Quality monitoring (QM) programs have been established to ensure that acceptable standards of care are delivered at US dialysis facilities. Global improvements have been made, but novel approaches are needed. The current QM reports do not account for differences in case-mix severity among dialysis facilities. Units caring for sicker patients are unfairly scrutinized while those with substandard care pass unnoticed. The assurances and objectives of a QM program are undermined if reporting methods are biased. The growing disparity between health care costs and resources threatens the quality of care delivered. As dialysis facilities are pressured to achieve results but struggle to balance budgets, facilities may select against high-risk patients. There is a great need to identify unbiased quality measures that promise improved outcomes. A unique strength of this proposal is the wealth of comorbidity and case-mix data collected in three contemporary and recent dialysis populations (n=5490). The three populations will be pooled. Case-mix severity scores for mortality and hospitalization use will be developed through multivariable regression modeling. Hierarchical models will be constructed to assess relationships of patient-specific and facility-specific factors with a number of process and outcome quality indicators, accounting for the clustering of patients within facilities. Also, through more complete accounting for of case-mix differences, relationships of facility-specific practices that are hypothesized to impact on process and outcomes, will be identified. A prospective study to test the feasibility of collecting comorbidity in routine practice and the validity of the predictive instruments will be initiated. The principal investigator holds a Master's degree in Epidemiology and the career development program will build on the candidate's skills and experience to enable her to lead future studies to improve outcomes patients with chronic kidney disease
