Commentary: The Journal of Pediatric Psychology Should Adopt the CONSORT Statement as a Way of Improving the Evidence Base in Pediatric Psychology

The CONSORT statement (www.consort-statement.org) was developed to improve standards of reporting of randomized clinical trials (RCTs) in medical journals. It currently includes 22 items used by reviewers and editors to ensure consistent and full reporting of the results of an RCT. It also aids researchers in the plan and execution of RCTs. The Journal of the American Medical Association was one of the first journals to adopt the CONSORT items as guidelines to ensure quality reporting of RCTs; since that time over 80 medical journals have adopted these standards for the reviewing and reporting of RCTs. Recent reviews of the quality of RCT reporting have demonstrated that the introduction of the CONSORT statement has a demonstrable impact on the completeness of reporting in numerous medical journals (Moher, Jones, & Lepage, 2001). Neither psychology as a field nor subspecialty areas of psychology have yet adopted CONSORT guidelines for the review and publication of RCTs. Division 12 (Society of Clinical Psychology) Task force on Promotion and Dissemination of Psychological Procedures of the American Psychological Association (Chambless et al., 1996) began reviewing evidence on psychological interventions and enjoined researchers to report their RCT results completely, so that evidence-based recommendations could be forwarded. Yet perennial concerns about the quality of reporting in psychology of evidence-based material available for review remain (Chambless & Ollendick, 2001). As outlined in the accompanying article (Stinson et al., this issue), most of the clinical trials published in the Journal of Pediatric Psychology( JPP) already meet some of the 22 CONSORT criteria. However, many important items are reported only infrequently or not at all. In fact, of the 22 items, 14 items were identified as reported in less than 25% of the trials in JPP between 1998 and 2001. Although this seems troubling, other psychology and behavioral medicine journals likely fare no better, and may fare worse (Davidson et al., in press). The CONSORT statement was devised for medical interventions and thus might be considered inappropriate for the reporting of psychological interventions. However, as noted by Stinson and others, all CONSORT items can be reported by psychosocial interventions, and only one item—successful blinding of treatment provider and participants—is difficult to attain. Interestingly, the reporting of the blinding status is attainable by psychosocial interventions; the presence or absence of the report of blinding meets or does not meet the CONSORT criteria, not the success of these procedures. It is of course possible, and highly desirable, to successfully blind the major assessor of the outcome of the RCT, and the test of the success of the blinding plan for assessors is not regularly reported in psychosocial RCTs. The CONSORT guidelines are strongest in reporting on key threats to internal validity; however, the CONSORT guidelines have been criticized for their relative inattention to external validity (Glasgow, McKay, Piette, & Reynolds, 2001). Only one vague item of the CONSORT asks for information on generalizability. On the other hand, 21 items request concrete, tangible report of internal validity issues. Some have argued for the augmentation of the CONSORT guidelines, with reporting criteria focused more exclusively on external validity issues (Glasow et al.). The representativeness of the setting, the sustainability of the intervention from the point of view of the caregiver and the patient, will effect eventual practice change and should be essential for effectiveness studies. Moreover, the CONSORT criteria do not address clinical versus statistical significance. In addition, the CONSORT criteria, though reasonable, are not sufficient to cover the unique reporting needs for psychosocial interventions. The Evidence-Based Be-