Is the emergency readmission rate a valid outcome indicator?

OBJECTIVES: The principal aim was to determine whether the emergency readmission rate varies between medical specialties, and to identify whether differences in emergency readmission rates between hospital trusts can be reduced by standardising for specialty. Possible factors influencing emergency readmission were also investigated, including frequency of previous admission and cause of readmission. DESIGN: Emergency readmission rates were obtained from the Scottish Morbidity Record scheme (SMR1) using record linkage, standardised for age and sex. Rates throughout Scotland were analysed by specialty, and rates for general medicine compared among teaching hospital trusts. Cause of emergency readmission was determined from hospital records in a random sample (177 patients). SETTING: Medical specialties throughout Scotland. SUBJECTS: All patients readmitted as an emergency within 28 days of discharge (October 1990 to September 1994). RESULTS: Emergency readmissions varied markedly between medical specialties, with highest rates in nephrology (24.2%, 95% CI 23.5 to 24.8) and haematology (20.4%, 95% CI 19.9 to 20.9), and the lowest in homeopathy (2.2%, 95% CI 1.6 to 2.7) and metabolic diseases (3.5%, 95% CI 2.4 to 4.5). The largest number of emergency readmissions was in general medicine, accounting for 63% of the total. Restricting emergency readmission rates to general medicine significantly altered previous rates. In the year preceding the emergency readmission, 59% of all patients had been admitted to hospital at least once, and most emergency readmissions (73.3%) resulted from a chronic underlying condition. CONCLUSIONS: Significant variations in emergency readmission rates occurred between medical specialties, suggesting that differences between hospital trusts are influenced by differences in specialties and thus case mix. The majority of emergency readmissions occurred in patients with an underlying chronic condition, and many had a history of multiple previous hospital admissions. The emergency readmission rate is therefore unlikely to be a valid outcome indicator reflecting quality of care until routine data are available for standardisation by case mix