An IV for the RCT: using instrumental variables to adjust for treatment contamination in randomised controlled trials

Although the randomised controlled trial is the “gold standard” for studying the efficacy and safety of medical treatments, it is not necessarily free from bias. When patients do not follow the protocol for their assigned treatment, the resultant “treatment contamination” can produce misleading findings. The methods used historically to deal with this problem, the “as treated” and “per protocol” analysis techniques, are flawed and inaccurate. Intention to treat analysis is the solution most often used to analyse randomised controlled trials, but this approach ignores this issue of treatment contamination. Intention to treat analysis estimates the effect of recommending a treatment to study participants, not the effect of the treatment on those study participants who actually received it. In this article, we describe a simple yet rarely used analytical technique, the “contamination adjusted intention to treat analysis,” which complements the intention to treat approach by producing a better estimate of the benefits and harms of receiving a treatment. This method uses the statistical technique of instrumental variable analysis to address contamination. We discuss the strengths and limitations of the current methods of addressing treatment contamination and the contamination adjusted intention to treat technique, provide examples of effective uses, and discuss how using estimates generated by contamination adjusted intention to treat analysis can improve clinical decision making and patient care.

[1]  N Segnan,et al.  Adjusting for non-compliance and contamination in randomized clinical trials. , 1997, Statistics in medicine.

[2]  I. White,et al.  Instrumental variables and interactions in the causal analysis of a complex clinical trial , 2007, Statistics in medicine.

[3]  J. Slattery,et al.  Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). 1994. , 1994, Atherosclerosis. Supplements.

[4]  David Chia,et al.  Mortality results from a randomized prostate-cancer screening trial. , 2009, The New England journal of medicine.

[5]  S. Zeger,et al.  On estimating efficacy from clinical trials. , 1991, Statistics in medicine.

[6]  B J McNeil,et al.  Does more intensive treatment of acute myocardial infarction in the elderly reduce mortality? Analysis using instrumental variables. , 1994, JAMA.

[7]  P. Glasziou,et al.  Meta-analysis adjusting for compliance: the example of screening for breast cancer. , 1992, Journal of clinical epidemiology.

[8]  M. Tobin,et al.  Mendelian Randomisation and Causal Inference in Observational Epidemiology , 2008, PLoS medicine.

[9]  Tosiya Sato.  A method for the analysis of repeated binary outcomes in randomized clinical trials with non‐compliance , 2001, Statistics in medicine.

[10]  M. Pfeffer,et al.  Adherence to candesartan and placebo and outcomes in chronic heart failure in the CHARM programme: double-blind, randomised, controlled clinical trial , 2005, The Lancet.

[11]  S. Yusuf.  MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20536 high-risk individuals: a randomised placebo-controlled trial. Commentary , 2002 .

[12]  B. G. Blijenberg,et al.  Screening and prostate-cancer mortality in a randomized European study. , 2009, The New England journal of medicine.

[13]  D. DeMets,et al.  Effect of Carvedilol on the Morbidity of Patients With Severe Chronic Heart Failure: Results of the Carvedilol Prospective Randomized Cumulative Survival (COPERNICUS) Study , 2002, Circulation.

[14]  Scandinavian Simvastatin Survival Study Group.  Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S) , 1994, The Lancet.

[15]  S. Schneeweiss,et al.  Risk of death associated with the use of conventional versus atypical antipsychotic drugs among elderly patients , 2007, Canadian Medical Association Journal.

[16]  Mark C. Wilson.  Evidence-Based Medicine: How to Practice and Teach EBM , 2001, ACP Journal Club.

[17]  S. Hulley,et al.  Delayed effects of the military draft on mortality. A randomized natural experiment. , 1986, The New England journal of medicine.

[18]  Daniel Eisenberg,et al.  Estimating the effect of smoking cessation on weight gain: an instrumental variable approach. , 2006, Health services research.

[19]  G. Guyatt,et al.  A new preference-based analysis for randomized trials can estimate treatment acceptability and effect in compliant patients. , 2006, Journal of clinical epidemiology.

[20]  W. Stolz,et al.  Differences in efficacy between intention‐to‐treat and per‐protocol analyses for patients with psoriasis vulgaris and atopic dermatitis: clinical and pharmacoeconomic implications , 2001, The British journal of dermatology.

[21]  AndrewJ. S. Coats.  MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebocontrolled trial , 2002, The Lancet.

[22]  Heejung Bang,et al.  On estimating treatment effects under non‐compliance in randomized clinical trials: are intent‐to‐treat or instrumental variables analyses perfect solutions? , 2007, Statistics in medicine.

[23]  Sander Greenland,et al.  Estimating effects from randomized trials with discontinuations: the need for intent-to-treat design and G-estimation , 2008, Clinical trials.

[24]  J. Manson,et al.  Calcium plus vitamin D supplementation and the risk of fractures. , 2006, The New England journal of medicine.

[25]  A. Walker,et al.  Drug Copayment and Adherence in Chronic Heart Failure: Effect on Cost and Outcomes , 2006, Pharmacotherapy.

[26]  Bartolome Celli,et al.  Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. , 2007, The New England journal of medicine.

[27]  Ross T Tsuyuki,et al.  A meta-analysis of the association between adherence to drug therapy and mortality , 2006, BMJ : British Medical Journal.

[28]  G. Imbens,et al.  Analyzing a randomized trial on breast self-examination with noncompliance and missing outcomes. , 2004, Biostatistics.

[29]  Douglas G Altman,et al.  Better reporting of randomised controlled trials: the CONSORT statement , 1996, BMJ.

[30]  J. Heckman,et al.  Use of instrumental variables in the presence of heterogeneity and self-selection: an application to treatments of breast cancer patients. , 2007, Health economics.

[31]  Jeanne S. Ringel,et al.  Can Higher Cigarette Taxes Improve Birth Outcomes? , 1997 .

[32]  Graham Dunn,et al.  Estimating treatment effects from randomized clinical trials with noncompliance and loss to follow-up: the role of instrumental variable methods , 2005, Statistical methods in medical research.

[33]  G. Dunn.  Estimating the causal effects of treatment , 2002, Epidemiologia e Psichiatria Sociale.

[34]  P. Glasziou,et al.  Cochrane Systematic Review of Colorectal Cancer Screening Using the Fecal Occult Blood Test (Hemoccult): An Update , 2008, The American Journal of Gastroenterology.

[35]  J. Stengård,et al.  Antibodies to glutamic acid decarboxylase as predictors of insulin-dependent diabetes mellitus before clinical onset of disease , 1994, The Lancet.

[36]  Joshua D. Angrist,et al.  Identification of Causal Effects Using Instrumental Variables , 1993 .

[37]  Brett Hanscom,et al.  Surgical versus nonsurgical treatment for lumbar degenerative spondylolisthesis. , 2007, The New England journal of medicine.

[38]  John D Seeger,et al.  Aprotinin during coronary-artery bypass grafting and risk of death. , 2008, The New England journal of medicine.

[39]  D. Rubin.  INFERENCE AND MISSING DATA , 1975 .