Effectiveness of clinic-based cardiovascular disease prevention: A randomized encouragement design experiment in the Philippines
Introduction
Efforts to improve primary prevention of CVDs in lower- and middle-income countries (LMICs) often rely on opportunistic screening for risk factors at health clinics (WHO, 2013). Most notably, the World Health Organization (WHO) Essential Package of Noncommunicable Disease (NCD) Interventions for Primary Health Care in Low Resource Settings (PEN) sets out protocols for clinics to routinely assess adult patients for CVD risks (WHO, 2010, 2020). In addition to CVD risk assessment and screening, this package specifies criteria for lifestyle counselling, referral, and prescription of medicines, such as antihypertensives and statins, of proven efficacy. Building on this, the HEARTS package (WHO, 2016) aims to strengthen the primary care response to CVDs by stipulating clinical protocols and advocating a risk-based management approach set out in the PEN.
These protocols specify efficacious interventions that modelling suggests are highly cost-effective (Lim et al., 2007; WHO, 2017). Simulations indicate that a package of interventions that includes the PEN-prescribed medicines for prevention and treatment of ischaemic heart disease and stroke would be almost sufficient to hit the Sustainable Development Goals (SDG) target of a one-third reduction in premature mortality from NCDs in the 20 countries (including the Philippines) with the largest NCD burdens (Bertram et al., 2018). There is no evidence, however, on the extent to which a visit to a clinic that should be routinely following the PEN protocols is effective in raising exposure to CVD prevention.
Observational studies have described (often deficient) PEN implementation (Aye et al., 2020; Wangchuk et al., 2014) and have compared outcomes at PEN and non-PEN clinics (AlHelo and Elessi, 2019). But without inducing random exposure to a PEN clinic, these studies were not capable of identifying effects of visiting such a clinic. A systematic review of experimental study evidence on clinic-based CVD prevention interventions in LMICs found that dissemination of clinical guidelines did not always lead to their implementation and had inconsistent effects on blood pressure, glucose, and lipid outcomes (Lee et al., 2016). A cluster-randomized trial in China and Nigeria found that a precursor of the PEN protocol for CVD risk management was efficacious in reducing the blood pressure of hypertension patients (Mendis et al., 2010). However, more than half of the patients exposed to the package still had uncontrolled hypertension after 12 months, and there was no consistent effect on behavioural risk factors (Mendis et al., 2010). These effects were obtained under experimental conditions that required clinics to follow the protocol. Effectiveness could be even lower where there are gaps in implementation of the protocol (Breda et al., 2019; Peters et al., 2019).
This study aimed to establish by how much a visit to a clinic that should have been screening opportunistically as per the PEN protocol raised exposure to primary prevention of CVD and, consequently, reduced risk factors. In randomly selected communities in one province of the Philippines, we used the conditional offer of a money-prize lottery ticket to induce random variation in visits to public health clinics responsible for operating the PEN. This allowed us to estimate effects of a check-up visit on exposure to CVD prevention processes (tests, diagnoses, medication, and medical advice), health behaviours, CVD risk factors, and predicted CVD risk. We did not aim to evaluate the impact of introducing the PEN. Nor did we seek to identify impediments to its implementation. Rather, we aimed to quantify effects of visiting a clinic operating the package under real-world conditions.
Section snippets
CVD prevention in the Philippines
CVDs are the main cause of death in the Philippines, accounting for around one third of all deaths in 2017 – just above the global average (GBDCN, 2018). Between 1990 and 2017, the age-standardized CVD death rate rose by 15 percent in the Philippines, while it fell by 20 percent in Southeast Asia on average (GBDCN, 2018, see Appendix A). CVD mortality risk in the country moved from being equal to the Southeast Asia average at the beginning of this period to 40 percent above the average at the
Evaluation approach
To identify effects of visits to clinics operating PhilPEN on exposure to CVD prevention processes, it would not have been sufficient to observe whether the protocol was followed when patients visited those clinics. These patients may have received tests, diagnoses, medicines, and medical advice even if they had not attended public clinics. Further, observation at clinics would not have identified effects on health behaviour and CVD risk factors, which depend on patient adherence to referrals,
Identification and estimation
The study aimed to estimate effects of visiting a public health clinic responsible for operating PhilPEN on exposure to CVD prevention processes, health behaviour, and CVD risk factors. This was achieved by identifying the variation in each outcome that was associated only with public health clinic visits that were induced by the random lottery offer. We estimated models with the following general structure,where is a latent outcome at
Preliminary analyses
Out of 1865 and 1861 households approached in the intervention and control groups, respectively, 168 and 137, respectively, did not respond, refused, or were ineligible (Fig. 1). There were 1697 participants in the intervention group and 1724 in the control group at baseline. Enumerator errors invalidated the (anthropometry) data of some participants (63 in intervention group, 76 in control group). Each was replaced by someone randomly selected from the same barangay. Attrition and item
Interpretation and implications
The estimated effects on aggregated outcomes suggest that going to a public health clinic for a check-up increased exposure to CVD prevention through some combination of measurement, diagnosis, and medication of risk factors, as well as medical advice. A check-up was estimated to increase the weighted average of effect sizes of clinic-centred prevention indicators by 0.16 of a standard deviation. Disaggregated analyses pinpointed increased likelihoods of having blood pressure measured (by 17
Conclusion
This study demonstrated that going for a check-up at a public health clinic responsible for conducting opportunistic CVD risk screening using a protocol that is widely adopted throughout the developing world increased the exposure of Filipinos to CVD prevention processes only modestly. They were more likely to have had their blood pressure measured and to have received medical advice on unhealthy habits, but their probability of living with undiagnosed and uncontrolled hypertension fell little,
Authors’ contributions
JC, AK and OO’D conceived the research objective and designed the study. JC and AK obtained approval from authorities, supervised the fieldwork and took responsibility for data collection and management. OO’D designed and conducted the data analysis, and drafted the paper. JC, AK and OO’D edited the draft. JC, AK and OO’D approved the final version of the paper.
Trial Registration
ClinicalTrials.gov NCT03512691, AEA Registry for Social Science experiments AEARCTR-0002867.
Data sharing
Data and code are available from the corresponding author.
Acknowledgements
We acknowledged the funder and research assistance in a separate file submitted to Editorial Manager. Here is the text from that file.Acknowledgements: Thanks to three referees for detailed comments, and also to Kayleen Calicdan, Benedict Evangelista Jr., Kristine Gloria, Jenny Kudymowa and Precious Montilla for excellent research assistance. The study was funded by the Swiss Agency for Development and Cooperation / National Science Foundation Programme for Research on Global Issues for
References (44)
- et al.
Investing in non-communicable diseases: an estimation of the return on investment for prevention and treatment services
Lancet
(2018) - et al.
Generalized reduced rank tests using the singular value decomposition
J. Econom.
(2006) - et al.
Prevention of cardiovascular disease in high-risk individuals in low-income and middle-income countries: health effects and costs
Lancet
(2007) - et al.
Trends in premature avertable mortality from non-communicable diseases for 195 countries and territories, 1990-2017: a population-based study
Lancet Glob. Health
(2020) - et al.
Efficient computation of adjusted p-values for resampling-based stepdown multiple testing
Stat. Probab. Lett.
(2016) - et al.
A community-based comprehensive intervention to reduce cardiovascular risk in hypertension (HOPE 4): a cluster-randomised controlled trial
Lancet
(2019) - et al.
Comparison of the Framingham Risk Score, SCORE andWHO/ISH cardiovascular risk prediction models in an Asian population
Int. J. Cardiol.
(2014) - et al.
Laboratory-based and office-based risk scores and charts to predict 10-year risk of cardiovascular disease in 182 countries: a pooled analysis of prospective cohorts and health surveys
Lancet Diabetes Endocrinol.
(2017) - et al.
Implementation of world health organization package of essential noncommunicable disease interventions for cardiovascular risk management in gaza/Palestine: a retrospective record review study
Dubai Med. J.
(2019) Multiple inference and gender differences in the effects of early intervention: a reevaluation of the Abecedarian, Perry Preschool, and Early Training Projects
J. Am. Stat. Assoc.
(2008)