Conclusiveness of the Cochrane Neonatal Reviews: a systematic analysis

Abstract

Aim: To assess the conclusiveness of the Cochrane Neonatal Reviews (CNRs). We tested the hypotheses that: 1) the majority of the reviews is inconclusive; 2) the majority of reviews recognizes the need for further studies; 3) the ability to reach a conclusion is dependent upon both the number of studies and the number of patients. We also aimed to determine whether the conclusiveness of the CNRs was affected by time.

Methods: We selected CNRs available in the Cochrane Library in June 2004. The number of randomized clinical trials (RCTs) found, number of RCTs included for analysis, number of patients enrolled, the stated need for further studies, and the conclusiveness of CNRs were recorded.

Results: Out of 170 CNRs, 67.7% were conclusive. The average number of articles was similar, but the total number of patients enrolled was three times higher in the conclusive CNRs. The percentage of articles included in conclusive studies was significantly higher than in inconclusive ones. The vast majority of CNRs recognized the need for further studies. The number of studies included correlated significantly with the total number of patients included. The percentage of conclusive CNRs correlated negatively with year of publication.

Conclusion: The majority of CNRs is conclusive, but emphasizes the need for further studies. The ability of a CNR to reach a conclusion is affected by the cumulative sample size and by the number of studies performed. The probability of a newer review to be conclusive is lower than that of an older review.