Social determinants of health (e.g. the income, education, and environment of patients) may exert greater influence on health outcomes than traditional clinical factors (e.g. lab results, diagnoses, and family history). Calls for integrating primary care and public health are therefore increasing, but merging these domains of care is logistically difficult. Research is lacking on the incremental benefit of adding public health data at the practice level-- in improving either health outcomes or care delivery. This proof of concept pilot will merge data from electronic health records (EHRs) with community vital signs, a set of metrics that describes key community resources that affect health. The investigators will identify resource poor communities, or cold spots, based on four variables (education, poverty, life expectancy, and access to healthy foods) at the census tract level - referred to as a community vital sign. The hypothesis is that patients coming from cold spots are more likely to have worse health outcomes and that clinicians will deliver better care if they know a patient's community context and his/her specific social needs. This study will involve 12 primary care practices in Northern Virginia that care for more than 170,000 patients. Patient addresses will be geocoded for each practice and determine which patients reside in cold spots for each community vital sign. The variation for each community vital sign for each practice's patients will be calculated and a bivariate and regression analyses will be used to determine whether coming from a cold spot is associated with worse clinical quality metrics. 15 clinicians will be alerted when they see a patient from a cold spot, patients will complete a social needs survey, and clinicians will prospectively document through surveys whether such knowledge affects interpersonal interactions (such as time spent with patients and the use of clearer language) or clinical management (such as referrals to care coordination or community resources). By pragmatically integrating community vital signs into care, this innovative proposal will seek to understand which community data clinicians value, how these data might influence care, and how best to incorporate these data into clinical and population care.

The investigators will integrate community vital signs into clinical data, identify cold spots, and determine prospectively how clinicians use these community data at the point of care and for populations of patients. The investigators hypothesize that (1) the patients in a primary care practice live in a range of communities with different social determinants of health, including cold spots that lack health-promoting resources, (2) community vital signs are associated with patient health outcomes (e.g., being up-to-date with preventive care, control of chronic conditions, morbidity), and (3) knowing that a patient resides in a cold spot will help clinicians improve patient care.

Aim 1a: Determine community vital sign cold spots in the catchment area of 12 primary care practices.

Aim 1b: Determine whether patients residing in cold spots experience worse outcomes, such as receiving inadequate care, worse chronic disease control, or higher morbidity.

Aim 2: Determine whether knowing a patient's community vital signs and social needs at the point of care will change clinical management; and explore how practices could use community vital signs for population care.

Aim 3: Identify other community vital signs that practices would find useful for managing patients.

Inclusion Criteria:

• Aims 1a and 1b: Patients seen at one of the 12 study practices
• Aims 2 and 3: Clinicians at one of the 12 study practices

Exclusion Criteria:

• Aims 1a and 1b: Patients outside of the age range
• Aims 2 and 3: Clinicians involved with the design of the research protocol.

Aims 1a and 1b: The participants include all unique patients seen at the 12 study practices (approximately 170,000 patients over the past two years). Aims 2 and 3: The clinicians in this analysis will include 15 participants recruited from the approximately 100 clinicians at the 12 practices - ideally 1-2 clinicians from each office. Because the study will characterize the upper bounds, or ideal outcomes, in the use of community vital signs, clinicians who are willing to participate and provide feedback will be selectively recruited.