Prospects for using DRGs in Swedish hospitals

Overall cost-containment seems satisfactory in Swedish health care. Health care's share of the Gross Domestic Product (GDP) has thus decreased from 9.7% to 9.1% between the years 1982 and 1986. A recent study indicates, however, a decreasing productivity, measured by traditional output measures. It may be argued that these measures are not adequate as they do not reflect the changing patterns of medical technology, case-mix and quality. The Swedish Planning and Rationalization Institute of the Health and Social Services (Spri) has started a project which aims at evaluating whether the Diagnosis Related Group (DRGs) could be used for planning, budgeting and follow-up in the Swedish health care system. The interest in these matters is at present great in Sweden. The feasibility and adequacy of DRGs in the Swedish context are being tested using more than 500,000 discharges from 6 counties. All the necessary variables for DRG assignment are already routinely collected in the Swedish discharge abstracts. In a comparison with an American database concerning individual DRGs, Swedish and American lengths of stay correlate rather well (r = 0.83). There is a difference in the number of diagnoses on each discharge record in the two databases. On average 1.4 diagnosis codes are used for each patient in the Swedish hospital sample compared to 2.8 codes in the United States database. This may indicate a systematic difference in coding practice of complications and co-morbidities.