Assessment of the current state of the consideration of specific features of time-to-event analyses in meta-analyses: re-analysis of systematic reviews and implementation of guidance

• Applicant: Professorin Dr. Nicole Skoetz
• Subject Area: Epidemiology and Medical Biometry/Statistics ; Public Health, Healthcare Research, Social and Occupational Medicine
• Term: from 2020 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 437030575

The measurement and comparison of event times in clinical research is done by means of time-to-event analysis, so for example in the assessment of overall survival. These analyses are challenging and contain specific methodical features. Their results are therefore not always straight forward to interpret and evaluate. While extensive guidance on time-to-event analysis at the study level and in presence of individual patient data is available, guidance for meta-analyses based on aggregated data, which are increasingly used in systematic reviews and guidelines, is lacking. Moreover, existing guidance is primarily focused on recovering result data to pool effect estimates while leaving the certainty in the included estimators largely untouched. Due to the particular difficulties and features of time-to-event analyses, however, the absence of guidance introduces a considerable threat to the quality of reviews when existing uncertainties are not taken into account or reported adequately. The main objective of this project is to improve the application and reporting of time-to-event meta-analyses sustainably. Therefore, this project includes an assessment of systematic reviews and their included studies to elicit how review authors currently deal with existing problems in primary studies when conducting time-to-event meta-analyses and deriving conclusions. This will then form a basis for future improvement efforts. For this purpose, the effect estimates and characteristics of time-to-event analyses from primary studies included in time-to-event meta-analyses of eligible systematic reviews will be examined. This then allows to evaluate whether and how the review authors consider methodological features and existing problems in their systematic reviews. The assessment will include, amongst other things, a re-analysis of the included time-to-event meta-analyses. In order to actively improve the quality of systematic reviews and to avoid future mistakes, the results of the systematic assessment will then be used to inform the development of targeted guidance materials. To enhance the user friendly design of the guidance, the development process will include multiple consultation rounds with review authors and guideline developers. The findings of this project are intended to provide a foundation for future initiatives that aim to improve and strengthen the quality of systematic reviews and guidelines which include time-to-event meta-analyses.

• DFG Programme: Research Grants