culturally educated. The whole Sandblad-family for giving a sense atherothrombosis and deep vein thrombosis. Pathology International. 2023/02/01 of belonging. 2023;73(2):65-80. doi:https://doi.org/10.1111/pin.13305 10. Brenner B. Haemostatic changes in pregnancy. Thromb Res. 2004;114(5-6):409-14. My brother, Lars, for always being there. For all great bike rides and runs. Great discus- doi:10.1016/j.thromres.2004.08.004 sions on life, research, and clinical work. And my sister-in-law, Josefi na, because you make things better for people around you. And all three of you (including Hedvig) for 11. Samad F, Ruf W. In ammation, obesity, and thrombosis. Blood. Nov 14 being there for my family when we need you. 2013;122(20):3415-22. doi:10.1182/blood-2013-05-427708 Staffan, my love. You are may rock. You remind me of what’s important in life and put 12. Frischmuth T, Hindberg K, Aukrust P, et al. Elevated plasma levels of plasminogen ac- things into perspective. You make me dare to make decisions that I would never have tivator inhibitor-1 are associated with risk of future incident venous thromboembolism. dared without you. J Thromb Haemost. Jul 2022;20(7):1618-1626. doi:10.1111/jth.15701 13. Galanaud JP, Laroche JP, Righini M. The history and historical treatments of deep vein Emma and Gustav, because the two of you are most important persons in my life, and I thrombosis. J Thromb Haemost. Mar 2013;11(3):402-11. doi:10.1111/jth.12127 love you both very, very much. 14. White C. An Inquiry Into the Nature and Cause of that Swelling, in One Or Both of the Lower Extremities, which Sometimes Happens to Lying-in Women: Together with an Examination Into the Propriety of Drawing the Breasts, of Those who Do, and Also of The work was supported by grants from the Swedish state under an agreement concerning Those who Do Not Give Suck. London: Warrington1784. research and education of doctors (ALF), Sahlgrenska University hospital funds, The Go- thenburg Medical Society, Elsa och Gustaf Lindhs stiftelse, Emelle fond, and the Swedish 15. Bauer G. Nine years’ Experience with Heparin in Acute Venous Thrombosis. Angiol-ogy. 1950;1(2):161-9. Heart and Lung association. 48 49 16. Holden WD. Treatment of deep venous thrombosis with reference to subcutaneous 28. Mulder FI, Horváth-Puhó E, van Es N, et al. Venous thromboembolism in cancer injection of heparin and use of dicumarol. Arch Surg (1920). Feb 1947;54(2):183-7. patients: a population-based cohort study. Blood. Apr 8 2021;137(14):1959-1969. doi:10.1001/archsurg.1947.01230070188006 doi:10.1182/blood.2020007338 17. Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-weight heparin ad- 29. Klovaite J, Benn M, Nordestgaard BG. Obesity as a causal risk factor for deep venous ministered primarily at home with unfractionated heparin administered in the hospi- thrombosis: a Mendelian randomization study. J Intern Med. May 2015;277(5):573-84. tal for proximal deep-vein thrombosis. N Engl J Med. Mar 14 1996;334(11):677-81. doi:10.1111/joim.12299 doi:10.1056/nejm199603143341101 30. (NCD-RisC) NRFC. Worldwide trends in body-mass index, underweight, over- 18. Partsch H, Blättler W. Compression and walking versus bed rest in the treatment of weight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based proximal deep venous thrombosis with low molecular weight heparin. Journal of measurement studies in 128·9 million children, adolescents, and adults. Lancet. Dec Vascular Surgery. 2000/11/01/ 2000;32(5):861-869. doi:https://doi.org/10.1067/ 2017;390(10113):2627-2642. doi:10.1016/S0140-6736(17)32129-3 mva.2000.110352 31. Khan F, Tritschler T, Kahn SR, Rodger MA. Venous thromboembolism. Lancet. Jul 3 19. Zondag W, Mos IC, Creemers-Schild D, et al. Outpatient treatment in patients with 2021;398(10294):64-77. doi:10.1016/s0140-6736(20)32658-1 acute pulmonary embolism: the Hestia Study. J Thromb Haemost. Aug 2011;9(8):1500- 7. doi:10.1111/j.1538-7836.2011.04388.x 32. Height and body-mass index trajectories of school-aged children and adolescents from 1985 to 2019 in 200 countries and territories: a pooled analysis of 2181 population- 20. Konstantinides SV, Meyer G, Becattini C, et al. 2019 ESC Guidelines for the diagnosis based studies with 65 million participants. Lancet. Nov 7 2020;396(10261):1511-1524. and management of acute pulmonary embolism developed in collaboration with the doi:10.1016/s0140-6736(20)31859-6 European Respiratory Society (ERS): The Task Force for the diagnosis and manage- ment of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur 33. Anderson FA, Jr., Spencer FA. Risk factors for venous thromboembolism. Circulation. Respir J. 08 2019;doi:10.1183/13993003.01647-2019 Jun 17 2003;107(23 Suppl 1):I9-16. doi:10.1161/01.cir.0000078469.07362.e6 21. Kearon C, Akl EA, Ornelas J, et al. Antithrombotic Therapy for VTE Disease: CHEST 34. Hales C, Carroll M, Fryar C, Ogden C. Prevalence of obesity and severe obesity among Guideline and Expert Panel Report. Chest. Feb 2016;149(2):315-352. doi:10.1016/j. adults: United States, 2017–2018. National Center for Health Statistics. Accessed 6 chest.2015.11.026 September, 2023. 22. Ortel TL, Neumann I, Ageno W, et al. American Society of Hematology 2020 guide- 35. Lyman GH, Carrier M, Ay C, et al. American Society of Hematology 2021 guide- lines for management of venous thromboembolism: treatment of deep vein thrombo- lines for management of venous thromboembolism: prevention and treatment in pa- sis and pulmonary embolism. Blood Adv. Oct 13 2020;4(19):4693-4738. doi:10.1182/ tients with cancer. Blood Adv. Feb 23 2021;5(4):927-974. doi:10.1182/bloodadvanc- bloodadvances.2020001830 es.2020003442 23. Raskob GE, Angchaisuksiri P, Blanco AN, et al. Thrombosis: a major contributor to 36. D’Astous J, Carrier M. Screening for Occult Cancer in Patients with Venous Thrombo- global disease burden. Arterioscler Thromb Vasc Biol. Nov 2014;34(11):2363-71. embolism. Journal of clinical medicine. 2020;9(8):2389. doi:10.3390/jcm9082389 doi:10.1161/ATVBAHA.114.304488 37. Streiff MB, Holmstrom B, Angelini D, et al. Cancer-Associated Venous Thromboem- 24. Liao S, Woulfe T, Hyder S, Merriman E, Simpson D, Chunilal S. Incidence of venous bolic Disease, Version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl thromboembolism in different ethnic groups: a regional direct comparison study. Jour- Compr Canc Netw. Oct 15 2021;19(10):1181-1201. doi:10.6004/jnccn.2021.0047 nal of Thrombosis and Haemostasis. 2014/02/01/ 2014;12(2):214-219. doi:https://doi. org/10.1111/jth.12464 38. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood. 25. Dentali F, Ageno W, Pomero F, Fenoglio L, Squizzato A, Bonzini M. Time trends and 2008;111(10):4902-4907. doi:10.1182/blood-2007-10-116327 case fatality rate of in-hospital treated pulmonary embolism during 11 years of obser- vation in Northwestern Italy. Thromb Haemost. Jan 2016;115(2):399-405. doi:10.1160/ 39. Mahajan A, Brunson A, Adesina O, Keegan THM, Wun T. The incidence of cancer- th15-02-0172 associated thrombosis is increasing over time. Blood Adv. 01 11 2022;6(1):307-320. doi:10.1182/bloodadvances.2021005590 26. Lehnert P, Lange T, Møller CH, Olsen PS, Carlsen J. Acute Pulmonary Embolism in a National Danish Cohort: Increasing Incidence and Decreasing Mortality. Thromb Hae- 40. Wattanakit K, Lutsey PL, Bell EJ, et al. Association between cardiovascular disease most. 03 2018;118(3):539-546. doi:10.1160/TH17-08-0531 risk factors and occurrence of venous thromboembolism. A time-dependent analysis. Thromb Haemost. Sep 2012;108(3):508-15. doi:10.1160/th11-10-0726 27. Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med. May 9 2011;171(9):831-7. 41. Sørensen HT, Horvath-Puho E, Lash TL, et al. Heart disease may be a risk factor for doi:10.1001/archinternmed.2011.178 pulmonary embolism without peripheral deep venous thrombosis. Circulation. Sep 27 2011;124(13):1435-41. doi:10.1161/circulationaha.111.025627 50 51 16. Holden WD. Treatment of deep venous thrombosis with reference to subcutaneous 28. Mulder FI, Horváth-Puhó E, van Es N, et al. Venous thromboembolism in cancer injection of heparin and use of dicumarol. Arch Surg (1920). Feb 1947;54(2):183-7. patients: a population-based cohort study. Blood. Apr 8 2021;137(14):1959-1969. doi:10.1001/archsurg.1947.01230070188006 doi:10.1182/blood.2020007338 17. Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-weight heparin ad- 29. Klovaite J, Benn M, Nordestgaard BG. Obesity as a causal risk factor for deep venous ministered primarily at home with unfractionated heparin administered in the hospi- thrombosis: a Mendelian randomization study. J Intern Med. May 2015;277(5):573-84. tal for proximal deep-vein thrombosis. N Engl J Med. Mar 14 1996;334(11):677-81. doi:10.1111/joim.12299 doi:10.1056/nejm199603143341101 30. (NCD-RisC) NRFC. Worldwide trends in body-mass index, underweight, over- 18. Partsch H, Blättler W. Compression and walking versus bed rest in the treatment of weight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based proximal deep venous thrombosis with low molecular weight heparin. Journal of measurement studies in 128·9 million children, adolescents, and adults. Lancet. Dec Vascular Surgery. 2000/11/01/ 2000;32(5):861-869. doi:https://doi.org/10.1067/ 2017;390(10113):2627-2642. doi:10.1016/S0140-6736(17)32129-3 mva.2000.110352 31. Khan F, Tritschler T, Kahn SR, Rodger MA. Venous thromboembolism. Lancet. Jul 3 19. Zondag W, Mos IC, Creemers-Schild D, et al. Outpatient treatment in patients with 2021;398(10294):64-77. doi:10.1016/s0140-6736(20)32658-1 acute pulmonary embolism: the Hestia Study. J Thromb Haemost. Aug 2011;9(8):1500- 7. doi:10.1111/j.1538-7836.2011.04388.x 32. Height and body-mass index trajectories of school-aged children and adolescents from 1985 to 2019 in 200 countries and territories: a pooled analysis of 2181 population- 20. Konstantinides SV, Meyer G, Becattini C, et al. 2019 ESC Guidelines for the diagnosis based studies with 65 million participants. Lancet. Nov 7 2020;396(10261):1511-1524. and management of acute pulmonary embolism developed in collaboration with the doi:10.1016/s0140-6736(20)31859-6 European Respiratory Society (ERS): The Task Force for the diagnosis and manage- ment of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur 33. Anderson FA, Jr., Spencer FA. Risk factors for venous thromboembolism. Circulation. Respir J. 08 2019;doi:10.1183/13993003.01647-2019 Jun 17 2003;107(23 Suppl 1):I9-16. doi:10.1161/01.cir.0000078469.07362.e6 21. Kearon C, Akl EA, Ornelas J, et al. Antithrombotic Therapy for VTE Disease: CHEST 34. Hales C, Carroll M, Fryar C, Ogden C. Prevalence of obesity and severe obesity among Guideline and Expert Panel Report. Chest. Feb 2016;149(2):315-352. doi:10.1016/j. adults: United States, 2017–2018. National Center for Health Statistics. Accessed 6 chest.2015.11.026 September, 2023. 22. Ortel TL, Neumann I, Ageno W, et al. American Society of Hematology 2020 guide- 35. Lyman GH, Carrier M, Ay C, et al. American Society of Hematology 2021 guide- lines for management of venous thromboembolism: treatment of deep vein thrombo- lines for management of venous thromboembolism: prevention and treatment in pa- sis and pulmonary embolism. Blood Adv. Oct 13 2020;4(19):4693-4738. doi:10.1182/ tients with cancer. Blood Adv. Feb 23 2021;5(4):927-974. doi:10.1182/bloodadvanc- bloodadvances.2020001830 es.2020003442 23. Raskob GE, Angchaisuksiri P, Blanco AN, et al. Thrombosis: a major contributor to 36. D’Astous J, Carrier M. Screening for Occult Cancer in Patients with Venous Thrombo- global disease burden. Arterioscler Thromb Vasc Biol. Nov 2014;34(11):2363-71. embolism. Journal of clinical medicine. 2020;9(8):2389. doi:10.3390/jcm9082389 doi:10.1161/ATVBAHA.114.304488 37. Streiff MB, Holmstrom B, Angelini D, et al. Cancer-Associated Venous Thromboem- 24. Liao S, Woulfe T, Hyder S, Merriman E, Simpson D, Chunilal S. Incidence of venous bolic Disease, Version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl thromboembolism in different ethnic groups: a regional direct comparison study. Jour- Compr Canc Netw. Oct 15 2021;19(10):1181-1201. doi:10.6004/jnccn.2021.0047 nal of Thrombosis and Haemostasis. 2014/02/01/ 2014;12(2):214-219. doi:https://doi. org/10.1111/jth.12464 38. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood. 25. Dentali F, Ageno W, Pomero F, Fenoglio L, Squizzato A, Bonzini M. Time trends and 2008;111(10):4902-4907. doi:10.1182/blood-2007-10-116327 case fatality rate of in-hospital treated pulmonary embolism during 11 years of obser- vation in Northwestern Italy. Thromb Haemost. Jan 2016;115(2):399-405. doi:10.1160/ 39. Mahajan A, Brunson A, Adesina O, Keegan THM, Wun T. The incidence of cancer- th15-02-0172 associated thrombosis is increasing over time. Blood Adv. 01 11 2022;6(1):307-320. doi:10.1182/bloodadvances.2021005590 26. Lehnert P, Lange T, Møller CH, Olsen PS, Carlsen J. Acute Pulmonary Embolism in a National Danish Cohort: Increasing Incidence and Decreasing Mortality. Thromb Hae- 40. Wattanakit K, Lutsey PL, Bell EJ, et al. Association between cardiovascular disease most. 03 2018;118(3):539-546. doi:10.1160/TH17-08-0531 risk factors and occurrence of venous thromboembolism. A time-dependent analysis. Thromb Haemost. Sep 2012;108(3):508-15. doi:10.1160/th11-10-0726 27. Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med. May 9 2011;171(9):831-7. 41. Sørensen HT, Horvath-Puho E, Lash TL, et al. Heart disease may be a risk factor for doi:10.1001/archinternmed.2011.178 pulmonary embolism without peripheral deep venous thrombosis. Circulation. Sep 27 2011;124(13):1435-41. doi:10.1161/circulationaha.111.025627 50 51 42. Hansson PO, Eriksson H, Welin L, Svardsudd K, Wilhelmsen L. Smoking and abdomi- 55. Roach REJ, Lijfering WM, Rosendaal FR, Cannegieter SC, le Cessie S. Sex Differ- nal obesity: risk factors for venous thromboembolism among middle-aged men: “the ence in Risk of Second but Not of First Venous Thrombosis. Circulation. 2014/01/07 study of men born in 1913”. Arch Intern Med. Sep 13 1999;159(16):1886-90. 2014;129(1):51-56. doi:10.1161/CIRCULATIONAHA.113.004768 43. Enga KF, Braekkan SK, Hansen-Krone IJ, le Cessie S, Rosendaal FR, Hansen JB. Ciga- 56. van Langevelde K, Flinterman LE, van Hylckama Vlieg A, Rosendaal FR, Cannegi- rette smoking and the risk of venous thromboembolism: the Tromsø Study. J Thromb eter SC. Broadening the factor V Leiden paradox: pulmonary embolism and deep-vein Haemost. Oct 2012;10(10):2068-74. doi:10.1111/j.1538-7836.2012.04880.x thrombosis as 2 sides of the spectrum. Blood. Aug 2 2012;120(5):933-46. doi:10.1182/ blood-2012-02-407551 44. Rinde LB, Småbrekke B, Mathiesen EB, et al. Ischemic Stroke and Risk of Venous Thromboembolism in the General Population: The Tromsø Study. J Am Heart Assoc. 57. Søgaard KK, Schmidt M, Pedersen L, Horváth-Puhó E, Sørensen HT. 30-year mortal- Nov 7 2016;5(11)doi:10.1161/jaha.116.004311 ity after venous thromboembolism: a population-based cohort study. Circulation. Sep 2 2014;130(10):829-36. doi:10.1161/circulationaha.114.009107 45. Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous throm- boembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol. Dec 15 58. Anderson DR, Morgano GP, Bennett C, et al. American Society of Hematology 2019 2005;96(12):1731-3. doi:10.1016/j.amjcard.2005.07.097 guidelines for management of venous thromboembolism: prevention of venous throm- boembolism in surgical hospitalized patients. Blood Adv. Dec 10 2019;3(23):3898- 46. Lankeit M, Held M. Incidence of venous thromboembolism in COPD: linking in am- 3944. doi:10.1182/bloodadvances.2019000975 mation and thrombosis? Eur Respir J. 2016:369-73. vol. 2. 59. Farge D, Frere C, Connors JM, et al. 2022 international clinical practice guidelines 47. Schneider C, Bothner U, Jick SS, Meier CR. Chronic obstructive pulmonary disease for the treatment and prophylaxis of venous thromboembolism in patients with can- and the risk of cardiovascular diseases. Eur J Epidemiol. Apr 2010;25(4):253-60. cer, including patients with COVID-19. Lancet Oncol. Jul 2022;23(7):e334-e347. doi:10.1007/s10654-010-9435-7 doi:10.1016/s1470-2045(22)00160-7 48. Borgstroem S, Greitz T, Van Der Linden W, Molin J, Rudics I. ANTICOAGULANT 60. Forgo G, Micieli E, Ageno W, et al. An update on the global use of risk assessment PROPHYLAXIS OF VENOUS THROMBOSIS IN PATIENTS WITH FRACTURED models and thromboprophylaxis in hospitalized patients with medical illnesses from NECK OF THE FEMUR; A CONTROLLED CLINICAL TRIAL USING VENOUS the World Thrombosis Day steering committee: Systematic review and meta-analysis. PHLEBOGRAPHY. Acta Chir Scand. May 1965;129:500-8. J Thromb Haemost. Feb 2022;20(2):409-421. doi:10.1111/jth.15607 49. Falck-Ytter Y, Francis CW, Johanson NA, et al. Prevention of VTE in orthopedic sur- 61. Sandström K, Guðnadóttir G, Wilhelmson K, Kristjánsdóttir H, Stigendal L. [Venous gery patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: Ameri- thromboembolism prophylaxis in medical patients at Sahlgrenska University Hospital]. can College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. Lakartidningen. May 30 2017;114Medicinpatienter behöver bättre trombosprofylax - Feb 2012;141(2 Suppl):e278S-e325S. doi:10.1378/chest.11-2404 12 procent av patienter med hög risk för venös tromboembolism  ck profylax – den internationella siffran är 40–60 procent. 50. Zöller B, Svensson PJ, Dahlbäck B, Lind-Hallden C, Hallden C, Elf J. Genetic risk fac- tors for venous thromboembolism. Expert Rev Hematol. Sep 2020;13(9):971-981. doi: 62. Samama MM, Cohen AT, Darmon JY, et al. A comparison of enoxaparin with pla- 10.1080/17474086.2020.1804354 cebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N Engl J Med. Sep 9 51. Skeith L. Anticoagulating patients with high-risk acquired thrombophilias. Hematol- 1999;341(11):793-800. doi:10.1056/nejm199909093411103 ogy. 2018;2018(1):439-449. doi:10.1182/asheducation-2018.1.439 63. Leizorovicz A, Cohen AT, Turpie AG, Olsson CG, Vaitkus PT, Goldhaber SZ. Ran- 52. Thienel M, Müller-Reif JB, Zhang Z, et al. Immobility-associated thromboprotec- domized, placebo-controlled trial of dalteparin for the prevention of venous throm- tion is conserved across mammalian species from bear to human. Science. 2023/04/14 boembolism in acutely ill medical patients. Circulation. Aug 17 2004;110(7):874-9. 2023;380(6641):178-187. doi:10.1126/science.abo5044 doi:10.1161/01.cir.0000138928.83266.24 53. Simon T, De Jonage-Canonico MBY, Oger E, et al. Indicators of lifetime endoge- 64. Hunt BJ. Preventing hospital associated venous thromboembolism. Bmj. 2019:l4239. nous estrogen exposure and risk of venous thromboembolism. Journal of Thrombosis and Haemostasis. 2006/01/01/ 2006;4(1):71-76. doi:https://doi.org/10.1111/j.1538- 65. Flanders SA, Greene MT, Grant P, et al. Hospital performance for pharmacologic venous 7836.2005.01693.x thromboembolism prophylaxis and rate of venous thromboembolism : a cohort study. JAMA Intern Med. Oct 2014;174(10):1577-84. doi:10.1001/jamainternmed.2014.3384 54. Sultan AA, West J, Tata LJ, Fleming KM, Nelson-Piercy C, Grainge MJ. Risk of  rst venous thromboembolism in and around pregnancy: a population-based cohort study. 66. Rowswell HR, Nokes TJC. Signi cant reduction in hospital-acquired thrombosis: impact Br J Haematol. Feb 2012;156(3):366-73. doi:10.1111/j.1365-2141.2011.08956.x of national risk assessment and real-time feedback. Open Heart. 2017;4(2):e000653. doi:10.1136/openhrt-2017-000653 52 53 42. Hansson PO, Eriksson H, Welin L, Svardsudd K, Wilhelmsen L. Smoking and abdomi- 55. Roach REJ, Lijfering WM, Rosendaal FR, Cannegieter SC, le Cessie S. Sex Differ- nal obesity: risk factors for venous thromboembolism among middle-aged men: “the ence in Risk of Second but Not of First Venous Thrombosis. Circulation. 2014/01/07 study of men born in 1913”. Arch Intern Med. Sep 13 1999;159(16):1886-90. 2014;129(1):51-56. doi:10.1161/CIRCULATIONAHA.113.004768 43. Enga KF, Braekkan SK, Hansen-Krone IJ, le Cessie S, Rosendaal FR, Hansen JB. Ciga- 56. van Langevelde K, Flinterman LE, van Hylckama Vlieg A, Rosendaal FR, Cannegi- rette smoking and the risk of venous thromboembolism: the Tromsø Study. J Thromb eter SC. Broadening the factor V Leiden paradox: pulmonary embolism and deep-vein Haemost. Oct 2012;10(10):2068-74. doi:10.1111/j.1538-7836.2012.04880.x thrombosis as 2 sides of the spectrum. Blood. Aug 2 2012;120(5):933-46. doi:10.1182/ blood-2012-02-407551 44. Rinde LB, Småbrekke B, Mathiesen EB, et al. Ischemic Stroke and Risk of Venous Thromboembolism in the General Population: The Tromsø Study. J Am Heart Assoc. 57. Søgaard KK, Schmidt M, Pedersen L, Horváth-Puhó E, Sørensen HT. 30-year mortal- Nov 7 2016;5(11)doi:10.1161/jaha.116.004311 ity after venous thromboembolism: a population-based cohort study. Circulation. Sep 2 2014;130(10):829-36. doi:10.1161/circulationaha.114.009107 45. Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous throm- boembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol. Dec 15 58. Anderson DR, Morgano GP, Bennett C, et al. American Society of Hematology 2019 2005;96(12):1731-3. doi:10.1016/j.amjcard.2005.07.097 guidelines for management of venous thromboembolism: prevention of venous throm- boembolism in surgical hospitalized patients. Blood Adv. Dec 10 2019;3(23):3898- 46. Lankeit M, Held M. Incidence of venous thromboembolism in COPD: linking in am- 3944. doi:10.1182/bloodadvances.2019000975 mation and thrombosis? Eur Respir J. 2016:369-73. vol. 2. 59. Farge D, Frere C, Connors JM, et al. 2022 international clinical practice guidelines 47. Schneider C, Bothner U, Jick SS, Meier CR. Chronic obstructive pulmonary disease for the treatment and prophylaxis of venous thromboembolism in patients with can- and the risk of cardiovascular diseases. Eur J Epidemiol. Apr 2010;25(4):253-60. cer, including patients with COVID-19. Lancet Oncol. Jul 2022;23(7):e334-e347. doi:10.1007/s10654-010-9435-7 doi:10.1016/s1470-2045(22)00160-7 48. Borgstroem S, Greitz T, Van Der Linden W, Molin J, Rudics I. ANTICOAGULANT 60. Forgo G, Micieli E, Ageno W, et al. An update on the global use of risk assessment PROPHYLAXIS OF VENOUS THROMBOSIS IN PATIENTS WITH FRACTURED models and thromboprophylaxis in hospitalized patients with medical illnesses from NECK OF THE FEMUR; A CONTROLLED CLINICAL TRIAL USING VENOUS the World Thrombosis Day steering committee: Systematic review and meta-analysis. PHLEBOGRAPHY. Acta Chir Scand. May 1965;129:500-8. J Thromb Haemost. Feb 2022;20(2):409-421. doi:10.1111/jth.15607 49. Falck-Ytter Y, Francis CW, Johanson NA, et al. Prevention of VTE in orthopedic sur- 61. Sandström K, Guðnadóttir G, Wilhelmson K, Kristjánsdóttir H, Stigendal L. [Venous gery patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: Ameri- thromboembolism prophylaxis in medical patients at Sahlgrenska University Hospital]. can College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. Lakartidningen. May 30 2017;114Medicinpatienter behöver bättre trombosprofylax - Feb 2012;141(2 Suppl):e278S-e325S. doi:10.1378/chest.11-2404 12 procent av patienter med hög risk för venös tromboembolism  ck profylax – den internationella siffran är 40–60 procent. 50. Zöller B, Svensson PJ, Dahlbäck B, Lind-Hallden C, Hallden C, Elf J. Genetic risk fac- tors for venous thromboembolism. Expert Rev Hematol. Sep 2020;13(9):971-981. doi: 62. Samama MM, Cohen AT, Darmon JY, et al. A comparison of enoxaparin with pla- 10.1080/17474086.2020.1804354 cebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N Engl J Med. Sep 9 51. Skeith L. Anticoagulating patients with high-risk acquired thrombophilias. Hematol- 1999;341(11):793-800. doi:10.1056/nejm199909093411103 ogy. 2018;2018(1):439-449. doi:10.1182/asheducation-2018.1.439 63. Leizorovicz A, Cohen AT, Turpie AG, Olsson CG, Vaitkus PT, Goldhaber SZ. Ran- 52. Thienel M, Müller-Reif JB, Zhang Z, et al. Immobility-associated thromboprotec- domized, placebo-controlled trial of dalteparin for the prevention of venous throm- tion is conserved across mammalian species from bear to human. Science. 2023/04/14 boembolism in acutely ill medical patients. Circulation. Aug 17 2004;110(7):874-9. 2023;380(6641):178-187. doi:10.1126/science.abo5044 doi:10.1161/01.cir.0000138928.83266.24 53. Simon T, De Jonage-Canonico MBY, Oger E, et al. Indicators of lifetime endoge- 64. Hunt BJ. Preventing hospital associated venous thromboembolism. Bmj. 2019:l4239. nous estrogen exposure and risk of venous thromboembolism. Journal of Thrombosis and Haemostasis. 2006/01/01/ 2006;4(1):71-76. doi:https://doi.org/10.1111/j.1538- 65. Flanders SA, Greene MT, Grant P, et al. Hospital performance for pharmacologic venous 7836.2005.01693.x thromboembolism prophylaxis and rate of venous thromboembolism : a cohort study. JAMA Intern Med. Oct 2014;174(10):1577-84. doi:10.1001/jamainternmed.2014.3384 54. Sultan AA, West J, Tata LJ, Fleming KM, Nelson-Piercy C, Grainge MJ. Risk of  rst venous thromboembolism in and around pregnancy: a population-based cohort study. 66. Rowswell HR, Nokes TJC. Signi cant reduction in hospital-acquired thrombosis: impact Br J Haematol. Feb 2012;156(3):366-73. doi:10.1111/j.1365-2141.2011.08956.x of national risk assessment and real-time feedback. Open Heart. 2017;4(2):e000653. doi:10.1136/openhrt-2017-000653 52 53 67. Mottier D, Girard P, Couturaud F, et al. Enoxaparin versus Placebo to Prevent Symptom- 80. Vedantham S, Goldhaber SZ, Julian JA, et al. Pharmacomechanical Catheter-Directed atic Venous Thromboembolism in Hospitalized Older Adult Medical Patients. NEJM Thrombolysis for Deep-Vein Thrombosis. N Engl J Med. Dec 7 2017;377(23):2240- Evidence. 2023/07/25 2023;2(8):EVIDoa2200332. doi:10.1056/EVIDoa2200332 2252. doi:10.1056/NEJMoa1615066 68. Horner D, Goodacre S, Davis S, Burton N, Hunt BJ. Which is the best model to assess risk 81. Pruszczyk P, Klok FA, Kucher N, et al. Percutaneous treatment options for acute pul- for venous thromboembolism in hospitalised patients? Bmj. May 27 2021;373:n1106. monary embolism: a clinical consensus statement by the ESC Working Group on Pul- doi:10.1136/bmj.n1106 monary Circulation and Right Ventricular Function and the European Association of Percutaneous Cardiovascular Interventions. EuroIntervention. Oct 7 2022;18(8):e623- 69. Carrier M, Abou-Nassar K, Mallick R, et al. Apixaban to Prevent Venous Thrombo- e638. doi:10.4244/eij-d-22-00246 embolism in Patients with Cancer. New England Journal of Medicine. 2019/02/21 2018;380(8):711-719. doi:10.1056/NEJMoa1814468 82. Goldberg JB, Giri J, Kobayashi T, et al. Surgical Management and Mechanical Circula- tory Support in High-Risk Pulmonary Embolisms: Historical Context, Current Status, 70. Khorana AA, Soff GA, Kakkar AK, et al. Rivaroxaban for Thromboprophylaxis and Future Directions: A Scienti c Statement From the American Heart Association. in High-Risk Ambulatory Patients with Cancer. New England Journal of Medicine. Circulation. Feb 28 2023;147(9):e628-e647. doi:10.1161/cir.0000000000001117 2019/02/21 2019;380(8):720-728. doi:10.1056/NEJMoa1814630 83. Casey ET, Murad MH, Zumaeta-Garcia M, et al. Treatment of acute iliofemoral deep 71. Agnelli G. Direct Oral Anticoagulants for Thromboprophylaxis in Ambulatory Patients vein thrombosis. J Vasc Surg. May 2012;55(5):1463-73. doi:10.1016/j.jvs.2011.12.082 with Cancer. New England Journal of Medicine. 2019/02/21 2019;380(8):781-783. doi:10.1056/NEJMe1816060 84. Wong PC, Chan YC, Law Y, Cheng SWK. Percutaneous mechanical thrombectomy in the treatment of acute iliofemoral deep vein thrombosis: a systematic review. Hong 72. Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in nonsurgical patients: An- Kong Med J. Feb 2019;25(1):48-57. doi:10.12809/hkmj187491 tithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of 85. Weitz JI. Low-molecular-weight heparins. N Engl J Med. Sep 4 1997;337(10):688-98. Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. Feb 2012;141(2 doi:10.1056/nejm199709043371007 Suppl):e195S-e226S. doi:10.1378/chest.11-2296 86. De Caterina R, Husted S, Wallentin L, et al. Vitamin K antagonists in heart disease: 73. Chaochankit W, Akaraborworn O. Phlegmasia Cerulea Dolens with Compartment Syn- current status and perspectives (Section III). Position paper of the ESC Working Group drome. Ann Vasc Dis. Sep 25 2018;11(3):355-357. doi:10.3400/avd.cr.18-00030 on Thrombosis--Task Force on Anticoagulants in Heart Disease. Thromb Haemost. Dec 2013;110(6):1087-107. doi:10.1160/th13-06-0443 74. Mumoli N, Invernizzi C, Luschi R, Carmignani G, Camaiti A, Cei M. Phlegmasia Ce- rulea Dolens. Circulation. 2012/02/28 2012;125(8):1056-1057. doi:10.1161/CIRCU- 87. Boutitie F, Pinede L, Schulman S, et al. In uence of preceding length of anticoagulant LATIONAHA.111.051912 treatment and initial presentation of venous thromboembolism on risk of recurrence af- ter stopping treatment: analysis of individual participants’ data from seven trials. BMJ. 75. Mazzolai L, Aboyans V, Ageno W, et al. Diagnosis and management of acute deep 2011;342:d3036. doi:10.1136/bmj.d3036 vein thrombosis: a joint consensus document from the European Society of Cardiology working groups of aorta and peripheral vascular diseases and pulmonary circulation and 88. Kearon C, Gent M, Hirsh J, et al. A comparison of three months of anticoagulation with right ventricular function. Eur Heart J. Dec 14 2018;39(47):4208-4218. doi:10.1093/ extended anticoagulation for a  rst episode of idiopathic venous thromboembolism. N eurheartj/ehx003 Engl J Med. Mar 25 1999;340(12):901-7. doi:10.1056/nejm199903253401201 76. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bed- 89. Agnelli G, Prandoni P, Becattini C, et al. Extended Oral Anticoagulant Therapy after side without diagnostic imaging: management of patients with suspected pulmonary a First Episode of Pulmonary Embolism. Ann Intern Med. 2003;139:19-25. doi:doi: embolism presenting to the emergency department by using a simple clinical model and 10.7326/0003-4819-139-1-200307010-00008 d-dimer. Ann Intern Med. Jul 17 2001;135(2):98-107. doi:10.7326/0003-4819-135-2- 200107170-00010 90. Khan F, Rahman A, Carrier M, et al. Long term risk of symptomatic recurrent venous thromboembolism after discontinuation of anticoagulant treatment for  rst unprovoked 77. Ultrasound Guidelines: Emergency, Point-of-Care and Clinical Ultrasound Guide- venous thromboembolism event: systematic review and meta-analysis. Bmj. Jul 24 lines in Medicine. Annals of Emergency Medicine. 2017;69(5):e27-e54. doi:10.1016/j. 2019;366:l4363. doi:10.1136/bmj.l4363 annemergmed.2016.08.457 91. Iorio A, Kearon C, Filippucci E, et al. Risk of recurrence after a  rst episode of symptom- 78. Marti C, John G, Konstantinides S, et al. Systemic thrombolytic therapy for acute atic venous thromboembolism provoked by a transient risk factor: a systematic review. pulmonary embolism: a systematic review and meta-analysis. Eur Heart J. Mar 7 Arch Intern Med. Oct 25 2010;170(19):1710-6. doi:10.1001/archinternmed.2010.367 2015;36(10):605-14. doi:10.1093/eurheartj/ehu218 92. Khan F, Tritschler T, Kimpton M, et al. Long-term risk of recurrent venous thrombo- 79. Meyer G, Vicaut E, Danays T, et al. Fibrinolysis for patients with intermediate-risk embolism among patients receiving extended oral anticoagulant therapy for  rst un- pulmonary embolism. N Engl J Med. Apr 10 2014;370(15):1402-11. doi:10.1056/NEJ- provoked venous thromboembolism: A systematic review and meta-analysis. J Thromb Moa1302097 Haemost. Aug 11 2021;doi:10.1111/jth.15491 54 55 67. Mottier D, Girard P, Couturaud F, et al. Enoxaparin versus Placebo to Prevent Symptom- 80. Vedantham S, Goldhaber SZ, Julian JA, et al. Pharmacomechanical Catheter-Directed atic Venous Thromboembolism in Hospitalized Older Adult Medical Patients. NEJM Thrombolysis for Deep-Vein Thrombosis. N Engl J Med. Dec 7 2017;377(23):2240- Evidence. 2023/07/25 2023;2(8):EVIDoa2200332. doi:10.1056/EVIDoa2200332 2252. doi:10.1056/NEJMoa1615066 68. Horner D, Goodacre S, Davis S, Burton N, Hunt BJ. Which is the best model to assess risk 81. Pruszczyk P, Klok FA, Kucher N, et al. Percutaneous treatment options for acute pul- for venous thromboembolism in hospitalised patients? Bmj. May 27 2021;373:n1106. monary embolism: a clinical consensus statement by the ESC Working Group on Pul- doi:10.1136/bmj.n1106 monary Circulation and Right Ventricular Function and the European Association of Percutaneous Cardiovascular Interventions. EuroIntervention. Oct 7 2022;18(8):e623- 69. Carrier M, Abou-Nassar K, Mallick R, et al. Apixaban to Prevent Venous Thrombo- e638. doi:10.4244/eij-d-22-00246 embolism in Patients with Cancer. New England Journal of Medicine. 2019/02/21 2018;380(8):711-719. doi:10.1056/NEJMoa1814468 82. Goldberg JB, Giri J, Kobayashi T, et al. Surgical Management and Mechanical Circula- tory Support in High-Risk Pulmonary Embolisms: Historical Context, Current Status, 70. Khorana AA, Soff GA, Kakkar AK, et al. Rivaroxaban for Thromboprophylaxis and Future Directions: A Scienti c Statement From the American Heart Association. in High-Risk Ambulatory Patients with Cancer. New England Journal of Medicine. Circulation. Feb 28 2023;147(9):e628-e647. doi:10.1161/cir.0000000000001117 2019/02/21 2019;380(8):720-728. doi:10.1056/NEJMoa1814630 83. Casey ET, Murad MH, Zumaeta-Garcia M, et al. Treatment of acute iliofemoral deep 71. Agnelli G. Direct Oral Anticoagulants for Thromboprophylaxis in Ambulatory Patients vein thrombosis. J Vasc Surg. May 2012;55(5):1463-73. doi:10.1016/j.jvs.2011.12.082 with Cancer. New England Journal of Medicine. 2019/02/21 2019;380(8):781-783. doi:10.1056/NEJMe1816060 84. Wong PC, Chan YC, Law Y, Cheng SWK. Percutaneous mechanical thrombectomy in the treatment of acute iliofemoral deep vein thrombosis: a systematic review. Hong 72. Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in nonsurgical patients: An- Kong Med J. Feb 2019;25(1):48-57. doi:10.12809/hkmj187491 tithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of 85. Weitz JI. Low-molecular-weight heparins. N Engl J Med. Sep 4 1997;337(10):688-98. Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. Feb 2012;141(2 doi:10.1056/nejm199709043371007 Suppl):e195S-e226S. doi:10.1378/chest.11-2296 86. De Caterina R, Husted S, Wallentin L, et al. Vitamin K antagonists in heart disease: 73. Chaochankit W, Akaraborworn O. Phlegmasia Cerulea Dolens with Compartment Syn- current status and perspectives (Section III). Position paper of the ESC Working Group drome. Ann Vasc Dis. Sep 25 2018;11(3):355-357. doi:10.3400/avd.cr.18-00030 on Thrombosis--Task Force on Anticoagulants in Heart Disease. Thromb Haemost. Dec 2013;110(6):1087-107. doi:10.1160/th13-06-0443 74. Mumoli N, Invernizzi C, Luschi R, Carmignani G, Camaiti A, Cei M. Phlegmasia Ce- rulea Dolens. Circulation. 2012/02/28 2012;125(8):1056-1057. doi:10.1161/CIRCU- 87. Boutitie F, Pinede L, Schulman S, et al. In uence of preceding length of anticoagulant LATIONAHA.111.051912 treatment and initial presentation of venous thromboembolism on risk of recurrence af- ter stopping treatment: analysis of individual participants’ data from seven trials. BMJ. 75. Mazzolai L, Aboyans V, Ageno W, et al. Diagnosis and management of acute deep 2011;342:d3036. doi:10.1136/bmj.d3036 vein thrombosis: a joint consensus document from the European Society of Cardiology working groups of aorta and peripheral vascular diseases and pulmonary circulation and 88. Kearon C, Gent M, Hirsh J, et al. A comparison of three months of anticoagulation with right ventricular function. Eur Heart J. Dec 14 2018;39(47):4208-4218. doi:10.1093/ extended anticoagulation for a  rst episode of idiopathic venous thromboembolism. N eurheartj/ehx003 Engl J Med. Mar 25 1999;340(12):901-7. doi:10.1056/nejm199903253401201 76. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bed- 89. Agnelli G, Prandoni P, Becattini C, et al. Extended Oral Anticoagulant Therapy after side without diagnostic imaging: management of patients with suspected pulmonary a First Episode of Pulmonary Embolism. Ann Intern Med. 2003;139:19-25. doi:doi: embolism presenting to the emergency department by using a simple clinical model and 10.7326/0003-4819-139-1-200307010-00008 d-dimer. Ann Intern Med. Jul 17 2001;135(2):98-107. doi:10.7326/0003-4819-135-2- 200107170-00010 90. Khan F, Rahman A, Carrier M, et al. Long term risk of symptomatic recurrent venous thromboembolism after discontinuation of anticoagulant treatment for  rst unprovoked 77. Ultrasound Guidelines: Emergency, Point-of-Care and Clinical Ultrasound Guide- venous thromboembolism event: systematic review and meta-analysis. Bmj. Jul 24 lines in Medicine. Annals of Emergency Medicine. 2017;69(5):e27-e54. doi:10.1016/j. 2019;366:l4363. doi:10.1136/bmj.l4363 annemergmed.2016.08.457 91. Iorio A, Kearon C, Filippucci E, et al. Risk of recurrence after a  rst episode of symptom- 78. Marti C, John G, Konstantinides S, et al. Systemic thrombolytic therapy for acute atic venous thromboembolism provoked by a transient risk factor: a systematic review. pulmonary embolism: a systematic review and meta-analysis. Eur Heart J. Mar 7 Arch Intern Med. Oct 25 2010;170(19):1710-6. doi:10.1001/archinternmed.2010.367 2015;36(10):605-14. doi:10.1093/eurheartj/ehu218 92. Khan F, Tritschler T, Kimpton M, et al. Long-term risk of recurrent venous thrombo- 79. Meyer G, Vicaut E, Danays T, et al. Fibrinolysis for patients with intermediate-risk embolism among patients receiving extended oral anticoagulant therapy for  rst un- pulmonary embolism. N Engl J Med. Apr 10 2014;370(15):1402-11. doi:10.1056/NEJ- provoked venous thromboembolism: A systematic review and meta-analysis. J Thromb Moa1302097 Haemost. Aug 11 2021;doi:10.1111/jth.15491 54 55 93. Khan F, Tritschler T, Kimpton M, et al. Long-Term Risk for Major Bleeding During Ex- Heart and Lung Transplantation (ISHLT) and the European Reference Network on rare tended Oral Anticoagulant Therapy for First Unprovoked Venous Thromboembolism : respiratory diseases (ERN-LUNG). European Heart Journal. 2022;43(38):3618-3731. A Systematic Review and Meta-analysis. Ann Intern Med. Sep 14 2021;doi:10.7326/ doi:10.1093/eurheartj/ehac237 m21-1094 105. Baldwin MJ, Moore HM, Rudarakanchana N, Gohel M, Davies AH. Post-thrombotic 94. Schulman S, Kearon C. De nition of major bleeding in clinical investigations of an- syndrome: a clinical review. J Thromb Haemost. May 2013;11(5):795-805. doi:10.1111/ tihemostatic medicinal products in non-surgical patients. J Thromb Haemost. Apr jth.12180 2005;3(4):692-4. doi:10.1111/j.1538-7836.2005.01204.x 106. Kahn SR, Shapiro S, Wells PS, et al. Compression stockings to prevent post-thrombotic 95. van der Hulle T, Kooiman J, den Exter PL, Dekkers OM, Klok FA, Huisman MV. Ef- syndrome: a randomised placebo-controlled trial. Lancet. Mar 8 2014;383(9920):880- fectiveness and safety of novel oral anticoagulants as compared with vitamin K antago- 8. doi:10.1016/s0140-6736(13)61902-9 nists in the treatment of acute symptomatic venous thromboembolism: a systematic re- view and meta-analysis. J Thromb Haemost. 2014;12(3):320-8. doi:10.1111/jth.12485 107. Ludvigsson JF, Berglind D, Sundquist K, Sundström J, Tynelius P, Neovius M. The Swedish military conscription register: opportunities for its use in medical research. 96. Larsen TB, Skjøth F, Kjældgaard JN, Lip GYH, Nielsen PB, Søgaard M. Effectiveness European Journal of Epidemiology. 2022/07/01 2022;37(7):767-777. doi:10.1007/ and safety of rivaroxaban and warfarin in patients with unprovoked venous throm- s10654-022-00887-0 boembolism: a propensity-matched nationwide cohort study. Lancet Haematol. May 2017;4(5):e237-e244. doi:10.1016/s2352-3026(17)30054-6 108. Bohman T, Tegern M, Halvarsson A, Broman L, Larsson H. Concurrent validity of an isokinetic lift test used for admission to the Swedish Armed Forces. PLOS ONE. 97. Bertoletti L, Gusto G, Khachatryan A, et al. Effectiveness and Safety of Oral Antico- 2018;13(11):e0207054. doi:10.1371/journal.pone.0207054 agulants in the Treatment of Acute Venous Thromboembolism: A Nationwide Com- parative Cohort Study in France. Thromb Haemost. Aug 2022;122(8):1384-1396. 109. Ludvigsson JF, Svedberg P, Olén O, Bruze G, Neovius M. The longitudinal integrated doi:10.1055/a-1731-3922 database for health insurance and labour market studies (LISA) and its use in medi- cal research. European Journal of Epidemiology. 2019/04/01 2019;34(4):423-437. 98. Weycker D, Li X, Wygant GD, et al. Effectiveness and Safety of Apixaban versus War- doi:10.1007/s10654-019-00511-8 farin as Outpatient Treatment of Venous Thromboembolism in U.S. Clinical Practice. Thromb Haemost. Nov 2018;118(11):1951-1961. doi:10.1055/s-0038-1673689 110. Patientregistret Tf. Det statistiska registrets framställning och kvalitet Patientregis- tret. 2 ed. https://www.socialstyrelsen.se/statistik-och-data/register/patientregistret/ 99. Pawar A, Gagne JJ, Gopalakrishnan C, et al. Association of Type of Oral Anticoagu- framstallning-och-kvalitet/2022. lant Dispensed With Adverse Clinical Outcomes in Patients Extending Anticoagulation Therapy Beyond 90 Days After Hospitalization for Venous Thromboembolism. Jama. 111. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swed- Mar 15 2022;327(11):1051-1060. doi:10.1001/jama.2022.1920 ish national inpatient register. BMC Public Health. Jun 2011;11:450. doi:10.1186/1471- 2458-11-450 100. Barco S, Mahmoudpour SH, Valerio L, et al. Trends in mortality related to pulmonary embolism in the European Region, 2000-15: analysis of vital registration data from the 112. Brooke HL, Talbäck M, Hörnblad J, et al. The Swedish cause of death register. Eur J WHO Mortality Database. Lancet Respir Med. Mar 2020;8(3):277-287. doi:10.1016/ Epidemiol. 09 2017;32(9):765-773. doi:10.1007/s10654-017-0316-1 s2213-2600(19)30354-6 113. Wettermark B, Hammar N, Fored CM, et al. The new Swedish Prescribed Drug Reg- 101. Barco S, Valerio L, Ageno W, et al. Age-sex speci c pulmonary embolism-related mor- ister--opportunities for pharmacoepidemiological research and experience from the tality in the USA and Canada, 2000-18: an analysis of the WHO Mortality Database and  rst six months. Pharmacoepidemiol Drug Saf. Jul 2007;16(7):726-35. doi:10.1002/ of the CDC Multiple Cause of Death database. Lancet Respir Med. Jan 2021;9(1):33- pds.1294 42. doi:10.1016/s2213-2600(20)30417-3 114. Ludvigsson JF, Almqvist C, Bonamy AK, et al. Registers of the Swedish total pop- 102. Klok FA, van der Hulle T, den Exter PL, Lankeit M, Huisman MV, Konstantinides S. ulation and their use in medical research. Eur J Epidemiol. Feb 2016;31(2):125-36. The post-PE syndrome: a new concept for chronic complications of pulmonary embo- doi:10.1007/s10654-016-0117-y lism. Blood reviews. 2014;28(6):221-226. doi:10.1016/j.blre.2014.07.003 115. Glise Sandblad K, Jern S, Åberg M, et al. Obesity in adolescent men increases the 103. Tavoly M, Wik HS, Sirnes PA, et al. The impact of post-pulmonary embolism syn- risk of venous thromboembolism in adult life. J Intern Med. Jun 2020;287(6):734-745. drome and its possible determinants. Thromb Res. Nov 2018;171:84-91. doi:10.1016/j. doi:10.1111/joim.13044 thromres.2018.09.048 116. Glise Sandblad K, Schulman S, Rosengren A, Sörbo J, Philipson J, Hansson PO. As- 104. Humbert M, Kovacs G, Hoeper MM, et al. 2022 ESC/ERS Guidelines for the diagnosis sociation of type of oral anticoagulation with risk of bleeding in 45,114 patients with and treatment of pulmonary hypertension: Developed by the task force for the diagnosis venous thromboembolism during initial and extended treatment-A nationwide register- and treatment of pulmonary hypertension of the European Society of Cardiology (ESC) based study. J Intern Med. Aug 28 2023;doi:10.1111/joim.13712 and the European Respiratory Society (ERS). Endorsed by the International Society for 56 57 93. Khan F, Tritschler T, Kimpton M, et al. Long-Term Risk for Major Bleeding During Ex- Heart and Lung Transplantation (ISHLT) and the European Reference Network on rare tended Oral Anticoagulant Therapy for First Unprovoked Venous Thromboembolism : respiratory diseases (ERN-LUNG). European Heart Journal. 2022;43(38):3618-3731. A Systematic Review and Meta-analysis. Ann Intern Med. Sep 14 2021;doi:10.7326/ doi:10.1093/eurheartj/ehac237 m21-1094 105. Baldwin MJ, Moore HM, Rudarakanchana N, Gohel M, Davies AH. Post-thrombotic 94. Schulman S, Kearon C. De nition of major bleeding in clinical investigations of an- syndrome: a clinical review. J Thromb Haemost. May 2013;11(5):795-805. doi:10.1111/ tihemostatic medicinal products in non-surgical patients. J Thromb Haemost. Apr jth.12180 2005;3(4):692-4. doi:10.1111/j.1538-7836.2005.01204.x 106. Kahn SR, Shapiro S, Wells PS, et al. Compression stockings to prevent post-thrombotic 95. van der Hulle T, Kooiman J, den Exter PL, Dekkers OM, Klok FA, Huisman MV. Ef- syndrome: a randomised placebo-controlled trial. Lancet. Mar 8 2014;383(9920):880- fectiveness and safety of novel oral anticoagulants as compared with vitamin K antago- 8. doi:10.1016/s0140-6736(13)61902-9 nists in the treatment of acute symptomatic venous thromboembolism: a systematic re- view and meta-analysis. J Thromb Haemost. 2014;12(3):320-8. doi:10.1111/jth.12485 107. Ludvigsson JF, Berglind D, Sundquist K, Sundström J, Tynelius P, Neovius M. The Swedish military conscription register: opportunities for its use in medical research. 96. Larsen TB, Skjøth F, Kjældgaard JN, Lip GYH, Nielsen PB, Søgaard M. Effectiveness European Journal of Epidemiology. 2022/07/01 2022;37(7):767-777. doi:10.1007/ and safety of rivaroxaban and warfarin in patients with unprovoked venous throm- s10654-022-00887-0 boembolism: a propensity-matched nationwide cohort study. Lancet Haematol. May 2017;4(5):e237-e244. doi:10.1016/s2352-3026(17)30054-6 108. Bohman T, Tegern M, Halvarsson A, Broman L, Larsson H. Concurrent validity of an isokinetic lift test used for admission to the Swedish Armed Forces. PLOS ONE. 97. Bertoletti L, Gusto G, Khachatryan A, et al. Effectiveness and Safety of Oral Antico- 2018;13(11):e0207054. doi:10.1371/journal.pone.0207054 agulants in the Treatment of Acute Venous Thromboembolism: A Nationwide Com- parative Cohort Study in France. Thromb Haemost. Aug 2022;122(8):1384-1396. 109. Ludvigsson JF, Svedberg P, Olén O, Bruze G, Neovius M. The longitudinal integrated doi:10.1055/a-1731-3922 database for health insurance and labour market studies (LISA) and its use in medi- cal research. European Journal of Epidemiology. 2019/04/01 2019;34(4):423-437. 98. Weycker D, Li X, Wygant GD, et al. Effectiveness and Safety of Apixaban versus War- doi:10.1007/s10654-019-00511-8 farin as Outpatient Treatment of Venous Thromboembolism in U.S. Clinical Practice. Thromb Haemost. Nov 2018;118(11):1951-1961. doi:10.1055/s-0038-1673689 110. Patientregistret Tf. Det statistiska registrets framställning och kvalitet Patientregis- tret. 2 ed. https://www.socialstyrelsen.se/statistik-och-data/register/patientregistret/ 99. Pawar A, Gagne JJ, Gopalakrishnan C, et al. Association of Type of Oral Anticoagu- framstallning-och-kvalitet/2022. lant Dispensed With Adverse Clinical Outcomes in Patients Extending Anticoagulation Therapy Beyond 90 Days After Hospitalization for Venous Thromboembolism. Jama. 111. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swed- Mar 15 2022;327(11):1051-1060. doi:10.1001/jama.2022.1920 ish national inpatient register. BMC Public Health. Jun 2011;11:450. doi:10.1186/1471- 2458-11-450 100. Barco S, Mahmoudpour SH, Valerio L, et al. Trends in mortality related to pulmonary embolism in the European Region, 2000-15: analysis of vital registration data from the 112. Brooke HL, Talbäck M, Hörnblad J, et al. The Swedish cause of death register. Eur J WHO Mortality Database. Lancet Respir Med. Mar 2020;8(3):277-287. doi:10.1016/ Epidemiol. 09 2017;32(9):765-773. doi:10.1007/s10654-017-0316-1 s2213-2600(19)30354-6 113. Wettermark B, Hammar N, Fored CM, et al. The new Swedish Prescribed Drug Reg- 101. Barco S, Valerio L, Ageno W, et al. Age-sex speci c pulmonary embolism-related mor- ister--opportunities for pharmacoepidemiological research and experience from the tality in the USA and Canada, 2000-18: an analysis of the WHO Mortality Database and  rst six months. Pharmacoepidemiol Drug Saf. Jul 2007;16(7):726-35. doi:10.1002/ of the CDC Multiple Cause of Death database. Lancet Respir Med. Jan 2021;9(1):33- pds.1294 42. doi:10.1016/s2213-2600(20)30417-3 114. Ludvigsson JF, Almqvist C, Bonamy AK, et al. Registers of the Swedish total pop- 102. Klok FA, van der Hulle T, den Exter PL, Lankeit M, Huisman MV, Konstantinides S. ulation and their use in medical research. Eur J Epidemiol. Feb 2016;31(2):125-36. The post-PE syndrome: a new concept for chronic complications of pulmonary embo- doi:10.1007/s10654-016-0117-y lism. Blood reviews. 2014;28(6):221-226. doi:10.1016/j.blre.2014.07.003 115. Glise Sandblad K, Jern S, Åberg M, et al. Obesity in adolescent men increases the 103. Tavoly M, Wik HS, Sirnes PA, et al. The impact of post-pulmonary embolism syn- risk of venous thromboembolism in adult life. J Intern Med. Jun 2020;287(6):734-745. drome and its possible determinants. Thromb Res. Nov 2018;171:84-91. doi:10.1016/j. doi:10.1111/joim.13044 thromres.2018.09.048 116. Glise Sandblad K, Schulman S, Rosengren A, Sörbo J, Philipson J, Hansson PO. As- 104. Humbert M, Kovacs G, Hoeper MM, et al. 2022 ESC/ERS Guidelines for the diagnosis sociation of type of oral anticoagulation with risk of bleeding in 45,114 patients with and treatment of pulmonary hypertension: Developed by the task force for the diagnosis venous thromboembolism during initial and extended treatment-A nationwide register- and treatment of pulmonary hypertension of the European Society of Cardiology (ESC) based study. J Intern Med. Aug 28 2023;doi:10.1111/joim.13712 and the European Respiratory Society (ERS). Endorsed by the International Society for 56 57 117. Bruzzi P, Green SB, Byar DP, Brinton LA, Schairer C. Estimating the population at- 129. Comparison of Bleeding Risk Between Rivaroxaban and Apixaban for the Treat- tributable risk for multiple risk factors using case-control data. Am J Epidemiol. Nov ment of Acute Venous Thromboembolism (COBRRA) ClinicalTrials.gov Identi er: 1985;122(5):904-14. doi:10.1093/oxfordjournals.aje.a114174 NCT03266783. Accessed December 6, 2022. https://clinicaltrials.gov/ct2/show/study/ NCT03266783 118. Glise Sandblad K, Rosengren A, Sörbo J, Jern S, Hansson P-O. Pulmonary embo- lism and deep vein thrombosis—comorbidities and temporary provoking factors in a 130. Simmonds M, Llewellyn A, Owen CG, Woolacott N. Predicting adult obesity from child- register-based study of 1.48 million people. https://doi.org/10.1002/rth2.12714. Re- hood obesity: a systematic review and meta-analysis. Obes Rev. Feb 2016;17(2):95- search and Practice in Thrombosis and Haemostasis. 2022/05/01 2022;6(4):e12714. 107. doi:10.1111/obr.12334 doi:https://doi.org/10.1002/rth2.12714 131. Pineda LA, Hathwar VS, Grant BJ. Clinical suspicion of fatal pulmonary embolism. 119. Khorana AA, Francis CW, Culakova E, Fisher RI, Kuderer NM, Lyman GH. Throm- Chest. Sep 2001;120(3):791-5. doi:10.1378/chest.120.3.791 boembolism in hospitalized neutropenic cancer patients. J Clin Oncol. Jan 20 2006;24(3):484-90. doi:10.1200/jco.2005.03.8877 132. Ossei PPS, Owusu IK, Owusu-Asubonteng G, Ankobea-Kokroe F, Ayibor WG, Nia- ko N. Prevalence of Venous Thromboembolism in Kumasi: A Postmortem-Based 120. Khorana AA, Dalal M, Lin J, Connolly GC. Incidence and predictors of venous throm- Study in a Tertiary Hospital in Ghana. Clin Med Insights Circ Respir Pulm Med. boembolism (VTE) among ambulatory high-risk cancer patients undergoing chemo- 2020;14:1179548420956364. doi:10.1177/1179548420956364 therapy in the United States. https://doi.org/10.1002/cncr.27772. Cancer. 2013/02/01 2013;119(3):648-655. doi:https://doi.org/10.1002/cncr.27772 133. Haug CJ, Drazen JM. Arti cial Intelligence and Machine Learning in Clinical Medi- cine, 2023. New England Journal of Medicine. 2023/03/30 2023;388(13):1201-1208. 121. Lyman GH, Eckert L, Wang Y, Wang H, Cohen A. Venous Thromboembolism Risk in doi:10.1056/NEJMra2302038 Patients With Cancer Receiving Chemotherapy: A Real-World Analysis. The Oncolo- gist. 2013;18(12):1321-1329. doi:10.1634/theoncologist.2013-0226 134. Nafee T, Gibson CM, Travis R, et al. Machine learning to predict venous thrombosis in acutely ill medical patients. Res Pract Thromb Haemost. Feb 2020;4(2):230-237. 122. Hansson PO, Eriksson H, Welin L, Svärdsudd K, Wilhelmsen L. Smoking and abdomi- doi:10.1002/rth2.12292 nal obesity: risk factors for venous thromboembolism among middle-aged men: “the study of men born in 1913”. Arch Intern Med. Sep 1999;159(16):1886-90. 135. Pavani A, Naushad SM, Kumar RM, Srinath M, Malempati AR, Kutala VK. Arti - cial neural network-based pharmacogenomic algorithm for warfarin dose optimization. 123. Severinsen MT, Kristensen SR, Johnsen SP, Dethlefsen C, Tjønneland A, Overvad K. Pharmacogenomics. 2016;17(2):121-31. doi:10.2217/pgs.15.161 Anthropometry, body fat, and venous thromboembolism: a Danish follow-up study. Cir- culation. Nov 2009;120(19):1850-7. doi:10.1161/CIRCULATIONAHA.109.863241 136. Hsu C, Hutt E, Bloom eld DM, Gailani D, Weitz JI. Factor XI Inhibition to Uncouple Thrombosis From Hemostasis: JACC Review Topic of the Week. J Am Coll Cardiol. 124. Hoyo C, Cook MB, Kamangar F, et al. Body mass index in relation to oesophageal Aug 10 2021;78(6):625-631. doi:10.1016/j.jacc.2021.06.010 and oesophagogastric junction adenocarcinomas: a pooled analysis from the Interna- tional BEACON Consortium. Int J Epidemiol. Dec 2012;41(6):1706-18. doi:10.1093/ 137. DeLoughery EP, Olson SR, Puy C, McCarty OJT, Shatzel JJ. The Safety and Ef cacy ije/dys176 of Novel Agents Targeting Factors XI and XII in Early Phase Human Trials. Semin Thromb Hemost. Jul 2019;45(5):502-508. doi:10.1055/s-0039-1692439 125. Wang F, Xu Y. Body mass index and risk of renal cell cancer: a dose-response me- ta-analysis of published cohort studies. Int J Cancer. Oct 1 2014;135(7):1673-86. doi:10.1002/ijc.28813 126. Öhman L, Johansson M, Jansson JH, Lind M, Johansson L. Positive predictive value and misclassi cation of diagnosis of pulmonary embolism and deep vein thrombosis in Swedish patient registries. Clin Epidemiol. 2018;10:1215-1221. doi:10.2147/CLEP. S177058 127. Abdul Sultan A, West J, Stephansson O, et al. De ning venous thromboembolism and measuring its incidence using Swedish health registries: a nationwide pregnancy cohort study. BMJ Open. Nov 2015;5(11):e008864. doi:10.1136/bmjopen-2015-008864 128. Chew HK, Wun T, Harvey D, Zhou H, White RH. Incidence of venous thromboembo- lism and its effect on survival among patients with common cancers. Arch Intern Med. Feb 2006;166(4):458-64. doi:10.1001/archinte.166.4.458 58 59 117. Bruzzi P, Green SB, Byar DP, Brinton LA, Schairer C. Estimating the population at- 129. Comparison of Bleeding Risk Between Rivaroxaban and Apixaban for the Treat- tributable risk for multiple risk factors using case-control data. Am J Epidemiol. Nov ment of Acute Venous Thromboembolism (COBRRA) ClinicalTrials.gov Identi er: 1985;122(5):904-14. doi:10.1093/oxfordjournals.aje.a114174 NCT03266783. Accessed December 6, 2022. https://clinicaltrials.gov/ct2/show/study/ NCT03266783 118. Glise Sandblad K, Rosengren A, Sörbo J, Jern S, Hansson P-O. Pulmonary embo- lism and deep vein thrombosis—comorbidities and temporary provoking factors in a 130. Simmonds M, Llewellyn A, Owen CG, Woolacott N. Predicting adult obesity from child- register-based study of 1.48 million people. https://doi.org/10.1002/rth2.12714. Re- hood obesity: a systematic review and meta-analysis. Obes Rev. Feb 2016;17(2):95- search and Practice in Thrombosis and Haemostasis. 2022/05/01 2022;6(4):e12714. 107. doi:10.1111/obr.12334 doi:https://doi.org/10.1002/rth2.12714 131. Pineda LA, Hathwar VS, Grant BJ. Clinical suspicion of fatal pulmonary embolism. 119. Khorana AA, Francis CW, Culakova E, Fisher RI, Kuderer NM, Lyman GH. Throm- Chest. Sep 2001;120(3):791-5. doi:10.1378/chest.120.3.791 boembolism in hospitalized neutropenic cancer patients. J Clin Oncol. Jan 20 2006;24(3):484-90. doi:10.1200/jco.2005.03.8877 132. Ossei PPS, Owusu IK, Owusu-Asubonteng G, Ankobea-Kokroe F, Ayibor WG, Nia- ko N. Prevalence of Venous Thromboembolism in Kumasi: A Postmortem-Based 120. Khorana AA, Dalal M, Lin J, Connolly GC. Incidence and predictors of venous throm- Study in a Tertiary Hospital in Ghana. Clin Med Insights Circ Respir Pulm Med. boembolism (VTE) among ambulatory high-risk cancer patients undergoing chemo- 2020;14:1179548420956364. doi:10.1177/1179548420956364 therapy in the United States. https://doi.org/10.1002/cncr.27772. Cancer. 2013/02/01 2013;119(3):648-655. doi:https://doi.org/10.1002/cncr.27772 133. Haug CJ, Drazen JM. Arti cial Intelligence and Machine Learning in Clinical Medi- cine, 2023. New England Journal of Medicine. 2023/03/30 2023;388(13):1201-1208. 121. Lyman GH, Eckert L, Wang Y, Wang H, Cohen A. Venous Thromboembolism Risk in doi:10.1056/NEJMra2302038 Patients With Cancer Receiving Chemotherapy: A Real-World Analysis. The Oncolo- gist. 2013;18(12):1321-1329. doi:10.1634/theoncologist.2013-0226 134. Nafee T, Gibson CM, Travis R, et al. Machine learning to predict venous thrombosis in acutely ill medical patients. Res Pract Thromb Haemost. Feb 2020;4(2):230-237. 122. Hansson PO, Eriksson H, Welin L, Svärdsudd K, Wilhelmsen L. Smoking and abdomi- doi:10.1002/rth2.12292 nal obesity: risk factors for venous thromboembolism among middle-aged men: “the study of men born in 1913”. Arch Intern Med. Sep 1999;159(16):1886-90. 135. Pavani A, Naushad SM, Kumar RM, Srinath M, Malempati AR, Kutala VK. Arti - cial neural network-based pharmacogenomic algorithm for warfarin dose optimization. 123. Severinsen MT, Kristensen SR, Johnsen SP, Dethlefsen C, Tjønneland A, Overvad K. Pharmacogenomics. 2016;17(2):121-31. doi:10.2217/pgs.15.161 Anthropometry, body fat, and venous thromboembolism: a Danish follow-up study. Cir- culation. Nov 2009;120(19):1850-7. doi:10.1161/CIRCULATIONAHA.109.863241 136. Hsu C, Hutt E, Bloom eld DM, Gailani D, Weitz JI. Factor XI Inhibition to Uncouple Thrombosis From Hemostasis: JACC Review Topic of the Week. J Am Coll Cardiol. 124. Hoyo C, Cook MB, Kamangar F, et al. Body mass index in relation to oesophageal Aug 10 2021;78(6):625-631. doi:10.1016/j.jacc.2021.06.010 and oesophagogastric junction adenocarcinomas: a pooled analysis from the Interna- tional BEACON Consortium. Int J Epidemiol. Dec 2012;41(6):1706-18. doi:10.1093/ 137. DeLoughery EP, Olson SR, Puy C, McCarty OJT, Shatzel JJ. The Safety and Ef cacy ije/dys176 of Novel Agents Targeting Factors XI and XII in Early Phase Human Trials. Semin Thromb Hemost. Jul 2019;45(5):502-508. doi:10.1055/s-0039-1692439 125. Wang F, Xu Y. Body mass index and risk of renal cell cancer: a dose-response me- ta-analysis of published cohort studies. Int J Cancer. Oct 1 2014;135(7):1673-86. doi:10.1002/ijc.28813 126. Öhman L, Johansson M, Jansson JH, Lind M, Johansson L. Positive predictive value and misclassi cation of diagnosis of pulmonary embolism and deep vein thrombosis in Swedish patient registries. Clin Epidemiol. 2018;10:1215-1221. doi:10.2147/CLEP. S177058 127. Abdul Sultan A, West J, Stephansson O, et al. De ning venous thromboembolism and measuring its incidence using Swedish health registries: a nationwide pregnancy cohort study. BMJ Open. Nov 2015;5(11):e008864. doi:10.1136/bmjopen-2015-008864 128. Chew HK, Wun T, Harvey D, Zhou H, White RH. Incidence of venous thromboembo- lism and its effect on survival among patients with common cancers. Arch Intern Med. Feb 2006;166(4):458-64. doi:10.1001/archinte.166.4.458 58 59