Research ArticleOriginal Research

Estimating the Cardiovascular Disease Risk Reduction of a Quality Improvement Initiative in Primary Care: Findings from EvidenceNOW

Stephan R. Lindner, Bijal Balasubramanian, Miguel Marino, K. John McConnell, Thomas E. Kottke, Samuel T. Edwards, Sam Cykert and Deborah J. Cohen
The Journal of the American Board of Family Medicine May 2023, 36 (3) 462-476; DOI: https://doi.org/10.3122/jabfm.2022.220331R1

Abstract

Background: This study estimates reductions in 10-year atherosclerotic cardiovascular disease (ASCVD) risk associated with EvidenceNOW, a multi-state initiative that sought to improve cardiovascular preventive care in the form of (A)spirin prescribing for high-risk patients, (B)lood pressure control for people with hypertension, (C)holesterol management, and (S)moking screening and cessation counseling (ABCS) among small primary care practices by providing supportive interventions such as practice facilitation.

Design: We conducted an analytic modeling study that combined (1) data from 1,278 EvidenceNOW practices collected 2015 to 2017; (2) patient-level information of individuals ages 40 to 79 years who participated in the 2015 to 2016 National Health and Nutrition Examination Survey (n = 1,295); and (3) 10-year ASCVD risk prediction equations.

Measures: The primary outcome measure was 10-year ASCVD risk.

Results: EvidenceNOW practices cared for an estimated 4 million patients ages 40 to 79 who might benefit from ABCS interventions. The average 10-year ASCVD risk of these patients before intervention was 10.11%. Improvements in ABCS due to EvidenceNOW reduced their 10-year ASCVD risk to 10.03% (absolute risk reduction: −0.08, P ≤ .001). This risk reduction would prevent 3,169 ASCVD events over 10 years and avoid $150 million in 90-day direct medical costs.

Conclusion: Small preventive care improvements and associated reductions in absolute ASCVD risk levels can lead to meaningful life-saving benefits at the population level.

Introduction

Cardiovascular disease (CVD) is the leading cause of mortality in the United States (US). In 2016, more than 750,000 deaths were attributed to the disease, and approximately 25 million adult Americans were living with the condition.1 The annual direct cost of the disease, including costs of health care services and prescription medications, currently exceeds $150 billion.

Preventive care in the form of (A)spirin prescribing for high-risk patients, (B)lood pressure control for people with hypertension, (C)holesterol management, and (S)moking screening and cessation counseling (the “ABCS”) is effective in reducing CVD.2,3 Yet, adoption of ABCS has been low.4,5 This is despite substantial attention to this issue directed by national improvement efforts such as the Million Hearts Initiative, increased use of electronic health records (EHR) and points of care decision support tools.

Improving ABCS may be especially challenging for smaller primary care practices. Although these practices serve a large number of people in the US, they often lack capacity to implement evidence-based care.6,7 In 2015, the Agency for Health care Research and Quality (AHRQ) launched EvidenceNOW, a large multi-state initiative to help small practices, with limited internal quality improvement resources, improve their ABCS by providing external support that primarily included facilitation, performance benchmarking, and audit and feedback.8 To accomplish its goal, AHRQ funded 7 regional cooperatives spanning 12 US states to recruit practices and provide external support (eg, facilitation, access to audit and feedback, performance benchmarking data). A study assessing the overall effectiveness of external support strategies across all cooperatives found, on average, moderate improvements in ABCS levels attributable to the initiative,9 which was consistent with cooperatives’ assessment of their own interventions.10⇓⇓⇓–14

This study estimates overall reductions in atherosclerotic cardiovascular disease (ASCVD) risk (defined as nonfatal myocardial infarction, coronary heart disease death, or fatal or nonfatal stroke15) that might be expected from improvements in the ABCS brought about by the external support of EvidenceNOW cooperatives. We did not have access to cardiovascular risk factor data for individual patients in many of the practices. To address this limitation, we developed a new analytic modeling approach that used patient-level information from the National Health and Nutrition Examination Survey (NHANES) in combination with EvidenceNOW practice-level data and 10-year ASCVD risk prediction equations to estimate the number of ASCVD events that might be prevented in response to the overall risk reduction observed in EvidenceNOW. We also assessed differences in risk reduction by population groups.

Methods

Data

We used the 2015 to 2016 NHANES as the primary data source to predict the impact of risk factor changes on ASCVD event risk. The NHANES is a national survey that reports respondents’ health and nutritional status. We started with an initial population of 9,971 respondents. We focused on individual respondents age 40 to 79 years because estimates of external ASCVD risk prediction equations were based on that age range (see below; sample size after exclusion: 3,390). We also excluded respondents with missing smoking status (sample size after exclusion: 1,552) and missing information about their blood pressure (sample size after exclusion: 1,362) or cholesterol levels (sample size after exclusion: 1,295) because this information was required for our calculations. Our final individual-level adult sample for this analysis included 1,295 of the 9,971 NHANES respondents.

The secondary data source included data from 1,278 primary care practices that participated in the EvidenceNOW initiative. Practice data included (1) average ABCS levels at baseline; (2) selected patient characteristics at the practice-level (eg, the fraction of black patients and the percentage of patients ages 60 to 75); and (3) information about the number of clinicians per practice and the number of patients per clinician. In-depth descriptions of EvidenceNOW practice-level data have been published previously.16,17

Analyses

To estimate the number of ASCVD events that might be prevented in response to overall ABCS improvements observed in EvidenceNOW, our approach proceeded in 3 steps: (1) we estimated 10-year ASCVD risk in the absence of ABCS treatment for each NHANES respondent; (2) we estimated 10-year ASCVD risk reduction among NHANES respondents, had they been exposed to ABCS improvements due to EvidenceNOW; and (3) we calculated weighted average 10-year ASCVD risk at baseline and 10-year ASCVD risk reduction due to the intervention. We present each step below, with further technical details described in the Appendix.

Step 1: Estimation of 10-Year ASCVD Risk Absent ABCS Treatment

For each NHANES respondent, we used an ASCVD risk prediction model developed by the American College of Cardiology and American Heart Association to estimate 10-year ASCVD risk in the absence of ABCS treatment.15,18 Model predictions are based on pooled cohort equations and multiple cohort studies of adults ages 40 to 79.

Step 2: Estimation of ASCVD Risk Reduction Due to EvidenceNOW

For this step, we identified NHANES patient groups that corresponded to the ABCS, then connected EvidenceNOW ABCS levels to these patient groups, and finally calculated implied ASCVD risk reductions.

First, we identified mutually exclusive patient groups in the NHANES using ABCS clinical quality metrics denominator definitions and information reported by NHANES respondents. ABCS denominator definitions were based on Centers for Medicaid and Medicare Services electronic clinical quality measure (eCQM) specifications used in EvidenceNOW (see Appendix Table A-3). Each denominator definition characterized a patient population for whom treatment was recommended, that is, patients who were eligible for this treatment. We applied these definitions to our NHANES sample using demographic and health information (eg, diagnosis of hypertension or coronary heart disease).

Because populations eligible for the ABCS metrics overlapped (eg, a person with diabetes and hypertension was eligible for both blood pressure and cholesterol interventions), we identified the following distinct, mutually exclusive patient groups: NHANES respondents eligible for (1) smoking screening and cessation counseling only (henceforth smoking screening/cessation counseling; denoted by Embedded Image (2) blood pressure control and smoking screening/cessation counseling (denoted by Embedded Image ); (3) cholesterol management and smoking screening/cessation counseling (denoted by Embedded Image ); (4) aspirin prescribing, cholesterol management, and smoking screening/cessation counseling (denoted by Embedded Image ); (5) blood pressure control, cholesterol management, and smoking screening/cessation counseling (denoted by Embedded Image ); and (6) all 4 treatment options (denoted by Embedded Image ). Every NHANES respondent in our sample was included in the smoking screening/cessation counseling denominator, because smoking screening applied to all adults ages 18 years and older. People were eligible for aspirin if they had an active diagnosis of an ischemic vascular disease or were discharged alive from for acute myocardial infarction, coronary artery bypass graft, or percutaneous coronary interventions (see Appendix Table A-3 for details).

Second, we connected ABCS treatment rates to the 6 patient groups by assigning possible treatment options and corresponding probabilities to them. For instance, NHANES patients eligible for cholesterol management and smoking intervention had 4 possible treatment options: (1) no treatment; (2) receiving a statin prescription; (3) smoking screening/cessation counseling; and (4) receiving both treatments.

Treatment probabilities were based on ABCS baseline levels, improvements due to EvidenceNOW, and postintervention levels, as follows: 61.9%, +3.4%, 65.3% (aspirin); 63.3%, +1.6%, 67.7% (blood pressure); 60.2%, +4.4%, 64.6% (cholesterol); 58.4%, 7.4%, 65.8% (smoking screening/cessation counseling). Improvements due to the intervention were based on an event study that assessed overall changes in ABCS across cooperatives.9

Improvements in the ABCS shifted the probability distribution to more intensive treatment. For instance, the probability of receiving both cholesterol management and smoking screening/cessation counseling for NHANES patients eligible for these treatments increased from 35.1% to 42.5%.

Third, we defined risk reduction factors, which specified how much ASCVD risk was reduced if a patient follows a certain treatment. Following literature, we assigned a number ranging from 0 (full risk reduction) to 1 (no risk reduction).2 For instance, if a patient had a 10-year ASCVD risk of 10%, then a risk reduction factor of 0.8 for a treatment option implied that 10-year ASCVD risk would be 8% if a patient consistently used the treatment, corresponding to a 20% risk reduction.

We used the following relative risk reduction factors: 0.75 for prescribing aspirin; 0.73 for controlling blood pressure, defined as less than 140 mm Hf systolic and less than 90 mmHg diastolic blood pressure; 0.75 for managing cholesterol with a statin; and 0.99 for smoking intervention. We obtained risk reduction factors for blood pressure control and cholesterol management from a systematic review.2 The relative risk reduction factor for aspirin prescribing was based on a meta-analysis of high-risk patients who were similar to our patient population.19 Risk reduction for the smoking screening/cessation counseling was small because screening included both smokers and nonsmokers, and evidence suggested that counseling was not very effective for patients who did smoke.20⇓–22

We assigned relative risk factors to the 6 mutually exclusive eligibility groups by first identifying all hypothetical treatment options for each group and then attributing corresponding relative risk factors to them. Relative risk factors of treatment combinations were obtained by multiplying relative risk reduction factors of single treatments components. For instance, people eligible for cholesterol management and smoking intervention had 4 possible treatment options: (1) no treatment; (2) receiving a statin prescription; (3) receiving smoking intervention; and (4) receiving both treatments. Corresponding relative risk factors were 1.0, 0.75, 0.99, and 0.75 × 0.99 = 0.7425.

Step 3: Calculation of Weights and Weighted Average ASCVD Risk

We created weights for each NHANES respondent in our sample so that NHANES-based risk and risk reduction calculations were representative of the EvidenceNOW population. We used an optimization algorithm that selected weights by minimizing the sum of squared differences of average standardized patient characteristics based on the NHANES and EvidenceNOW sample (see Appendix, section A.2.4, for details). Patient characteristics for the 2 populations were similar after reweighing.

Next, we calculated the weighted average 10-year ASCVD risk at baseline across all NHANES respondents, repeated this calculation for postintervention ASCVD risk levels, and calculated the estimated reduction in 10-year ASCVD risk by subtracting the average postintervention ASCVD risk from the average baseline ASCVD risk.

To account for uncertainty in ABCS reduction estimates, we calculated bootstrapped standard errors of ASCVD risk and risk reduction using 1000 iterations. For each iteration, we first sampled ABCS improvement estimates using a normal distribution with mean equal to the respective point estimate (e.g., +3.39 for aspirin and the full sample) and standard deviation equal to the estimate’s respective standard error. We then calculated postintervention ASCVD risk and risk reduction using sampled ABCS estimates. After repeating these steps 1000 times, we calculated standard deviation of the simulated 1000 postintervention ASCVD risk and risk reduction estimates to obtain standard errors.

As a sensitivity check, we repeated these step 1 to 3 calculations with average changes for practices that had higher than median changes in outcomes, because ABCS improvements varied widely across practices. Respective changes were 12.9 percentage points (aspirin prescribing); 9.4 percentage points (blood pressure control); 12.0 percentage points (cholesterol management); and 20.0 percentage points (smoking intervention). These calculations provide an estimate of ASCVD reductions associated with high-performing practices.

Validation: Using NHANES for ASCVD Risk Calculations

Our calculations required that our approach for constructing weights accurately estimated 10-year ASCVD risk of the EvidenceNOW patient population. Although we did not have access to individual-level risk factors of EvidenceNOW patients that would have permitted us to directly calculate ASCVD risk for them, we were able to work together with 2 cooperatives who did have such patient-level information. These cooperatives calculated 10-year ASCVD risk absent treatment for their practices’ patient population using individual risk factors and the same risk prediction model as we used in our calculations. They then provided us with average 10-year ASCVD risk levels absent treatment for the full patient population at the practice level. They also shared practice-level estimates of 10-year ASCVD risk for those eligible for each of the 4 ABCS interventions. We validated our weighting approach by comparing our estimated ASCVD risk to theirs. For this validation, we created separate weights for each cooperative using respective patient characteristics. All calculations were performed in R, version 3.5.1. The Institutional Review Board of Oregon Health & Science University approved this study.

Results

NHANES respondents corresponding to EvidenceNOW patients tended to be female, white, and less than 60 years old (Table 1). Patient population group sizes ranged from 1.6% (people only eligible for cholesterol management and smoking intervention) to 49.8% (people only eligible for smoking intervention).

View this table:
Table 1.

Sample Characteristics

EvidenceNOW practices provided care to patients with an average 10-year ASCVD baseline risk of 10.11% (Table 2). Improvements in ABCS measures due to EvidenceNOW reduced the average 10-year ASCVD risk to 10.03%, corresponding to an absolute reduction of 0.08 percentage points (P ≤ .01), or a risk reduction relative to the baseline risk of 0.79%. This risk reduction implied an expected prevention of 3,169 ASCVD events among the 3,961,384 EvidenceNOW patients over a 10-year period. A recent study estimated that the average direct 90-day medical cost of a major cardiovascular event was $47,433.23 Thus, the prevention of 3,169 ASCVD events would save approximately $150 million in direct medical costs.

View this table:
Table 2.

Average ASCVD Risk and ASCVD Risk Reductions Due to Improvements in the ABCS

Improvements in blood pressure control or cholesterol management alone would each reduce 10-year ASCVD risk among the EvidenceNOW population by approximately 0.03 percentage points, or 0.30 and 0.28% relative to the baseline risk, respectively. Improvements in aspirin prescribing and smoking intervention alone would have resulted in smaller ASCVD risk reductions among the EvidenceNOW population (absolute 10-year ASCVD risk reduction: 0.01; relative 10-year ASCVD risk reduction: 0.14 and 0.07, respectively). Corresponding p-values for these estimates were not statistically significant.

In the sensitivity analyses that focused on practices that demonstrated improvements at the median or higher, we estimated that there would have been an absolute reduction in 10-year ASCVD risk of 0.32 (P ≤ .001). This absolute risk reduction corresponds to a 3.28% decline in relative risk.

Approximately 50% percent of individuals were only eligible for smoking screening/cessation counseling of smokers (ie, individuals with no health conditions that would have made them eligible for aspirin, blood pressure control, or statin management; Table 3). Approximately 1/3 were eligible only for blood pressure control, and 7.8% were eligible for all 4 treatments. The average 10-year ASCVD risk by eligibility group ranged from 7.7% (people only eligible for smoking intervention) to 24.6% (people eligible for all 4 treatment options). The contribution to the overall ASCVD reduction varied from 3.6% (those only eligible for smoking intervention) to 31.1 and 41.2% for those eligible for blood pressure control and all treatment options, respectively. Validation estimates of 10-year ASCVD risk levels at baseline were similar to estimates based on our approach for both cooperatives (see Appendix, section A.3).

View this table:
Table 3.

Relative Risk Factors and Risk Reduction Due to Improvements in ABCS

Discussion

This study calculated reductions in 10-year risk of an ASCVD event (which includes nonfatal myocardial infarction, coronary heart disease death, or fatal or nonfatal stroke) associated with EvidenceNOW, a large multi-state initiative to improve cardiovascular risk prevention among smaller primary care practices that served approximately 4 million adult patients. We developed a novel method to estimate that adult EvidenceNOW patients had an average 10-year ASCVD risk of 10.11 percentage points at baseline, and that improvements in the ABCS due to EvidenceNOW reduced 10-year ASCVD risk by 0.08 percentage points. This risk reduction implied that EvidenceNOW would prevent approximately 3,169 ASCVD events over 10 years if improvements in the ABCS were sustained, saving $150 million in direct, 90-day medical costs alone. We found that this risk reduction is greater among higher performing practices, and for patients with multiple risk factors.

Although other EvidenceNOW cooperatives have conducted assessments of their practice-based interventions, most did not include assessments of change in cardiovascular disease risk as outcomes.10,12⇓–14 Identifying small improvements in health at the population level requires collecting granular, comprehensive information as precisely as possible. Obtaining such data often necessitates prior investment in data infrastructure (eg, practices’ EHR system) that is currently not in place in primary care.7 One of the unique attributes of our work was that we were able to provide estimates of cardiovascular disease risk reduction due to the initiative despite a general lack of such data infrastructure.

One cooperative, North Carolina, was able to estimate cardiovascular disease reduction due to their intervention. This cooperative reported a larger reduction in 10-year ASCVD risk than we found.11 Several reasons can explain this discrepancy. First, North Carolina implemented an informatics tool to calculate ASCVD risk for all patients aged 40 to 79 years and to focus statin and aspirin preventive care improvement efforts on patients with 10-year ASCVD risk above 10%. Only 147,000 of the 430,000 patients included in their study met this definition of high-risk patients. As a result, patients in the North Carolina study population had a much higher average 10-year ASCVD risk score at baseline than our study population (23.4 vs 10.13 percentage points in our study), which resulted in a correspondingly higher absolute ASCVD risk reduction. Second, the North Carolina study identified the patient population eligible for statin based on ASCVD risk as well as more traditional factors used in EvidenceNOW (e.g., presence of high cholesterol or ASCVD). They were able to achieve strong cholesterol management improvements for this patient population. Third, they used exact blood pressure levels based on electronic health records (EHR). This allowed them to calculate 10-year ASCVD reductions due to any reduction in systolic blood pressure, whereas our study was limited to estimating the effects of reductions below a specific threshold. Our imprecision in estimating the full effect of EvidenceNOW interventions, especially with regard to blood pressure, suggests that our results understate the full effect of the initiative on cardiovascular disease reduction.

It is also important to note that North Carolina was among the most experienced cooperatives in the EvidenceNOW cohort. An ability to leverage regional health information exchange and target high-risk patients, among other attributes of their cooperative’s work, is evidence of that. This experience led to larger ABCS changes than less experienced cooperatives, and suggests that effectiveness of initiatives, such as EvidenceNOW, could be further strengthened by investing in cooperatives’ expertise and infrastructure, and at very least needs to take experience into account when contextualizing outcomes.24

Our study contributes to a growing body of work examining the effectiveness of initiatives to improve cardiovascular health. One such initiative was the Cardiovascular Health Awareness Program (CHAP) in Ontario, Canada, which focused on outreach and educational effort at the community level.25,26 This intervention led to a 9% relative improvement in a composite measure of hospital admissions for acute myocardial infarction, stroke, and congestive heart failure.27 Related hospitalization costs declined by 14% relative to baseline, which offset the costs of the intervention.28 These studies, together with our findings, suggest that intervening at the community and practice level are promising and potentially complementary strategies to improve cardiovascular health at the population level.

Limitations

Our study had several limitations. First, our risk calculation was based on 2 separate samples, and we might not have been fully successful in balancing them along all ASCVD risk factors. However, our validation suggested that our approach was relatively accurate. Second, our calculations were based on an external risk prediction model that might not be accurate for some population groups. Third, estimates of ABCS improvements were based on an observational study and might reflect unrelated trends; however, other studies have shown no general improvements in the ABCS during our study period.29⇓–31 Fourth, our calculations assumed that improvements would persist for 10 years, and we overstate the number of ASCVD events prevented otherwise. Conversely, we focused on patients 40 to 79 years old and assumed that only the current, but not future patient cohorts, would be affected by the intervention. This assumption by itself implied that we underestimated the number of prevented ASCVD events. Fifth, calculations for above-median practices may suffer from regression to the mean. Sixth, we did not include measures of morbidity and well-being, and therefore may not have captured the full health effects of the intervention. Seventh, qualitative data from EvidenceNOW shows that documentation changes, such as improving documentation of smoking interventions in the EHR, may explain some of the improvements observed in ABCS.32 Although we considered improving documentation a quality improvement, it was not one that is likely to have an impact on ASCVD events and thus by itself would lead to an overstatement of ASCVD reductions. Eighth, interventions for lower risk patients may have included lifestyle changes (e.g., diet and exercise), the effects of which were imperfectly captured by the ABCS, and thus not fully accounted for in our calculations. Finally, we only included 13% of NHANES respondents in our analysis, partially due to missing information about smoking, blood pressure, and cholesterol.

Conclusion

This study showed how external support strategies for smaller primary care practices, when implemented at a large-scale via regional cooperatives, could meaningfully reduce cardiovascular population health risk levels while being nearly cost neutral.33,34 Findings from this study, and the overall EvidenceNOW initiative,7⇓–9,16,17,24,32,35⇓⇓⇓–39 suggest that policy-makers should support long-term investment in organizations, such as the primary care cooperative extension, that can reach large numbers of smaller practices and provide care enhancing external support.

Acknowledgments

The authors thank Chunliu Zhan, Leif Solberg, William Miller, and Benjamin Crabtree for helpful comments.

Appendix


Embedded Image

Embedded Image

Embedded Image

Embedded Image

Embedded Image

Embedded Image

Notes

  • This article was externally peer reviewed.

  • Funding: Agency for Healthcare Research and Quality (R01HS023940).

  • Conflict of interest: The authors have no conflicts of interest to declare.

  • To see this article online, please go to: http://jabfm.org/content/36/3/462.full.

  • Received for publication September 22, 2022.
  • Revision received January 31, 2023.
  • Accepted for publication February 1, 2023.

References

  1. ↵
    1. Benjamin EJ,
    2. Muntner P,
    3. Alonso A
    , On behalf of the American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommitteeet al. Heart disease and stroke statistics-2019 update: A report from the American Heart Association. Circulation 2019;139:e56–e528.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Lloyd-Jones DM,
    2. Huffman MD,
    3. Karmali KN,
    4. et al
    . Estimating longitudinal risks and benefits from cardiovascular preventive therapies among Medicare patients. Journal of the American College of Cardiology 2017;69:1617–36.
    OpenUrlFREE Full Text
  3. ↵
    1. Karmali KN,
    2. Lloyd-Jones DM,
    3. Berendsen MA,
    4. et al
    . Drugs for primary prevention of atherosclerotic cardiovascular disease. JAMA Cardiol 2016;1:341–9.
    OpenUrl
  4. ↵
    1. Farley TA,
    2. Dalal MA,
    3. Mostashari F,
    4. Frieden TR
    . Deaths preventable in the U.S. by improvements in use of clinical preventive services. American Journal of Preventive Medicine 2010;38:600–9.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Frieden TR,
    2. Berwick DM
    . The “million hearts” initiative - preventing heart attacks and strokes. N Engl J Med 2011;365:e27.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Hing E,
    2. Rui P,
    3. Palso K
    . National ambulatory medical care survey: 2013 state and national summary tables. 2013. Available at: https://www.cdc.gov/nchs/data/ahcd/namcs_summary/2013_namcs_web_tables.pdf.
  7. ↵
    1. Cohen DJ,
    2. Dorr DA,
    3. Knierim K,
    4. et al
    . Primary care practices’ abilities and challenges in using electronic health record data for quality improvement. Health Affairs 2018;37:635–43.
    OpenUrlCrossRef
  8. ↵
    1. Cohen DJ,
    2. Balasubramanian BA,
    3. Gordon L,
    4. et al
    . A national evaluation of a dissemination and implementation initiative to enhance primary care practice capacity and improve cardiovascular disease care: The Escalates study protocol. Implementation Sci 2015;11:86.
    OpenUrl
  9. ↵
    1. Balasubramanian BA,
    2. Lindner S,
    3. Marino M,
    4. et al
    . Improving delivery of cardiovascular disease preventive services in small-to-medium primary care practices. J Am Board Fam Med 2022.
  10. ↵
    1. Persell SD,
    2. Liss DT,
    3. Walunas TL,
    4. et al
    . Effects of 2 forms of practice facilitation on cardiovascular prevention in primary care. Med Care 2020;58:344–51.
    OpenUrl
  11. ↵
    1. Cykert S,
    2. Keyserling TC,
    3. Pignone M,
    4. et al
    . A controlled trial of dissemination and implementation of a cardiovascular risk reduction strategy in small primary care practices. Health Serv Res 2020;55:944–53.
    OpenUrlCrossRef
  12. ↵
    1. Parchman ML,
    2. Anderson ML,
    3. Dorr DA,
    4. et al
    . A randomized trial of external practice support to improve cardiovascular risk factors in primary care. Ann Fam Med 2019;17:S40–S49.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Shelley DR,
    2. Gepts T,
    3. Siman N,
    4. et al
    . Cardiovascular disease guideline adherence: An RCT using practice facilitation. American Journal of Preventive Medicine 2020;58:683–90.
    OpenUrl
  14. ↵
    1. Dickinson WP,
    2. Nease DE,
    3. Rhyne RL,
    4. et al
    . Practice transformation support and patient engagement to improve cardiovascular care: From EvidenceNOW Southwest (ENSW). J Am Board Fam Med 2020;33:675–86.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Goff DC,
    2. Lloyd-Jones DM,
    3. Bennett G
    , American College of Cardiology/American Heart Association Task Force on Practice Guidelineset al. 2013 acc/aha guideline on the assessment of cardiovascular risk. Circulation 2014;129:S49–73.
    OpenUrlFREE Full Text
  16. ↵
    1. Balasubramanian BA,
    2. Marino M,
    3. Cohen DJ,
    4. et al
    . Use of quality improvement strategies among small to medium-size us primary care practices. Ann Fam Med 2018;16:S35–S43.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Lindner S,
    2. Solberg L,
    3. Miller W,
    4. et al
    . Does ownership make a difference in primary care practice? J Am Board Fam Med 2019;32:398–407.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Muntner P,
    2. Colantonio LD,
    3. Cushman M,
    4. et al
    . Validation of the atherosclerotic cardiovascular disease pooled cohort risk equations. JAMA 2014;311:1406–15.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Collaboration.
    AT Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002;324:71–86.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Cummings SR
    . The cost-effectiveness of counseling smokers to quit. JAMA 1989;261:75.
    OpenUrlCrossRefPubMed
  21. ↵
    Tobacco use and dependence guideline panel. Treating tobacco use and dependence: 2008 update. US Department of Health; Human Services; 2008.
  22. ↵
    1. Wray JM,
    2. Funderburk JS,
    3. Acker JD,
    4. Wray LO,
    5. Maisto SA
    . A meta-analysis of brief tobacco interventions for use in integrated primary care. Nicotine & Tobacco Research 2018;20:1418–26.
    OpenUrlPubMed
  23. ↵
    1. Bonafede M,
    2. Johnson B,
    3. Gandra S,
    4. Richhariya A
    . Medical costs associated with cardiovascular events among high-risk patients with hyperlipidemia. ClinicoEconomics and Outcomes Research 2015;7:337–45.
    OpenUrl
  24. ↵
    1. Cohen DJ,
    2. Balasubramanian BA,
    3. Lindner S,
    4. et al
    . How does prior experience pay off in large-scale quality improvement initiatives? The J Am Board Fam Med 2022;35:1115–27.
    OpenUrl
  25. ↵
    1. Chambers LW,
    2. Kaczorowski J,
    3. Dolovich L,
    4. et al
    . A community-based program for cardiovascular health awareness. Canadian Journal of Public Health 2005;96:294–8.
    OpenUrlPubMed
  26. ↵
    1. Kaczorowski J,
    2. Chambers LW,
    3. Karwalajtys T,
    4. et al
    . Cardiovascular health awareness program (CHAP): A community cluster-randomised trial among elderly Canadians. Preventive Medicine 2008;46:537–44.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Kaczorowski J,
    2. Chambers LW,
    3. Dolovich L,
    4. et al
    . Improving cardiovascular health at population level: 39 community cluster randomised trial of cardiovascular health awareness program (CHAP). BMJ 2011;342:d442–2.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Goeree RKC,
    2. von Burke N,
    3. et al
    . Economic appraisal of a community-wide cardiovascular health awareness program. Value in Health 2013;16:39–45.
    OpenUrl
  29. ↵
    1. Muntner P,
    2. Hardy ST,
    3. Fine LJ,
    4. et al
    . Trends in blood pressure control among US adults with hypertension, 1999–2000 to 2017–2018. JAMA 2020;324:1190.
    OpenUrlPubMed
  30. ↵
    1. Yao X,
    2. Shah ND,
    3. Gersh BJ,
    4. Lopez-Jimenez F,
    5. Noseworthy PA
    . Assessment of trends in statin therapy for secondary prevention of atherosclerotic cardiovascular disease in US adults from 2007 to 2016. JAMA Network Open 2020;3:e2025505–5.
    OpenUrl
  31. ↵
    1. Salami JA,
    2. Warraich H,
    3. Valero-Elizondo J,
    4. et al
    . National trends in statin use and expenditures in the US adult population from 2002 to 2013: insights from the Medical Expenditure Panel Survey. JAMA Cardiology 2017;2:56–65.
    OpenUrl
  32. ↵
    1. Cohen DJ,
    2. Sweeney SM,
    3. Miller WL,
    4. et al
    . Improving smoking and blood pressure outcomes: The interplay between operational changes and local context. Ann Fam Med 2021;19:240–8.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Rose G
    . Strategy of prevention: Lessons from cardiovascular disease. Br Med J (Clin Res Ed) 1981;282:1847–51.
    OpenUrlFREE Full Text
  34. ↵
    1. Kadakia KT,
    2. Song Z
    . Saving money or improving health? Reconsidering payment reform. Milbank Quarterly Opinion 2022.
  35. ↵
    1. Ono SS,
    2. Crabtree BF,
    3. Hemler JR,
    4. et al
    . Taking innovation to scale in primary care practices: the functions of health care extension. Health Affairs 2018;37:222–30.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Sweeney SM,
    2. Hemler JR,
    3. Baron AN,
    4. et al
    . Dedicated workforce required to support large-scale practice improvement. The J Am Board Fam Med 2020;33:230–9.
    OpenUrl
  37. ↵
    1. Marino M,
    2. Solberg L,
    3. Springer R,
    4. et al
    . Cardiovascular disease preventive services among smaller primary care practices. American Journal of Preventive Medicine 2021.
  38. ↵
    1. Perry CK,
    2. Lindner S,
    3. Hall J,
    4. Solberg LI,
    5. Baron A,
    6. Cohen DJ
    . How type of practice ownership affects participation with quality improvement external facilitation: findings from EvidenceNOW. Journal of General Internal Medicine 2022;37:793–801.
    OpenUrl
  39. ↵
    1. Sweeney SM,
    2. Baron A,
    3. Hall JD,
    4. et al
    . Effective facilitator strategies for supporting primary care practice change: a mixed methods study. Ann Fam Med 2022;20:414–22.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Keywords