Do public and private hospitals differ in quality? Evidence from Italy

We investigate whether public and private providers differ in quality in Lombardy, a large Italian region. This region has adopted an “internal market” model where public and private providers are paid by DRG and compete for publicly-funded patients for both elective and emergency treatments. Using a large administrative sample in 2012–14, we measure clinical quality with 30-day mortality for the following emergency conditions: heart attack (AMI), stroke (ischemic and haemorrhagic) and hip fracture. For elective care, where mortality is negligible, we measure 30-day emergency readmission rates for hip replacement and knee replacement. Public and private hospitals may compete not only on clinical quality, but also on non-clinical aspects of patients' experience. We investigate whether private providers have shorter waiting times for hip and knee replacements. To control for unobserved differences in casemix between public and private providers we pursue an instrumental variable approach based on the distance between patient's residence and the closest public and private provider: longer distances to the closest private and public hospital are highly significant determinants of whether the patient is treated by a private versus a public provider. We find, with few exceptions, that public and private providers generally do not differ in elective and emergency quality, neither in waiting times. The only exception is AMI for which mortality risk is lower in private providers, and hip replacement for which readmission risk is higher in private providers.