Length of stay has minimal impact on the cost of hospital admission11No competing interests declared.

Most strategies proposed to control the rising cost of health care are aimed at reducing medical resource consumption rates. These approaches may be limited in effectiveness because of the relatively low variable cost of medical care. Variable costs (for medication and supplies) are saved if a facility does not provide a service while fixed costs (for salaried labor, buildings, and equipment) are not saved over the short term when a health care facility reduces service. To determine the relative variable and fixed costs of inpatient and outpatient care for a large urban public teaching hospital. Cost analysis. A large urban public teaching hospital. All expenditures for the institution during 1993 and for each service were categorized as either variable or fixed. Fixed costs included capital expenditures, employee salaries and benefits, building maintenance, and utilities. Variable costs included health care worker supplies, patient care supplies, diagnostic and therapeutic supplies, and medications. In 1993, the hospital had nearly 114000 emergency department visits, 40000 hospital admissions, 240000 inpatient days, and more than 500000 outpatient clinic visits. The total budget for 1993 was $429.2 million, of which $360.3 million (84%) was fixed and $68.8 million (16%) was variable. Overall, 31.5% of total costs were for support expenses such as utilities, employee benefits, and housekeeping salaries, and 52.4% included direct costs of salary for service center personnel who provide services to individual patients. The majority of cost in providing hospital service is related to buildings, equipment, salaried labor, and overhead, which are fixed over the short term. The high fixed costs emphasize the importance of adjusting fixed costs to patient consumption to maintain efficiency.

A key strategy used to contain hospital costs during the 1980s was to reduce the total number of admissions and average lengths of stay. We assessed the magnitude of the savings achieved, the effect of the reductions on the rate of increase in costs, and the prospects for future savings through reductions in the number of days patients spend in the hospital (inpatient days). Using data from the American Hospital Association and the Health Care Financing Administration, we calculated the savings in the total number of inpatient days as the deviation from the historical increase in the number of inpatient days per year. We then estimated the real increase in costs that would have been observed if the reduction in the number of inpatient days had not occurred; we defined this value as the "underlying" rate of increase in costs. Finally, we compared the rates of increase in hospital reimbursement for Medicare beneficiaries and patients not covered by Medicare (non-Medicare patients). The total number of inpatient days per year decreased by 28 percent, in aggregate, between 1981 and 1988. The annual reduction was greatest in 1984 and 1985 and became progressively smaller in each subsequent year; by 1988 there was virtually no further reduction in the total number of inpatient days. The brief slowing of the increase in costs in the mid-1980s can be attributed entirely to the reduction in the number of inpatient days per year. The underlying rate of increase in costs was thus unaffected by efforts to contain spending. An increased number of outpatient visits partially offset the savings that resulted from the reduction in the number of inpatient days. This increase persisted even when the savings due to the lower number of inpatient days dwindled, and it virtually eliminated any dollar savings during the latter part of the 1980s. Between 1976 and 1982, Medicare spending on services provided by acute care hospitals rose by 9.2 percent per year in real terms, whereas non-Medicare expenditures rose by only 4.6 percent. This pattern has been reversed in recent years; in 1987-1988, Medicare spending rose by only 0.6 percent per year, whereas non-Medicare spending rose by 9 percent. Our findings suggest that the era of easy reductions in the number of inpatient days, with the associated attenuation of rising costs, is largely over. If further reductions in inpatient days are accompanied by an increase in the amount of ambulatory care similar to that during the past few years, the net savings will probably be negligible. Once the potential savings due to reductions in the number of inappropriate inpatient days has been exhausted, real hospital costs can be expected to rise, unless other effective measures to contain costs are implemented.