Advertisement

Clinical Benefit, Harm, and Cost Effectiveness of Screening Men for Peripheral Artery Disease: A Markov Model Based on the VIVA Trial

  • Jes S. Lindholt
    Correspondence
    Corresponding author. Department of Cardiothoracic and Vascular Surgery, Odense University Hospital, Sdr. Boulevard 29, 5000, Odense C, Denmark.
    Affiliations
    Department of Cardiothoracic and Vascular Surgery, Odense University Hospital, Odense, Denmark

    Department of Vascular Surgery, Viborg Hospital, Viborg, Denmark
    Search for articles by this author
  • Rikke Søgaard
    Affiliations
    Department of Clinical Medicine, University of Southern Denmark, Odense, Denmark

    Department of Public Health, Aarhus University, Aarhus, Denmark
    Search for articles by this author
Open ArchivePublished:April 09, 2021DOI:https://doi.org/10.1016/j.ejvs.2021.02.039

      Objective

      Although screening for peripheral arterial disease (PAD) seems obvious due to its two to three times increased mortality, high prevalence in the elderly, ease of detection, and relatively harmless prevention, the evidence is sparse.

      Methods

      A Markov decision model was created to model the lifetime effectiveness and cost effectiveness of general population PAD screening and relevant intervention in 65 year old men. The model was informed by original estimates from the VIVA trial data except for ankle brachial systolic blood pressure index test accuracy, quality of life, and background mortality, which were adopted from the literature. A Markov model was designed for 65 year old men, who were distributed in the starting states of no/detected/undetected PAD. The main outcomes were life years, quality adjusted life years, and costs of healthcare.

      Results

      Screening for PAD reduced the rates of amputations and stroke by 10.9% and 2.4%, respectively, while it increased the rates of revascularisation, acute myocardial infarction, and major bleeding by 5.5%, 7.1%, and 4.3% respectively. The overall life expectancy was increased by 14 days per invited subject. The cost per life year/quality adjusted life year was estimated at €16 717/€20 673. On the addition of low dose rivaroxaban reduced the costs per life year gained by 40%. If the model ran for only five follow up years, screening reduced relative mortality by 1.71%, suggesting PAD screening accounts for one fourth of the reported overall 7% relative mortality risk reduction of combined abdominal aortic aneurysm, PAD, and hypertension screening.

      Conclusion

      Screening of men for PAD is likely to be both clinically effective and cost effective in a lifetime perspective.

      Keywords

      This decision model represents the first examination of the value of male general population screening for peripheral artery disease (PAD), which is based on original data of a randomised screening trial. The model shows that screening is likely to be both clinically effective and cost effective in a lifetime perspective, and that low dose rivaroxaban could further reduce the costs per life year gain by 40%. After the first five years, screening for PAD reduced overall mortality by 1.71% corresponding to around one fourth of the 7% relative mortality risk reduction generated by triple vascular screening in the VIVA trial.

      Introduction

      International studies indicate that up to 20% of men over the age of 65 have peripheral arterial disease (PAD), a proportion which increases with age, and is mainly asymptomatic.
      • Diehm C.
      • Kareem S.
      • Lawall H.
      Epidemiology of peripheral arterial disease.
      ,
      • McDermott M.M.
      The magnitude of the problem of peripheral arterial disease: epidemiology and clinical significance.
      Approximately 25% – 30% of these men will die from cardiovascular disease within a five year period, and an even higher proportion will need hospitalisation due to cardiovascular disease. Cholesterol lowering, smoking cessation, low dose acetylsalicylic acid (aspirins), exercise, a healthy diet, and blood pressure control reduce the increased risks of cardiovascular disease by at least 20% – 25%.
      • Aboyans V.
      • Ricco J.B.
      • Bartelink M.E.L.
      • Björck M.
      • Brodmann M.
      • Cohnert T.
      • et al.
      Editor's Choice – 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European Society for Vascular Surgery (ESVS).
      ,
      • Conte M.S.
      • Bradbury A.W.
      • Kolh P.
      • White J.V.
      • Dick F.
      • Fitridge R.
      • et al.
      Global vascular guidelines on the management of chronic limb-threatening ischemia.
      For the case of symptomatic PAD, diagnosed by relevant symptoms and an ankle brachial systolic blood pressure index (ABI) < 0.9, professional vascular societies recommend preventive actions in order to prevent not just amputation but also general cardiovascular events, such as myocardial infarction (MI) and stroke, as well as excess all cause mortality.
      • Rooke T.W.
      • Hirsch A.T.
      • Misra S.
      • Sidawy A.N.
      • Beckman J.A.
      • Findeiss L.
      • et al.
      Management of patients with peripheral artery disease (compilation of 2005 and 2011 ACCF/AHA Guideline Recommendations): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.
      A relatively extensive literature has examined the cost effectiveness of diagnostic work up and treatment for populations with intermittent claudication.
      • Vaidya A.
      • Joore M.A.
      • ten Cate-Hoek A.J.
      • Kleinegris M.C.
      • ten Cate H.
      • Severens J.L.
      A systematic review of model-based economic evaluations of diagnostic and therapeutic strategies for lower extremity artery disease.
      Except for two studies focusing on the cost effectiveness of pharmacological therapy, the literature generally focuses on non-peripherally targeted and invasive interventions such as angiography, angioplasty, or bypass surgery. One recent study has assessed the cost effectiveness of screening and eventual prescription of low dose aspirin or clopidogrel for men at high risk of cardiovascular disease and found screening to be cost effective in a Dutch public healthcare sector perspective.
      • Vaidya A.
      • Joore M.A.
      • Ten Cate-Hoek A.J.
      • Ten Cate H.
      • Severens J.L.
      Screen or not to screen for peripheral arterial disease: guidance from a decision model.
      In a general population, ABI testing and subsequent prescription of aspirin therapy has been shown not to reduce cardiovascular events and mortality.
      • Fowkes F.G.
      • Price J.F.
      • Stewart M.C.
      • Butcher I.
      • Leng G.C.
      • Pell A.C.
      • et al.
      Aspirin for prevention of cardiovascular events in a general population screened for a low ankle brachial index: a randomized controlled trial.
      This is in line with the most recent recommendation of the US Preventive Services Task Force, concluding that there is insufficient evidence for the benefits of screening asymptomatic individuals for PAD.
      • Lin J.S.
      • Olson C.M.
      • Johnson E.S.
      • Whitlock E.P.
      The ankle-brachial index for peripheral artery disease screening and cardiovascular disease prediction among asymptomatic adults: a systematic evidence review for the U.S. Preventive Services Task Force.
      However, recent results from the population based VIVA screening trial demonstrate a substantial and cost effective 7% relative risk reduction in all cause mortality of combined screening for hypertension, abdominal aortic aneurysm (AAA), and PAD in the male general population.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      ,
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      The extent to which this effect can be attributed to the focus on PAD is unknown, although the effect has been linked to higher initiation rates for aspirin and lipid lowering and antihypertensive therapy, all of which have been found to reduce cardiovascular events and mortality in patients with intermittent claudication.
      • Wong P.F.
      • Chong L.Y.
      • Mikhailidis D.P.
      • Robless P.
      • Stansby G.
      Antiplatelet agents for intermittent claudication.
      The aim of this study was to model the lifetime effectiveness and cost effectiveness of general population PAD screening and relevant intervention in men. The contribution of the study is a first examination of the effectiveness and cost effectiveness of general population screening from a national healthcare system perspective.

      Materials and Methods

      A Markov model was built and informed, whenever possible, by estimates of the VIVA trial, which is currently the only randomised screening trial on PAD.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Original re-analysis in order for the VIVA data to fit the requirements of the model was conducted, and these were supplemented with estimates from the literature.

      Target population and screening programme

      The target population consisted of general population males aged 65 years to be invited for ABI measurement. For ABI < 0.9 or > 1.4, participants would be offered a consultation for confirmation of the diagnosis, information, and preventive measures such as walking instruction, smoking cessation, healthy diet, and initiation of 75 mg of low dose aspirin and 40 mg simvastatin, if not already initiated.

      Model structure

      The model structure was inspired by that of Vaidya et al.,
      • Vaidya A.
      • Joore M.A.
      • Ten Cate-Hoek A.J.
      • Ten Cate H.
      • Severens J.L.
      Screen or not to screen for peripheral arterial disease: guidance from a decision model.
      which was considered to accommodate the main consequences of screening and intervention except for a state for surgical repair and exclusive states for no/detected/undetected disease; in accordance, these were added for the present purpose (see Figs. S1–S3). Particular consideration was given to eventual interaction with non-PAD but as no effects of screening could be demonstrated in the VIVA trial
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      this was disregarded.
      Generally, model transitions occur annually for staying in the no PAD (and no major event) state or transition to undetected PAD, detected PAD, major events or death. A simplifying assumption was that transition between major events or to a specific state for repeated major events of the same type was not possible (this was instead built in the parameter estimates). Similarly, the model structure does not distinguish between different levels of severity within the individual states for the major events.
      The Markov part of the model structure was preceded by a tree structure distributing non-symptomatic individuals in the starting states of no/detected/undetected disease based on screening participation, the observed “prevalence” in VIVA based on two consecutive positive ABI tests, and the true prevalence
      • Rogan W.J.
      • Gladen B.
      Estimating prevalence from the results of a screening test.
      assuming sensitivity of 0.90 and specificity of 0.95.
      • Lijmer J.G.
      • Hunink M.G.
      • van den Dungen J.J.
      • Loonstra J.
      • Smit A.J.
      ROC analysis of noninvasive tests for peripheral arterial disease.
      False positives were assumed to incur costs and outcomes similar to those of true negatives (i.e. they were referred to the state of no disease). False negatives were referred for undetected disease. Symptomatic individuals were referred to the starting state of detected disease irrespective of scenario of screening or no screening.

      Parameter estimates

      Parameter estimates are the results of original analyses of data from the VIVA trial
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      ,
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      ,
      • Grondal N.
      • Sogaard R.
      • Henneberg E.W.
      • Lindholt J.S.
      The Viborg Vascular (VIVA) screening trial of 65-74 year old men in the central region of Denmark: study protocol.
      ,
      • Grondal N.
      • Sogaard R.
      • Lindholt J.S.
      Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65-74 years from a population screening study (VIVA trial).
      except for ABI test accuracy, quality of life, and background mortality. Table 1 shows parameter estimates and whether these have been published previously or whether they are based on original analysis for the present work, as described in the following and detailed in supplementary material.
      Table 1Parameters and their sources in the Markov decision model for the lifetime effectiveness and cost effectiveness of general population peripheral arterial disease (PAD) screening and relevant intervention in 65 year old men
      ParameterValue (95% CI)Probability distribution
      TypeMoments
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Reference
      Prevalence and incidence
      Annual probability.
       Detected after repeated ABI0.079 (0.067–0.091)Beta2051; 16624
      • Grondal N.
      • Sogaard R.
      • Lindholt J.S.
      Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65-74 years from a population screening study (VIVA trial).
       Incidence per year0.002 (0.002–0.003)Beta309; 127091
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Proportion with symptoms0.334 (0.031–0.036)Beta672; 1340
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Accept screening invitation0.748 (0.737–0.770)Beta18748; 6330
      • Grondal N.
      • Sogaard R.
      • Lindholt J.S.
      Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65-74 years from a population screening study (VIVA trial).
      Test accuracy
       Sensitivity0.90 (0.84–0.96)Beta Pert0.68; 1
      • Lijmer J.G.
      • Hunink M.G.
      • van den Dungen J.J.
      • Loonstra J.
      • Smit A.J.
      ROC analysis of noninvasive tests for peripheral arterial disease.
       Specificity0.95 (0.91–0.99)Beta Pert0.71; 1
      • Lijmer J.G.
      • Hunink M.G.
      • van den Dungen J.J.
      • Loonstra J.
      • Smit A.J.
      ROC analysis of noninvasive tests for peripheral arterial disease.
      Major event
      Annual probability.
       Revascularisation0.002 (0.002–0.003)Beta296; 128359
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Amputation0.002 (0.001–0.002)Beta220; 128654
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Myocardial infarction0.006 (0.005–0.006)Beta747; 127263
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Stroke0.008 (0.007–0.008)Beta970; 126717
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Intracerebral bleeding0.001 (0.001–0.002)Beta173; 128795
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Hazard ratio event
      Detected vs. undetected PAD
      Revascularisation1.055 (0.881–1.264)Lognormal0.054; 0.092
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Amputation0.891 (0.708–1.120)Lognormal–0.116; 0.117
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Myocardial infarction1.071 (0.956–1.199)Lognormal0.068; 0.058
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Stroke0.976 (0.879–1.083)Lognormal–0.024; 0.054
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Intracerebral bleeding1.043 (0.817–1.331)Lognormal0.042; 0.124
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Hazard ratio mortality
      PAD or major events vs. no PAD
      PAD without major events2.935 (2.670–3.201)Lognormal1.077; 0.136
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Revascularisation1.607 (1.386–1.828)Lognormal0.475; 0.113
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Amputation3.653 (3.457–3.849)Lognormal1.230; 0.100
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Myocardial infarction1.811 (1.677–1.944)Lognormal0.594; 0.068
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Stroke1.811 (1.687–1.935)Lognormal0.594; 0.063
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Intracerebral bleeding5.834 (5.658–6.011)Lognormal1.764; 0.090
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Detected vs. undetected PAD
      Pharmacological therapy0.881 (0.813–0.950)Lognormal–0.127; 0.035
      • Heart Protection Study Collaborative G.
      MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.
      Post revascularisation0.948 (0.611–1.469)Lognormal–0.054; 0.224
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post amputation0.977 (0.662–1.442)Lognormal–0.023; 0.199
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post myocardial infarction0.812 (0.625–1.054)Lognormal–0.208; 0.134
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post stroke0.852 (0.669–1.086)Lognormal–0.160; 0.124
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Intracerebral bleeding0.992 (0.699–1.407)Lognormal–0.008; 0.178
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Quality of life decrements
       Pharmacological therapy0.010 (–0.005 to 0.025)Beta1.680; 166.350
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
       Post revascularisation0.010 (–0.005 to 0.025)Beta1.680; 166.350
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
       Post amputation0.122 (0.069–0.175)Beta17.750; 127.742
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
       Post myocardial infarction0.026 (0.005–0.048)Beta5.446; 204.013
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
       Post stroke or intracerebral bleeding0.099 (0.081–0.117)Beta104.605; 952.009
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
      Costs – €
       Screening26Fixed
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Event year
      Pharmacological therapy2 183 (2 091–2 275)Gamma2157.306; 1.012
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Revascularisation25 932 (21 465–30 399)Gamma129.474; 200.287
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Amputation47 152 (31 560–62 744)Gamma35.133; 1324.086
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Myocardial infarction36 265 (31 722–40 808)Gamma244.765; 148.163
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Stroke21 402 (19 368–23 436)Gamma425.122; 50.343
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Intracerebral bleeding48 676 (32 141–65 211)Gamma33.293; 1462.037
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Subsequent years
      Pharmacological therapy1 711 (1 629–1 793)Gamma1659.592; 1.031
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post-revascularisation3 983 (2 866–5 100)Gamma48.828; 81.572
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post-amputation4 432 (1 966– 6 898)Gamma12.412; 357.077
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post-myocardial infarction5 152 (3 304–7 000)Gamma29.849; 172.603
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post-stroke2 420 (1 916–2 924)Gamma88.668; 27.293
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
      Post-intracerebral bleeding1 957 (895–3 019)Gamma13.037; 150.109
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Tables 2 and 3 and supplementary material for details).
       Community cost after amputation74 623Fixed
      • Kruse M.
      • Nørregaard J.
      Community Costs of Selected Late Complications to Diabetes.
       Community cost after stroke32 819Fixed
      • Portelli R.
      • Lowe D.
      • Irwin P.
      • Pearson M.
      • Rudd A.G.
      Intercollegiate Stroke Working P. Institutionalization after stroke.
      ,
      • Skolarus L.E.
      • Freedman V.A.
      • Feng C.
      • Wing J.J.
      • Burke J.F.
      Care received by elderly US stroke survivors may be underestimated.
      Discounting
       Costs0.035Fixed
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
       Life years0.035Fixed
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      CI = confidence interval; ABI = ankle brachial index.
      Annual probability.
      Additional original analysis to what was initially reported has been conducted for the present purpose (see Table 2, Table 3 and supplementary material for details).
      Event probabilities were analysed by estimating rates in the subgroup of the VIVA population who were not screened. Time to first event was used for estimating rates that were mathematically converted to annual probabilities using the conventional approach.
      • Gidwani R.
      • Russell L.B.
      Estimating transition probabilities from published evidence: a tutorial for decision modelers.
      The National Patient Registry
      • Lynge E.
      • Sandegaard J.L.
      • Rebolj M.
      The Danish National Patient Register.
      was used to identify events defined by International Classification of Disease 10th edition codes for PAD diagnosis (I700, I702, I708, I709, I740, I741, I743, I744, I748, I749), acute myocardial infarction (I21-3), stroke, or transient ischaemic attack (I63-7), and intracerebral haemorrhage (I60-1). Vascular repair was identified by procedure codes in the national clinical database National Vascular Registry Karbasen (indication 1, 16-9) and amputations were identified in the National Patient Registry by procedure codes (KNFQ, KNGQ, and KNHQ). Hazard ratios were estimated by Cox regression comparing the rates between relevant subgroups groups.
      Excess mortality for the health states with PAD was estimated for relevant subpopulations of VIVA. The relative risk of undetected disease was assumed to be equal to the risk of detected disease treated by pharmacological therapy only.
      The effects of prophylactic pharmacological therapy on events of MI and stroke were informed from the British Heart Protection Collaborative Group.
      • Heart Protection Study Collaborative G.
      MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.
      The screening programme cost was based on the original microcosting in the VIVA trial,
      • Søgaard R.
      • Lindholt J.S.
      Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
      which here deducted five minutes of personnel time per participant and the ultrasonography equipment cost, which was related solely to the AAA test. Follow up costs were estimated as incidence costs (1) during the first 12 months after an event and (2) per 12 months for subsequent years in the relevant subpopulations of VIVA. Estimates were based on the registries used for identification of events and valued using national average tariffs of collaborative agreements (general practice), market prices (prescription medication), and the Danish Ambulant or Diagnosis Related Grouping system (hospital based activity).
      Community based, formal care was estimated from the literature. For stroke patients, 14% (not institutionalised before the event) became institutionalised after discharge.
      • Portelli R.
      • Lowe D.
      • Irwin P.
      • Pearson M.
      • Rudd A.G.
      Intercollegiate Stroke Working P. Institutionalization after stroke.
      The remainder required, on average, 10.5 hours more home care per week than a matched control group without stroke.
      • Skolarus L.E.
      • Freedman V.A.
      • Feng C.
      • Wing J.J.
      • Burke J.F.
      Care received by elderly US stroke survivors may be underestimated.
      These estimates were valued using unit costs of €26 854 per year for home care and €69 462 per year for nursing home placement. For amputation patients, a total annual cost of €74 623 has been estimated to cover wound care, prosthesis, assistive aids, physiotherapy, home care, etc.
      • Kruse M.
      • Nørregaard J.
      Community Costs of Selected Late Complications to Diabetes.

      Model validation

      The model was first validated by various logical checks, e.g., by confirming that fewer (and ultimately zero) individuals would end up in event states if event rates were reduced towards zero. Next, trackers were defined for all of the transitions in order to ensure that the total population was accounted for by probabilities summing to one, and that the overall life expectancy matched national statistics. Finally, graphical illustration of events over time was made for all states in order to visually ensure consistency and confirm the expected differences between scenarios.

      One way sensitivity analysis

      Deterministic sensitivity analyses were conducted individually for all model parameters. Confidence limits (95%) of the individual parameters were used as the lowest/highest plausible value. The absolute deviation in euros from the incremental costs per life year/quality adjusted life year are reported in tornado diagrams.

      Alternative scenario cases

      • 1.
        Restricting the cost perspective to healthcare (excluding community care costs after amputation and stroke).
      • 2.
        Updating the cost of pharmacological therapy and the relative risks of major events in accordance with a recent report of improved effectiveness of low dose rivaroxaban as add on to aspirin therapy.
        • Anand S.S.
        • Bosch J.
        • Eikelboom J.W.
        • Connolly S.J.
        • Diaz R.
        • Widimsky P.
        • et al.
        Rivaroxaban with or without aspirin in patients with stable peripheral or carotid artery disease: an international, randomised, double-blind, placebo-controlled trial.
        Based on the Danish market price for a daily dose of 2 × 2.5 mg, an annual therapy cost of €980 was assigned. Relative risk reductions of events were: MI 0.76, stroke 0.54, amputation 0.40, revascularisation 0.63 (weighted according to proportion with acute vs. chronic ischaemia in the VIVA trial), major bleeding 1.61, and death 0.91.
        • Anand S.S.
        • Bosch J.
        • Eikelboom J.W.
        • Connolly S.J.
        • Diaz R.
        • Widimsky P.
        • et al.
        Rivaroxaban with or without aspirin in patients with stable peripheral or carotid artery disease: an international, randomised, double-blind, placebo-controlled trial.
      • 3.
        Revising the relative risks of MI and stroke in the screened arm according to the Medical Research Council/British Heart Foundation Heart Protection Study where initiation of statin therapy reduced the risks of events by 0.74 for MI and 0.76 for stroke.
        • Heart Protection Study Collaborative G.
        MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.

      Results

      The Markov model was informed mainly by the VIVA trial of which some parameter estimates required original re-analysis.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      Table 1 provides a complete overview of the parameters used in the model and their sources. Table 2, Table 3 detail key original re-analyses of data from the VIVA trial. In terms of major events, screening increased the rate of revascularisation by 5.5% while it reduced the rate of amputation by 10.9%. Major bleeding was increased by 4.3% as a consequence of initiation of aspirin therapy and MI by 7.1% while stroke was reduced by 2.4%. In terms of all cause mortality, the preventive effects of screening are most pronounced after stroke and MI, followed by revascularisation and amputation, due to the earlier intervention.
      Table 2Hazard ratios (HR) of major events for screened vs. not screened for peripheral arterial disease in the VIVA trial (n = 50 152)
      Person time – yEvents – nRate per 1000 person yearsHR (95% CI)
      Revascularisation1.055 (0.881–1.264)
       No screening103 7672292.207
       Screening104 9342442.325
      Amputation0.891 (0.708–1.120)
       No screening103 9921541.481
       Screening105 2541391.321
      Myocardial infarction1.071 (0.956–1.199)
       No screening103 1305775.595
       Screening104 1526245.991
      Stroke0.976 (0.879–1.083)
       No screening102 8497116.913
       Screening104 0837026.745
      Intracerebral bleeding1.043 (0.817–1.331)
       No screening104 0671261.211
       Screening105 2881331.263
      Table 3Hazard ratios (HR) of mortality for screened vs. not screened for peripheral arterial disease in the VIVA trial (n = 50 152)
      nPerson time – yEvents – nRate per 1000 person yearsHR (95% CI)
      Post-revascularisation4730.948 (0.611–1.469)
       No screening9493941.105
       Screening1 0364139.593
      Post-amputation2930.977 (0.662–1.442)
       No screening5975490.451
       Screening5314890.302
      Post-myocardial infarction1 2030.812 (0.625–1.054)
       No screening2 37911849.610
       Screening2 64010740.536
      Post-stroke1 4130.852 (0.669–1.086)
       No screening2 93314148.073
       Screening2 95712241.251
      Post-intracerebral bleeding2590.992 (0.699–1.407)
       No screening42461143.746
       Screening45965141.460
      Note that relative mortality rate ratios for individuals detected with peripheral artery disease (PAD, before eventual major event) cannot be estimated because no control group can be identified (only incidentally detected individuals will have a PAD diagnosis in the National Patient Registry, statin and aspirin users registered in the Prescription Registry do not necessarily have PAD).
      Figure 1, Figure 2, Figure 3 illustrate the modelled mechanisms of screening by the incremental number of events between scenarios with vs. without screening for a population of 100 000 men. In Fig. 1, the rates of detected and undetected PAD are illustrated with the main difference between scenarios being that asymptomatic individuals are detected, which, in the first year, is perfectly mirrored by fewer being undetected. As time passes, the effect of screening with respect to fewer deaths means that marginally more individuals live with PAD.
      Figure 1
      Figure 1The impact of screening for peripheral arterial disease (PAD) on the timing of PAD detections.
      Figure 2
      Figure 2The impact of screening for peripheral arterial disease on the timing of pharmacological (Pharm.) prevention and associated harm.
      Figure 3
      Figure 3The impact of screening for peripheral arterial disease on major cardiovascular events.
      In Fig. 2, initiation of pharmacological preventive therapy is illustrated. The rate of initiation aggregates the rate of PAD detection by screening plus detections due to symptoms or major events (MI or stroke), and the rate of undertreatment, again mirrors that of therapy initiated minus the additional potential because more individuals live with PAD. The most important relationship in the figure is the ratio between benefit and harm, i.e., the ratio of initiations of intracerebral haemorrhage as a consequence of aspirin therapy. In absolute numbers per 100 000 population, screening results in one or two more events per year (maximum number of events per year is 53 for screening and 51 without screening, see Fig. S4).
      Fig. 3 illustrates that the revascularisation rate is increased up to 30 years after screening and that this results in fewer amputations during the first 10 years. A similar pattern is seen for stroke where the rate is reduced during the first four years but then gradually increases and reaches its maximum 19 years after screening due to events being delayed and more individuals living at risk because they have survived competing events. An increase in MI appears to a possible harm of screening throughout the entire follow up time.
      The main results are listed in Table 4. From a lifetime perspective for a 65 year old man, the clinical effectiveness of screening increases the life expectancy from 11.170 years to 11.209 years, which corresponds to 14 days. The quality adjusted analogues, which take into account the disutility of, e.g., pharmacological therapy, as well as the general decrease in quality of life as individuals age, are from 9.510 years to 9.478 years. The net cost of screening, which includes the screening programme as well as the cost consequences flowing from more or less healthcare over time, sums to €654 per invited person. The additional cost of a life year gain thus ends at €16 717 and the quality adjusted analogue at €20 673, which is likely to be considered cost effective at conventional thresholds for cost effectiveness.
      Table 4The overall clinical effectiveness and cost effectiveness of screening vs. no screening for peripheral arterial disease (PAD) for the reference case and three alternative scenarios
      ScreeningNo screeningDifferenceICERProbability cost effective at threshold of €1 000
      102030
      Reference case
       Cost – €4 0163 362654
       Life year11.20911.1700.03916 7170.0220.7840.982
       QALY9.5109.4780.032206730.0050.4350.921
      Restricted cost perspective
       Cost – €1 7801267513
       Life year11.20911.1700.039131120.0700.9470.992
       QALY9.5109.4780.032162140.0040.8140.983
      Rivaroxaban add on to aspirin
       Cost – €3 9743 323650
       Life year11.23711.1730.064101800.469>0.999>0.999
       QALY9.5349.4810.052123970.0690.999>0.999
      MRC/BHF Heart Protection Study relative risks for MI and stroke
       Cost – €4 4383 949489
       Life year11.24911.2060.04311 4380.1580.9360.996
       QALY9.5409.5060.034144410.0600.8060.979
      ICER = incremental cost effectiveness ratio, MI = myocardial infarction; QALY = quality adjusted life years.
      The first alternative case scenario illustrates how disregard of the community care costs marginally reduced the net cost of screening and thus increase the likelihood of screening being cost effective. The next reflects a scenario where rivaroxaban is initiated as an add on to aspirin for all men detected with PAD. Due to rivaroxaban effectively protecting against mortality after all of the major events, the costs per quality adjusted life year decreased to €12397 – a 40% reduction. The final alternative scenario case adopts relative effects of initiation of statin therapy on the rates of MI and stroke from the MRC/BHF Heart Protection Study, demonstrating increased clinical effectiveness and leading to a cost per quality adjusted life year of €14 441.
      One way sensitivity analysis was additionally conducted to provide an estimate of the importance of individual model parameters for the results, as well as to give an indication of the value of further information. The most influential parameters were the mortality ratios associated with screening after major events (stroke, amputation, myocardial infarction) or PAD onset, i.e., the advantage of early detection (full results shown in Figs. S8 and S9). The main reason for these large influences is the relatively small sample sizes behind the estimates. One exception is the mortality ratio associated with early detection of PAD, which is influential because of scale rather than imprecision; an alternative ratio of 0.95 instead of 0.88 almost doubled the cost per quality adjusted life year.
      If the Markov model is run for only five follow up years in order to produce a relative mortality risk reduction of PAD screening, which is comparable to the overall relative mortality risk reduction of the VIVA trial, the result is 1.71%, which suggests that PAD screening accounts for around one fourth of the overall 7% relative mortality risk reduction of combined AAA, PAD and hypertension screening.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.

      Discussion

      In this first, trial based model on the clinical effectiveness and cost effectiveness of screening for PAD in a lifetime perspective, it was found that screening is warranted and, overall, could increase the general life expectancy for 65 year old men by 14 days at a healthcare system cost of €654 per invited person. In the original VIVA trial, a relative risk reduction in mortality of 7% for combined screening for AAA, PAD, and hypertension was reported and caused much debate as to the relative contribution of the individual screening targets.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      ,
      • Ayoub C.
      • Murad M.H.
      Population-based screening for vascular disease.
      It has previously been suggested that the hypertension focus accounts for approximately two percentage points and now it is suggested that the PAD focus accounts for 1.71 percentage points. Consequently, the remaining 3.29 percentage points would be attributable to the AAA focus. This is considerably higher than what can be explained by AAA specific mortality, which accounts for 1.5% of all deaths in the VIVA control group.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      We believe that the interval between 1.5% and up to the 3.29% can be explained by the initiation of modern cardiovascular prevention when an AAA is diagnosed. This is in line with observational studies reporting up to 60% reduced overall mortality when AAA patients use statins.
      • Kertai M.D.
      • Boersma E.
      • Westerhout C.M.
      • van Domburg R.
      • Klein J.
      • Bax J.J.
      • et al.
      Association between long-term statin use and mortality after successful abdominal aortic aneurysm surgery.
      Previous attempts to estimate the benefit of PAD screening have been based on data from non-screening contexts. Itoga et al. estimated the incremental cost of PAD screening at US$338 and the incremental quality adjusted life year (QALY) at 0.00380, which generated an incremental cost per QALY at $88 758 in the US setting.
      • Itoga N.K.
      • Minami H.R.
      • Chelvakumar M.
      • Pearson K.
      • Mell M.M.
      • Bendavid E.
      • et al.
      Cost-effectiveness analysis of asymptomatic peripheral artery disease screening with the ABI test.
      They estimated event probabilities from a large number of different sources where specialised care related selection mechanisms could mean that severity is different from the population average. Another main difference to the present is that the effect of screening was defined by three relative risks only (25% decrease in probability of development of PAD symptoms and 50% decrease in probability of MI or stroke, respectively) as opposed to more widespread effects of screening in the present model (including also increased probabilities of revascularisation and intracerebral bleeding, and reduced probabilities of amputation).
      The results can be interpreted as conservative. First, parameter estimates are based on use of Simvastatin although more efficient statins are available today at comparable costs, e.g., atorvastatin. Similarly, first line choice use of clopidogrel would probably also increase cost effectiveness, as CAPRIE reported it to reduce the relative risk of cardiovascular events in the subgroup with PAD.
      CAPRIE Steering Committee
      A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events (CAPRIE).
      However, when VIVA was designed, clopidogrel was still very expensive, and low dose aspirin remains the first line choice today for many vascular surgeons. Furthermore, the benefits of reacting to potential untreated hypertension exposed by ABI screening, smoking cessation advice, and monitoring of whether the goal of low density lipoprotein < 1.4 mmol/L was achieved with and without use of PSK9 inhibitors were not included in the model. In all, these caused the model to be conservative, so even more attractive benefits and cost effectiveness could most likely be achieved.
      In the second alternative scenario, the consequences of adding low dose rivaroxaban, which is relatively expensive, to low dose aspirin led to increased clinical effectiveness. It was surprising to observe that it turned out to be the most cost effective scenario, especially as the modelled scenario could be too pessimistic, as a hazard ratio of 1.61 was used for the risk of intracranial bleeding. That is in direct contrast to a recent report actually reporting lowered risks of bleeding when used in a clinical cohort.
      • Garcia Rodriguez L.A.
      • Vora P.
      • Brobert G.
      • Soriano-Gabarro M.
      • Cea Soriano L.
      Bleeding associated with low-dose aspirin: comparison of data from the COMPASS randomized controlled trial and routine clinical practice.
      In fact, this attractive result may also be conservative, as the PAD population in COMPASS also included carotid stenosis, where the benefit of low dose rivaroxaban is smaller than the benefits recently published regarding those with symptomatic lower limb PAD, but the overall PAD results were chosen as two thirds of screening diagnosed PAD are asymptomatic with assumed lower benefit than symptomatic PAD.
      • Kaplovitch E.
      • Eikelboom J.W.
      • Dyal L.
      • Aboyans V.
      • Abola M.T.
      • Verhamme P.
      • et al.
      Rivaroxaban and aspirin in patients with symptomatic lower extremity peripheral artery disease: a subanalysis of the COMPASS randomized clinical trial.
      Further, the elevated risk of MI after screening was maintained, which was observed in VIVA, although this seems illogical and in contrast to the British Heart Protection Study.
      • Heart Protection Study Collaborative G.
      MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.
      On the one hand, the observed elevated risk in VIVA could be due to random variation, but, on the other hand, it could also be a consequence of competing risks; screening reduces mortality so the beneficiaries, who by nature have increased risk of MI, live longer. Consequently, the conservative scenario was chosen as the reference case and the third alternative scenario was the risks observed in the British Heart Protection Study, which reduced the costs per QALY by 30%. Finally, the model disregarded the fact that quality of life was observed to improve after detection of PAD in the VIVA trial, probably due to explanation and instructions or revascularisation due to intermittent claudication.
      • Lindholt J.S.
      • Sogaard R.
      Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
      The main strength of this work is that the model is based on original data from a screening trial, and that a lifetime perspective is estimated. As always, model results depend not only on the structure of the model (simplified version of real world) but also on the parameter estimates that may vary from context to context, and which will reflect higher levels of precision as time passes.
      The model structure is based on a number of simplifying assumptions that are partly compensated for by the samples underlying the parameter estimates. For example, the model has no specific states for repeated or mixed major events such as revascularisation or myocardial infarction but instead assigns average mortality risks and health utilities according to the event occurring. These averages can be seen as weighted according to the occurrence of repeated and mixed major events based on the follow up in the VIVA trial. The most obvious bias that could result from this is an underestimation of severity in the longer run. A related issue is that the major event risks are constant throughout life and have no further age effects built in because the VIVA data so far cover only five years of follow up and events are too rare to analyse differences in rates between different ages. But as this affects both the screening and the no screening scenarios, this is not considered a major threat to the validity of the conclusion.
      It is difficult to provide alternative estimates for other contexts as no other general population based trial than VIVA is currently available. Generally, the lower the disease burden, the poorer the rationale for screening. Stakeholders might thus consider the comparability of their respective contexts with the CVD burden in Denmark; in a comparison of age standardised CVD related mortality in men, Denmark has the eighth lowest burden amongst 50 European countries.
      • Townsend N.
      • Wilson L.
      • Bhatnagar P.
      • Wickramasinghe K.
      • Rayner M.
      • Nichols M.
      Cardiovascular disease in Europe: epidemiological update 2016.
      If decision makers are to await perfect information from randomised controlled trials with lifetime follow up, lives will be lost if effective programmes are neglected. That is a key rationale for a decision analytical model such as the present one. If the model result of a 1.71% relative reduction in mortality five years after PAD screening should be replicated in a 1:1 randomised trial, around 600 000 men would have to be included, using 5% significance level, 80% power, and assuming two years of reaching full recruitment. Such a volume corresponds to all 65 year old men and women in the Nordic countries (Finland, Norway, Sweden, Denmark, Svalbard, Iceland, and Faroe Islands) and would incur a cost of at least €8 million for the screenings alone and without considering costs for follow up and analysis. This is hardly going to happen.
      In conclusion, screening of men for PAD is likely to be both clinically effective and cost effective in a lifetime perspective. Whether or not original trials should be conducted to further explore these results, must be a decision for health policy makers. However, in that case, the value of additional information should be weighed against the lives possibly lost while delaying a decision.

      Conflict of interest

      None.

      Funding

      None.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Diehm C.
        • Kareem S.
        • Lawall H.
        Epidemiology of peripheral arterial disease.
        VASA. 2004; 33: 183-189
        • McDermott M.M.
        The magnitude of the problem of peripheral arterial disease: epidemiology and clinical significance.
        Cleve Clin J Med. 2006; 73: S2-7
        • Aboyans V.
        • Ricco J.B.
        • Bartelink M.E.L.
        • Björck M.
        • Brodmann M.
        • Cohnert T.
        • et al.
        Editor's Choice – 2017 ESC guidelines on the diagnosis and treatment of peripheral arterial diseases, in collaboration with the European Society for Vascular Surgery (ESVS).
        Eur J Vasc Endovasc Surg. 2018; 55: 305-368
        • Conte M.S.
        • Bradbury A.W.
        • Kolh P.
        • White J.V.
        • Dick F.
        • Fitridge R.
        • et al.
        Global vascular guidelines on the management of chronic limb-threatening ischemia.
        Eur J Vasc Endovasc Surg. 2019; 58: S1-S109
        • Rooke T.W.
        • Hirsch A.T.
        • Misra S.
        • Sidawy A.N.
        • Beckman J.A.
        • Findeiss L.
        • et al.
        Management of patients with peripheral artery disease (compilation of 2005 and 2011 ACCF/AHA Guideline Recommendations): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines.
        J Am Coll Cardiol. 2013; 61: 1555-1570
        • Vaidya A.
        • Joore M.A.
        • ten Cate-Hoek A.J.
        • Kleinegris M.C.
        • ten Cate H.
        • Severens J.L.
        A systematic review of model-based economic evaluations of diagnostic and therapeutic strategies for lower extremity artery disease.
        Thromb Haemost. 2014; 111: 19-28
        • Vaidya A.
        • Joore M.A.
        • Ten Cate-Hoek A.J.
        • Ten Cate H.
        • Severens J.L.
        Screen or not to screen for peripheral arterial disease: guidance from a decision model.
        BMC Public Health. 2014; 14: 89
        • Fowkes F.G.
        • Price J.F.
        • Stewart M.C.
        • Butcher I.
        • Leng G.C.
        • Pell A.C.
        • et al.
        Aspirin for prevention of cardiovascular events in a general population screened for a low ankle brachial index: a randomized controlled trial.
        JAMA. 2010; 303: 841-848
        • Lin J.S.
        • Olson C.M.
        • Johnson E.S.
        • Whitlock E.P.
        The ankle-brachial index for peripheral artery disease screening and cardiovascular disease prediction among asymptomatic adults: a systematic evidence review for the U.S. Preventive Services Task Force.
        Ann Intern Med. 2013; 159: 333-341
        • Lindholt J.S.
        • Sogaard R.
        Population screening and intervention for vascular disease: Five-year results of the randomised Viborg Vascular (VIVA) screening trial of Danish men.
        Lancet. 2017; 390: 2256-2265
        • Søgaard R.
        • Lindholt J.S.
        Cost-effectiveness of population-based vascular disease screening and intervention in men from the Viborg Vascular (VIVA) trial.
        Br J Surg. 2018; 105: 1283-1293
        • Wong P.F.
        • Chong L.Y.
        • Mikhailidis D.P.
        • Robless P.
        • Stansby G.
        Antiplatelet agents for intermittent claudication.
        Cochrane Database Syst Rev. 2011; : CD001272
        • Rogan W.J.
        • Gladen B.
        Estimating prevalence from the results of a screening test.
        Am J Epidemiol. 1978; 107: 71-76
        • Lijmer J.G.
        • Hunink M.G.
        • van den Dungen J.J.
        • Loonstra J.
        • Smit A.J.
        ROC analysis of noninvasive tests for peripheral arterial disease.
        Ultrasound Med Biol. 1996; 22: 391-398
        • Grondal N.
        • Sogaard R.
        • Henneberg E.W.
        • Lindholt J.S.
        The Viborg Vascular (VIVA) screening trial of 65-74 year old men in the central region of Denmark: study protocol.
        Trials. 2010; 11: 67
        • Grondal N.
        • Sogaard R.
        • Lindholt J.S.
        Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65-74 years from a population screening study (VIVA trial).
        Br J Surg. 2015; 102: 902-906
        • Gidwani R.
        • Russell L.B.
        Estimating transition probabilities from published evidence: a tutorial for decision modelers.
        Pharmacoeconomics. 2020; 38: 1153-1164
        • Lynge E.
        • Sandegaard J.L.
        • Rebolj M.
        The Danish National Patient Register.
        Scand J Public Health. 2011; 39: 30-33
        • Heart Protection Study Collaborative G.
        MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial.
        Lancet. 2002; 360: 7-22
        • Portelli R.
        • Lowe D.
        • Irwin P.
        • Pearson M.
        • Rudd A.G.
        Intercollegiate Stroke Working P. Institutionalization after stroke.
        Clin Rehabil. 2005; 19: 97-108
        • Skolarus L.E.
        • Freedman V.A.
        • Feng C.
        • Wing J.J.
        • Burke J.F.
        Care received by elderly US stroke survivors may be underestimated.
        Stroke. 2016; 47: 2090-2095
        • Kruse M.
        • Nørregaard J.
        Community Costs of Selected Late Complications to Diabetes.
        Danish Institute for Health, Copenhagen2011 ([In Danish])
        • Anand S.S.
        • Bosch J.
        • Eikelboom J.W.
        • Connolly S.J.
        • Diaz R.
        • Widimsky P.
        • et al.
        Rivaroxaban with or without aspirin in patients with stable peripheral or carotid artery disease: an international, randomised, double-blind, placebo-controlled trial.
        Lancet. 2018; 391: 219-229
        • Ayoub C.
        • Murad M.H.
        Population-based screening for vascular disease.
        Lancet. 2017; 390: 2218-2220
        • Kertai M.D.
        • Boersma E.
        • Westerhout C.M.
        • van Domburg R.
        • Klein J.
        • Bax J.J.
        • et al.
        Association between long-term statin use and mortality after successful abdominal aortic aneurysm surgery.
        Am J Med. 2004; 116: 96-103
        • Itoga N.K.
        • Minami H.R.
        • Chelvakumar M.
        • Pearson K.
        • Mell M.M.
        • Bendavid E.
        • et al.
        Cost-effectiveness analysis of asymptomatic peripheral artery disease screening with the ABI test.
        Vasc Med. 2018; 23: 97-106
        • CAPRIE Steering Committee
        A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events (CAPRIE).
        Lancet. 1996; 348: 1329-1339
        • Garcia Rodriguez L.A.
        • Vora P.
        • Brobert G.
        • Soriano-Gabarro M.
        • Cea Soriano L.
        Bleeding associated with low-dose aspirin: comparison of data from the COMPASS randomized controlled trial and routine clinical practice.
        Int J Cardiol. 2020; 318: 21-24
        • Kaplovitch E.
        • Eikelboom J.W.
        • Dyal L.
        • Aboyans V.
        • Abola M.T.
        • Verhamme P.
        • et al.
        Rivaroxaban and aspirin in patients with symptomatic lower extremity peripheral artery disease: a subanalysis of the COMPASS randomized clinical trial.
        JAMA Cardiol. 2021; 6: 21-29
        • Townsend N.
        • Wilson L.
        • Bhatnagar P.
        • Wickramasinghe K.
        • Rayner M.
        • Nichols M.
        Cardiovascular disease in Europe: epidemiological update 2016.
        Eur Heart J. 2016; 37: 3232-3245

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.