The paper by Kay Dickersin, Roberta Scherer, and Carol Lefebvre examines the sensitivity and precision of Medline searches for randomized clinical trials. The study compares the results of Medline searches to a "gold standard" of known randomized clinical trials in ophthalmology published in 1988. The authors also conduct a systematic review of similar studies from various fields of medicine. The main outcome measures are sensitivity (proportion of known trials identified) and precision (proportion of retrieved publications that are actually randomized clinical trials). The sensitivity of searching for ophthalmology randomized clinical trials published in 1988 was 82% when the gold standard included any journal, 87% for Medline-indexed journals, and 88% for selected Medline journals. Weighted means for sensitivity across all studies were 51%, 77%, and 63%, respectively. The weighted mean for precision was 8% (median 32-5%). The authors conclude that while Medline indexing terms have improved, sensitivity remains unsatisfactory. They recommend a mechanism to "register" known trials, preferably by retrospective tagging of Medline entries, and incorporating trials published before 1966 and in journals not indexed by Medline. They also suggest improving terminology, indexing, and search strategies to enhance retrieval of randomized clinical trials through Medline.The paper by Kay Dickersin, Roberta Scherer, and Carol Lefebvre examines the sensitivity and precision of Medline searches for randomized clinical trials. The study compares the results of Medline searches to a "gold standard" of known randomized clinical trials in ophthalmology published in 1988. The authors also conduct a systematic review of similar studies from various fields of medicine. The main outcome measures are sensitivity (proportion of known trials identified) and precision (proportion of retrieved publications that are actually randomized clinical trials). The sensitivity of searching for ophthalmology randomized clinical trials published in 1988 was 82% when the gold standard included any journal, 87% for Medline-indexed journals, and 88% for selected Medline journals. Weighted means for sensitivity across all studies were 51%, 77%, and 63%, respectively. The weighted mean for precision was 8% (median 32-5%). The authors conclude that while Medline indexing terms have improved, sensitivity remains unsatisfactory. They recommend a mechanism to "register" known trials, preferably by retrospective tagging of Medline entries, and incorporating trials published before 1966 and in journals not indexed by Medline. They also suggest improving terminology, indexing, and search strategies to enhance retrieval of randomized clinical trials through Medline.