Cochrane Methods

Methodology Register

Background: Health-related quality of life (HRQoL) has become an important patient relevant outcome measure, which increasingly is used in clinical trials. One step further is to assess HRQoL as an outcome measure in systematic reviews. So far, there is little guidance on how to incorporate HRQoL in these documents. We aimed to assess the current state of including HRQoL as an outcome measure and the quality of reporting results on HRQoL in Cochrane reviews. Objectives: To assess the extent and quality of reporting HRQoL as an outcome in Cochrane reviews. Methods: We aimed to identify all reviews presenting data on HRQoL. We searched The Cochrane Database of Systematic Reviews of Issue 1, 2007 of The Cochrane Library using different search terms for HRQoL to identify reviews addressing HRQoL. In a second step we selected reviews which included HRQoL as an outcome measure. These reviews were examined further regarding the reporting and analysis of HRQoL. Results: Out of the 2997 reviews in the database, 1718 reviews (57% of the total database) were retrieved through the search strategy. 283 (9% of the reviews in the database) listed HRQoL as an outcome measure in the methods section. Out of these, 93 did not report on HRQoL, i.e. there was no further information on whether HRQoL data were available from primary studies. 34 of these 93 reviews did not include any study. 8 reviews did not report on HRQoL but listed HRQoL as an outcome measure in at least one primary study in the "Characteristics of Included Studies" table. 182 reviews (65% out of 283 reviews listing HRQoL as an outcome measure) reported on HRQoL. Out of these, 91 reviews stated that none of the included primary studies provided HRQoL data. Of the 90 reviews with results on HRQoL in the primary studies, 88 reviews presented results on the basis of single studies, in 12 reviews a pooling of the results in a meta-analysis was performed and only 4 reviews gave a statement on clinical relevance of the results presented. 1 review was inconsistent with regard to information on HRQoL by stating there were no data on HRQoL in primary studies but presenting results in the "Characteristics of Included Studies" table. The presentation of the results on HRQoL of the 90 reviews with primary studies reporting HRQoL was in general poor, since statements on questionnaires used, validation status of questionnaires as well as scaling and clinical relevance of results was missing in most cases. In a few reviews the outcome reported by the authors as quality of life was questionable since in these cases this seemed not to represent the multidimensional approach which is usually required for quality of life. Conclusions: In Cochrane reviews the reporting of HRQoL as an important patient relevant outcome measure is in general poor. To support authors and to improve the usability of information on HRQoL from systematic reviews, the definition of a HQRoL specific core data set (e.g. information on questionnaires used, validation status of questionnaires, scaling and scoring, definition of clinically relevant changes) might be helpful. Furthermore, specific approaches to quality assessment of HRQoL studies, data presentation and integration of data from the variety of questionnaires need to be discussed.