Measuring Health-Related Quality of Life in Randomised Controlled Trials: Expected and Reported Results Do Not Match

Felicitas Wiedemann,1,2 Franz Porzsolt2 1Department of General and Visceral Surgery, Stuttgart, 70176, Germany; 2Institute of Clinical Economics (ICE) e.V., Ulm, 89081, GermanyCorrespondence: Felicitas Wiedemann, Department of General and Visceral Surgery, Diakonie-Klinikum Stuttgart, Markelstraße 44, Stuttgart, Baden-Württemberg, Germany, Tel +491781731296, Email felicitas.wiedemann@gmail.comPurpose: The assessment of health-related quality of life (hrQoL) may need to be reconsidered due to important differences between efficacy (the effect of a treatment under experimental study conditions) and effectiveness (the effect of a treatment under real-world conditions). We presume that most researchers intend to describe effects under real-world conditions when investigating hrQoL as an endpoint. Unfortunately, most studies are designed to confirm two theories: the efficacy of a new intervention under experimental study conditions and the real-world effectiveness of this intervention on hrQoL under non-experimental study conditions. Conflicting information emerges when the outcomes are supposed to describe effects under real-world conditions, but the assessment generates results obtained under experimental conditions. This paper examines the existing conflict between efficacy and effectiveness in a sample of 100 studies investigating hrQoL.Methods: We analysed a sample of freely available publications of clinical studies listed in PubMed between April 2015 and August 2016 which assessed quality of life as an outcome. We assessed the following four characteristics that should differ in studies measuring either efficacy or effectiveness: 1) specification of the study as a randomised controlled trial or not, 2) description of the study design as pragmatic or not, 3) classification of the study as an efficacy or an effectiveness study and 4) number of selected inclusion and exclusion criteria.Results: 91% of the studies assessed hrQoL under experimental conditions (in a r