Utilizing EHR Data to Study and Measure the Relationship Between Primary Care Teams and Patient Activation

Team-based care has been linked to key outcomes associated with the Quadruple Aim including improving the health of populations, patient and provider experience and lowering healthcare costs. Less is understood about the connection between team-based care and the patient experience. Emerging evidence connects team-based care with patient activation, a component of the patient experience. Use of the Electronic Health Record (EHR) and machine learning have significant potential to overcome previous barriers in how teams are studied to better understand their impact on critical care delivery outcomes, such as patient activation. This research program included a systematic review of the literature to analyze the relationship between team-based care and patient satisfaction, a proxy for the patient experience. Overall, this review found a positive relationship between team-based care and patient satisfaction, including 57% of studies with improved patient satisfaction with team-based care implementation. Secondary findings included a relationship between team composition and patient satisfaction, with larger teams (three or more disciplines) associated with improved patient satisfaction. A methodological paper was then prepared to describe the process in which primary care teams were identified within EHR data utilizing a common definition for team-based care supported by prominent team theorists. This novel approach provides a roadmap for the health services researcher to leverage EHR data to study the impact teams may have on critical patient outcomes in the real-world practice environment. The final study in this work utilized a large EHR data set (n = 316,542) from an urban health system to examine the relationship between team composition and patient activation. Patient Activation was measured using the Patient Activation Measure (PAM). Results from mixed-level model analysis were compared to machine learning analysis using multinomial logistic regression to calculate propensity scores for the multiple effect of team composition. After controlling for confounding variables in both analyses, more diverse, multidisciplinary teams were associated with improved patient activation scores. Implications for this research program include the feasibility of identifying teams within the EHR and utilize big data analytics with machine learning to measure the impact of teams and real-world patient related outcomes.