What Randomized Clinical Trials Are Possible / Necessary In Phlebology Mark H. Meissner, MD Professor of Surgery University of Washington School of Medicine

Levels of Evidence for Therapeutic Studies Straus SE, Evidence-Based Medicine 3rd Ed Level of Evidence Studies of therapy, prevention, etiology, harm 1a Systematic review with homogeneity of RCTs 1b Individual RCT with narrow confidence intervals 1c All or none 2a Systematic review with homogeneity of cohort studies 2b Individual cohort study or low quality RCT 3a Systematic review with homogeneity of case-control studies 3b Individual case-control study 4 Case series 5 Expert opinion without explicit critical appraisal What Do We Really Care About? l Incorporation of societal values l Societal costs l Comparative effectiveness of different technologies l The best available estimate of benefits and harms (estimate of treatment effect) l Application of the evidence to the individual patient (generalizeabilty) (generalizeabilty)

Where Does Clinical Evidence Come From? How Do We Measure the Magnitude of Effect? l Semi – experimental l Comparison with historical controls l Fatally biased l Observational studies (all with concurrent controls) l Cross sectional - Compares proportion with disorder based on exposure at one point in time l Cohort studies – Prospective evaluation of outcome based on exposure l Case - control studies – Retrospective evaluation of exposure based on outcome l Randomized, controlled clinical trials

Determinants of Evidence Quality DeterminantDefinitionQualityBias Treatment Effect Systematic Review of RCTs HighLowPrecise Randomized Clinical Trials Observational Studies Methodology Cohort Studies Case-Control Studies Case Series Unknown Expert Opinion LowHighUnknown Consistency Directness Similarity of treatment effect across studies Appropriateness of groups and outcomes

RCTs – The Holy Grail l Require true clinical equipoise (RR 0.4 – 0.9) l Difficult to justify if observational studies show l Large harmful effects l Large (risk ratio < 0.4) beneficial effects l Small beneficial effects (risk ratio ) l Are expensive l May be difficult to generalize (Restrictive inclusion criteria) l Usually not powered to detect harms of treatment l May be better, worse, or complimentary to observational studies l Why are RCTs the holy grail? l Comparison to standard of care l Minimizes bias & confounders l Provides a precise estimate of effect But …

Not All Questions Require RCTs “We think that everyone might benefit if the most radical protagonists of evidence-based medicine organised and participated in a double blind, placebo controlled, crossover trial of the parachute” This is Nonsense Magnitude of effect is important All or None Phenomenon

Nor Is There An RCT For Every Question Ioannidis et al: JAMA 2001 l 48 interventions with randomized and observational trials l Results highly correlated (correlation coefficient ) l Larger treatment effect in nonrandomized trials

Trial Design A Continuum Rather Than A Hierarchy Treatmen t Effect Example Huge (All or None) Parachutes Epinephrine/An aphylaxis UFH/DVT LargeBypass for CLI Moderate Statins HCSE Case Series Observational Studies Case Series Observational Studies RCTs Standard of Care Established

What Are The Important Questions? Chronic Venous Disease l Is the use of compression prior to intervention cost effective ? l What is the best treatment for C2 & C3 disease? l Interventions Compression Pharmacotherapy (HCSE, MPFF) Ablation (RF, laser, foam) l Outcomes Patient important benefits – Pain, quality of life, recurrence Costs to health care system l Perforating veins l The pathological perforator – Which are clinically important? l C5, 6 disease – Healing and recurrence l Is 1 st rib resection after a first effort thrombosis warranted? l What is the accuracy of CTV / MRV for iliac obstruction … And Many Others

What Are The Important Questions? Chronic Venous Disease l Is there a role for extended prophylaxis other than THR and malignancy? l Are there ANY prophylactic indications for IVC Filters? l The treatment of acute DVT l Pharmacomechanical thrombolysis Iliofemoral DVT Femoropopliteal DVT l Isolated calf vein thrombosis l Is there any role for U/S (using US protocols) in determining the duration of anticoagulation? … And Many Others

How Do We Answer the Questions? Clinical QuestionRCTObservationalOutcomes Value of Compression (C2)√√QoL, Cost (ICER) Comparative effectiveness of different technologies √QoL, Cost (ICER, cost-consequence) Definition of the pathologic perforator √Ulcer healing / recurrence Extended prophylaxis√Recurrent DVT, Bleeding Pharmacomechanical lysis√QoL, Bleeding, Cost Calf vein thrombosis√Recurrent thrombosis, Bleeding, cost U/S & anticoagulation√Recurrent thrombosis Role of 1 st rib resection in effort thrombosis √Recurrent thrombosis

The CLASS Trial l HTA (UK) funded randomized clinical trial l 1000 C 2-6 patients (6 centers) l Saphenous surgery l Foam sclerotherapy l Laser ablation with adjuvant foam sclerotherapy l 1º outcomes (6 months, possible 5 yr) l Disease specific – Aberdeen VV Questionnaire l Generic – EuroQol, SF-36 l 2º outcomes l Validated return to function instrument l Incremental cost effectiveness

ATTRACT TRIAL l 692 patients l 28 North American centers l Randomized to l Best medical therapy l Pharmacomechanical lysis Trellis 8 Angiojet powerpulse l Iliofemoral & femoropopliteal arms l Clinically relevant endpoints l Objective PTS (Villalta) l Quality of life

The DiVeTAS Trial – Specific Aims DIstal VEnous Thrombosis: Anticoagulation vs Surveillance l To compare the short-term efficacy and safety of standard anticoagulation versus duplex ultrasound surveillance for a first episode of acute symptomatic DVT confined to the calf veins. The primary endpoint will be a composite of proximal propagation, symptomatic pulmonary embolism (PE), major bleeding, and all-cause mortality occurring during the first 3 months of treatment. l To compare the short-term efficacy and safety of standard anticoagulation versus duplex ultrasound surveillance for a first episode of acute symptomatic DVT confined to the calf veins. The primary endpoint will be a composite of proximal propagation, symptomatic pulmonary embolism (PE), major bleeding, and all-cause mortality occurring during the first 3 months of treatment. l To evaluate the relationship between baseline characteristics, including D- Dimer and other biomarkers, and the risk of proximal propagation and other endpoints, with the goal of identifying high risk and low risk sub- groups which may differ in treatment efficacy. l To compare long-term outcomes of calf DVT after treatment with standard anticoagulation versus duplex ultrasound surveillance with respect to the development of objectively defined PTS and quality of life. l To compare the cost and cost-effectiveness of standard anticoagulation versus duplex ultrasound surveillance for the management of isolated calf vein thrombosis.

Comparative Effectiveness Research The “New” Holy Grail l Background l Interventional technology – 50% of healthcare resources (50 million procedures / yr) l Clinical data in < 15% of 510k approvals l Adoption after only 10-20% perceived implementation l Practice integration before value, risks, and costs established l Comparative effectiveness l “a rigorous evaluation of different treatment options” (Congressional Budget Office) (Congressional Budget Office) l May focus on benefits/risks or cost/benefit l > $1 billion dollars appropriated by Congress

CDRH Device Classification l Class I l Low risk devices (tongue depressors, scalpels) l General controls Good manufacturing practices Quality systems regulation l Class II l Venous lasers, RF devices l Special controls - Performance standards, registries, postmarket surveillance l Most approved through Premarket Notification (510k) l Safety / effectiveness equivalent to predicate device l Class III l Insufficient information to ensure safety & effectiveness l Most approved through Premarket Application (PMA)

Growth in Varicose Vein Treatment Courtesy of John Mauriello

Economic Analysis* ModelDescription Pros / Cons Economic Quantitative, statistical analysis of economics only Simple, but neglects clinical outcomes Cost- consequence Economic & clinical outcomes evaluated in common Allows evaluation of “trade offs” Model-based Previously reported data used as model input Flexible, but relies on high quality data * All require data from comparative trials

The REACTIV Trial Ratcliffe, Br J Surg 2006 l 246 patients extensive vv and saphenous reflux randomized to l Conservative measures (n = 122) l Saphenous stripping / phlebectomy (n = 124) l 24 mo cost effectiveness of £4682 per QALY gained l Below NHS threshold of £20,000 per QALY ConservativeSurgery Mean Difference Mean NHS Cost £344.53£733.10£ AUC SF-6D ICER * £4682 * Incremental cost effectiveness ratio

Conclusions l The questions are important and need prioritization l The goals, not the methods, are most important l Precise estimates of harms, risks, and benefits l Minimizing bias and unknown confounders l Every question requires a comparison group l An RCT is not necessary, feasible, or even desirable for every question But …

Developing Phlebology as a Clinical Science l Demands for industry l Clinical evidence prior to marketing l Research with patient important endpoints l Demands for ourselves l Avoid herd mentality in the absence of data l Pay attention to costs to the health care system l Consider comparative effectiveness of technology l Demands for phlebology l Raise the bar for presentation / publication l Fellowships in epidemiology & health systems research