Ann Intern Med 2002 Jul;137(1):10-6
Yale University School of Medicine, Yale-New Haven Hospital Center for Outcomes Research and Evaluation, PO Box 208025, 333 Cedar Street, New Haven, Connecticut 06520, USA.
Background: A common criticism of randomized, controlled trials (RCTs) is that the enrollment process may be highly selective and those who enroll may not represent persons in the general population. The recruitment process reported in published RCTs has not been systematically evaluated.
Objective: To determine whether published RCTs report information about how their study sample was assembled and to describe the proportion of potential study participants who were actually enrolled.
Design: Cross-sectional explicit review of RCTs published in four high-impact medical journals between 1 April 1999 and 1 April 2000. All RCTs involved interventions in humans.
Measurements: The number of persons who were screened for eligibility, the number who were eligible, and the number who were enrolled in each RCT.
Results: A total of 172 RCTs were reviewed. Ninety (52%) reported the number of persons who were evaluated by the investigators for eligibility, and 74 (43%) reported the number of persons who were actually eligible for participation. Of the studies that reported quantitative recruitment information, the median proportion of screened persons who were eligible for participation was 65% (interquartile range, 41% to 82%) and the median proportion of eligible persons who enrolled was 93% (interquartile range, 79% to 100%). Some trials reportedly enrolled every person screened for eligibility; others screened as many as 68 people for each person finally enrolled.
Conclusions: Many RCTs published in major medical journals do not provide information about the patient recruitment process. As a result, it is difficult for readers to gauge the extent to which participants may represent a highly selected subgroup.