Engaging Cardiology Providers in Quality Measurement.

See Article Segal et al.

The call to hold hospitals, provider groups, and individuals accountable for quality and value, through payment or other incentives, continues to grow. Most of the quality focus has been on hospitals with national rankings such as the publically available US News and World Report, Health Grades, and the private VIZIENT (formerly the University Health Care Consortium) rankings. These hospital rankings vary in how they incorporate multiple factors such as patient experience, safety, and structural measures of quality, but the observed‐to‐expected mortality rate is a major component of each. The Hospital Readmissions Reduction Program (HRRP) of the Center for Medicare and Medicaid Services (CMS) is another highly visible program that financially penalizes hospitals that are below average in the hospital rankings of 30‐day all‐cause readmission.1 Many of the patient populations that are targeted by these outcome metrics include those with cardiovascular disease (acute myocardial infarction, and heart failure). Accordingly, cardiology providers are often asked by their hospitals to improve mortality and readmission. Since these are adjusted rates or observed‐to‐expected ratios, documenting all patient illnesses is an important part of lowering observed‐to‐expected mortality and readmission ratios. Beyond better documentation, it is unclear whether providers know how to directly improve outcome measures.

In this issue of the Journal of the American Heart Association (JAHA), Segal and colleagues help address the question of which types of performance measures are useful to providers.2 They report on the views on performance measurement of cardiovascular clinicians from the Department of Veterans Affairs (VA). The VA has a long history of quality measurement, primarily at the hospital level. For many years the VA has invested a large amount of resources using trained chart abstracters to review medical records to be sure that process of care measures are accurate.3 For example, assessment of the left ventricular ejection fraction and any provider‐documented contraindication or intolerance to medication allowed the VA to determine a valid numerator and denominator for many heart failure medication use measures. These process of care measures have been routinely reported at the VA facility level. More recently, the VA has developed Strategic Analytics for Improvement and Learning Value Model (SAIL). The SAIL metrics include 25 quality measures covering areas of mortality, readmission, complications, patient satisfaction, efficiency, and physician capacity. The VA uses these metrics to rank all facilities with a Star Rating (1 the worst, to 5 the best). By definition, a set percentage of VA facilities always receive the worst rating.4 These metrics emphasize outcome over process.

Through interviews, Segal and colleagues found that providers were most familiar with process of care measures and most importantly, providers used these data to modify their practice. This was in contrast to clinical outcomes where providers did not use the data to change practice. The lack of interest in outcome data by individual providers should not be surprising. While outcome metrics from clinical care and cost (value) may help leaders determine where quality improvement efforts are needed, the measures themselves are impractical for clinicians who change care via processes. Providers need to know how to improve care in order to alter an outcome measure. To fully engage providers, one needs to translate these outcome measures into actions providers feel capable of implementing. Breaking down a mortality metric so clinicians can compare specific dimensions of their performance to peer groups—to understand which patient subgroup is having higher mortality, the cause of increased death, and the timing and setting of these events—may make these metrics actionable.

Importantly, VA cardiovascular care providers appear to be favorable to the VA's current systems of quality measurement, indicating a helpful degree of enthusiasm for improvement efforts. This degree of enthusiasm will be important as the VA continues national efforts at modernizing the healthcare system and improving patient access.

In addition to hospitals passing on outcome measures to providers, payers also wish to provide incentives directly to individual providers to improve mortality and hospitalization. Such measures are currently under development by the Center for Medicare and Medicaid Services for the Merit‐based Incentive Payment System (MIPS). The work by Segal suggests such measures may be less effective than those focused on process of care.

Can outcome measures be successfully attributed to a single provider? This is highly unlikely given the growing use of team care. Similarly, it is a challenge to attribute cost of care to an individual provider. It is notable that Segal found that cost and value were not used by cardiovascular providers to inform their practice. While this may be because of less emphasis on hospital billing, collections, and profit margins in the VA system, it is likely that VA providers felt that cost of care is much more dependent on patient characteristics than on their practice choices. Cost metrics may be useful if they can be separated into components that suggest a specific process for improvement.

Cardiovascular clinicians face an increasing number of performance metrics, with many of these now focused on outcome and cost. In order to keep providers engaged, these measures need to be viewed as relevant and actionable. The work by Segal and colleagues shows us that such measures will be more accepted and acted upon if they focus on process of care measures. Outcome and cost measures may be useful if the results can be displayed in ways that suggest process of care changes for clinicians.

Disclosures

None.