PROMISing development for standardising the assessment of self-reported health worldwide? An exploration of the potential application of U.S. developed patient-reported outcome measures to Germany

点击翻译按钮获取中文摘要

Patient-reported outcomes (PROs) are of growing importance in the assessment of benefits and harms of medical interventions. As observers (proxy) do not accurately capture the patient perspective, regulatory authorities worldwide recommend the application of PROs in clinical trials. However, for these data to be used for comparative purposes across studies and crucial subgroups, it is imperative to develop standards for the assessment of self-reported health. Recent publications in this field recommend crosswalk tables or a calibration of different instruments on the same, a common, metric, which is a ground-breaking step towards standardised assessment of PROs. A major pioneer in the context of standardised PRO assessment is the Patient-Reported Outcomes Measurement Information System (PROMIS®) initiative, launched in 2004 by the U.S. National Institutes of Health. PROMIS is based on the idea that items measuring the same construct can be merged in one item bank and calibrated on the same metric. Since its launch, PROMIS researchers have developed and largely validated over 50 item banks. In 2012, PROMIS started expanding its efforts to other countries, which is a unique opportunity to standardise PROs worldwide. However, it comes with a range of challenges that need to be explored before item banks can be launched globally. These challenges relate to (1) cultural adequacy, (2) unidimensionality and generalisability, (3) transferability of U.S. item parameters to other language versions, and (4) applicability of reference data based on the U.S. general population for norming purposes. As a result, this proposal aims at a thorough testing of PROMIS items, using the PROMIS Physical Function (PF) Item Bank as an example. Specifically, to answer research question (1), the PF items are tested qualitatively. To explore (2) and (3), PROMIS items are tested in n=600 patients from three clinical samples (rheumatology, cardiology, psychosomatic medicine). First, confirmatory factor analyses explore unidimensionality of the construct. Second, tests of differential item functioning (DIF) explore generalisability (disease group-DIF), and third, new item parameter estimates derived from the German data explore transferability of the U.S. item parameters to the translated version (language-DIF). Finally, to answer (4), reference data from the German general population are obtained (n=1,000). The overarching aim of the present project is to make a fundamental contribution to answering some of the remaining psychometric challenges involved with the translation and cultural adaptation of items aimed at measuring self-reported health. This study shall make an important contribution to the standardisation of PRO assessment not only in Germany, but it shall contribute to the international body of knowledge of the assessment of subjective well-being.