Providers and policy-makers rely on population estimates of the attributable mortality risk associated with chronic diseases to make decisions regarding therapeutic choices and to justify funding for research. Disease mortality estimates also provide researchers with tools to identify disparities in care or improvements in treatment over time, and provide a more accurate estimate that can be compared across populations and over time.
Previous studies have developed methods to estimate disease-specific survival. However, according to Margaret R. Stedman, PhD, MPH, and colleagues, those methods have been underutilized in research focused on kidney disease. Relative survival is often used to measure disease-specific survival where information on cause of death is unavailable; however, there are few publications using US data that implement relative survival methods in kidney disease research.
Noting that estimates of mortality from kidney failure are misleading because mortality in patients with kidney failure is connected to mortality attributed to comorbid conditions, the researchers conducted a longitudinal cohort study to develop an alternative method to reduce the bias in estimating mortality due to kidney failure using life table methods. Results of the study were reported in the American Journal of Kidney Diseases.
The study utilized data from the US Renal Data System and the Medicare 5% sample to identify an incident cohort of patients 66 years of age and older who had a first diagnosis of kidney failure in 2009 and a similar population cohort without kidney failure. The study exposure was kidney failure, and the outcome of interest was death.
The researchers estimated relative survival of patients with incident failure by creating comorbidity-, age-, sex-, race-, and year-specific life tables. The tables also provided an estimate of excess deaths related to kidney failure. The estimates were compared with those based on standard life tables (not adjusted for comorbidity).
A total of 53,612 patients with incident kidney failure had complete information on age, sex, and race. Following application of exclusion criteria, the final cohort with kidney failure totaled 31,944. Of those, 0.7% (n=251) had received a preemptive kidney transplant on or before the index date. Following similar exclusions and exclusion of patients with kidney failure, the 5% sample of Medicare beneficiaries yielded a cohort of 1,237,540 individuals without kidney failure.
Comorbidity scores were derived from a Cox proportional hazards regression model, predicting time to all-cause death from age, sex, designated race, and individual comorbidities. Designated race was grouped as Black, White, and other, and age was grouped into 5-year increments. Comorbidity models were stratified by diabetes. Log scaler comorbidity scores were categorized as low (≥0), medium (>0 to 0.5), or high comorbidity (≥0.5).
The proportion of Black patients in the kidney failure population was more than twice as high as that in the population without kidney failure (18% vs 7%, respectively). There were substantially more comorbidities in the kidney failure cohort compared with the cohort without kidney failure (71% in the high category vs 24% in the high category, respectively). Over follow-up of 8 years, 11% of the population with kidney failure survived compared with 70% of the population without kidney failure. Two percent of patients received a transplant during the follow-up period; this event was not censored.
Using standard life tables adjusted for age, sex, race, and year, without comorbidities, the 5-year relative survival was 31%. Using life tables adjusted for those factors, as well as for comorbidities, the 5-year relative survival was 36%. Relative survival from kidney failure was higher among patients with fewer comorbidities (47% for low comorbidity). Older age groups had lower relative survival (22% for patients 91 to 95 years of age). Men and women had similar relative survival (36%). Compared with White patients, relative survival was highest for Black patients and patients of other races (33% vs 48% vs 44%, respectively).
Compared with other chronic diseases, relative survival was lowest for patients with kidney failure. Kidney failure, dementia, and heart failure had the lowest relative survival (36%, 50%, and 61%, respectively) compared with stroke, lung disease, and peripheral arterial disease (78%, 79%, and 83%, respectively).
Patients with incident kidney failure ages 66 to 70 years of age had a survival comparable with adults without kidney failure roughly 86 to 90 years of age and 91 to 95 years of age, respectively.
The researchers cited some limitations to the study findings, including not including younger patients, the lack of data on patients with kidney failure not treated with dialysis or preemptive transplant, and the inability to compare relative survival to patients with kidney failure who pursue conservative, nonanalytic therapy. In addition, the authors said that relative survival estimates can be improved by narrowing the specificity of the covariates collected.
In summary, the researchers said, “Relative survival is an estimate of cause-specific survival, a measure that is adjusted for the differences in survival due to natural aging process and other causes of death. Relative survival using comorbidity-adjusted life tables gives us a less biased estimate of the mortality burden from kidney failure that may be easier to communicate to policy-makers and patients. Furthermore, it can be directly compared across age, sex, race, year, and comorbidity groups, which makes it ideal for national and international research where demographics vary. Our estimates help quantify the immense mortality burden due to kidney failure and underscore the importance of disease prevention efforts for older adults.”
© 2025 Mashup Media, LLC, a Formedics Property. All Rights Reserved.