Protease inhibitor-based therapy is associated with decreased HIV-related health care costs in men treated at a Veterans Administration Hospital

Background: Protease inhibitor (PI) therapy for HIV infection is associated with decreased rates of opportunistic infections and death. Statistical mo dels predict that decreased complications will be associated with decreased hospitalization costs. A recent report suggested that the decrease in the HIV hospitalization costs were offset by increases in demand for outpatient services. We performed a study of hospital use and HIV-associated health c are costs in our center to determine the following: whether PI therapy is a ssociated with decreased inpatient use; whether PI therapy is associated wi th decreased outpatient use and costs; whether decreased HIV health care co sts are associated with increased use of nucleoside analogues. Methods: The Dallas Veteran Affairs Medical Center provides comprehensive i npatient and outpatient HIV care and thus can evaluate the relation between inpatient and outpatient costs. The mean monthly number of hospital days, Infectious Diseases clinic visits, emergency department visits, other outpa tient clinic visits, inpatient costs, outpatient costs, and PI costs were d etermined from January 1, 1995 through July 31, 1997. This time period was then divided into three intervals. Comparisons of PI use and HIV-related he alth care costs were during the three intervals was performed using analysi s of variance (ANOVA). Significant differences between the baseline charact eristics were further analyzed through multiple linear regression. Results: A decrease in hospital days, and all outpatient visits including e mergency visits, and HIV clinic visits was determined. No difference was fo und in the rate of use of other outpatient services. The per patient costs of HIV care decreased from a monthly average of $1905 U.S. in the first int erval to $1122 U.S. in the last interval (p < .01). Linear regression demon strated an inverse relation between PI use and total HIV costs (B = -0.67, p = .00, adjusted R-2 = 0.52) but no relation between nucleoside use, stage of disease or financial class. Conclusions: PI therapy is associated with decreased hospital days and use of outpatient services. Total patient costs decreased, but a concomitant ri se in outpatient costs took place. This increase was primarily a result of increased costs of acquiring PI. Increases in the number of nucleoside agen ts prescribed were not associated with decreased coals.