Profiling dialysis facilities for adverse recurrent events.

Area of Science:

• Health Services Research
• Biostatistics
• Nephrology

Background:

• Healthcare provider profiling traditionally uses binary patient outcomes.
• Dialysis patients require regular monitoring for adverse events, including anemia.
• Existing profiling methods are insufficient for the complex count-based outcomes in dialysis care.

Purpose of the Study:

• To develop advanced statistical models for profiling dialysis facilities based on patient count/rate outcomes.
• To introduce a standardized event ratio for comparing facility-specific event rates to national benchmarks.
• To address challenges of high-dimensional parameters and overdispersion in dialysis patient profiling.

Main Methods:

• Application of high-dimensional Poisson and negative binomial regression models.
• Development of a standardized event ratio for relative performance assessment.
• Investigation of overdispersion's impact on statistical inference for facility profiling.

Main Results:

• The proposed methods effectively profile dialysis facilities using recurrent anemia events as an example outcome.
• The standardized event ratio provides a robust measure for comparing facility performance.
• The study demonstrates overcoming high-dimensional parameter challenges in large-scale healthcare data.

Conclusions:

• The new statistical framework enhances the evaluation of healthcare providers, specifically dialysis facilities.
• Accurate profiling of dialysis facilities can lead to improved patient care and outcomes.
• The methods are adaptable for profiling other healthcare settings with count-based outcomes.