Hospital Effectiveness from administrative data: the Lombardy case

Purpose-The aim of this paper is the discussion and the dissemination of initiatives promoted by the Lombardy region for the construction of benchmarking systems between regional health structures of care utilizing administrative archives. Design/methodology/approach-The paper focuses on relative effectiveness (specific effect of care on patients) in a benchmarking framework, considering the dimension of sentinel outcomes. From Lombardy Hospital Discharge Cards proxies of sentinel outcomes are identified, defined as "context indicators" useful for a benchmarking analysis. Findings-First, the authors present outcomes and covariates at different levels (patient and healthcare structure) extracted from the Lombardy Hospital Discharge Cards for a benchmarking analysis. Second, empirical results show a consistent quota of outcome variability between structures of care and weak agreement between estimated rankings for context indicators. Finally, a slicing approach is suggested in order to apply an equitable comparison among healthcare structures. Practical implications-The paper provides regional stakeholders with practical implications regarding available strategies (outcomes, statistical methodology, risk adjustment) for consistent processes of evaluation, in a benchmarking framework, based on existing regional administrative data. Originality/value-After having presented available information contained in regional archives for a benchmark analysis, empirical results were discussed about context indicators, presenting indications and strategies for a refinement of the approach. From a methodological point of view, the utilization of multilevel models (improving methodological strategies adopted by international agencies) in large administrative databases is proposed. © 2010 Emerald Group Publishing Limited. All rights reserved.