Clinical and Sociodemographic Predictors of Mortality in End-Stage Renal Disease Inpatients in Rural Areas of the USA: Evidence From the Nationwide Inpatient Sample.

backgroundEnd-stage renal disease (ESRD) has been associated with an increase in all-cause mortality among patients. The accumulation of comorbidities appears to be a contributing factor. This study set out to identify the effect of comorbidity severity and other predictors of mortality among ESRD inpatients in rural America.

methodsThis is a cross-sectional study that used the 2016-2018 Nationwide Inpatient Sample (NIS) from the Healthcare Cost and Utilization Project (HCUP). The study included patients aged 18 years or older with ESRD hospitalized in rural hospitals in America. Independent variables used in the survey include age, gender, race, type of admission (elective versus nonelective), type of hospital control, expected primary payer, and severity of comorbidities. The dependent variable was death during hospitalization. All analyses were weighted. Univariate (frequencies), bivariate (Chi-square), and logistic regression analyses were done using the SAS Studio (SAS Institute Inc., Cary, NC, USA).

resultsThere were 144,575 weighted ESRD hospitalizations, and 5% of the hospitalized patients died. In the bivariate analysis, significant variables include age group, race, type of hospital admission, expected primary payer, type of hospital control, and severity of comorbidities, and all had a significant P-value of <0.0001. On multivariable logistic regression analysis, middle-aged and elderly patients had 40% (adjusted odds ratio (AOR): 1.40, 95% confidence interval (CI): 1.20-1.62) and 201% (AOR: 3.01, 95% CI: 2.61-3.48) more odds of mortality while hospitalized, respectively, compared to the young. Compared to whites, blacks had 19% (AOR: 0.81, 95% CI: 0.77-0.86) reduced odds of mortality, Hispanics had 47% (AOR: 0.53, 95% CI: 0.46-0.61) reduced odds of mortality, Native Americans had 27% (AOR: 0.73, 95% CI: 0.63-0.84) reduced odds of mortality, and Asian or Pacific Islanders had 30% (AOR: 0.70, 95% CI: 0.54-0.90) reduced odds of mortality. ESRD patients on nonelective hospitalizations had 16% (AOR: 0.84, 95% CI: 0.79-0.90) reduced odds of mortality while hospitalized versus those on elective hospitalization. ESRD patients with severe comorbidities had 40% (AOR: 1.40, 95% CI: 1.26-1.54) more odds of mortality compared to those with mild comorbidities, and those with moderate comorbidities had 22% (AOR: 1.22, 95% CI: 1.10-1.36) compared to those with mild comorbidities. Compared to patients on Medicare, ESRD hospitalizations on Medicaid had 19% (AOR: 1.19, 95% CI: 1.06-1.32) higher odds of mortality, hospitalizations on private insurance had 26% (AOR: 1.26, 96% CI: 1.15-1.37) higher odds of mortality, self-pay patients had 99% (AOR: 1.99, 95% CI: 1.61-2.45) higher odds of mortality, and no charge patients had over 1400% (AOR: 15.61, 95% CI: 7.09-34.35) higher odds of mortality. The area under the curve (AUC) for the model was 62%.

conclusionThe severity of comorbidities and expected primary payer are the modifiable predictors identified to predict ESRD inpatient mortality. From this study, the findings suggest that strategies aimed at preventing the severity of comorbidities and ensuring universal health coverage might help reduce ESRD inpatient mortality in rural America.