This study is designed to test three hypotheses: 1) that co- payment obligations for Medicare-only insurance patients are a cause of non-compliance and a risk factor for graft survival; 2) that rates of compliance-related late graft failure (after year 1) differ by insurance coverage; and 3) that differences in health insurance coverage partially explain racial differences in graft survival. Hypotheses will be tested using the U.S. Renal Data System (USRDS), which includes 47,681 first transplants performed between 1989 and 1994. Three cohorts will be defined as Medicare only, private Medicare, and Medicaid Medicare. For each cohort, actuarial graft survival, compliance-related graft failure rates, and within-race actuarial graft survival will be determined. Cadaveric and living-related grafts will be analyzed separately, and outcome measures adjusted for 16 known risk factors. The data base will be analyzed using the Cox regression model with graft failure as the dependent variable. Insurance coverage, subject to epochal changes in financial liabilities, will be treated as a time-dependent explanatory covariate. Difference in distribution of graft failure times will be tested using the log likelihood ratio test. Univariate comparisons will be done by Mantel-Haenszel chi-square and ANOVA for unbalanced design.
| Ojo, A O; Wolfe, R A; Held, P J et al. (1997) Delayed graft function: risk factors and implications for renal allograft survival. Transplantation 63:968-74 |
