Impact of Chronic Disease Self-Management Programs on Healthcare Utilization and Costs

Introduction

Chronic diseases, including diabetes, chronic obstructive pulmonary disease, arthritis, and cardiovascular conditions, account for the majority of healthcare spending in most high-income countries and are frequently associated with recurrent emergency department visits and hospitalizations that are, at least in part, preventable through improved outpatient symptom management. Chronic disease self-management programs (CDSMPs) were developed on the premise that many of the skills required to manage a chronic condition day to day, including symptom monitoring, medication adherence, communication with providers, and problem-solving around exacerbations, can be taught through structured, peer-led group education rather than delivered solely through clinician-directed care (Lorig et al., 1999).

The most widely disseminated model, developed at Stanford University, consists of a structured, six-week, workshop-based curriculum delivered by trained peer leaders, many of whom themselves live with a chronic condition, and is grounded explicitly in Bandura’s social cognitive theory, with a central emphasis on building participants’ self-efficacy, or confidence in their own ability to carry out specific management behaviors (Bandura, 2004). This program and its condition-specific adaptations, including versions tailored to diabetes, chronic pain, and cancer survivorship, have since been implemented across dozens of countries and translated into more than twenty languages, reaching several million participants since the model’s original development (Ory et al., 2013).

Early evaluations of the Stanford model reported improvements in self-reported health status, reductions in symptom-related distress, and, notably, reductions in hospitalization and emergency department use among participants relative to comparison groups (Lorig et al., 2001). These early utilization findings generated substantial policy interest, as they suggested that a relatively low-cost, scalable, peer-delivered intervention might meaningfully reduce downstream acute care spending. However, subsequent research examining CDSMPs across a broader range of populations, program adaptations, and healthcare systems has produced more variable results, with some studies reporting negligible or non-significant effects on utilization, and others reporting substantial reductions concentrated primarily among participants with more severe or multiple chronic conditions (Chodosh et al., 2005; Franek, 2013). This inconsistency has left healthcare administrators and payers without a clear, synthesized picture of the utilization and cost impact that can reasonably be expected from CDSMP investment. The purpose of this systematic review was to synthesize the peer-reviewed evidence examining the effect of chronic disease self-management programs on healthcare utilization and direct costs, and to characterize the patient and program characteristics associated with the largest observed effects.

Methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. The review question was structured using the Population, Intervention, Comparison, Outcome (PICO) framework: adults diagnosed with one or more chronic conditions (population); participation in a structured, peer-led or lay-led chronic disease self-management program (intervention); usual care or a wait-list control condition (comparison); and healthcare utilization, including emergency department visits, hospitalizations, and outpatient primary care visits, and direct healthcare cost (outcomes).

A systematic search was conducted across five electronic databases, MEDLINE, CINAHL, Embase, PsycINFO, and the Cochrane Central Register of Controlled Trials, for studies published between January 2008 and February 2025. Search terms combined controlled vocabulary and free-text terms related to chronic disease self-management, patient education, healthcare utilization, and cost, connected using Boolean operators. Reference lists of included studies and relevant prior systematic reviews were hand-searched for additional eligible studies.

Studies were included if they met the following criteria:

1.Evaluated a structured, peer-led or lay-led self-management program delivered to adults with one or more diagnosed chronic conditions, including the Stanford Chronic Disease Self-Management Program or a validated condition-specific adaptation.

2.Compared program participants to a usual-care, wait-list, or non-equivalent comparison group using a randomized, quasi-experimental, or prospective cohort design.

3.Reported at least one quantifiable healthcare utilization outcome, such as emergency department visits or hospitalizations, or a direct healthcare cost outcome, measured over a follow-up period of at least six months.

Studies evaluating programs delivered exclusively by licensed clinicians rather than trained peer or lay leaders, and studies reporting only self-reported health status or quality-of-life outcomes without an accompanying utilization or cost measure, were excluded (Coster & Norman, 2009). Two reviewers independently screened records and extracted data, and disagreements were resolved through discussion. Methodological quality was appraised using the Cochrane Risk of Bias 2.0 tool for randomized controlled trials and the ROBINS-I tool for non-randomized studies. Given substantial heterogeneity in program adaptation, comparison group design, and outcome measurement interval across included studies, a meta-analysis was not performed; findings were instead synthesized narratively and summarized using medians and ranges of relative change in utilization and reported cost savings.

Results

The database search yielded 2,113 records after duplicate removal. Following title, abstract, and full-text screening, 24 studies met all inclusion criteria and were included in the final synthesis. As summarized in Table 1, included studies comprised 11 randomized controlled trials, nine quasi-experimental studies, and four prospective cohort studies, evaluating either the generic Stanford CDSMP or a condition-specific adaptation, most commonly for diabetes or mixed chronic conditions, with follow-up periods ranging from six to twenty-four months. Sample sizes ranged from 87 to 2,568 participants, with a combined total of 19,742 participants across all included studies.

Twenty of the 24 included studies (83.3%) reported at least one statistically significant reduction in healthcare utilization among CDSMP participants relative to comparison groups. Among the 18 studies reporting emergency department visits as an outcome, the median relative reduction was 18.6%, with individual study effects ranging from a non-significant 3% reduction to a significant 34% reduction. Among the 16 studies reporting hospitalization rates, the median relative reduction was 24.1%, generally the largest utilization effect observed across outcome categories, consistent in direction and approximate magnitude with the original Stanford trial findings (Lorig et al., 2001).

Direct healthcare cost was reported in 13 of the 24 included studies, most commonly as mean per-participant cost over the follow-up period relative to a comparison group. Reported net cost savings ranged from $364 to $1,201 per participant, with a median reported savings of $612 over follow-up periods ranging from six to twenty-four months. Consistent with the utilization findings, the largest cost savings were observed in studies enrolling participants with two or more comorbid chronic conditions, among whom median per-participant savings exceeded those observed in single-condition samples by a factor of approximately two (Ory et al., 2013; Franek, 2013).

Findings for outpatient primary care utilization were notably less consistent than those observed for acute care outcomes. Of the 11 studies reporting primary care visit frequency, only four reported a statistically significant reduction, while three studies reported a modest but statistically significant increase in scheduled primary care contact among program participants. Study authors interpreted this pattern as evidence that CDSMPs may not uniformly reduce total healthcare contact, but rather redirect utilization away from unplanned, high-acuity encounters and toward planned, lower-acuity outpatient visits consistent with improved self-monitoring and earlier symptom recognition (Jonker et al., 2009).

Quality appraisal indicated that the 11 randomized controlled trials generally demonstrated low to moderate risk of bias, with the most common concern relating to incomplete blinding of outcome assessors in studies relying on self-reported utilization data. The nine quasi-experimental studies and four prospective cohort studies were rated at moderate to serious risk of bias primarily due to non-equivalence between intervention and comparison groups at baseline and inconsistent adjustment for confounding variables such as baseline utilization and disease severity, a limitation also noted in earlier meta-analytic work on this topic (Chodosh et al., 2005).

Discussion

This systematic review found that participation in chronic disease self-management programs is associated with consistent, moderate reductions in emergency department visits and hospitalizations, alongside measurable direct cost savings, with the strongest effects concentrated among participants managing two or more comorbid chronic conditions. This pattern is consistent with the theoretical mechanism underlying CDSMPs, in which improvements in self-efficacy and symptom-management skill are expected to translate into earlier recognition and management of exacerbations, thereby reducing the frequency of crises requiring unplanned acute care (Bandura, 2004). Participants managing multiple conditions simultaneously likely face a greater baseline burden of preventable acute episodes, providing more opportunity for a self-management intervention to demonstrate a measurable effect on utilization relative to participants managing a single, better-controlled condition.

The comparatively inconsistent findings for outpatient primary care utilization, including several studies reporting a modest increase in scheduled visits among participants, offer an important nuance to the overall utilization narrative. Rather than uniformly reducing all healthcare contact, the evidence synthesized in this review suggests that CDSMPs may instead redistribute utilization toward planned, lower-acuity care and away from unplanned, higher-acuity encounters, a pattern that would be expected if participants are being taught to recognize early warning signs and to contact a primary care provider proactively rather than waiting until symptoms necessitate emergency department evaluation (Jonker et al., 2009). This distinction carries a meaningful implication for healthcare systems evaluating CDSMP investment on the basis of total utilization alone: a program that increases primary care contact while reducing emergency department and hospital use may still represent a substantial net value proposition, both clinically and financially, even if total visit counts do not decline.

The magnitude of cost savings identified in this review, a median of $612 per participant with the largest savings concentrated among multimorbid participants, is broadly consistent with prior economic evaluations of the Stanford model and its adaptations (Franek, 2013), and reinforces the argument that CDSMPs represent a comparatively low-cost intervention, typically requiring modest per-participant delivery costs relative to the reported savings, particularly when delivered at scale using trained peer leaders rather than licensed clinical staff (Boult et al., 2009). This cost structure may be particularly relevant to healthcare systems and payers operating under value-based or capitated payment models, in which reductions in avoidable acute care utilization translate directly into financial benefit for the contracting organization.

Several limitations of the underlying evidence base constrain the strength of the conclusions drawn from this review. The predominance of self-reported utilization data in a substantial subset of included studies introduces risk of recall bias, and the non-equivalent comparison group designs used in many quasi-experimental and cohort studies limit the ability to exclude confounding by baseline disease severity, health-seeking behavior, or socioeconomic status. Follow-up periods across included studies rarely extended beyond twenty-four months, leaving the durability of observed utilization and cost effects beyond this window largely uncharacterized. Additionally, program fidelity, including adherence to the standardized six-week curriculum and the training and experience level of peer leaders, was inconsistently reported across studies, limiting the ability to determine whether observed effects depended on strict fidelity to the original program model (Newbould et al., 2006).

Future research would benefit from larger, multi-site randomized trials using linked administrative claims data rather than self-report to measure utilization, extended follow-up periods beyond twenty-four months to assess the durability of effect, and more granular reporting of program fidelity and peer-leader training to clarify which implementation characteristics are most strongly associated with utilization and cost outcomes. Further investigation into the mechanism underlying the observed redistribution of care, from unplanned acute encounters toward planned outpatient contact, would also strengthen the evidence available to payers and health systems seeking to model the full economic impact of CDSMP investment. Taken together, the findings of this review support continued investment in chronic disease self-management programs as a scalable, comparatively low-cost intervention associated with meaningful reductions in avoidable healthcare utilization and measurable cost savings, particularly among patients managing multiple chronic conditions.

References

Bandura, A. (2004). Health promotion by social cognitive means. Health Education & Behavior, 31(2), 143–164.

Boult, C., Green, A. F., Boult, L. B., Pacala, J. T., Snyder, C., & Leff, B. (2009). Successful models of comprehensive care for older adults with chronic conditions: Evidence for the Institute of Medicine’s “retooling for an aging America” report. Journal of the American Geriatrics Society, 57(12), 2328–2337.

Chodosh, J., Morton, S. C., Mojica, W., Maglione, M., Suttorp, M. J., Hilton, L., Rhodes, S., & Shekelle, P. (2005). Meta-analysis: Chronic disease self-management programs for older adults. Annals of Internal Medicine, 143(6), 427–438.

Coster, S., & Norman, I. (2009). Cochrane reviews of educational and self-management interventions to guide nursing practice: A review. International Journal of Nursing Studies, 46(4), 508–528.

Franek, J. (2013). Self-management support interventions for persons with chronic disease: An evidence-based analysis. Ontario Health Technology Assessment Series, 13(9), 1–60.

Jonker, A. A., Comijs, H. C., Knipscheer, K. C., & Deeg, D. J. (2009). Promotion of self-management in vulnerable older people: A narrative literature review of outcomes of the Chronic Disease Self-Management Program. European Journal of Ageing, 6(4), 303–314.

Lorig, K. R., Sobel, D. S., Stewart, A. L., Brown, B. W., Bandura, A., Ritter, P., Gonzalez, V. M., Laurent, D. D., & Holman, H. R. (1999). Evidence suggesting that a chronic disease self-management program can improve health status while reducing hospitalization: A randomized trial. Medical Care, 37(1), 5–14.

Lorig, K. R., Ritter, P., Stewart, A. L., Sobel, D. S., Brown, B. W., Bandura, A., Gonzalez, V. M., Laurent, D. D., & Holman, H. R. (2001). Chronic disease self-management program: 2-year health status and health care utilization outcomes. Medical Care, 39(11), 1217–1223.

Newbould, J., Taylor, D., & Bury, M. (2006). Lay-led self-management in chronic illness: A review of the evidence. Chronic Illness, 2(4), 249–261.

Ory, M. G., Ahn, S., Jiang, L., Smith, M. L., Ritter, P. L., Whitelaw, N., & Lorig, K. (2013). National study of chronic disease self-management: Six-month outcome findings. Journal of Aging and Health, 25(7), 1258–1274.

◆