S. Iyer1, M. Jarosinski2, J. N. Kennedy1, K. E. Rudd1, C. W. Seymour1,3, E. Tzeng2, M. M. Marron4, K. M. Reitz2 1University Of Pittsburg, Critical Care Medicine, Pittsburgh, PA, USA 2University Of Pittsburg, Surgery, Pittsburgh, PA, USA 3University Of Pittsburg, Emergency Medicine, Pittsburgh, PA, USA 4University Of Pittsburg, Epidemiology, Pittsburgh, PA, USA
Introduction: Sepsis is common, deadly, and exacerbated by comorbid conditions. Peripheral artery disease (PAD) is one such condition, affecting >230 million people worldwide. PAD can lead to de novo ischemic wounds and retarded wound healing that increase the risk of sepsis. We aimed to establish the prevalence of PAD among sepsis hospitalizations and hypothesized that PAD would be associated with higher risk of in-hospital mortality and amputation among sepsis hospitalizations.
Methods: We examined adult hospital discharge data with survey-weights using the 2018 National Inpatient Sample, generating national sepsis admission population prevalence estimates. We included hospitalizations with a primary diagnosis of sepsis and excluded non-adult patients (<18 years) and those with missing outcome (i.e., in-hospital mortality) and demographic (i.e., age, sex, race/ethnicity) data. Associations between PAD and in-hospital mortality or amputation among sepsis hospitalizations were evaluated using log-binomial regression, adjusting for demographics (age, sex, race/ethnicity, income) and a modified Elixhauser Comorbidity Index, which excluded PAD. Subgroup analyses to determine moderators of each PAD-outcome relationship were also conducted.
Results: Of 35,527,481 hospitalizations (age mean ± standard error: 49.9±0.2, 44% male, 65% White), 1,955,275 (5.5%, 95% confidence interval (CI): 5.4-5.6%) had a primary diagnosis of sepsis (age: 68.8±0.1, 50% male, 70% White) of which 9,105 (0.5%, 95% CI 0.44-0.49%) had a secondary diagnosis of PAD (age: 71.1±0.3, 62% male, 67% White). Among sepsis hospitalizations, in-hospital mortality was 10% and 0.3% underwent a major or transmetatarsal amputation. Comorbid PAD was associated with 14% higher in-hospital mortality (95% CI: 1.01-1.29) and 24 times the risk of major or transmetatarsal amputation (95% CI: 19.7-29.0). Sex and a modified Elixhauser Comorbidity Index were important moderators of the PAD-mortality relationship.
Conclusion: Among sepsis hospitalizations, diagnosed and coded PAD was infrequent. Those with PAD were significantly older, more likely to be men, individuals of Black race, and in the lowest quartile for median income. They were also more likely to be transferred to a skilled nursing or similar facility. PAD was associated with significantly higher in-hospital mortality and major or transmetatarsal amputation, even after adjusting for demographics and a modified Elixhauser Comorbidity Index. These findings highlight the markedly higher risk of mortality and limb loss in sepsis in the setting of PAD, especially for men and those with worse scores on a modified Elixhauser Comorbidity Index.
