The potential impact of policies and structural interventions in reducing cardiovascular disease and mortality: a systematic review of simulation based studies

Article history

The contractual start date for this research was in February 2019. This article began editorial review in June 2022 and was accepted for publication in April 2023. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright statement

Copyright © 2023 Uthman et al. This work was produced by Uthman et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2023 Uthman et al.

Eligibility criteria

Information sources and search strategy

A systematic search was performed to identify all relevant studies that satisfied our selection criteria within the following databases from inception to March 2021: Medline, Embase, Web of Science and the Cochrane Library. Additionally, we checked the reviews that we identified for further studies. The search strategy is included as Report Supplementary Material 1.

Selection process

After removing duplicates, two authors independently screened titles and abstracts followed by full-text articles. Any disagreements between the two reviewers were resolved through discussions. However, if a disagreement persisted, it was resolved through consultations involving a third author.

Data extraction

Two reviewers independently extracted data using these forms. Disagreements were resolved through discussions between the two reviewers, and in cases in which no agreement was reached, a third author intervened.

Data items

The following data were extracted from each study: (1) overview of study characteristics, (2) model structure and (3) sources of evidence for model parameters.

Risk of bias in individual studies

The validation of a simulation model is an important determinant of the risk of bias and applicability of a simulation model. All models were assessed in accordance with the guidelines of the International Society for Pharmacoeconomics and Outcomes Research-Society for Medical Decision Making report. 7 For each study, we assessed face validity (wherein experts evaluate model structure, data sources, assumptions and results), verification or internal validity (check accuracy of coding), cross validity (comparison of results with other models analysing the same problem), external validity (comparing model results with real-world results) and predictive validity (comparing model results with prospectively observed events). 7

Data synthesis

We performed a systematic narrative synthesis to present a critical appraisal of the methodological quality and risk of bias of the selected modelling studies and of their results.

Study selection and characteristics

Figure 1 shows the study selection flow diagram. The literature search yielded 1997 articles. After review, 99 articles were selected for critical reading. Forty-five studies did not meet the inclusion criteria and were excluded. The other 54 studies were included. Table 1 presents the characteristics of the included studies. The studies were published between 2007 and 2021. Most of the studies were from the USA (n = 23, 42.6%), followed by the UK (n = 10, 18.5%). Most of the studies were based on the USA the Cardiovascular Disease Policy Model for Risk, Events, Detection, Interventions, Costs, and Trends (CVD PREDICT)13–17 and the UK Infliximab Multinational Psoriatic Arthritis Controlled Trial (IMPACT) Coronary Heart Disease (CHD) Policy Model. 18–24 The majority of the studies were full-text articles (n = 48, 88.9%), and only six were conference abstracts. When reported, the time horizon ranged from 3 years to as much as 50 years. Microsimulation (n = 32, 59.3%) was the most frequently used method for developing the policy simulation models, followed by Markov chain models (n = 11, 20.4%). Most studies conducted probabilistic sensitivity analysis (PSA) (n = 30, 56%). Fourteen studies conducted one-way sensitivity analysis. The policy simulation models can be broadly categorised into diet and nutrition, tobacco and alcohol control initiatives and other initiatives. Diet and nutrition initiatives included: national salt reduction initiatives, fruit and vegetable subsidies, sugar-sweetened beverage (SSB) taxes, elimination of industrial trans-fatty acids, food taxes and subsidies, healthy food and lifestyles policies, palm oil taxes, processed meat taxes, reduction in ultra-processed foods, a supplementary nutrition assistance program, stricter food policy and subsidised community-supported agriculture. Tobacco and alcohol control initiatives included smoking and alcohol, smoke-free policies, national tobacco control initiatives and alcohol pricing policies. Other initiatives included the NHS Health Check, the WHO 25 × 25 initiative and air quality management policies.

FIGURE 1.

Study selection flow.

Percentage reduction in cardiovascular disease mortality

Six studies reported a percentage reduction in CVD mortality (see Figure 2). 30,35,36,53,55,64 Three studies were from the USA, one each from the Republic of South Korea, Argentina and Costa Rica. Face validation was adequate for all the studies. Internal validation was adequate in four studies and unclear in the remaining two studies (see Table 2). None of the studies reported cross-validation. External validation of the models was performed in only study and none of the studies did not perform predictive validation. The studies were published between 2010 and 2018 and the following policy-level interventions were modelled: national salt reduction initiatives (n = 3), air quality management policy (n = 1), food and tax subsidy (n = 1) and one multicomponent intervention (n = 1). The reported potential percentage reduction in CVD mortality varied across interventions modelled, from 0.1% for air quality management policy in the Republic of Korea to 39% for a national salt reduction initiative in Costa Rica. The potential benefit of a national salt reduction initiative varied across countries and periods, from a 2.5% reduction in CVD mortality in Argentina to a 39% reduction in Costa Rica. Three studies modelled different interventions among the US population. 30,36,53 Multicomponent interventions demonstrated the largest percentage reduction in CVD mortality (20%),36 followed by food tax and subsidy (8.0%)53 and a national salt reduction initiative (5.4%). 30

FIGURE 2.

Effects of policies and structural interventions on percentage reduction in CVD mortality.

Percentage reduction in cardiovascular disease cases

Four studies reported a percentage reduction in CVD cases (see Figure 3). 28,31,44,52 Three studies were from the USA31,44,52 and one from India. 28 Face validation was adequate for all the studies. Internal validation was adequate in three studies and unclear in one study (see Table 2). None of the studies reported cross-validation. External validation of the models was performed in three studies and none of the studies did not perform predictive validation. The studies were published between 2013 and 2019 and the following policy-level interventions were modelled: national tobacco control initiative (n = 1), smoking-free policy (n = 1), SSB tax (n = 1) and supplementary nutrition assistance programme (n = 1). The reported potential percentage reduction in CVD cases ranged from 1% for a SSB tax in the USA to 25% for a national tobacco control initiative in India. Three studies modelled different interventions among the US population. 31,44,52 A smoke smoking-free policy demonstrated the largest percentage reduction in CVD cases (4%),52 followed by a supplementary nutrition assistance programme (1.4%)31 and a SSB tax (1.0%). 31

FIGURE 3.

Effects of policies and structural interventions on percentage reduction in MACE cases.

Number of deaths prevented or postponed per year

Twenty-six studies reported the number of DPP per year (see Figure 4). 19,21–24,27,29,32,33,38–42,45–50,54,57,59,60,65,66 There were seven studies each from the USA and the UK, three from Argentina, and one each from Australia, Brazil, China, India, Ireland, New Zealand, South Africa and Turkey. One was a multicountry study. Face validation was adequate for all the studies. Internal validation was adequate in most studies (n = 25, 96%) and unclear in one study (see Table 2). None of the studies reported cross-validation. External validation of the models was performed by 14 studies (54%), not performed by 4 studies (15%) and unclear in the remaining 8 studies (31%). Most of the studies did not perform predictive validation (n = 24, 92%). The studies were published between 2008 and 2020, and the following policy-level interventions were modelled: food tax and subsidy (n = 7), national salt reduction initiative (n = 7), food and beverage tax (n = 3), air quality management policy (n = 1), eliminating industrial trans-fatty acids (n = 1), fruit and vegetable subsidy (n = 1), healthy food and lifestyle policy (n = 1), national tobacco control initiative (n = 1), palm oil tax (n = 1), reduction in ultra-processed foods (n = 1), stricter food policy (n = 1), Subsided Community Supported Agriculture (n = 1) and WHO 25 × 25 initiative (n = 1).

FIGURE 4.

Effects of policies and structural interventions on number of deaths prevented or postponed per year.

The number of DPP depended on the size of population affected but ranged from just 23 per year for eliminating industrial trans-fatty acids in Australia42 to as many as 48,000 per year for a national salt reduction initiative in the USA. 29 Evidence from Argentina38,39,60 showed that a national tobacco control initiative could prevent or postpone the largest number of deaths (15,500 deaths per year), followed by a SSB tax (2700 deaths per year) and a national salt reduction initiative (1900 deaths per year). Evidence from the USA29,33,49,50,54,57,66 showed that a national salt reduction initiative could prevent or postpone the largest number of deaths (48,000 deaths per year), while air quality management could prevent or postpone the least number of deaths (290 deaths per year). Evidence from the UK19,21,22,24,32,47,48 showed that the WHO 25 × 25 initiative could prevent or postpone the largest number of deaths (20,000 deaths per year), while a SSB tax could prevent or postpone the least number of deaths (370 deaths per year).

Number of CVD cases prevented or postponed per year

Ten studies reported the number of CVD cases prevented or postponed per year (see Figure 5). 14,15,20,37,51,54,58,61–63 Five studies were from the USA, two from the UK, and one each from Australia, Germany and Mexico. Face validation was adequate for all the studies. Internal validation was adequate in most studies (n = 9, 90%) and unclear in one study (see Table 2). None of the studies reported cross-validation. External validation of the models was performed by six studies (60%), not performed by one study (10%) and unclear in the remaining three studies (30%). Most of the studies did not perform predictive validation (n = 9, 90%). The studies were published between 2008 and 2020, and the following policy-level interventions were modelled: national salt reduction initiative (n = 2), alcohol pricing policy (n = 1), fruit and vegetable subsidy (n = 1), national menu calorie labelling law (n = 1), NHS Health Check (n = 1), processed meat tax (n = 1), SSB tax (n = 1) and supplementary nutrition assistance program (n = 1). The number of CVD cases prevented or postponed ranged from just 750 per year for the NHS Health Check in the UK20 to as many as 45,000 per year for the national salt reduction initiative in the USA. 54 Five studies modelled different interventions among the US population. 14,15,51,54,63 The national salt reduction initiative demonstrated the largest number of CVD cases prevented or postponed per year (45,000 CVD cases per year prevented), followed by supplementary nutrition assistance programmes (7756 CVD cases per year prevented), fruit and vegetable subsidies (2690 CVD cases per year prevented) and the national menu calorie labelling law (2690 CVD cases per year prevented). In the UK, a larger number of CVD cases could be prevented by alcohol pricing policy than through NHS Health Check (2631 vs. 750 CVD cases per year prevented).

FIGURE 5.

Effects of policies and structural interventions on number of CVD cases prevented or postponed per year.

Main findings

The importance of population-level policies is becoming more widely recognised in the public health community. 67 When compared to individual-level interventions, they are more effective, cost-effective and equitable. 67 Policy-makers and the public health community must make further efforts towards shifting the focus of prevention away from individuals and towards populations. 67 Evidence about the potential benefits of a particular intervention is frequently available from a variety of sources. We included well-conducted simulation studies to examine the potential impact of structural and policy interventions at the national level. We identified 54 simulation-based studies finding that there is good evidence that population‐wide strategies could reduce the burden of CVD. The modelling studies we included suggest that these interventions could bring population‐wide reductions in the burden of CVD.

Diet and nutrition, tobacco and alcohol control, and other programmes are among the policy simulation models explored. The diet and nutrition initiatives included food taxes and subsidies, healthy food and lifestyles policies, palm oil tax, processed meat tax, reduction in ultra-processed foods, supplementary nutrition assistance programmes, stricter food policy and subsidised community-supported agriculture. Initiatives to reduce tobacco and alcohol use included a smoking ban, a national tobacco control initiative and a tax on alcohol. The other programmes we identified included the NHS Health Check, WHO 25 × 25 and air quality management policy. Our review shows that taxation on tobacco, salt, sugar and alcohol is the most potential impactful structural intervention to reduce CVD events and mortality at the population level.

Simulation and modelling studies are important tools in supporting evidence for the effectiveness of primary prevention of CVD. 68–71 These types of studies allow researchers to investigate the potential impact of interventions on population health, without having to conduct large, costly and time-consuming RCTs. 72 One advantage of simulation and modelling studies is that they can be used to estimate the potential impact of an intervention on a population level. 68–71 For example, a simulation study could be used to estimate the number of CVD cases that would be prevented by increasing physical activity levels in a population. This information can be used to inform public health policy and decision-making. Another advantage of simulation and modelling studies is that they can be used to investigate the impact of interventions in subpopulations that may be difficult to study in RCTs. For example, a modelling study could be used to investigate the impact of a dietary intervention on CVD risk in older adults. Simulation and modelling studies also can combine multiple interventions and assess their joint effect on population health; this is a critical aspect for primary prevention of CVD as it often involves multifactorial interventions. Overall, simulation and modelling studies are an important complementary tool to RCTs in providing evidence for the effectiveness of primary prevention of CVD. 68–71

Strengths and limitations

This systematic review has several advantages. Firstly, our study’s main strength was that we conducted a thorough up-to-date overview of the status of policy-level intervention models internationally, using search of the published literature to undertake a comprehensive review of their potential impact on primary prevention of CVD. Second, using appropriate validated tools, two reviewers independently checked all papers and assessed quality. Third, we documented the effect size used in each modelling paper, as well as the source reference. Finally, the studies included a wide range of interventions across a wide range of countries, providing highly useful estimates of potential impact and information to design future policies and action.

This review however has some limitations. Simulation and modelling studies are important tools in supporting evidence for the effectiveness of primary prevention of CVD, but they also have limitations that should be considered. 73–77 One limitation of simulation and modelling studies is that they rely on assumptions and estimates. For example, a simulation study may assume that a certain percentage of the population will respond to an intervention in a certain way. If these assumptions are not accurate, the results of the simulation may not be valid. Additionally, the accuracy of the simulation is dependent on the quality and availability of data used to input the model. Another limitation of simulation and modelling studies is that they may not capture all the complex interactions that occur in real-world settings. Additionally, simulation and modelling studies are mostly based on observational data, which is subject to the same limitations as observational studies. Observational studies are often excluded in systematic reviews of primary prevention of CVD due to their inherent limitations. One of the main limitations is the potential for bias, such as selection bias and confounding variables, which can make it difficult to determine causality between the intervention and the outcome. 73–77 This can make it challenging to draw valid conclusions from observational studies and make them less informative compared to RCTs which have a higher level of evidence. Moreover, observational studies are often subject to measurement errors, which can lead to inaccurate results. This can further decrease the validity and generalisability of the results.

We have been unable to undertake a formal meta-analysis because the studies were so diverse. Publication bias is still a possibility, which could lead to an overestimation of the real effects of some interventions. Furthermore, the positive effects of policy changes may appear to be greater when favourable underpinning secular trends are not formally considered. There is significant heterogeneity in simulation models, making output data significantly incomparable between models; thus, the studies in the analysis may not be necessarily comparable. While policy interventions typically include a variety of strategies, none of the models considered interrelations between multiple policies and potential interactions. Different aspects of these policies may have varying effects.

Another important limitation of the studies reviewed is that they primarily reported on major cardiovascular events, such as heart attacks and strokes, but did not report on other CVD outcomes such as cardiomyopathies. By not reporting on these outcomes, the studies may not provide a comprehensive understanding of the impact of the interventions on CVD. Furthermore, it may make it difficult to draw conclusions about the effectiveness of the interventions in preventing different types of CVD. This limitation should be taken into consideration when interpreting the results of the studies and considering the implications for public health policy and clinical practice.

Implications for future research and policy

Our research uncovered a number of evidence gaps that merit additional investigation. First, no real-world policy impact studies were uncovered. Though we discovered that policies and structural interventions aimed at reducing CVD could reduce the burden of CVD and save money if implemented over time, the feasibility of implementing policy changes also warrants further investigation. Many elements, including political feasibility and stakeholder engagement, can help or hinder successful policy formulation.

In a number of other areas, more research is required. Longitudinal studies, for example, are required to increase the evidence base and to aid in the identification of precise combinations of interventions that have the greatest impact on deaths avoided or postponed. These longitudinal studies need to be broader and more representative of target populations likely to benefit. In addition, the adoption of common measures to compare different initiatives and communities would help to boost such evaluation efforts. A standardised evaluation system across various policy-level interventions would be immensely valuable and might serve as a policy lever at the regional or national levels, inspiring greater action. Future study on the effectiveness of policy-level interventions in various populations and settings, as well as intervention effectiveness among different communities or populations with differing demography, would be beneficial. Finally, research that investigates dose–response and interactions between numerous modifications as well as longer-term clinical outcomes can help us better understand the potential impact of policy-level interventions.

Patient and public involvement

Drawing on INVOLVE guidance and support for best practice, we worked closely with three dedicated patient and public involvement advisors, and we welcomed guidance and support from our advisors during the preparatory phase of the project.

Contributions of authors

Olalekan A. Uthman (https://orcid.org/0000-0002-8567-3081) (Professor, Evidence Synthesis) contributed to the protocol, study selection, data extraction, validity assessments and synthesis of the included studies. He developed the classifiers and undertook the analyses. He also contributed to the interpretation of the results and the writing of the report and had overall responsibility for the project.

Rachel Court (https://orcid.org/0000-0002-4567-2586) (Information Specialist) contributed to the protocol development, developed the search strategies, and wrote the sections of the report relating to the literature searches. She also contributed to the protocol and interpretation of the results and commented on drafts of the report.

Seun Anjorin (https://orcid.org/0000-0003-0187-6410) (Research Associate) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. He also contributed to the interpretation of the results and the writing of the report.

Jodie Enderby (https://orcid.org/0000-0002-1446-7512) (Research Associate) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. She also contributed to the interpretation of the results and the writing of the report.

Lena Al-Khudairy (https://orcid.org/0000-0003-0638-583X) (Associate Professor, Evidence Synthesis) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. She also contributed to the interpretation of the results and the writing of the report.

Chidozie Nduka (https://orcid.org/0000-0001-7031-5444) (Senior Research Fellow, Evidence Synthesis) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. He also contributed to the interpretation of the results and the writing of the report.

Hema Mistry (https://orcid.org/0000-0002-5023-1160) (Associate Professor, Health Economics) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. She also contributed to the interpretation of the results and the writing of the report.

G. J. Melendez-Torres (https://orcid.org/0000-0002-9823-4790) (Professor, Evidence Synthesis) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. He also contributed to the interpretation of the results and the writing of the report.

Sian Taylor-Phillips (https://orcid.org/0000-0002-1841-4346) (Professor, Evidence Synthesis) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. She also contributed to the interpretation of the results and the writing of the report.

Aileen Clarke (https://orcid.org/0000-0001-8299-3146) (Professor, Evidence Synthesis) contributed to the protocol, study selection, validity assessments and synthesis of the included studies. She also contributed to the interpretation of the results and the writing of the report.

Ethics statement

This work is a systematic review of accessing, processing and analysing data that has already been published and is available to the public. As a result, no patient data were processed; and patient consent and/or registration via human research ethics committees were, therefore, not relevant.

Data-sharing statement

No new data have been created in the preparation of this article and therefore there is nothing available for access and further sharing. All queries should be submitted to the corresponding author.

Consent for publication

Not applicable.

This article

The contractual start date for this research was in February 2019. This article began editorial review in June 2022 and was accepted for publication in April 2023. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Disclaimer

This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number 17/148/05. The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the Health Technology Assessment programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, the Health Technology Assessment programme or the Department of Health and Social Care.