Mapping variation in intervention design: a systematic review to develop a program theory for patient navigator programs

Abstract

Background: There is a great deal of variation in the design and delivery of patient navigator (PN) programs, making it difficult to design or adopt these interventions in new contexts. We (1) systematically reviewed the literature to generate a preliminary program theory to describe how patient navigator interventions are designed and delivered; and (2) describe how the resulting program theory was applied in context to inform a prototype for a patient navigator program.

Methods: The current study includes a secondary review of a larger systematic review. We reviewed studies included in the primary review to identify those that designed and evaluated programs to assist patients in accessing and/or adhering to care. We conducted a content analysis of included publications to describe the barriers targeted by PN interventions and the navigator activities addressing those barriers. A program theory was constructed by mapping patient navigator activities to corresponding constructs within the capability-opportunity-motivation model of behavior change (COM-B) model of behavior change. The program theory was then presented to individuals with chronic disease, healthcare providers, and system stakeholders, and refined iteratively based on feedback.

Results: Twenty one publications describing 19 patient navigator interventions were included. A total of 17 unique patient navigator activities were reported. The most common included providing education, facilitating referrals, providing social and emotional support, and supporting self-management. The majority of navigator activities targeted barriers to physical opportunity, including facilitating insurance claims, assistance with scheduling, and providing transportation. Across all interventions, navigator activities were designed to target a total of 20 patient barriers. Among interventions reporting positive effects, over two thirds targeted knowledge barriers, problems with scheduling, proactive re-scheduling following a missed appointment, and insurance. The final program design included a total of 13 navigator activities-10 informed by the original program theory and 3 unique activities informed by stakeholders.

Conclusions: There is considerable heterogeneity in intervention content across patient navigator interventions. Our results provide a schema from which to develop PN interventions and illustrate how an evidence-based model was used to develop a real-world PN intervention. Our findings also highlight a critical need to improve the reporting of intervention components to facilitate translation.

Systematic review registration: PROSPERO CRD42013005857.

Keywords: Behaviour Change Wheel; Chronic disease; Intervention design; Patient navigator.

Fig. 1

Study flow diagram

Fig. 2

Composite program theory underlying patient navigator interventions (n = 19). Potential patient navigator activities (a) are linked to the corresponding behavior construct that they target (b). These constructs are components of more generalized sources (c) that influence individual behavior. These sources directly influence the overall health behavior targeted by patient navigator interventions (d), which has a direct impact on patient-centered outcomes (e). Note: All outcomes listed were shown to be significantly impacted by patient navigator interventions. ED = emergency department; HCP = healthcare professional

Fig. 3

Final applied program theory of patient navigator intervention