Selection bias in clinical trials.

Abstract

Of 90 patients with intermediate or high-grade sarcoma eligible for a randomized trial of adjuvant doxorubicin (Adriamycin, Adria Laboratories, Columbus, Ohio), 48 were not entered: 24 (27%) by physician's choice and 24 refused randomization. Sixty-five percent of lower stage patients were randomized compared with 37% of those with higher stage (P = .02). Patients with extremity lesions were more frequently offered participation in the study (P = .07). Patients with lower stage lesions accepted randomization more readily than those with higher stage lesions (P = .01). As predicted by the higher stage and percentage of central lesions, the disease-free survival of nonrandomized patients was inferior to that of randomized patients (P = .15). Thus, patients at high risk appeared to avoid randomization and adjuvant doxorubicin in this trial, resulting in an inferior disease-free survival for the nonrandomized control group. Important questions generally require randomized trials that reliably determine relative treatment differences. If, however, the patients in a clinical trial are not representative of the entire patient population because of patient and physician selection biases, the generalizability of the results to the entire patient population may be compromised. For example, the prognosis of the general population cannot necessarily be inferred from the selected group in the study. In this study, the randomized and nonrandomized series yielded differing conclusions regarding treatment efficacy, even when an adjustment was made for known prognostic facts.

Selection of Patients may Limit the Generalizability of Results from Cancer Trials

“…However, for most trials, no records concerning non-included eligible and ineligible patients are kept (3), and assessments of the generalizability of trial results are thus not possible or at least limited. Better survival for trial patients as compared to eligible non-participants has, however, been demonstrated in a nephroblastoma trial (12), a sarcoma trial (1), and a pooled analysis of several non-small cell lung cancer trials (11). These observations are supported by our analysis.…”

“…The pretreatment characteristics of randomized and excluded patients may differ substantially with respect to clinical variables that potentially affect outcome (29). Patients who are willing to be entered into clinical trials differ from those who refuse randomization (1,5,6,30), as con rmed by our study. However, for most trials, no records concerning non-included eligible and ineligible patients are kept (3), and assessments of the generalizability of trial results are thus not possible or at least limited.…”

“…Secondarily, it may be of interest to know what proportion of all available patients with the malignancy studied was in fact eligible for the trial. It is often stated that patients enrolled in a trial have a more favourable outcome than those not included, irrespective of the speci c treatment that is evaluated (1,11,12). Even though trial patients and non-trial patients receive the same standard treatment, differences in outcome are observed.…”

“…Generalizability may be reduced if a large number of patients seen in daily practice are ineligible, or if many eligible patients are not entered in the study (1). Eligible patients may refuse to be randomized because they prefer one therapy to another, or the physician may omit randomization because of some treatment preference.…”

“…Eligible patients may refuse to be randomized because they prefer one therapy to another, or the physician may omit randomization because of some treatment preference. This is sometimes referred to as non-consent bias (1). Often the number of patients included in a randomized trial is thus only a limited proportion of eligible patients (2)(3)(4)(5)(6)(7)(8).…”