Use of existing systematic reviews for evidence assessments in infectious disease prevention: a comparative case...

Given limited resources and time constraints, the use of existing systematic reviews (SR) for the development of evidence-based public health recommendations has become increasingly important. Recently, a five-step approach for identifying, analyzing, appraising and using existing SRs based on recent guidance by the US Agency for Healthcare Research and Quality (AHRQ) was proposed within the Project on a Framework for Rating Evidence in Public Health (PRECEPT). However, case studies are needed to test whether this approach is useful, what challenges arise and how problems can be solved.

In two case studies, the five-step approach was applied to integrate existing SRs in the development of evidence-based public health recommendations. Case study A focused on the role of neonatal sepsis as a risk factor for adverse neurodevelopmental outcome. Case study B examined the efficacy, effectiveness and safety of influenza vaccination during pregnancy. For each step, we report the approach of the review team, discuss challenges and describe solutions.

For case study A, one existing SR was identified, while in case study B four SRs were eligible for analysis. We found that comparison of inclusion criteria alone was sufficient to judge on relevance of SRs in case study A, but not B. Although methodological quality of all identified SRs was acceptable, risk of bias assessments of individual studies included in the SRs had to be repeated in both case studies. Particular challenges appeared in case study B where multiple SRs addressed the same research question. With the help of spreadsheets comparing the characteristics of the existing SR we decided to use the most comprehensive one for our evidence synthesis and supplemented the results with those from the other SRs.

In both case studies using the complete SR was not possible. The five-step approach provided useful and structured guidance and should be routinely applied when using existing SRs as a basis for evidence-based recommendations in public health. In situations where more than one SR has to be considered, the development of spreadsheets comparing characteristics, inclusion criteria, risk of bias, included studies and outcomes seems useful.