The impact of hospital teaching status in atrial fibrillation hospital admission outcomes: a national inpatient analysis in the united states

Abstract

Abstract Background Teaching hospitals are are usually larger and have a higher level of complexity. Admissions for atrial fibrillation are commonly encountered by residents and fellows and they carry a hight cost, morbidity and mortality. Purpose Our aim is to do a retrospective analysis of the National Impatient Sample (NIS) from 2016 to 2018 and determine the impact of the hospital teaching status and rural Vs urban location in the atrial fibrillation admission outcomes. Methods We combined the NIS from 2016–2018. A primary diagnosis of atrial fibrillation was identified identified. We stratified the admissions based on the hospital teaching status. We defined as primary outcomes the length of stay (LOS), in hospital mortality and total charges/cost per admission. The secondary outcomes were the rates acute kidney injury (AKI), acute blood loss anemia (ABLA) and embolic stroke (ES). We used the Charlson comorbidity index. We used univariate and multivariate linear regressions for the continuous outcomes and logistic regressions for the binary outcomes. Results We identified 981,739 admissions with a primary diagnosis of atrial fibrillation; 513,914 (47%) were female patients, the average age was 70.5 years. In the teaching hospitals, the LOS was 0.3 days longer [95% CI: 0.22–0.37]; p=0.000 and the total hospital charges were higher (8406.32 dollar more per admission [95% CI: 6510.89–10301.76]; p=0.000); there was no difference in mortality compared to the non-teaching institutions (OR 1.03 [95% CI: 0.87–1.21]; p=0.628). In teaching hospitals, the rates of AKI, ABLA and ES were higher (OR 1.2, 95% CI [1.12–1.27]; p=0.000), (OR 1.29, 95% CI [1.09–1.53]; p=0.003) and (OR 1.72, 95% CI [1.11–2.66]; p=0.014)respectively. In a subgroup analysis comparing urban teaching hospitals Vs urban non-teaching Vs rural non-teaching hospitals, we found that the primary outcomes were similar than the above but the results in the secondary outcomes were driven mainly by a marked difference between rural and urban institutions regardless of teaching status, which makes us think that there might be under diagnosis of atrial fibrillation related complications in lower complexity rural hospitals. The difference in costs per admission was not as high when comparing urban teaching with non-teaching hospitals (5103.96 dollar higher, 95% [CI: 3008.70–7199.23]; p=0.000). Conclusion Amongst admissions for atrial fibrillation, the total cost per admission was higher in teaching hospitals which was concordant with a longer LOS. Although our results show higher rates of AKI, ABLA and ES in teaching hospitals, this might be caused by underdiagnosis of such conditions in non-teaching rural institutions compared to large teaching hospital systems and these were not associated with an increase rate of mortality. There is need for more research to determine the factors impacting the higher costs of atrial fibrillation admissions amongst teaching institutions. Funding Acknowledgement Type of funding sources: None.