Comparing health outcomes among hospitals: the experience of the Lombardy Region

Abstract

In recent years, governments and other stakeholders have increasingly used administrative data for measuring healthcare outcomes and building rankings of health care providers. However, the accuracy of such data sources has often been questioned. Starting in 2002, the Lombardy (Italy) regional administration began monitoring hospital care effectiveness on administrative databases using seven outcome measures related to mortality and readmissions. The present study describes the use of benchmarking results of risk-standardized mortality from Lombardy regional hospitals. The data usage is part of a general program of continuous improvement directed to health care service and organizational learning, rather than at penalizing or rewarding hospitals. In particular, hierarchical regression analyses - taking into account mortality variation across hospitals - were conducted separately for each of the most relevant clinical disciplines. Overall mortality was used as the outcome variable and the mix of the hospitals' output was taken into account by means of Diagnosis Related Group data, while also adjusting for both patient and hospital characteristics. Yearly adjusted mortality rates for each hospital were translated into a reporting tool that indicates to healthcare managers at a glance, in a user-friendly and non-threatening format, underachieving and over-performing hospitals. Even considering that benchmarking on risk-adjusted outcomes tend to elicit contrasting public opinions and diverging policymaking, we show that repeated outcome measurements and the development and dissemination of organizational best practices have promoted in Lombardy region implementation of outcome measures in healthcare management and stimulated interest and involvement of healthcare stakeholders.

Exploring the Effect of Network Governance Models on Health-Care Systems Performance

“…Pathways vary in volume, with the total number of patients moving among hospitals (Iwashyna, Christie, Moody, Kahn, & Asch, 2009). According to Berta, Seghieri, and Vittadini (2013), transfer between Hospitals A and B occurs when a patient is registered at Hospital A up to a certain day, and then at Hospital B on the same day or the following day. Two different typologies of network have been identified: cardiac disease departments and neurological disease departments.…”

“…Following a consistent literature on health-care research (Jencks et al, 1988;Rosenthal et al, 2000;Thomas & Hofer, 1998), the 30-day mortality rate was considered as a proxy for clinical outcome performance. Previously adopted in relation to resource allocation (Gowrisankaran & Town, 2003), hospital governance models, and hospital markets conditions (Tomal, 1998), mortality rates are measures for the quality of care (Palangkaraya & Yong, 2013) in both wards (Rosenthal, Shah, Way, & Harper, 1998) and groups of diagnosis levels (Berta et al, 2013).…”

Insights on the effectiveness of reward schemes from 10‐year longitudinal case studies in 2 Italian regions.

“…This evaluation system, based on benchmarking but without disclosing the names of providers, has been in place until 2014 when Lombardy joined the Italian collaborative network and adopted their shared PES …”

The coxian phase-type distribution as a contribution to the multilevel model of in-hospital mortality

“…The use of in-hospital mortality rates in the comparative evaluation of quality of care has been proposed by several authors (for an depth analysis see for example Berta et al, 2013).…”