Random or predictable?: Adoption patterns of chronic care management practices in physician organizations

Abstract

Background: Theories, models, and frameworks used by implementation science, including Diffusion of Innovations, tend to focus on the adoption of one innovation, when often organizations may be facing multiple simultaneous adoption decisions. For instance, despite evidence that care management practices (CMPs) are helpful in managing chronic illness, there is still uneven adoption by physician organizations. This exploratory paper leverages this natural variation in uptake to describe inter-organizational patterns in adoption of CMPs and to better understand how adoption choices may be related to one another.

Methods: We assessed a cross section of national survey data from physician organizations reporting on the use of 20 CMPs (5 each for asthma, congestive heart failure, depression, and diabetes). Item response theory was used to explore patterns in adoption, first considering all 20 CMPs together and then by subsets according to disease focus or CMP type (e.g., registries, patient reminders). Mokken scale analysis explored whether adoption choices were linked by disease focus or CMP type and whether a consistent ordering of adoption choices was present.

Results: The Mokken scale for all 20 CMPs demonstrated medium scalability (H = 0.43), but no consistent ordering. Scales for subsets of CMPs sharing a disease focus had medium scalability (0.4 < H < 0.5), while subsets sharing a CMP type had strong scalability (H > 0.5). Scales for CMP type consistently ranked diabetes CMPs as most adoptable and depression CMPs as least adoptable. Within disease focus scales, patient reminders were ranked as the most adoptable CMP, while clinician feedback and patient education were ranked the least adoptable.

Conclusions: Patterns of adoption indicate that innovation characteristics may influence adoption. CMP dissemination efforts may be strengthened by encouraging traditionally non-adopting organizations to focus on more adoptable practices first and then describing a pathway for the adoption of subsequent CMPs. Clarifying why certain CMPs are "less adoptable" may also provide insights into how to overcome CMP adoption constraints.

Keywords: Adoption; Care management practices; Chronic care; Physician organizations.

Fig. 1

Overall distribution of care management practice adoption by physician organizations (N = 1398)

Fig. 2

Adoption frequencies for care management practices, grouped by CMP type

Fig. 3

Mokken scale analysis results for scalability. Legend: asterisks indicate reminders

Fig. 4

Ordering within disease focus and CMP type scales. Legend: asterisks indicate reminders note: Diabetes scale and overarching scales are not in the figure given their lack of ordering present. Each scale’s most adoptable practices is in the largest box and the least adoptable in the smallest. For disease-focused scales, feedback and education were both the least adoptable practices when run in separate scales and thus have equal rankings.