Evidence-based practice in pediatric surgery

Background/Purpose: The current medical environment demands the provision of quality healthcare at an affordable cost. Both payers and regulators are committed to lowering cost through initiation of best practice strategies that include practice guidelines, clinical pathways, and standards of care. The on ly practical way to join this debate is th rough the use of objective, unbiased clinical data. This study was undertaken to review the current slate of the pediatric surgery literature and its value in determining best clinical practice. Methods: The National Library of Medicine Medline database was accessed using the Ovid Internet client software. All references, abstracts, and keyword indexes from the core pediatric surgery literature, the Journal of Pediatric Surgery, the European Journal of Pediatric Surgery, Pediatric Surgery international, Zeitschrift fur Kinderchirurgie, and Seminars in Pediatric Surgery were downloaded and reviewed. Search criteria were defined to identify prospective, randomized, controlled studies. References were then categorized as case reports; retrospective case series; prospective case series; randomized, controlled studies; laboratory studies; review articles; or miscellaneous studies. Results: As of March 1, 1998, there are 9,373 references, excluding citations of letters or comments, contained in the core pediatric surgery literature, as provided through Medline. Of these, 485 were identified as studies for review, possible prospective case series or prospective, randomized, controlled studies. After review, 34 studies (0.3%) were classified as prospective, randomized, controlled studies, whereas 139 (1.48%) were classified as prospective studies. There were 3,241 (34.6%) case reports, 5,619 (59.9%) retrospective case series, 1,109 (11.8%) laboratory studies, 195 (2.1%) review articles, and 36 (0.3%) miscellaneous studies that did not fit into other categories. When analyzed by decade of publication, prospective studies and prospective, randomized, controlled studies (n = 173) numbered 103 in the 1990s, 63 in the 1980s, and seven in the 1970s. Conclusions: There is a paucity of scientifically rigorous data on which to base clinical practice in pediatric surgery. The increasing numbers of prospective, case-controlled studies or the more sou nd prospective, randomized, controlled trials in the 1990s suggests that pediatric surgeons are aware of the need to generate unbiased data to support current clinical practice and the development of practice guidelines. Limitations exist in conducting prospective, randomized, controlled trials because of the rare nature of many pediatric surgical conditions and the lack of clinical "equipoise" over available treatment options. The authors encourage the use of multiinstitutional trials and the prospective, randomized, controlled study methodology to develop data that can be used to guide clinical practice in our evolving healthcare environment. Copyright (C) 1999 by W.B. Saunders Company.