Resolving Ethical Issues in Randomised Controlled Trials

• Published date: 21 May 2012
• DOI: https://doi.org/10.1108/S1474-7863(2012)0000013009
• Pages: 95-127
• Date: 21 May 2012
• Author: Robert F. Boruch,Joe S. Cecil,Herb Turner,Timothy Victor,Jordan M. Hyatt

RESOLVING ETHICAL ISSUES IN

RANDOMISED CONTROLLED

TRIALS

Robert F. Boruch, Joe S. Cecil,

Herb Turner III, Timothy Victor

and Jordan M. Hyatt

ABSTRACT

The chapter considers the ethical problems engendered by random

assignment and privacy concerns in randomised controlled experiments

and cluster randomised trials. The particular focus is on procedural,

legislative and technical approaches to reducing or avoiding the problems.

Examples are given from a variety of disciplines including health and

education, though the main emphasis is on research in crime and

delinquency.

INTRODUCTION

Randomised trials, like all experiments in the social sciences, ‘are exposed to

epistemological, social and ethical threats to their validity and, in turn, that

the outcomes of any experiment are open to more than one interpretation’.

Perspectives on Evaluating Criminal Justice and Corrections

Advances in Program Evaluation, Volume 13, 95–127

Copyright r2012 by Emerald Group Publishing Limited

All rights of reproduction in any form reserved

ISSN: 1474-7863/doi:10.1108/S1474-7863(2012)0000013009

95

However, they offer the best chance, given the current state of experimental

design and statistical know-how, to isolate the causal influences from

exogenous noise. In affirmation, Tukey declared that ‘the only source of

reliable evidence about the usefulness of almost any sort of therapy or

surgical intervention is that obtained from well-planned and carefully

conducted randomised yclinical trial’. Despite decades of research spent

refining the statistical methodologies that underlie randomised trials, ethical

issues persist.

Learning about the ethics of randomised controlled trials (RCTs) is not

always easy. For instance, a recent Google search using the phrase

‘randomized controlled trial ethics’ yielded 1,800,000 results in 0.07 seconds.

This chapter is intended to make it a bit easier for experimenters, trial

participants, research policy people and the thoughtful citizen to understand

the topic.

The focus here is on ethical issues encountered in RCTs and especially on

how the issues can be or might be resolved. Boundary conditions such as

law, local culture, institutional rules and government regulations are

important, of course, and considered here. This is partly because they

may embody ethical standards. Some ingredients of this chapter are

dedicated to the future. This is because the interpretation of ethics can

change with time and with new kinds of research just as the boundary

conditions may.

In generating good evidence through RCTs, the main ethical issues

usually emerge in two areas. The first concerns the random assignment

feature that is common to any randomised trial. Put simply, individual

people or entities, such as villages, crime hot spots or schools, that are

randomly allocated to different interventions might be put at a disadvantage

as a consequence of the attempt to compare interventions in a randomised

trial. The ethical and legal issues engendered by such trials, real or

ostensible, occur in many disciplines. Examples in research on crime and

delinquency, education, health, welfare and other areas are included in what

follows.

The second issue area concerns individual privacy. Here, we regard the

two arms of a randomised trial as parallel surveys. The privacy imperative

affects researchers’ efforts to elicit information in probability sample surveys

and ethnographic research, as well as in randomised trials. Privacy issues

are also pertinent in efforts to linking research records on individuals

who participate in experiments (or surveys) with administrative records

from different sources on the same individuals, a form of integrated data

system.

ROBERT F. BORUCH ET AL.96

The issue areas and approaches to resolutions are discussed seriatim. The

discussion depends heavily on actual trials. We also identify topics that

deserve more attention from people who want to advance understanding of

the topic based on empirical research at the interstices of ethics, law and

RCTs.

THE RCT

The power of experimental trials to do certain things well has been clearly

established. While the label may be something of a misnomer or misguided,

randomised experiments are often referred to as the ‘gold standard’ of

research methods within criminology, in medicine, education and other

scientific disciplines. RCTs have provided evidence regarding the effective-

ness of significant scientific advances.

Randomly assigning individuals or entities, such as organisations, to one

of two or more interventions is a key element of a randomised trial. From

this ostensibly simple process, a number of significant advantages are

derived; we speak of only two here. First, RCTs allow for the elimination of

bias in the sample by eliminating all potential systematic differences between

the groups that are involved in the interventions being compared. This

ensures that all known and unknown variables have the same expected

values and that any post-experiment differences can be attributed to the

interventions. Second, randomisation permits the measurement of uncer-

tainty within a study.

In particular, random assignment of people or entities to different

programmes avoids the need to depend on assumptions, sometimes heroic

ones, about what would have happened to people or entities in the absence

of the intervention they actually received. The random allocation to the two

or more interventions, one of which may be a control, assures fairness in the

sense of generating groups that do not differ systematically. It also permits

legitimate probabilistic statements about confidence intervals for the

differences in outcomes. Both products are important in science and, at

times, in legislative and judicial settings (Boruch & Cecil, 1979;Kaye &

Freedman, 1994).

The thoughtful ethicist might construe at least two basic types of

randomised trials, classifying them by their intention: experimental trials are

designed to test an intervention for which the outcome is unknown; and

demonstration trials are designed to illustrate the benefits of a treatment,

either in relation to an absence of treatment, or some other standard of care.

Resolving Ethical Issues in RCTs 97