3091 — Lack of Structured Family History Capture in CPRS Limits Opportunities for Risk Assessment and Prevention

Scheuner MT (VA Greater Los Angeles, Center for the Study of Healthcare Provider Behavior), Rubenstein LV (Center for the Study of Healthcare Provider Behavior), Oishi S (Center for the Study of Healthcare Provider Behavior), Simon B (Center for the Study of Healthcare Provider Behavior), Austin C (Center for the Study of Healthcare Provider Behavior), Goldzweig C (Center for the Study of Healthcare Provider Behavior), Yano EM (Center for the Study of Healthcare Provider Behavior)

Objectives:
Health professionals often fail to document relevant family history that can be used to assess risk and influence management and prevention. Electronic health records (EHRs) could facilitate effective use of family history if they standardized data collection and organization, and if they included decision support systems that educated clinicians at the point of care about the utility of familial risk assessment. We sought to ascertain how well VA’s computerized patient record system (CPRS) captures family history by evaluating existing electronic mechanisms in use at one large VAMC.

Methods:
We reviewed all shared templates that could be imported into progress notes (PNs), as well as any PN with an embedded template used at least once between August 15, 2007 and August 14, 2008 at VA Greater Los Angeles. We evaluated the templates and PNs for family history content and format. We excluded PNs and templates if they pertained to procedures, examinations, treatment plans, administrative processes, education/training, letters, research, or were notes from services/specialties without direct patient care responsibilities.

Results:
We identified 1,416 eligible shared templates and PNs, and found reference to family history content in only 7.9%. The most common format was the disease checklist (48%), with an average of 1.6 diseases listed per template (range 1-16). A family history text box without prompts was found in 38%, and the remaining 14% listed first-degree relatives with an accompanying text box to input additional information. None of the formats captured information about specific diseases for specific relatives, second-degree or more distant relatives, ethnicity/race of relatives, relatives’ current age or age at death, or age at disease onset.

Implications:
At one VA, PNs and templates in CPRS lack standards for family history data elements and format, and the data captured with current formats was not adequate for assessing and stratifying familial risk.

Impacts:
If family history is to be used effectively to identify patients at risk and inform medical decisions, more robust EHR tools may need to be developed so that family history information can be easily identifiable in the electronic record. This may require changes to CPRS organization and functionalities.