Interhospital Transfers to MUSC Marc Heincelman, MD
Study Question At MUSC, is IHT status independently associated with inpatient mortality after adjusting for more detailed patient-level clinical characteristics?
Hypothesis IHT status will remain independently associated with inpatient mortality after adjusting for more detailed patient-level clinical characteristics?
Study Design Retrospective cohort study of adults admitted to MUSC internal medicine services from 2013-2014 Primary outcome= in-hospital mortality Independent variable of interest= IHT status Covariates: admit service, patient demographics, disease-specific conditions, labs, vitals Analysis= Cox proportional hazard regression analysis Cox regression analysis
Retrospective Cohort IHT In-hospital Mortality No in-hospital ED/Clinic No in-hospital Mortality
Data Collection Using MUSC’s Enterprise Data Warehouse, data were extracted from 2 local databases: the Medical University Hospital Authority (MUHA) inpatient database, and the hospital’s patient accounting system. Exposure: IHT status coded as a categorical variable (IHT vs non-IHT) Outcome: In-hospital mortality was the primary outcome variable and coded as a categorical variable
Data Analysis 4 Cox proportional hazard regression analyses will be performed to examine the independent association between IHT status and in-hospital mortality, controlling for covariates that may be potential confounders to the relationship between transfer and death. Model 1- IHT status and admit service Model 2- Model 1 + patient demographics Model 3- Model 2 + comorbidities Model 4- Model 3 + clinical variables Measure of association: Hazard Ratio
Sample Size Estimate Using an alpha level of 0.05 with 80% power, a sample size estimate was obtained. Assuming a 4% inpatient mortality rate at MUSC among non-IHT patients (admitted via the ED or clinic), we assumed we would be able to detect a difference if the inpatient mortality rate at MUSC among IHTs is 8%. Thus, the minimal detectable risk ratio for inpatient mortality among IHTs compared to non-IHTs is 2.0. Based on this, the sample size has been calculated as 2320.
Problems, Pitfalls, Limitations Residual confounding cannot be completely excluded including factors associated with transfer Time of day, day of week, transfer mode, etc Generalizability Single-center study in the southeast Does not meet criteria for causality Problems/Pitfalls Old data set
References Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project: an overview. Effective clinical practice : ECP 2002;5:143-51. Mueller S, Zheng J, Orav EJP, Schnipper JL. Inter-hospital transfer and patient outcomes: a retrospective cohort study. BMJ quality & safety 2018. Nathens AB, Maier RV, Brundage SI, Jurkovich GJ, Grossman DC. The effect of interfacility transfer on outcome in an urban trauma system. The Journal of trauma 2003;55:444-9. Sokol-Hessner L, White AA, Davis KF, Herzig SJ, Hohmann SF. Interhospital transfer patients discharged by academic hospitalists and general internists: Characteristics and outcomes. Journal of hospital medicine 2016;11:245-50. Hernandez-Boussard T, Davies S, McDonald K, Wang NE. Interhospital Facility Transfers in the United States: A Nationwide Outcomes Study. Journal of patient safety 2017;13:187-91.
Questions?