Public reporting of hospital quality shows inconsistent ranking results

Evidence from the US has demonstrated that hospital report cards might generate confusion for consumers who are searching for a hospital. So far, little is known regarding hospital ranking agreement on German report cards as well as underlying factors creating disagreement.

This study examined the consistency of hospital recommendations on German hospital report cards and discussed underlying reasons for differences.

We compared hospital recommendations for three procedures on four German hospital report cards. The agreement between two report cards was determined by Cohen's-Kappa. Fleiss' kappa was applied to evaluate the overlap across all four report cards.

Overall, 43.40% of all hospitals were labeled equally as low, middle, or top performers on two report cards (hip replacement: 43.2%; knee replacement: 42.8%; percutaneous coronary intervention: 44.3%). In contrast, 8.5% of all hospitals were rated a top performer on one report card and a low performer on another report card. The inter-report card agreement was slight at best between two report cards (κ_max=0.148) and poor between all four report cards (κ_max=0.111).

To increase the benefit of public reporting, increasing the transparency about the concept of - medical - "quality" that is represented on each report card seems to be important. This would help patients and other consumers use the report cards that most represent one's individual preferences.