[Identification of Control Hospitals for the Implementation of the Nationwide and Standardized Evaluation of Model...

Health care needs of mentally ill patients make special demands on cross-sectoral health care structures. § 64b SGB V enables care of mentally ill patients through model projects that are multi-professional, work across treatment periods and sectors and implement new forms of financing. These model projects in their hospitals (case hospitals) need to be evaluated and compared with standard treatment methods.

The aim of this analysis is to identify matching hospitals according to a priori defined criteria for the establishment of a control group (control hospitals) using secondary data.

A systematic analysis was conducted based on structured quality reports according to §+137 SGB V and matched data from the Federal Institute for Research on Building, Urban Affairs and Spatial Development (BBSR). Based on a priori defined knock-out criteria, criteria based on patients (weighting 50%), structural features of hospitals (25%) and environmental factors (25%), a weighted similarity score was calculated for each of the 13 case hospitals, which could reach the maximum of 100 points (perfect match).

10 control hospitals per case hospital were identified according to the weighted similarity score. The median of the total deviation of potential control hospitals from the case hospitals was 34.3 (range: 17.6-66.7). The median of the 10 selected control hospitals per case hospital was 30.9 (range: 17.6-40.8).

The defined algorithm could be used to identify similar control hospitals. The method using the mentioned databases and derivation of specific criteria of structural similarity are generally suitable in controlled designs for the evaluation of complex interventions based on routine data.