OUTCOMES OF INSTABILITY ON DISCHARGE IN URBAN HOSPITALS

20世纪80年代引入的医疗保险预付费制度

The introduction of prospective payment for Medicare in the 1980s resulted in shorter hospital length of stay, higher proportions of patients being discharged prior to reaching medical stability, and greater risks of post-discharge death. This has been called the quicker and sicker phenomenon. The growth of managed care over the past 10 years has brought about even more dramatic declines in length of stay. We propose to develop generic and condition-specific measures of instability on discharge, and examine its prevalence and associated consequences in urban hospitals during the 1990s. Our proposal focuses on pneumonia, hip fracture and asthma because they are common, costly conditions with significant morbidity and mortality. We target urban hospitals because inner city populations (poor, minorities, uninsured, and elderly) are especially vulnerable to the consequences of early discharge. In NY, there is the additional problem of a relatively underdeveloped, post-hospital safety net of expanded primary and subacute care services. We propose an observational, multiple cohort study of instability on discharge and its associated risk-adjusted post-hospital outcomes (mortality, readmission, functional status, symptom resolution, and patient satisfaction). The feasibility and affordability of this proposal is aided by our use of 2 existing databases. For pneumonia, we will use the AHCPR-funded Pneumonia Patient Outcomes Research Team database, a prospective study of 686 inpatients from 4 hospitals in 3 cities. For hip fracture, we will supplement data from the AHCPR-funded Hip Fracture Study, a prospective study of 600 patients in 4 NY hospitals in the Mount Sinai Health System. For asthma, we propose a prospective, observational cohort study of 500 adults admitted with asthma to the Mount Sinai Medical Center, NY over 1 year. All databases will have baseline patient data, stability on discharge, discharge location, and 30 day outcomes. Our analyses will assess: 1) rates of instability on discharge; 2) associations between instability on discharge and post-hospital outcomes after adjusting for baseline patient characteristics, disease severity, comorbidities, and functional status; 3) predictive validity of generic versus condition-specific measures of instability; and 4) effects of post-discharge location on these risk-adjusted outcomes. We will perform the study of pneumonia in the 1st year, hip fracture in the 2nd year, and asthma in the 3rd year. The final product of this work will be an operational set of generic and condition-specific measures of instability on discharge that can serve as evidence-based indicators of inpatient quality of care.