Participant demographics reported in "Table 1" of randomised controlled trials: a case of "inverse evidence"?

The contents lists of four leading general medical journals were hand searched to identify 25 consecutive papers reporting RCT results in each journal (n = 100). Data on demographic characteristics were extracted from each paper's Table 1 only (or equivalent).Authors infrequently reported key demographic characteristics relating to SES of RCT participants. Age and gender of participants were commonly reported. Less than 10% reported occupational group, employment status, income or area based measures of disadvantage.Without adequate reporting of key indicators of SES in trial participants it is unclear if lower SES groups are under-represented. If such groups are systematically under-recruited into trials, this may limit the external validity and applicability of study findings to these groups. This is in spite of the higher health-care need in more disadvantaged populations. Under-representation of low SES groups could underestimate the reported effect of an intervention for those with a higher baseline risk. The marginal benefit identified in a trial with poor or no representation of lower SES participants could significantly underestimate the potential benefit to a low SES community. More transparency in this reporting and greater attention to the impact of SES on intervention outcomes in clinical trials is needed. This could be considered in the next revision of the CONSORT statement.The CONSORT statement [1] encourages authors reporting randomised controlled trial (RCT) results to include "baseline demographic and clinical characteristics for each group" of participants. This baseline information should be presented in a table, effectively becoming the "Table 1" in published reports of RCTs.Demographic, from "demography", refers to characteristics related to the "life-conditions of communities of people" [2]. Significantly, the CONSORT statement is not prescriptive as to which demographic features should be reported, giving no indication as to which are importa