Using the Prevention Impacts Simulation Model to Estimate Long-Term Impacts of Multisector Community Partnerships’ Efforts to Address Social Determinants of Health
RESEARCH BRIEF — Volume 20 — July 20, 2023
Amanda A. Honeycutt, PhD1; Benjamin Yarnoff, PhD2; Zohra Tayebali, BA1; LaShawn Glasgow, DrPH, MPH1; Karen Hacker, MD, MPH3 (View author affiliations)
Suggested citation for this article: Honeycutt AA, Yarnoff B, Tayebali Z, Glasgow L, Hacker K. Using the Prevention Impacts Simulation Model to Estimate Long-Term Impacts of Multisector Community Partnerships’ Efforts to Address Social Determinants of Health. Prev Chronic Dis 2023;20:220327. DOI: http://dx.doi.org/10.5888/pcd20.220327.
PEER REVIEWED
What is already known on this topic?
Public health plays a key role in addressing social determinants of health (SDOH), including supporting multisector community partnerships (MCPs); however, little is known about the longer-term impact of MCPs’ SDOH initiatives.
What is added by this report?
System dynamics modeling is an underused tool for informing and refining public health interventions. This report demonstrates how the Prevention Impacts Simulation Model (PRISM) can be incorporated in evaluations to estimate cumulative longer-term impacts of SDOH initiatives.
What are the implications for public health practice?
If sustained, the initiatives we studied could avert hundreds of deaths and avoid half a billion dollars in costs over 20 years. As a validated model that estimates impact using available implementation data, PRISM is a useful tool for rapid evaluation of SDOH initiatives.
Abstract
Public health plays a key role in addressing social determinants of health (SDOH) through multisector community partnerships (MCPs), which contribute to community changes that promote healthy living; however, little is known about the longer-term impact of MCP-driven interventions. We used the Prevention Impacts Simulation Model (PRISM) in a rapid evaluation to better understand the implementation and potential impact of MCPs’ SDOH initiatives. Results suggest that, if sustained, initiatives implemented by the 27 included MCPs may prevent 880 premature deaths and avert $125.7 million in medical costs over 20 years. As a validated model that estimates impact by using available implementation data, PRISM is a useful tool for evaluating SDOH initiatives.
Objective
Chronic diseases are leading causes of illness, death, and health care costs in our nation (1–3). To reduce the chronic disease burden, it is essential to address underlying social determinants of health (SDOH) (4). However, addressing SDOH is challenging; it requires various intervention approaches across multiple sectors (2,4). Public health’s role in addressing SDOH includes supporting multisector community partnerships (MCPs), but little is known about the longer-term impact of the interventions MCPs promote (5,6).
Improving Social Determinants of Health — Getting Further Faster (GFF) is a rapid retrospective evaluation designed to better understand the implementation and outcomes of 42 MCPs’ SDOH initiatives in 5 domains: 1) built environment (BE), 2) community–clinical linkages (CCL), 3) food and nutrition security (FNS), 4) social connectedness (SC), and 5) tobacco-free policies (TFP). The Centers for Disease Control and Prevention’s National Center for Chronic Disease Prevention and Health Promotion (NCCDPHP) selected these domains based on their links to chronic disease. Although the NCCDPHP framework consists of a more-focused set of SDOH domains than other frameworks commonly used in public health, such as the Healthy People 2030 SDOH Framework (7), it is well-aligned with broader conceptualizations of SDOH. For example, NCCDPHP’s built environment, social connectedness, and community–clinical linkages domains are components of Healthy People 2030’s neighborhood and built environment, social and community context, and health care access and quality domains, respectively (8).
NCCDPHP launched GFF in partnership with the Association of State and Territorial Health Officials (ASTHO) and the National Association of County and City Health Officials (NACCHO). ASTHO and NACCHO conducted a competitive application process to select 42 MCPs from almost 100 applicants. Selection criteria included past success in implementing interventions in 1 or more of the 5 NCCDPHP SDOH domains and partnerships with local or state health departments. ASTHO and NACCHO also contracted with RTI International to conduct the rapid retrospective evaluation, which included virtual discussions with GFF partnerships, document review and abstraction of outcomes data, and simulation modeling. Additional GFF details are in Glasgow et al (9). This brief focuses on simulation modeling of MCP-driven interventions. Because rapid evaluations preclude long-term data collection, we used the existing cardiovascular disease (CVD)-focused Prevention Impacts Simulation Model (PRISM) to estimate the cumulative longer-term impacts of MCPs’ interventions.
Methods
We used PRISM to estimate the potential impacts of selected MCPs’ initiatives on CVD events, deaths, and medical and productivity costs. PRISM is a system dynamics model that simulates the impacts and costs of 32 strategies to improve CVD-related health behaviors and outcomes; other publications describe the model in detail (10,11). Prior projects have used PRISM to support strategic planning for chronic disease prevention and to evaluate the potential longer-term impacts of community-level strategies to address chronic disease risk factors (eg, tobacco use, obesity, limited access to clinical services) (12,13). The strategies included in PRISM align closely with GFF focus areas. For example, CCL strategies in PRISM address hypertension, diabetes, and high cholesterol management through the provision of high-quality clinical care.
To incorporate PRISM analyses, we first reviewed the 42 MCPs’ GFF applications and identified which partnerships had implemented interventions in 1 or more of the 32 PRISM strategy areas (Table). For example, the mobile farmers markets intervention aligns with the PRISM strategy Fruit and Vegetable Access. Some interventions, such as outdoor smoking bans, are not modeled in PRISM and were therefore not included in our analysis. We asked MCPs to confirm that they had implemented the interventions we had assigned to them; we also requested data on the implementation start date and number of people reached for each intervention. For all interventions that could be modeled in PRISM, we calculated the PRISM lever movement based on established PRISM lever-setting processes (15). We modeled all of an MCP’s interventions in a single PRISM analysis run, conducting 1 analysis run for each MCP.
We obtained cumulative results through 20 years for each MCP, then summed results across MCPs. Results are relative to status quo trends. We analyzed impacts on coronary heart disease events, strokes, deaths, medical costs, and productivity costs; analyses did not include intervention costs. The costs and deaths are for CVD and non-CVD conditions resulting from risk factors included in PRISM. All data were obtained in 2021 for interventions that MCPs had implemented in the previous 7 years.
Results
From application review, we determined that 32 of 42 GFF MCPs had implemented at least 1 SDOH initiative that could be linked to a strategy modeled in PRISM. Applications for the 10 excluded partnerships described only implementing strategies that are not modeled in PRISM, such as providing cancer screening. Twenty-seven of the remaining MCPs provided data on intervention timing and reach for the PRISM analysis. MCPs included in the analysis had delivered interventions that focus on at least 1 of the 5 SDOH domains, with 6 partnerships working on BE, 10 on CCL, 11 on FNS, 1 on SC, and 7 on TFP. Further information on how we translated the efforts of MCPs into PRISM is available in the Appendix.
The Figure displays the aggregate potential impact of the 27 MCPs’ efforts. We estimated that MCPs’ interventions could potentially avert 150 deaths in 5 years, 340 deaths in 10 years, and 880 deaths in 20 years if sustained among the 1.5 million people reached across the 27 partnerships. Impacts on costs are also meaningful; focusing on 10-year results, we estimate that potential health improvements could lead to averted medical costs of $45.4 million and averted productivity costs of $193.7 million. Costs are reported in 2021 dollars and incorporate no discounting. We saw a large increase in the potential health and economic impacts between the 5- and 10-year and the 10- and 20-year results because PRISM is a system dynamics model that incorporates time delays in the impacts of interventions that address SDOH on health outcomes and mortality.
Figure.
Estimated cumulative potential impacts of efforts implemented by Getting Further Faster partnerships (N = 27) at 5, 10, and 20 years. Coronary heart disease events, strokes, and deaths averted were rounded to the nearest 10. Medical costs, productivity costs, and total costs averted were rounded to the nearest $1,000. Medical costs and productivity costs averted include the costs of cardiovascular disease and risk factors of cardiovascular disease. [A tabular version of this figure is available.]
Discussion
PRISM was developed to support planning and evaluation of 32 broad strategies to prevent or manage CVD by addressing CVD risk factors. Leveraging close alignment between PRISM and GFF strategies, we used PRISM to analyze the potential longer-term impacts of SDOH initiatives implemented by 27 MCPs. We found that MCPs’ efforts, if sustained, could potentially avert 880 deaths and $633.4 million in costs over 20 years. Because it can take years for SDOH interventions to have a measurable impact on health outcomes, the average annual impact of MCPs’ interventions increased considerably over time. Within the GFF cohort, the potential average number of deaths per year averted was 1.5 times larger after 20 years of implementation compared with 5 years (44 averted deaths per year versus 30 deaths per year).
Our analysis has limitations. These results likely provide conservative impact estimates for the GFF cohort because some MCPs implemented interventions that are not modeled in PRISM. PRISM focuses primarily on CVD; although it captures costs and deaths from non-CVD conditions that are attributable to CVD risk factors, such as smoking, it does not fully capture impacts on non-CVD conditions. Additionally, the retrospective nature of the evaluation meant that some MCPs were unable to provide the data needed for PRISM analyses for all the interventions that they implemented. However, our estimates may overstate impact if MCPs’ efforts are not sustained with at least the current numbers of people reached. Another limitation is the lack of standardization in intervention implementation across MCPs. Unless all interventions were implemented in a manner that has been shown to be effective, our estimates may overstate the long-term impact.
Despite limitations, incorporating PRISM analysis in our rapid retrospective evaluation provided helpful information about the potential longer-term impact of MCPs’ SDOH initiatives. PRISM analysis was used to overcome common obstacles to MCP-driven intervention evaluation, including the challenges of working within evaluation timeframes that are much shorter than the time required for interventions to yield health and other salient outcomes (5). Our work also helps address a key gap in the literature around the use of modeling to inform decision-making in the public health sector (13). As public health mobilizes to better address SDOH and advance health equity (16), mathematical models like PRISM could be integrated into program planning and evaluation. This could be an excellent accompaniment to other evaluation efforts.
Acknowledgments
The findings and conclusions of this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. This project and publication were supported by cooperative agreement no. CDC-RFA-OT18-1802. The authors thank the GFF partnerships for sharing data and practice-based insights from their community-driven work to address social determinants of health and to advance health equity. They also thank Claire Korzen and Rebecca Brophy for editorial and graphics support and Matthew Dempsey for analytic support.
Author Information
Corresponding Author: Amanda A. Honeycutt, PhD, RTI International, 3040 E Cornwallis Rd, Research Triangle Park, NC 27709 (honeycutt@rti.org).
Author Affiliations: 1RTI International, Research Triangle Park, North Carolina. 2Evidera, Bethesda, Maryland. 3National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia.
References
- Xu J, Murphy SL, Kockanek KD, Arias E. Mortality in the United States, 2018. NCHS Data Brief 2020;(355):1–8. PubMed
- Braveman P, Gottlieb L. The social determinants of health: it’s time to consider the causes of the causes. Public Health Rep 2014;129(suppl 2):19–31. PubMed doi:10.1177/00333549141291S206
- Centers for Disease Control and Prevention. Health and economic costs of chronic diseases. Accessed January 20, 2021. https://www.cdc.gov/chronicdisease/about/costs/index.htm
- Frieden TR. A framework for public health action: the health impact pyramid. Am J Public Health 2010;100(4):590–5. PubMed doi:10.2105/AJPH.2009.185652
- Kegler MC, Halpin SN, Butterfoss FD. Evaluation methods commonly used to assess effectiveness of community coalitions in public health: results from a scoping review. New Dir Eval 2020;2020(165):139–57.
- Sadana R, Blas E. What can public health programs do to improve health equity? Public Health Rep 2013;128(suppl 3):12–20. PubMed doi:10.1177/00333549131286S303
- US Department of Health and Human Services. Social determinants of health. Accessed May 1, 2023. https://health.gov/healthypeople/priority-areas/social-determinants-health
- Gómez CA, Kleinman DV, Pronk N, Wrenn Gordon GL, Ochiai E, Blakey C, et al. . Addressing health equity and social determinants of health through Healthy People 2030. J Public Health Manag Pract 2021;27(suppl 6):S249–57. PubMed doi:10.1097/PHH.0000000000001297
- Glasgow L, Clayton M, Honeycutt A, Bayer EM, Plescia M, Holtgrave PL, et al. . Key insights on multisector community partnerships from real-world efforts to address social determinants of health. Eval Program Plann 2023;99:102298. PubMed doi:10.1016/j.evalprogplan.2023.102298
- Yarnoff B, Honeycutt A, Bradley C, Khavjou O, Bates L, Bass S, et al. . Validation of the Prevention Impacts Simulation Model (PRISM). Prev Chronic Dis 2021;18:E09. PubMed doi:10.5888/pcd18.200225
- Honeycutt AA, Wile K, Dove C, Hawkins J, Orenstein D. Strategic planning for chronic disease prevention in rural America: looking through a PRISM lens. J Public Health Manag Pract 2015;21(4):392–9. PubMed doi:10.1097/PHH.0000000000000062
- Yarnoff B, Bradley C, Honeycutt AA, Soler RE, Orenstein D. Estimating the relative impact of clinical and preventive community-based interventions: an example based on the Community Transformation Grant program. Prev Chronic Dis 2019;16:E87. PubMed doi:10.5888/pcd16.180594
- Jadeja N, Zhu NJ, Lebcir RM, Sassi F, Holmes A, Ahmad R. Using system dynamics modelling to assess the economic efficiency of innovations in the public sector — a systematic review. PLoS One 2022;17(2):e0263299. PubMed doi:10.1371/journal.pone.0263299
- Harner LT, Kuo ES, Cheadle A, Rauzon S, Schwartz PM, Parnell B, et al. . Using population dose to evaluate community-level health initiatives. Am J Prev Med 2018;54(5 suppl 2):S117–23. PubMed doi:10.1016/j.amepre.2018.01.026
- Centers for Disease Control and Prevention. PRISM: Prevention Impacts Simulation Model. Accessed May 1, 2003. https://prism-simulation.cdc.gov/app/cdc/prism/#/resources
- Bipartisan Policy Center. Public health forward: modernizing the U.S. public health system. 2021. Accessed May 1, 2023. https://bipartisanpolicy.org/download/?file=/wp-content/uploads/2021/12/BPC_Public-Health-Forward_R01_WEB.pdf
Table
GFF focus area | MCP intervention | PRISM lever | PRISM intervention | Intervention intensity applied in PRISMb |
---|---|---|---|---|
Built environment | Parks | PA access | Parks | Medium |
Safe streets | PA access | Safe street promotions | Low | |
Street design, land use, zoning, active transportation policy | PA access | Street design, land use, zoning, active transportation policy | High | |
Walking clubs | PA access | Walking/jogging trail | Low | |
Walking trails | PA access | Walking/jogging trail | Low | |
Childcare playground equipment | PA in childcare | Installation of portable playground equipment | Medium | |
PA in schools | PA in school | PA school requirements | High | |
Safe routes to school | PA in school | Safe routes to school | Medium | |
Clinical–community linkages | Pharmacist program | Quality CVD care | Pharmacist intervention | Medium |
High cholesterol self-management | Quality high cholesterol care | Chronic disease self-management programs (high cholesterol) | Low | |
Community health workers | Quality high cholesterol care | Community health workers (high cholesterol) | Medium | |
Health IT for chronic disease management | Quality high cholesterol care | Health IT (high cholesterol) | Low | |
Community health workers | Quality hypertension care | Community health worker model (hypertension) | Medium | |
Culturally tailored interventions for chronic disease management | Quality hypertension care | Culturally tailored interventions | Low | |
Health IT for chronic disease management | Quality hypertension care | Health IT (hypertension) | Minimal | |
Hypertension self-management | Quality hypertension care | Home blood pressure monitoring | High | |
Chronic disease self-management | Quality type 2 diabetes care | Chronic disease self-management programs (diabetes) | Low | |
Clinical information system with patient registry to track clinical measures and generate performance reports; includes referral mechanism | Quality type 2 diabetes care | Clinical information system with patient registry to track clinical measures and generate performance reports (diabetes) | Low | |
Community Health Worker | Quality type 2 diabetes care | Community health workers (diabetes) | High | |
Target underserved populations to increase number of people with access to care | Quality type 2 diabetes care | Target underserved populations to increase number of people with access to care (diabetes) | Low | |
Social support for chronic disease management | Quality high cholesterol care | Social support from family and friends | Low | |
Food insecurity | Fruit and vegetable price reduction | Energy-dense food pricing | Fruit and vegetable price reduction | High |
Community garden | Fruit and vegetable access | Community gardens | Low | |
Farmers markets | Fruit and vegetable access | Farmers markets and stands | Medium | |
SNAP at farmers markets | Fruit and vegetable access | Farmers markets accepting SNAP/EBT, outreach and transportation for farmers markets | Medium | |
Community supported agriculture | Fruit and vegetable access | Food hubs | Low | |
Healthy vending machines | Fruit and vegetable access | Healthy vending machines | Medium | |
Mobile farmers markets accepting SNAP/EBT | Fruit and vegetable access | Mobile farmers markets accepting SNAP/EBT | Medium | |
New grocery stores in underserved areas | Fruit and vegetable access | New grocery stores in underserved areas | Minimal | |
Nutrition standards and guidelines in childcare | Fruit and vegetable access | Nutrition standards and guidelines in childcare | Minimal | |
Salad bars in schools | Fruit and vegetable access | Salad bars in school cafeterias | Medium | |
School gardens | Fruit and vegetable access | School fruit and vegetable gardens | Medium | |
NACCHO Food Service Guidelines Local Action Institute | Fruit and vegetable access | School nutrition standards | Low | |
Farm to foodbank | Fruit and vegetable access | Worksites: farm-to-site programs, healthy food procurement | Low | |
Worksites: NACCHO Food Service Guidelines Local Action Institute | Fruit and vegetable access | Worksites: nutrition standards and guidelines in work cafeterias | Medium | |
Nutrition education | Fruit and vegetable access | Educational outreach and awareness of food consumption | Medium | |
Promotion in school | Fruit and vegetable access | Promotion in school: food service intervention | Medium | |
Social connectedness | Depression management | Support services for distressed | Referral to community resources | High |
Tobacco-free policies | Smoke-free MUH | Smoke-free MUH | Smoke-free MUH | Medium |
Referrals for smoking cessation services | Smoke quit services | Physician sending patient directly to a counselor, increase physician referrals, increase provider contact | Medium | |
Proactive quitlines | Smoke quit services | Proactive tobacco quitlines | Low | |
Smoking cessation classes | Smoke quit services | Smoking cessation counseling or motivational interviewing | Low | |
Telephone- or cell phone–based cessation intervention | Smoke quit services | Telephone- or cell phone–based cessation intervention | Low | |
Smoke-free bars | Workplace smoking bans | Smoke-free bars | Medium | |
Smoke-free campuses | Workplace smoking bans | Smoke-free campuses | Medium | |
Smoke-free restaurants | Workplace smoking bans | Smoke-free restaurants | Medium | |
Smoke-free worksites | Workplace smoking bans | Smoke-free workplaces | High |
Appendix. Example of Using PRISM to Analyze Long-Term Impacts of MCPs’ Interventions
This appendix provides details on how we translated the efforts of multisector community partnerships (MCPs) into the Prevention Impacts Simulation Model (PRISM). Appendix Table A1 shows, for 2 MCPs (denoted as A and B), their efforts to address social determinants of health (SDOH) factors, their reported reach for each intervention effort, and how we translated those efforts into PRISM lever movements. We indicate the PRISM intervention that our study team assigned to each MCP effort, the PRISM lever affected by that intervention, and the PRISM lever movement associated with the MCPs’ interventions. PRISM levers were not allowed to move beyond the best possible level for a given lever (eg, the percentage of workplaces that allow smoking cannot go below 0, as reflected for MCP B in their efforts to reduce indoor smoking in workplaces). The information in Appendix Table A1 reflects the MCP-specific inputs that were used in the PRISM analysis.
Appendix Table A2 provides intermediate and longer-term results from PRISM for the 2 MCPs shown in Appendix Table A1. The results are based on the PRISM lever movements shown in Appendix Table A1. The baseline mean column shows the mean in 2021 for selected outcomes from PRISM sensitivity analysis runs for the scenarios shown in Appendix Table A1. In 2021, only about 39% of those with diabetes had the condition adequately controlled, 50% of those with high blood pressure had achieved control, and 53% of those with high cholesterol had the condition controlled. About 18% of US adults were smokers. The results in the columns for MCPs A and B depict how the efforts shown in Appendix Table A1 affect the intermediate outcomes in 10 and 20 years, respectively. Because MCP A’s work focuses on managing chronic disease, the disease management outcomes for MCP A improve considerably, whereas those for MCP B show little improvement beyond current trends (eg, expected reductions in adult smoking, even without additional targeted intervention).
Appendix Table A2 also shows that impacts tend to more than double between 10 and 20 years because these interventions take time to affect chronic disease and mortality-related outcomes. Additionally, the higher reach of MCP B means greater impact than for MCP A, despite the large impact that MCP A’s efforts produce per person reached. Our analysis calculated results for each MCP by using the same approach and then summed outcomes across all MCPs to demonstrate the expected impact of these 27 MCPs’ efforts on health-, mortality-, and cost-related outcomes. Ten-year PRISM per capita outcomes aggregated by the numbers reached in each MCP are shown for all 27 MCPs in Appendix Table A3.
MCP | SDOH initiative reported by MCP | Reach reported by MCP, no. | PRISM intervention assignment | PRISM lever | Intervention intensity | Initial lever value | PRISM lever movement | PRISM lever value for analysis |
---|---|---|---|---|---|---|---|---|
A | Community health workers | 167 | Community health workers (high cholesterol) | Quality high cholesterol care, no previous CVD event | Medium | 0.565 | 0.175 | 0.74 |
A | Community health workers | 167 | Community health workers (high cholesterol) | Quality high cholesterol care, post-CVD event | Medium | 0.825 | 0.175 | 1.00 |
A | Community health workers | 167 | Community health workers (diabetes) | Quality type 2 diabetes care, no previous CVD event | High | 0.51 | 0.35 | 0.86 |
A | Community health workers | 167 | Community health workers (diabetes) | Quality type 2 diabetes care, post-CVD event | High | 0.515 | 0.35 | 0.865 |
A | Community health workers | 167 | Community health worker model (hypertension) | Quality hypertension care, no previous CVD event | Medium | 0.685 | 0.175 | 0.86 |
A | Community health workers | 167 | Community health worker model (hypertension) | Quality hypertension care, post-CVD event | Medium | 0.69 | 0.175 | 0.865 |
B | Fruit and vegetable price reduction | 1,800 | Fruit and vegetable price reduction | Energy-dense food pricing | High | 0.01 | 0.14 | 0.15 |
B | Smoke-free bars | 1,579 | Smoke-free bars | Workplace smoking bans | Medium | 0.04 | 0.14 | 0 |
B | Smoke-free campuses | 1,579 | Smoke-free campuses | Workplace smoking bans | Medium | 0.04 | 0.14 | 0 |
B | Smoke-free restaurants | 98,012 | Smoke-free restaurants | Workplace smoking bans | Medium | 0.04 | 0.14 | 0 |
B | Smoke-free workplaces | 21,700 | Smoke-free workplaces | Workplace smoking bans | High | 0.04 | 0.28 | 0 |
Outcomes | Baseline mean, 2021 | MCP A, events or costs averted | MCP B, events or costs averted | ||
---|---|---|---|---|---|
10-Year results | 20-Year results | 10-Year results | 20-Year results | ||
Intermediate outcomes | |||||
Controlled fraction of diabetes subpopulation | 39.2% | 78.6% | 78.7% | 39.7% | 40.1% |
Controlled fraction of high blood pressure subpopulation | 50.0% | 74.6% | 74.7% | 50.2% | 50.4% |
Controlled fraction of high cholesterol subpopulation | 53.3% | 71.4% | 72.1% | 55.2% | 56.2% |
Smoker fraction of total population | 18.3% | 15.2% | 13.2% | 15.2% | 13.2% |
Longer-term outcomes | |||||
Cumulative number of strokes | 4.36 per thousand | 8.44 | 18.07 | 41.27 | 103.57 |
Cumulative number of coronary heart disease events | 6.86 per thousand | 8.39 | 17.52 | 82.86 | 187.00 |
Cumulative medical costsa | $1,070 per person | $513,370 | $1,349,290 | $3,479,390 | $10,468,080 |
Cumulative productivity costsa | $2,950 per person | $2,241,080 | $5,641,280 | $9,494,090 | $28,000,410 |
Cumulative deaths | 5.03 deaths per thousand | 4.61 | 11.12 | 16.73 | 47.45 |
MCP (sorted by impact) | No. of strokes averted | No. of coronary heart disease events averted | No. of deaths averted | Medical costs averted, in 2021 dollars | Productivity costs averted, in 2021 dollars |
---|---|---|---|---|---|
1 | 166.52 | 281.49 | 135.45 | $14,517,284 | $70,153,868 |
2 | 106.22 | 136.05 | 55.66 | $7,965,325 | $30,589,261 |
3 | 20.54 | 39.40 | 39.73 | $3,125,567 | $26,888,457 |
4 | 27.15 | 251.81 | 23.09 | $7,045,020 | $15,380,845 |
5 | 41.27 | 82.86 | 16.73 | $3,479,392 | $9,494,093 |
6 | 24.26 | 24.85 | 13.05 | $1,502,098 | $6,356,518 |
7 | 22.96 | 22.82 | 12.53 | $1,396,739 | $6,097,349 |
8 | 22.82 | 20.44 | 10.98 | $1,226,953 | $7,102,645 |
9 | 12.29 | 11.29 | 5.77 | $648,997 | $3,745,502 |
10 | 12.21 | 18.17 | 4.97 | $941,151 | $2,450,398 |
11 | 8.44 | 8.39 | 4.61 | $513,371 | $2,241,078 |
12 | 7.04 | 9.15 | 4.09 | $562,064 | $2,220,555 |
13 | 7.94 | 7.11 | 3.82 | $427,051 | $2,473,194 |
14 | 6.15 | 5.55 | 2.94 | $328,856 | $1,900,900 |
15 | 5.33 | 4.79 | 2.55 | $285,487 | $1,651,425 |
16 | 4.35 | 8.02 | 1.78 | $265,886 | $1,183,909 |
17 | 2.76 | 2.75 | 1.51 | $168,040 | $733,566 |
18 | 3.65 | 5.48 | 1.48 | $280,148 | $729,432 |
19 | 2.96 | 6.71 | 1.24 | $229,550 | $807,657 |
20 | 2.08 | 3.12 | 0.84 | $159,715 | $415,856 |
21 | 1.52 | 1.69 | 0.53 | $95,307 | $286,469 |
22 | 1.10 | 2.90 | 0.47 | $85,971 | $338,805 |
23 | 1.10 | 1.64 | 0.44 | $84,028 | $218,786 |
24 | 0.28 | 1.62 | 0.18 | $46,166 | $116,825 |
25 | 0.37 | 0.54 | 0.15 | $28,543 | $74,351 |
26 | 0.05 | 0.06 | 0.04 | $4,339 | $20,719 |
27 | 0.03 | 0.05 | 0.01 | $1,735 | $7,733 |
Totala | 511 | 959 | 344 | $45,414,800 | $193,680,200 |
The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors’ affiliated institutions.