Multi-Sectoral Partnerships and Patient-Engagement Strategies in Accountable Care Organizations

Multi-sector partnerships and patient engagement are increasingly encouraged approaches to improve population health. The “Accountable Health Communities” initiative ^1,2 proposed by the Centers for Medicare and Medicaid (CMS) and “Building a Culture of Health”³ investments by the Robert Wood Johnson Foundation (RWJF) are two large scale efforts to bolster multi-sector partnerships and patient engagement. Patient engagement strategies aim to support patient self-care and health decision-making. Health care systems traditionally have not addressed upstream health factors and social needs. Partnerships between medical practices and other organizations such as faith-based organizations, local public health departments, and social services/community development agencies may support medical practices’ efforts to better engage patients, thereby promoting health and reducing downstream costs of care.⁴

Accountable Care Organizations (ACOs) may be well-positioned to cultivate multi-sector partnerships and patient engagement strategies given their responsibilities to meet triple aim targets: lowering costs, increasing quality and satisfaction, and improving population health.⁵ To our knowledge, no research has examined associations between multi-sector partnerships and implementation of patient engagement strategies in ACO-affiliated practices.

Methods:

One clinical or administrative leader was surveyed from each of 71 adult primary care practices in two ACOs (October–December 2014, response rate =100%). One ACO includes a multispecialty medical group with multiple hospital affiliations that serve the greater Los Angeles area; the other ACO includes a multispecialty medical group and a hospital system that serves the greater Chicago area. The ACOs were selected because they are examples of different large ACOs with risk-based contracting and have real incentives to engage patients for both quality and cost containment purposes.

Questions assessed use of 12 different partnership sectors (see Table 1) and 14 patient engagement strategies based on a comprehensive review of the patient engagement literature. The 12 partnership sectors represent medical, social, and community-based organizations that could influence patient engagement reach and population health. The 14 patient engagement strategies were selected from 39 total strategies because they may necessitate interface and coordination with outside organizations. Respondents reported full, partial, irregular, or no implementation for each patient engagement strategy, recoded for interpretability to “yes” if full or partial implementation and “no” if irregular or no implementation.

Table 1. Mean number of patient engagement strategies among practices with and without multi-sector partnerships.

	% Practices with Partner (N)	Strategies a with Partner	Strategies a without Partner	Difference (95%CI)
Medical Partnerships
Hospital	74.7% (53)	7.6	5.8	1.8 (−0.6 – 3.6)
Insurer	64.8% (46)	7.4	6.6	0.8 (−1.2 – 2.6)
Home Health	63.4% (45)	7.3	7.0	0.3 (−1.8 – 2.0)
Non-Medical Partnerships
Faith	26.8% (19)	8.9	6.5	2.4 (0.9 – 3.9) **
Public Health Department	23.9% (17)	8.9	6.6	2.3 (0.6 – 4.0) **
Transportation	21.1% (15)	8.3	6.8	1.5 (−0.2 – 3.3)
School	14.1% (10)	8.3	6.9	1.4 (−0.8 – 3.6)
Housing	9.9% (7)	10	6.8	3.2 (0.6 – 5.9) *
Parks & Recreation	7.0% (5)	7.4	7.1	0.3 (−3.8 – 4.4)
University	7.0% (5)	8.2	7.0	1.2 (−3.1 – 5.4)
Utility	7.0% (5)	9.0	7.0	2 (−0.5 – 4.6)
Library	1.4% (1)	Too few to calculate

^*= p-value <0.05,

^**<0.01,

^***<0.001

^a.List of 14 patient engagement strategies analyzed is described in results

Specific patient engagement strategies included: 1) practice refers patients for disease prevention/health promotion based on a health risk assessment; 2) practice encourages patient participation in a program for healthy eating, 3) physical activity, or 4) employee wellness; 5) practice sponsors or participates in school health clinic interventions; 6) clinicians encourage patients to discuss work, home, and social life; 7) select staff serve as “health coaches” for patients seeking to modify their lifestyle; 8) an organized follow-up program assists patients in managing medications at home; 9) practice has implemented group visits for patients with diabetes or 10) cardiovascular disease; 11) practice has implemented patient-to-patient programs for diabetes or 12) cardiovascular disease; and 13) programs exist to improve family participation and support for patients with diabetes or 14) cardiovascular disease.

Partnerships (12 overall or 9 non-medical partnerships) and patient engagement strategies (14 strategies) were each summed to create composite scores. To examine bivariate associations, X²-tests compared use of each partnership with use of each patient engagement strategy (data not shown) and t-tests compared use of each partnership sector with the number of overall patient engagement strategies (Table 1). Finally, multivariate linear regression models estimated the extent to which 1) overall partnerships and 2) non-medical partnerships were associated with greater use of engagement strategies, controlling for the ACO with which each practice was affiliated since geography, contracting and organizational structures, or other ACO-based characteristics may influence the relationship between practices’ partnerships and patient engagement strategies. The study was approved by the XXX Committee for Protection of Human Subjects (2014–086613).

Results:

Practices reported partnering with a mean of 3.2 of the 12 sectors (SD=2.1; range = 1–9). Traditional medical partnerships—hospitals (74.7%), insurers (64.8%), and home health agencies (63.4%)—were most common. Partnerships with parks and recreation (7.0%) universities (7.0%), utility companies (7.0%) and libraries (1.4%) were most rare (Table 1).

Practices also reported implementing a mean of 7.1 of 14 patient engagement strategies (SD=3.4; range = 0–14). The most commonly implemented strategies were encouraging participation in a physical activity program (88.7%) or a healthy eating program (85.9%). The least implemented strategies were group visits for cardiovascular disease (16.9%), school health clinic interventions (16.9%), and patient-to-patient programs for cardiovascular disease (11.3%).

In bivariate analyses, each partnership sector was associated with implementation of 1 to 4 individual patient engagement strategies (p<0.05). For example, hospital partnership was significantly associated with 4 individual patient engagement strategies: disease prevention/health promotion referrals, healthy eating program participation encouraged, physical activity program participation encouraged, and discussion of work, home, and social life encouraged (p<0.01 for each strategy). The total number of patient engagement strategies implemented did not significantly differ between practices with and without hospital partnerships, but was significantly greater for practices partnering with faith-based organizations, local public health departments, or housing agencies (p<0.05) (Table 1).

In multivariate analyses, controlling for ACO affiliation, each additional partnership was significantly associated with implementing more patient engagement strategies (β =0.59; 95% CI 0.23–0.95). When analysis was restricted to non-medical partnerships, the association increased (β =0.92; 95% CI 0.41–1.43) (Table 2).

Table 2. Association between total partnerships and patient engagement strategies, controlling for ACO affiliation.

All Partnerships
		Observations = 71
		R2 = 0.15
		F-statistic = 5.8, p-val = 0.005
	B-Coefficient	95% CI
All Partnerships Composite	0.59	0.23 – 0.95 **
ACO Affiliation	1.36	−0.23 – 2.95
Non-Medical Partnerships
		Observations = 71
		R2 = 0.15
		F-statistic = 7.1, p-value = 0.002
	B-Coefficient	95% CI
Non-Medical Partnerships Composite	0.92	0.41 – 1.43 ***
ACO Affiliation	1.43	−0.13 – 3.00

^*= p-value <0.05,

^**<0.01,

^***<0.001

Implications:

Our results provide the first evidence on the breadth of multi-sector partnerships that ACO-affiliated practices are developing. We found that a greater number of multi-sector partnerships developed by practices was significantly associated with more extensive implementation of patient engagement strategies, supporting hypotheses that multi-sector partnerships may enable and support patient engagement strategies.

We also found that restricting analyses to non-medical partnerships strengthened the association between multi-sector partnerships and the implementation of patient engagement strategies. In particular, practices partnering with faith-based organizations, local public health departments, or housing agencies tended to use more patient engagement strategies than non-partnering practices. These associations support the importance of non-medical partnerships in promoting opportunities for patients to more fully engage in improving their health.

Our results should be considered in light of some limitations. First, we are unable to make causal conclusions based on the cross-sectional study data. While the survey response was 100%, the sample was small (n=71 practices), and since all practices were affiliated with one of two ACOs, results may not generalize to other ACOs. Also, in effort to keep the survey brief, information such as practice size, payer mix, patient demographics, and other organizational characteristics were not collected. The inclusion of these omitted variables in multivariate analyses might alter the association between multi-sector partnerships and patient engagement factors. Future research examining the relation of multi-sector partnerships and patient engagement should assess practice-level factors that might confound the associations examined. In addition, qualitative research could explore how multi-sector partnerships and patient engagement activities are interwoven and where the greatest opportunities to address social needs can leverage partnerships to better activate and engage patients.

The movement to bridge traditional medical care and public health might be accelerated by efforts to support primary care practices in developing and using multi-sector partnerships to improve patient engagement. More evidence is needed to inform approaches for primary care practices to integrate multi-sector partnerships. The wide range in number of partnership sectors reported in our analyses (1–9) suggests that practices with fewer partnerships may be able to learn from highly partnered practices to increase partnership development. In addition, it may be that combinations of specific partnerships can be complementary and resources can be pooled for greater impact. Researchers, policy makers, and practice-based implementers have much to learn about optimizing multi-sector partnerships in a way that more effectively engages patients in managing their own care, improves the quality of care, and ultimately improves population health.

Summary Box.

What is already known about the topic?

Multi-sector partnerships have potential to help bridge medical care and population health improvement efforts. Accountable Care Organizations may be well positioned to leverage multi-sector partnerships for patient engagement, but little is known about the connection between practice use of partnerships and implementation of patient engagement strategies.

What is added by this report?

ACO affiliated practices vary in their use of multi-sector partnerships and implementation of patient engagement activities. Non-medical partnerships are rare, yet several sectors including faith-based organizations, local public health departments, and housing agencies were associated with greater use of patient engagement strategies in bivariate analysis. Practices with a greater number of partnerships implemented a greater number of patient engagement strategies, particularly for practices with more non-medical partnerships.

What are the implications for public health practice/policy/research?

The movement to bridge traditional medical care and public health might be accelerated by efforts to support primary care practice use of multi-sector partnerships to improve patient engagement. More evidence is needed to inform how primary care practices should integrate multi-sector partnerships to improve the reach and effectiveness of patient engagement efforts.