This study reaffirms the importance and potential of community-based self-management interventions rooted in public health to control healthcare costs among adults with chronic conditions. Extrapolating the estimated $364 cost saving per CDSMP participant results in meaningful national savings (i.e., ~$3.3 billion), if program penetration reaches only 5% of all individuals with one or more chronic condition. Assuming a $350 average CDSMP cost per participant, we could achieve potential national healthcare savings from $0.7 billion to $65.7 billion by averting from ER visits and hospitalizations if CDSMP reaches a minimum of 1% of adults having at least 1 chronic condition (i.e., the range of national savings depends on the level of program penetration). The cost savings achieved herein among heterogeneous populations served by diverse organizations were substantial; however, they were slightly lower than those estimated in the original, more tightly controlled, randomized trial which included controls whose hospital-related healthcare costs increased [4].
Previous studies have documented the value of CDSMP in improving participants’ health behaviors, disease-related symptoms, communications with providers, and overall health status [4, 5]. With the addition of findings from the current study, it is clear that this intervention can influence all aspects of the Triple Aim (i.e., enhanced care, improved health, and better value) [6, 13]. Modest past investments by the U.S. Administration on Aging, the Centers for Disease Control and Prevention, and other agencies have established a viable foundation for scaling up this intervention. Within the past 5 years, over 150,000 people have participated in CDSMP workshops through the Communities Putting Prevention to Work Initiative and other public-private collaborations [12]. This highlights the probability of high-level CDSMP penetration to reach populations with chronic conditions as long as strong support and funding sources are available for this initiative. As such, additional public sector resources are needed to continue the momentum and leverage the existing infrastructure. At the same time, private insurers are highly encouraged to provide benefits to their patients with chronic conditions by discounting their premiums as well as CDSMP workshop delivery agencies by providing generous reimbursements.
A policy issue of interest surrounds how efficiencies in CDSMP workshop delivery can increase overall savings. In the current study, we have used an estimated $350 per participant cost in our calculations assuming 10 participants in a workshop and $3,500 workshop costs. To estimate this average cost, we relied on experts’ opinions including the program developers, field reports (ranging from $204 to $375) [14, 15], and an unpublished survey among state CDSMP implementers conducted by the National Council on Aging (ranging from $150 to $750) [16]. However, it is worth noting that the cost estimation should vary by the number of CDSMP participants in a workshop and the administrative capacity of the delivering agencies. Nevertheless, we would expect costs to decrease with the efficiencies gained through capacity building accompanying widespread dissemination. A prior study projected lower average CDSMP costs for agencies with higher numbers of participants over time compared to the other agencies with lower numbers of participants, which emphasizes the importance of strong recruitment efforts and collaborating with community partners [15]. Further studies are warranted to identify how costs associated with marketing and administration would be affected when scaling up for widespread program delivery.
It is important to ensure that the cost-saving benefits of CDSMP equitably reach various populations despite geographic location and demographic factors. The small enrollment in some workshops highlights potential difficulties of scaling up in rural areas as these areas typically have smaller class sizes due to population dispersion and lack of infrastructure supports. However, on the positive side, CDSMP is being widely disseminated throughout the United States (e.g., between 2010 and 2012 more than 100,000 participants enrolled in CDSMP programs sponsored by the Administration on Aging) [12]. A prior study analyzing demographic factors and disease profiles among more than 100,000 CDSMP participants (between 2010 and 2012) documented how representative CDSMP participants are of the adult population [12]. When comparing to 2010 U.S. Census, CDSMP participants tended to be more female (77.7%) and older (mean age = 67 years) compared to the Census (51%, 37 years) [11]. However, there were similarities in terms of rural residence and race/ethnicity. Approximately 25% of CDSMP participants resided in rural areas (compared to 19.3% of Census) and had a similar racial/ethnic composition (white of CDSMP: 66.4% vs. 63.7% of Census; African American: 21.5% vs. 12.2%; Hispanic: 17.0% vs. 16.3%; Asian/Pacific Islanders: 4.5% vs. 4.9%; American Indians: 1.6% vs. 0.7%) [17]. These statistics are encouraging, especially when considering the capacity of evidence-based interventions to reach various populations at risk of chronic conditions.
Additionally, the potential of CDSMP to contribute to cost savings while improving health status provides a strong incentive for alignment with Accountable Care Organizations, models of enhanced primary care, initiatives for dually eligible beneficiaries, and State Innovation Models. To better integrate and leverage CDSMP to improve healthcare organization and financing, new initiatives are needed to design and test ways to: 1) strengthen collaboration among healthcare organizations, community partners, and public health agencies; 2) establish useful quality measures related to self-management; and 3) incentivize providers to further support evidence-based approaches to self-management.
Study limitations
First, data were drawn from a national study with a pre-post design appropriate for addressing translational research questions. While the current study lacks a comparison group, improvements were generally similar as those reported in the original randomized trial with some attenuated cases [4]. Our current study design does not permit the elimination of potentially confounding factors influencing study outcomes. Second, healthcare utilization was self-reported resulting in the possibility of recall bias. Nevertheless, a prior study found high concordance between self-reported and objectively measured ER visits and inpatient use [18]. Third, the current 12-month study may require a longer study duration to conclude definitively the healthcare cost-saving effects of CDSMP; however, we expect to see sustainable effects of reducing ER visits based on prior 2-year study [4]. Last, the current study is based on a critical assumption that we can extrapolate the healthcare cost savings of the National CDSMP to the national level using census data. Therefore, the cost-saving effects of CDSMP should be further studied to account for demographic changes in the Americans population over time and variations by population’s disease profiles. Nevertheless, we caution that it will be difficult to accurately estimate cost-saving effects by specific chronic condition types given the presence of multiple chronic conditions and the multitude of different disease clusters. To provide context, a previous study reported that CDSMP participants have on average 2.2 chronic conditions (e.g., hypertension = 43.0%; arthritis = 40.8%; diabetes = 30.3%; depression = 19.5%) [12]. Future study is needed to examine the average per participant cost of CDSMP based on geographic locations and capacities of agencies to deliver CDSMP.
