Systematic reviews with time-to-event (TTE) analyses provide fundamental insights in many areas of research. TTE data indicate not only whether events (e.g. deaths) have occurred, but also how much time has elapsed between the start of a study and the occurrence of an event. Well-known examples are survival curves, which show how the proportion of patients who survive changes over time, or hazard ratios, which compare the mortality rates of two groups.
Analyses from 100 systematic reviews evaluated
TTE analyses are methodologically challenging. Previous research has already highlighted shortcomings in the presentation of TTE results and analyses in clinical trial publications. In this paper, the authors for the first time evaluated the characteristics and methods of meta-analytic TTE analyses in systematic reviews. To this end, they used a systematic search to identify 50 Cochrane reviews with meta-analyses based on hazard ratios and 50 other systematic reviews from the most important clinical journals.
Their evaluation shows that current systematic reviews containing meta-analyses based on hazard ratios provide very different and often inadequate information on TTE characteristics and methods. In a previous publication, the authors had already found this for publications on individual clinical trials. The problem therefore persists in systematic reviews.
The authors' conclusion: review authors should consistently use the available methodological guidelines for TTE meta-analyses. Moreover, additional reporting standards for TTE analyses in systematic reviews can improve the quality of the reviews and thus the usability of their results.