Impact of a modified data capture period on Liu comorbidity index scores in Medicare enrollees initiating chronic dialysis

Abstract

Background: The Liu Comorbidity Index uses the United States Renal Data System (USRDS) to quantify comorbidity in chronic dialysis patients, capturing baseline comorbidities from days 91 through 270 after dialysis initiation. The 270 day survival requirement results in sample size reductions and potential survivor bias. An earlier and shorter time-frame for data capture could be beneficial, if sufficiently similar comorbidity information could be ascertained. Methods: USRDS data were used in a retrospective observational study of 70,114 Medicare- and Medicaid-eligible persons who initiated chronic dialysis during the years 2000–2005. The Liu index was modified by changing the baseline comorbidity capture period to days 1–90 after dialysis initiation for persons continuously enrolled in Medicare. The scores resulting from the original and the modified comorbidity indices were compared, and the impact on sample size was calculated. Results: The original Liu comorbidity index could be calculated for 75% of the sample, but the remaining 25% did not survive to 270 days. Among 52,937 individuals for whom both scores could be calculated, the mean scores for the original and the modified index were 7.4 ± 4.0 and 6.4 ± 3.6 points, respectively, on a 24-point scale. The most commonly calculated difference between scores was zero, occurring in 44% of patients. Greater comorbidity was found in those who died before 270 days. Conclusions: A modified version of the Liu comorbidity index captures the majority of comorbidity in persons who are Medicare-enrolled at the time of chronic dialysis initiation. This modification reduces sample size losses and facilitates inclusion of a sicker portion of the population in whom early mortality is common. Keywords: Comorbidity, Kidney failure, Chronic, Renal dialysis, Epidemiologic research design