Acute hospital medical care for the elderly: is there a valid performance indicator?

Many clinicians have doubts about the validity of using activity data to measure the efficiency of health services, since they take no account of outcome. In elderly patients presenting with acute illness, minimising the time spent in hospital or institutional care is a justifiable aim in itself. If allowance is made for deaths and readmissions, valid performance indicators might be derived from the distribution of the resulting 'Adjusted Length of Stay'. By surveying all medical patients aged 65 and over, admitted as emergencies to the hospitals of one large health district, this study investigated whether such an index (the Adjusted length of Stay) could be of practical value. The influence of the following characteristics on the patient's length of stay were assessed; age, social circumstances, premorbid disability and the nature of the presenting complaint. Patients admitted to general medical and health care of the elderly wards were compared and investigated, to determine whether their outcome was similar after adjusting for differences in these characteristics. Although there was some overlap, marked differences emerged in age distribution, nature of the presenting illness and prevalence of previous disability. These factors were all found to have a substantial effect on the time spent in hospital. The Adjusted Length of Stay Index could be calculated with little change to existing methods of data collection. It provides a fairer basis for overall evaluation of acute medical services for the elderly than current performance indicators.