45. Resource Use and Length of Stay in ICU Patients in VA and Private Sector Hospitals
PJ Kaboli, Iowa City VAMC, VA National Quality Scholars Program and the Univ of Iowa, Iowa City, IA; MJ Barnett, Iowa City VAMC, Iowa City, IA; SM Fuehrer, Cleveland VAMC, Cleveland, OH; GE Rosenthal, Iowa City VAMC, VA National Quality Scholars Program and the Univ of Iowa, Iowa City, IA
Objectives: Compare length of stay (LOS) and treatment intensity in intensive care units (ICUs) in a VA hospital and private sector hospitals in the same metropolitan region.
Methods: We studied consecutive ICU patients admitted to a university-affiliated VA hospital (N=1,142) and 27 private sector hospitals (N=52,249) in the same region during the period 1/94-12/95. Hospital LOS, ICU LOS, and first-day ICU treatment intensity were compared in the two cohorts, adjusting for admission severity of illness. Severity was measured using the APACHE III methodology, which accounts for age, comorbidity, admission source, admission diagnosis, and abnormalities in 17 physiologic variables during the first 24 hours of ICU care. First-day ICU treatment intensity was measured using a modified version of the Therapeutic Intervention Scoring System (TISS), which is a weighted scale that considers the use of 30 diagnostic and therapeutic interventions.
Results: VA patients were more likely to be male (98% vs. 52%; p<.001) and had higher admission APACHE III scores (54 vs. 50; p<.001), but were similar in age (62.5 vs. 62.9; p=.50). Unadjusted hospital LOS was higher in the VA (28.3 vs. 11.3 days; p<.001), as was ICU LOS (4.3 vs. 3.9 days; p<.001). In a linear regression model adjusting for admission severity of illness, VA hospital LOS remained significantly longer (17.4 days; 95% CI, 16.5-18.3; p=.01). However, the difference in ICU LOS was no longer significant (0.16 days; 95% CI -0.08-0.40; p=.19). Treatment intensity, as measured by mean TISS scores, was higher in VA patients (2.42 vs. 1.50; p<.001). After adjusting for admission severity, TISS scores remained higher in VA patients (0.77 points; 95% CI 0.58-0.95; p<.001). The differences in adjusted TISS scores were consistent in stratified analysis of medical and surgical ICU patients. Lastly, differences in adjusted TISS scores were less in comparisons limited to the five major teaching hospitals (0.32; 95% CI 0.07-0.57; p=.01) than to non-teaching hospitals (0.95; 95% CI 0.77-1.13; p<.001).
Conclusions: In an analysis of ICU care in a VA hospital and private sector hospitals serving the same metropolitan region, resource utilization, as measured by hospital LOS and first-day ICU treatment intensity, was higher in the VA hospital; however, ICU LOS was similar. Although we did not directly examine the appropriateness of the ICU interventions measured, these findings highlight the complexity of measuring resource utilization in ICU populations.
Impact: The current findings provide important information on patterns of ICU resource utilization in VA patients and the importance of severity-adjustment in making comparisons across systems that care for different types of patients. While ICU LOS was similar, the longer hospital LOS in VA patients may represent utilization and a significant opportunity for re-organizing the delivery of post-ICU care in a more cost-efficient manner. Lastly, our findings highlight the need for future evaluation of factors underlying differences in treatment intensity and the impact of these differences on patient outcomes.