Health Insurance and the Use of Peritoneal Dialysis in the United States

Abstract

Background: Many patients in the United States have limited or no health insurance at the time they develop end-stage renal disease (ESRD). We examined whether health insurance limitations affected the likelihood of peritoneal dialysis (PD) use.

Study design: Retrospective cohort analysis of patients from the US Renal Data System initiating dialysis therapy in 2006 through 2012.

Setting & participants: We identified socioeconomically similar groups of patients to examine the association between health insurance and PD use. Patients aged 60 to 64 years with "limited insurance" (defined as having Medicaid or no insurance) at ESRD onset were compared with patients aged 66 to 70 years who were dually eligible for Medicare and Medicaid at ESRD onset.

Predictor: Type of insurance coverage at ESRD onset.

Outcomes: The likelihoods of receiving PD before dialysis month 4, when all patients qualified for Medicare due to ESRD, and of switching to PD therapy following receipt of Medicare.

Results: After adjusting for observable patient and geographic differences, patients with limited insurance had an absolute 2.4% (95% CI, 1.1%-3.7%) lower probability of PD use by dialysis month 4 compared with patients with Medicare at ESRD onset. The association between insurance and PD use reversed when patients became Medicare eligible; patients with limited insurance had a 3-fold higher rate of switching to PD therapy between months 4 and 12 of dialysis (HR, 2.9; 95% CI, 1.8-4.6) compared with patients with Medicare at ESRD onset.

Limitations: Because this study was observational, there is a potential for bias from unmeasured patient-level factors.

Conclusions: Despite Medicare's policy of covering patients in the month that they initiate PD therapy, insurance limitations remain a barrier to PD use for many patients. Educating providers about Medicare reimbursement policy and expanding access to pre-ESRD education and training may help overcome these barriers.

Keywords: Health insurance; Medicaid; Medicare; PD use; US Renal Data System (USRDS); dialysis modality; end-stage renal disease (ESRD); health disparities; health economics; hemodialysis (HD); insurance coverage; peritoneal dialysis (PD); renal replacement therapy (RRT).

Figure 1. Unadjusted Hazard of Switching to Peritoneal Dialysis among Patients Receiving Hemodialysis by the 4^th Month of Dialysis

Note: Limited Insurance includes patient who are uninsured and who have Medicaid at the onset of dialysis. Cumulative hazards calculated using the Kaplan Meier method.

Figure 2. Age-Stratified Predicted Probabilities of Peritoneal Dialysis by the 4^th Dialysis Month before versus after Enactment of the ESRD Prospective Payment System

Note: Patients aged 60 to 64 years have Limited Insurance (i.e. are uninsured or have Medicaid-only) at dialysis initiation, while patients aged 66 to 70 years have Medicare at dialysis initiation. Predicted probabilities are derived from our logistic model estimates in a model that allowed for the effect of insurance to vary following enactment of ESRD Prospective Payment System (PPS). Predicted probabilities of peritoneal dialysis use among patients with Limited Insurance are plotted both before and after the PPS. Standard errors are obtained using the delta method. Residual confounding by age may exist as a result of non-overlap of age groups by design. This residual confounding is most likely to induce a bias towards the null, thus leading to an underestimation of true effect.

Figure 3. Unadjusted Rate of Switching to Peritoneal Dialysis before and after Medicare Eligibility based on the 90-day Waiting Period among Patients with Limited Insurance and Medicare at the onset of End-Stage Renal Disease

Note: This figure illustrates daily rates of switching to peritoneal dialysis averaged over 10-day intervals. The shaded area represents time after patients would be eligible for Medicare based on the 90-day waiting period. For a more detailed description of how these rates were obtained, see Item S2.