S789 The Differences of Hospitalization Outcomes Between Teaching and Non-Teaching Hospitals in Taking Care of Patients With Inflammatory Bowel Disease in the United Sates

Abstract

Introduction: IBD costs continue to rise in the US.1 Teaching hospitals tend to have more resources for care, but the effect of these resources on cost and utilization is unknown.2 We therefore investigated the differences in characteristics and outcomes of patients admitted to teaching hospitals vs non-teaching hospitals. Methods: We utilized the national inpatient sample data (NIS) 2016-20183 and hospitalized patients with a 1ry discharge diagnosis of IBD were identified. We excluded all patients who were younger than 18 years and classified our population to those who were admitted to teaching and non-teaching hospitals. Our outcomes included hospital length of stay (LOS), hospital charges, mortality, and discharge disposition. Survey analysis was applied to account for discharge weight and multistage sampling of the NIS data.4 We calculated the propensity score for patients based on their demographics, hospital bed size and region, Charlson comorbidity index,5 obesity, and tobacco, alcohol, or drug use. We estimated the average treatment effect weight based on the propensity score and used it to adjust differences between the two groups. Results: 281,440 patients (weighted number) were eligible for our study and 203,815 patients were admitted to teaching hospitals. These patients were more likely to be from larger counties, had private insurance, higher median household income, and their hospitals were more likely to be located at the Northeast or the Midwest (Table). In the adjusted analyses, patients who were admitted to teaching facilities had a higher mean difference of LOS (0.82 days, 95% CI 0.7 – 0.94, P < 0.001), hospital charges ($9,482.75, 95%CI 7,336.2 – 11,629.3, P=0.14), higher odds ratio (OR) of discharge with visiting nurse (2.02, 95%CI 1.81 – 2.26, P=< 0.001), and lower OR to be transferred to other hospitals (0.4, 95%CI 0.34 – 0.48, P< 0.001). (Figure) Conclusion: In this national analysis of hospitalized patients with IBD, patients who were admitted to teaching hospitals were associated with higher utilization of healthcare resources even after adjusting for demographic, hospitals differences, and comorbidities. Further studies are needed to understand the differences in care at teaching and non-teaching hospitals to explain the cause for the increased utilization of resources in teaching hospitals compared to non-teaching hospitals.Figure 1.: The hospitalization outcomes for patients with Inflammatory Bowel Disease based on their hospital type, teaching vs non-teaching facilities. Table 1. - Baseline characteristics for hospitalized patients with IBD comparing patients who were admitted to teaching vs non-teaching hospitals Characteristics Teaching Hospitals Non- teaching Hospitals P value Total number (Weighted) 203,815 77,625 Age (mean (SD)) 44.56 (17.81) 47.74 (18.33) < 0.001 Female (%) 108,075 (53) 42,730 (55.1) < 0.001 Race (%) < 0.001 •F020White 139,595 (71.0) 59,149.9 (78.7) •F020Black 30,525 (15.5) 7,785 (10.4) •F020Hispanic 17,065 (8.7) 5,185 (6.9) •F020Others 9,505 (4.8) 3,075 (4.1) Patient Location: NCHS Urban-Rural Code (%) < 0.001 •F020"Central" counties of ≥1 million population 68,235 (33.6) 12,725 (16.4) •F020"Fringe" counties of ≥1 million population 57,420 (28.3) 20,360 (26.3) •F020Counties of 250,000-999,999 population 42,580 (21.0) 13,740 (17.7) •F020Counties of 50,000-249,999 population 16,120 (7.9) 8,890 (11.5) •F020Not metropolitan or micropolitan counties 18,685 (9.2) 21,739.9 (28.1) Expected primary payer (%) < 0.001 •F020Medicare 48,645 (23.9) 23,385 (30.2) •F020Medicaid 39,290 (19.3) 14,000 (18.1) •F020Private insurance 98,510 (48.4) 32,045 (41.4) •F020Self-pay/ Others/No charge 17,100 (8.4) 8,025 (10.4) Median household income for patients ZIP Code (%) < 0.001 •F0200-25th percentile 48,885 (24.3) 22,435 (29.4) •F02026th to 50th percentile 47,765 (23.8) 22,730 (29.8) •F02051st to 75th percentile 53,400 (26.6) 17,320 (22.7) •F02076th to 100th percentile 50,780 (25.3) 13,845 (18.1) Weighted Charlson Comorbidity (mean (SD)) 0.64 (1.30) 0.68 (1.30) 0.003 Alcohol use disorder (%) 3,655 (1.8) 1,670 (2.2) 0.007 Drug use disorder (%) 12,795 (6.3) 4,600 (5.9) 0.16 Tobacco use (history and current) (%) 34,295 (16.8) 15,345 (19.8) < 0.001 Patients with Obesity (%) 18,370 (9) 6545 (8.4) < 0.001 Weekend admissions (%) 40,890 (20.1) 17,940 (23.1) < 0.001 Bed size of the hospital (%) < 0.001 •F020Small 40,970 (20.1) 12,300 (15.8) •F020Medium 55,340 (27.2) 23,860 (30.7) •F020Large 107,505 (52.7) 41,465 (53.4) Region of the hospital (%) < 0.001 •F020Northeast 49,705 (24.4) 10,870 (14.0) •F020Midwest 51,420 (25.2) 17,025 (21.9) •F020South 71,525 (35.1) 35,110 (45.2) •F020West 31,165 (15.3) 14,620 (18.8)