Patient noncompliance with study regimens may greatly impact the interpretation of results from randomized clinical trials. Although much has been written about issues related to noncompliance, there has been little done to provide an overall framework for designing efficacy trials in a way that accommodates possible noncompliance. This paper is intended to helpful this gap. First, decisions directly related to the randomized treatment comparison, such as sample size and the number of study arms, are discussed. Next, the secondary or observational component of the design involving compliance monitoring is examined. Alternative methods for monitoring compliance, with their relative advantages and disadvantages, are reviewed. Subsampling of patients to be monitored, and the construction of a compliance score, are also addressed. Much work remains in developing and understanding methods of analysis that incorporate compliance data; nevertheless, compliance information may be crucial for optimizing the use of drugs in clinical practice and in developing more effective therapeutic regimens.
FoulkesMAEllenbergSS. Large, simple trials of HIV therapies. In: FinkelsteinDMSchoenfeldDA, Eds. AIDS Clinical Trials.New York, NY. Wiley-Liss, Inc.; 1995.
5.
PullarT. Compliance with drug therapy. Br J Clin Pharmacol.1991;32:535–539.
6.
EllenbergSSFinkelsteinDMSchoenfeldDA. Statistical issues arising in AIDS clinical trials (with comments and rejoinder). J Am Stat Assoc.1992; 87:562–583.
7.
FreedmanLS. The effect of partial noncompliance on the power of a clinical trial. Control Clin Trials.1990;11:157–168.
8.
LachinJM. Introduction to sample size determination and power analysis for clinical trials. Control Clin Trials.1981;2:93–114.
9.
LachinJMFoulkesMA. Evaluation of sample size and power for analysis of survival with allowance for nonuniform patient entry, losses to follow-up, noncompliance, and stratification. Biometrics.1986; 42:507–519.
10.
PledgerGW. Compliance in clinical trials: Impact on design, analysis, and interpretation. In: SchmidtDLeppikIE, Eds. Compliance in Epilepsy.Amsterdam: Elsevier Science Publishers BV; 1988.
11.
EisenSAMillerDKWoodwardRSSpitznagelEPryzbeckTR. The effect of prescribed daily dose frequency on patient medication compliance. Arch Intern Med.1990;150:1881–1884.
12.
TaggartAJJohnsonGDMcDevittDG. Does the frequency of daily dosage influence compliance with digoxin therapy?. Br J Clin Pharmacol.1991;1: 31–34.
13.
LangJ. The use of a run-in to enhance compliance. Stat Med.1990;9:87–95.
14.
BrittainEWittesJ. The run-in period in clinical trials: The effect of misclassification on efficiency. Control Clin Trials.1990;11:327–338.
15.
FletcherSWPappiusEMHarperSJ. Measurement of medication compliance in a clinical setting. Arch Intern Med.1979;139:635–638.
16.
CramerJA. Overview of methods to measure and enhance patient compliance. In: CramerJASpilkerB, Eds. Patient Compliance in Medical Practice and Clinical Trials.New York, Raven Press; 1991.
17.
KruseWNikolausTRampmaierJWeberESchlierfG. Actual versus prescribed timing of lovastatin doses assessed by electronic compliance monitoring. Eur J Clin Pharmacol.1993;45:211–215.
18.
YoungLMHaakensonCMLeeKKvan EeckhoutJP. Riboflavin use as a drug marker in Veterans Administration cooperative studies. Control Clin Trials.1984;5:497–504.
19.
HardyEKumarSPeakerSFeelyMPullarT. A comparison of a short half-life marker (low-dose isoniazid), a long half-life pharmacological indicator (low-dose phenobarbitone) and measurements of a controlled release ‘therapeutic drug’ (metroprolol, Meteros) in reflecting compliance by volunteers. Br J Clin Pharmacol.1990;30:437–441.
20.
PullarTKumarSChrystynHRicePPeakerSFeelyM. The prediction of steady-state plasma phenobarbitone concentrations (following low-dose phenobarbitone) to refine its use as an indicator of compliance. Br J Clin Pharmacol.1991;32:329–333.
21.
RichardsonDLiouS-HKahnJO. Uric acid and didanosine compliance in AIDS clinical trials: An analysis of AIDS Clinical Trials Group protocols 116A and 116B/117. J Acquir Immune Defic Syndr.1993;6:1212–1223.
22.
PullarTKumarSTindallHFeelyM. Time to stop counting the tablets?. Clin Pharmacol Ther.1989; 46:163–168.
23.
KishL. Survey Sampling.New York, NY, John Wiley & Sons; 1965.
24.
Coronary Drug Research Group.Influence of adherence to treatment and response of cholesterol on mortality in the Coronary Drug Project. N Engl J Med.1980;302:1038–1041.
25.
PeduzziPWittesJDetreKHolfordT. Analysis as-randomized and the problem of nonadherence: An example from the Veterans Affairs randomized trial of coronary artery bypass surgery. Stat Med.1993; 12:1185–1195.
26.
LagakosSWLimLL-YRobinsJM. Adjusting for early treatment termination in comparative clinical trials. Stat Med.1990;9:1417–1424.
27.
MarkSDRobinsJM. A method for the analysis of randomized trials with compliance information: An application to the multiple risk factor intervention trial. Control Clin Trials.1993;14:79–97.
28.
RobinsJMTsiatisAA. Correcting for non-compliance in randomized trials using rank preserving structural failure time models. Comm Stat—Theory Methods.1991;20:2609–2631.
29.
EfronBFeldmanD. Compliance as an explanatory variable in clinical trials. J Am Stat Assoc.1991;86: 9–26.
30.
AlbertJMDeMetsDL. On a model-based approach to estimating efficacy in clinical trials. Stat Med.1995;13:2323–2335.
31.
NewcombeRG. Explanatory and pragmatic estimates of the treatment effect when deviations from allocated treatment occur. Stat Med.1988;7:1179–1186.
32.
SommerAZegerSL. On estimating efficacy from clinical trials. Stat Med.1991;10:45–52.
33.
DixonDOAlbertJM. Issues in the design and analysis of AIDS clinical trials. In: ThallPF, Ed. Recent Advances in Clinical Trial Design and Analysis.Norwell, MA: Kluwer Academic Publishers; 1995.
34.
LeeYJEllenbergJHHirtzDGNelsonKB. Analysis of clinical trials by treatment actually received: Is it really an option?. Stat Med.1991;10:1595–1605.
35.
MayGSDeMetsDLFriedmanLM. The randomized clinical trial: Bias in analysis. Circulation.1981;64:669–673.
