Inpatient burden and association with comorbidities of polyarteritis nodosa: National Inpatient Sample 2014

Abstract

© 2019 Elsevier Inc. Objectives: To characterize inpatient burden, expenditures and association with comorbidities of polyarteritis nodosa (PAN). Methods: Patients with PAN were identified from the Nationwide Inpatient Sample (NIS) database for the year 2014 using ICD-9 diagnostic codes. The primary outcome was determining the inpatient prevalence of PAN in hospitalized patients in the US. Secondary outcomes included determining inpatient mortality, morbidity, comorbidities, hospital length of stay (LOS) and total hospital costs and charges. A cohort of patients without PAN was also identified from the same database to serve as comparators for analysis of comorbidities. Multivariate regression analysis was used to adjust for age, gender, ethnicity, comorbidities and hospital characteristics. Results: A total of 4,110 patients with PAN were included in the study. The mean age was 59.5 years and 61% were female. The inpatient prevalence of PAN was 11.6 cases per 100,000 discharges. Patients with PAN displayed increased adjusted odds of mortality (OR:1.35, p = 0.13), shock (OR:1.75, p<0.01), ICU admission (OR:1.88, p<0.01) and multiorgan failure (OR:3.12, p<0.01) compared to patients without PAN. Patients with PAN also displayed significantly higher hospital costs (additional adjusted mean [aAM]: $9,693, p<0.01), hospitalization charges (aAM: $34,273, p<0.01) and LOS (aAM: 4.1 days, p<0.01) compared to patients without PAN. Analysis of comorbidities found a significant association between PAN and venous thromboembolism, renal injury and sepsis. The main limitation of this study was reliance on accuracy of diagnostic coding. The high inpatient prevalence of PAN might have been inflated and we cannot be certain that the higher risk of comorbidities and expenditures were entirely attributable to PAN as some patients in this cohort may have other vasculitides. Conclusions: The inpatient prevalence of PAN is higher than what would be expected from the overall general prevalence. Hospitalizations of patients with PAN are associated with significantly higher rates of morbidity and expenditures.