Public release of hospital quality data for referral practices in Germany: results from a cluster-randomised controlled trial

OBJECTIVE: To evaluate the impact of different dissemination channels on the awareness and usage of hospital performance reports among referring physicians, as well as the usefulness of such reports from the referring physicians’ perspective. DATA SOURCES/STUDY SETTING: Primary data collected from a survey with 277 referring physicians (response rate = 26.2%) in Nuremberg, Germany (03–06/2016). STUDY DESIGN: Cluster-randomised controlled trial at the practice level. Physician practices were randomly assigned to one of two conditions: (1) physicians in the control arm could become aware of the performance reports via mass media channels (Mass Media, [Formula: see text] =132, [Formula: see text] =147); (2) physicians in the intervention arm also received a printed version of the report via mail (Mass and Special Media, [Formula: see text] =117; [Formula: see text] =130). PRINCIPAL FINDINGS: Overall, 68% of respondents recalled hospital performance reports and 21% used them for referral decisions. Physicians from the Mass and Special Media group were more likely to be aware of the performance reports (OR 4.16; 95% CI 2.16–8.00, p < .001) but not more likely to be influenced when referring patients into hospitals (OR 1.73; 95% CI 0.72–4.12, p > .05). On a 1 (very good) to 6 (insufficient) scale, the usefulness of the performance reports was rated 3.67 (±1.40). Aggregated presentation formats were rated more helpful than detailed hospital quality information. CONCLUSIONS: Hospital quality reports have limited impact on referral practices. To increase the latter, concerns raised by referring physicians must be given more weight. Those principally refer to the underlying data, the design of the reports, and the lack of important information. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi: 10.1186/s13561-017-0171-5) contains supplementary material, which is available to authorized users.