Kaiser Permanente

CREATE Wellness


As a research assistant at Kaiser Permanente, I had the opportunity to design and manage a research study that examined various qualitative data sources collected during their program, CREATE Wellness. This program was meant to teach patients how to better engage in their chronic disease management, and included 3 group sessions facilitated by a health educator.

The goal of this study was to identify patient barriers to engaging with the health system for chronic disease management. We additionally sought to define intervention opportunities that could address these barriers. Study conception and data analysis was conducted over a period of 6 months (March 2020 - August 2020).

Identifying Data Sources and Methodology

In collaboration with a team of physicians, psychologists, and data analysts, the first step was to identify what existing qualitative sources we could investigate. Factors for consideration included whether or not data sources were cohesive enough for analysis, relevant to our study aims, and could provide a unique perspective of patient behavior. After I conducted an initial data assessment and delivered recommendations to the greater research team, we decided to focus on the following:

  • Secure Messages from Patients to Clinicians (n = 95)

  • Interviews with patients for post-program feedback (n = 320)

  • Fieldnotes taken by the health educator during group sessions (n = 332)

  • Physician notes taken during clinical visits (n = 95)

Analytic Process

With these sources in mind, I proposed an analytical plan that used an ethnographic approach to triangulate findings from these various sources. This additionally the use of inductive thematic analysis and the application of a psychological framework that assists with identifying behavioral barriers (COM-B framework). Specific steps included:

1. Using a general inductive analytic approach, our research team read the raw data and identified an initial set of thematic categories related to engaging with chronic disease management (e.g., “patient-provider interactions,” “patient preferences”).

2. Forming a codebook which was used by two analysts to assign text excerpts to each category using QSR International’s NVivo 12 qualitative data analysis software.

3. Meeting regularly with other analysts and the larger study team to develop a set of specific themes.

4. Mapping the qualitative themes onto the COM-B model (Capability, Opportunity, Motivation).

This process took 2 months to complete, as it involved several rounds of coding for thorough thematic analysis. This also involved the time I spent in training another research assistant on how to conduct qualitative coding and several weeks of comparing our findings to improve coding accuracy.  


After data analysis, we identified several behavioral determinants and described what motivational and environmental factors were crucial for patients to successfully engage in their chronic disease management. Given the complexity and volume of psychosocial factors, I additionally created conceptual maps and tables to clearly communicate the relationship between these various components. An example of this is provided below:

create com-b diagram.jpg

Barriers for engagement were related to communication skills and activation, care team relationship processes, and emotional factors; such as feelings of shame and distrust of the health system.

**The full description of results are limited as these findings are currently being reviewed for publication. Please contact for more information. **


After discussing with the larger study team on potential clinical practice implications, we recommended the following actions:

  • Support for prolonged goal-setting and review of personal health goals

  • Training or tools to aid physicians in having more empathetic and personalized interactions with patients.

  • Providing patients with the necessary skills to commincate concerns and build rapport with their physician.

  • Promoting shared decision-making practices and non-judgement support.


One of the biggest challenges for me was knowing how to effectively and clearly communicate my research findings with a diverse team of psychologists, physicians, and data analysts. It was also especially challenging trying to triangulate such a diverse set of qualitative resources, and I learned several best practices about ethnographic methodologies.

However, with the help of my mentors and team members, I was able to gain skills in these areas and successfully conduct a study from beginning to end. This experience inspired me to continue investigating behavioral factors within healthcare, and I hope to gain further expertise on ethnographic methodologies for various audiences.