Less AND More Are Needed to Assess Primary Care

• Community Health Care
• Patient-centered Care
• Primary Health Care

Significant federal investment is now being directed toward lessening the burden of clinical quality measurement.^1,2 At the same time, there is growing recognition that current measures are inadequate to assess the domains of primary care that result in improved person and population health at sustainable cost.^3⇓–5 Our study reveals a significant gap between the universe of what is measured and those elements most critical to good quality primary care, indicating that efforts to reduce measurement burden must be accompanied by efforts to increase the relevance of measures to domains of care affecting population health outcomes.

Exemplified by the Institute of Medicine's Vital Signs report last year,¹ the United States has experienced an increase in calls to identify an effective means to assess and pay for health care. The foundational role of primary care draws attention to the need for broad assessment and support of accessible, coordinated, whole-person, relationship-based care.^3,5 Yet, funders, physicians, and policy makers agree: we have too many measures, creating tremendous administrative burden, leading to high cost and limited return.^1,2 In addition, most measures used share a myopic focus on clinical processes and limited short-term outcomes.³ National efforts to fix this problem have focused on reducing the number of measures on which primary care is required to report.² Although that effort may result in reduced administrative burden, it fails to address systemic gaps in the assessment of primary care characteristics most responsible for its added value and its ability to avoid the pitfalls associated with fragmented care.³

We administered an open-ended, electronic survey to primary care clinicians, allowing 1 to 5 free text responses to each of 2 questions, paraphrased: 1) how do you know good primary care when you see it, and 2) what questions would you ask a practice to know if they are helping to deliver health and wellness?⁶ Questions were first pilot-tested and then vetted among 30 multi-disciplinary primary care experts before fielding. The survey was distributed among 4 groups: practice-based research networks (n = 167), listservs (n = 8), a national cohort of innovating practices (n = 190),⁷ and a purchased list of 10,000 physicians evenly distributed among family medicine, internal medicine, and pediatrics. 412 clinicians provided 3524 unique survey responses.

Responses were coded using 92 codes, 27 of which were based on commonly used measures and 65 of which were based on code groups emergent from the data (see Table 1). Three coauthors reached agreement on code definitions and coded independently, using consensus to resolve any discrepancies. Forty-two percent of clinician responses could not be assigned measure-based codes, indicating a significant gap between how primary care is assessed and what those on the frontlines of its delivery identify as valuable. Concepts reflected among code groups using the (nonmeasure-based) emergent codes include ability to prioritize care, accurate problem recognition, management of patient complexity, focus on patient preferences and goals, investment in longitudinal relationships, and ability to adapt care based on personal and communal social determinants of health.

None of the emergent concepts share an overlap with current measurement focus on clinical processes and outcomes. Such misalignment risks inadequate reporting of the work of primary care, and chronic undermining of the role of primary care within the larger health care system.^4,5 Policies able to support both reduction in number of measures and creation of measures specific to primary care would allow for improved assessment of primary care and appropriate identification of areas on which to focus quality improvement.

• Received for publication June 28, 2016.
• Revision received September 9, 2016.
• Accepted for publication September 13, 2016.