Reply to letter from I. Ekman, K. Swedberg, and M. Böhm

We thank Drs Ekman, Swedberg, and Böhm for their comments related to the search strategy used in our recent meta-analysis published in the European Journal of Heart Failure that examines the role of disease-specific health status as a predictor of mortality in patients with heart failure.1

We have been able to establish why the Systolic Heart failure treatment with the I_f inhibitor ivabradine Trial (SHIFT) study was not identified in our original search. The original search was conducted as follows: (quality of life OR health status OR self-rated health OR functional status OR functioning) AND (hospitalization OR readmission OR mortality OR survival OR prognosis) AND (heart failure OR CHF), with ‘hospitalization’ as ‘MeSH’ term and other search terms using ‘Title/Abstract’. With this strategy, we did not identify the SHIFT study. As the SHIFT study is of major importance, and given that it is one of the largest studies to date to have used the Kansas City Cardiomyopathy Questionnaire (KCCQ), based on the comment of Ekman and colleagues we have re-examined the results with an adapted search strategy using ‘All fields’ (instead of ‘Title/Abstract’ or ‘MeSH’) while adding ‘death*’, ‘hospital*’, and ‘admission’ to the original search.

This new search strategy resulted in 5382 hits (1980 to May 2014) instead of the original 1420 hits (1980 to March 2013). After checking all 5382 hits on title and abstract, we extracted 190 hits which were evaluated using full text. This resulted in the identification of 25 articles already listed in our paper and four additional articles (including the SHIFT study) that seem to fulfil all inclusion criteria,2-5 but were not found with the original search.

Unfortunately, we cannot update the meta-analysis based solely on the information provided in these four articles. We would first need to ask all authors to rerun their adjusted Cox regression analysis on all-cause mortality with the health status variable dichotomized using the evidence-based cut-off value. Hence, we also cannot make any inferences about the potential impact of these studies on the pooled effect size in our meta-analysis, although the results of two out of four studies (including the SHIFT study) seem to support the conclusion of the meta-analysis that poor disease-specific health status is associated with an increased risk of mortality in patients with heart failure.