Health Affairs has published a study that demonstrates little impact of Medicare’s mortality data on improving outcomes. No surprise here. Once again, we learn that data without context have modest motivational effects. While the data imply an urgent message, too often the response generates two questions that lack obvious answers: “So what?” and “Now what?”
What are the limitations of mortality data? All-cause mortality can contaminate the picture pertaining to a particular disease state or procedure, particularly in populations who are of advanced age, poor socioeconomic status, or limited health literacy. The complexities of adherence, care coordination, social support, and comorbidities result in a multivariable equation that can defy root cause analysis for a population that is not homogeneous.
While many process measures are unsatisfying, judicious linking of evidence-based care activities to outcomes provides end users a framework to design one or several interventions that can improve the broader metric, such as mortality. Clearly, any effort to improve mortality requires identification of components of care that should have priority attention to effect change. Progress tracked on sub metrics is more immediate and gives reassurance on eventual success or suggests need for further thought on effective improvement strategies to address the outcome of concern.
Focus on all-cause mortality and 30-day readmission rates presents a “Back to the Future” moment. Papers written in the early 90’s used Monte Carlo modeling to demonstrate the limited predictive value of readmission events in chronic disease. Twenty years later, on the eve of pay-for-performance penalties for heart failure, papers demonstrate limited change in global readmission rates, despite hospital QI initiatives. “Compelling data” do not always suggest solutions based on their existence alone
Clearly, mortality and readmission rates can improve, but mostly at the margins. Data should promote creative thought. However, without a breakout conceptual leap in processes, outcomes will remain the same and we’ll still be in need of a broad-based strategy for change.