Survival effect of early nephrology care prior to dialysis initiation in elderly patients with end-stage renal disease

Early nephrology care may reduce mortality in patients who later progress to endstage renal disease (ESRD) and start dialysis. The survival effect of early nephrology care prior to dialysis initiation, however, is little evident. Previous studies using claims data may bias the estimation of the survival effect of early nephrology care due to omitted confounders. Under certain assumptions, instrumental variables analysis may yield consistent estimates of the survival effect for a subset of ESRD patients, the marginal patients whose treatment choices are affected by instrumental variables.; The research objectives were to explore whether non-clinical factors affect elderly Medicare patients' choices of early nephrology care defined as seeing a nephrologists for disease evaluation and management between 24 and 13 months and between 12 and 4 months prior to dialysis initiation as well as late nephrology care defined as seeing a nephrologist within 3 months prior to dialysis initiation, and to examine whether greater use of early and late nephrology care improves the first and second half-year survival of dialysis for the marginal patients.; Multivariate regression models showed that the use of early and late nephrology care was significantly affected by three instrumental variables, non-nephrologist physicians per 10,000 population and nephrologists per 1,000 prevalent elderly ESRD patients in a local area as well as the distance from a patient's residence to the nearest nephrologist. After controlling patient demographics, comorbidities and other covariates, multivariate regressions indicated significantly positive survival effects of early and late nephrology care on the short-term survival for the patients who actually received nephrology care but may differ considerably from the patients who might get early and late nephrology care. Instrumental variables models showed no significant survival benefits in the first and second half year of dialysis among the marginal patients whose selection of early and late nephrology care prior to dialysis initiation was affected by the three instrumental variables. Increasing the use of early nephrology care in the marginal patients may not significantly enhance the short-term dialysis survival rates. Treatment benefits of early nephrology care prior to dialysis initiation need to be further explored.