Healthcare Costs, Resource Use, and Mortality Rates for Sepsis in Teaching Versus Non-Teaching Hospitals
Methods: We conducted a retrospective study of hospitalized children with the diagnosis of sepsis. The Agency for Healthcare Research and Quality (AHRQ) 2009 Kids’ Inpatient Database (KID) provided the data for analysis. The data set consisted of 7,370,203 weighted discharges in 4,121 hospitals over 44 states. Diagnosis of sepsis was characterized based on an All Patient Refined Diagnosis-Related Groups (APR-DRG) of 720: Septicemia & Disseminated Infections. APR-DRG validated severity classes segmented the illness severity. Teaching hospital status was determined based on the Accreditation Council for Graduate Medical Education (ACGME) approved residency training, presence in the Council of Teaching Hospitals (COTH), or a ratio of full-time equivalent interns and residents to beds of 0.25 or higher. Statistical analysis was conducted using STATA 11.2. Institutional Review Board approval obtained prior to research study.
Results: There were 11,893 patients discharged with an APR-DRG of 720, 5,085 in non-teaching hospitals and 6,808 in teaching hospitals. In comparing non-teaching versus teaching hospitals, patients were on average 8.28 versus 7.23 years old and 54.25% versus 49.08% female (both p values < 0.00). Non-teaching hospital patients (versus teaching hospital patients) had a mortality rate of 1.63% versus 4.66%, average length of stay of 4.86 versus 8.13 days, average number of procedures of 0.89 versus 2.04, and average total hospitalization charges of $29,829 versus $65,639 (all p values < 0.00). When examining the highest severity of illness class (discharges=2,626 with 592 in non-teaching versus 2,034 in teaching), those in non-teaching hospitals had a mortality rate of 12.50% versus 14.01% (p value=0.35), average length of stay of 10.91 versus 14.19 days, average number of procedures of 3.31 versus 4.30, and average total hospitalization charges of $113,384 versus $143,999 (all p values < 0.00).
Conclusion: Our study demonstrated that teaching hospitals, in caring for pediatric inpatients diagnosed with sepsis, had overall higher charges, length of stay, procedures performed, and mortality rates. Even when accounting for a higher level of severity, these trends mostly remain. Our findings were contrary to previous studies attributing higher costs and resource utilization in teaching hospitals but having similar clinical outcomes. Thus, our study suggested that teaching hospitals, when compared to non-teaching hospitals, provide care for sepsis at greater costs and resources without an improvement in mortality rates.