Patient pathways for cardiovascular diseases in Malaysia and the Philippines: a systematic review [version 1; peer review: awaiting peer review]

1Centre for Global Chronic Conditions, London School of Hygiene & Tropical Medicine, London, WC1E 7HT, UK 2Department of Anthropology, University of the Philippines Diliman, Quezon City, Philippines 3Institute for Social Science Studies, University Putra Malaysia, Serdang, Malaysia 4College of Medicine, University of the Philippines Manila, Manila, Philippines 5Department of Social and Environmental Health Research, London School of Hygiene & Tropical Medicine, London, WC1H 9SH, UK 6Universiti Teknologi MARA, Sungai Buloh, Selangor, Malaysia 7UCSI University, Kuala Lumpur, Malaysia


Introduction
Cardiovascular diseases (CVD) are now the leading contributor to the burden of disease in South East Asia and lost Disability Adjusted Life Years per 100,000 attributed to these diseases have increased by 22.2% between 1970 and 2017 1 . Yet much of this burden is avoidable. Western countries have seen steep falls in the CVD burden in this period, as in Canada (-25.8%), Finland (-35.6%), and the United Kingdom (-53.3%) 1 . Typically, between a quarter and a third of these declines can be attributed to improvements in acute disease treatment and secondary prevention, and a further third to a half to reductions in risk factors, such as hypertension, smoking, and high cholesterol, although the contributions varied depending on the initial risk factor profile [2][3][4] . If these successes are to be replicated in middle-income countries, such as those in South East Asia, concerted action will be needed to reduce risk factors in the population and ensure that those who can benefit from treatment and secondary prevention receive it. This paper has been written to inform a study on barriers to effective care of CVDs in two countries in this region, Malaysia and the Philippines 5 , and how they might be overcome. These are neighbouring countries in South East Asia that have much in common geographically and culturally. However, there are also important differences, such as economic performance (US$ 9,951 vs. 2,988 per capita, respectively 6 ), but especially with respect to their health systems. While both have pluralistic systems, offering both mainstream and traditional treatments, Malaysia has near universal coverage provided by public sector services and lower out of pocket expenditure than in the Philippines (38% vs. 53% respectively in 2016) 7 . Yet both face considerable challenges. In Malaysia, only 7.6% of patients with CVD are taking three or more of the drugs recommended for secondary prevention, and only 2.3% in The Philippines 8 . Corresponding figures for the percentage of individuals with hypertension that is controlled are 12.5% and 13.5% 9 . Effective policy responses are clearly needed but must be informed by a detailed understanding of the barriers that patients with CVD face on their journey through the health system in each country, alongside accounting for socio-cultural contexts that also influence patients' conceptions of -and responses to -illness 10 .
As we explore opportunities to intervene in Malaysia and the Philippines, we must first: (1) identify and analyze existing research on how patients with CVD move through each health system and (2) summarize what is known about barriers and enabling factors that are specific to each stage of the patient pathway. Here we report a systematic review of the published evidence, drawing on our two earlier systematic reviews, where we categorized barriers to effective management of hypertension into those at the level of the health system and in the clinical encounter 11,12 and then analyzed barriers in Malaysia 13 and Colombia 14 . Although the findings are of immediate relevance to Malaysia and the Philippines, we envisage that our review can be used to inform similar approaches elsewhere. The paper presents the barriers we identified according to different stages in the patient pathway, and discusses the implications of the findings in the context of existing literature.

Eligibility criteria
We included primary research (both quantitative and qualitative) undertaken in Malaysia or the Philippines, excluding editorials and commentaries, which focused on the patient pathways (and related barriers and facilitators) related to any CVD (including hypertension, ischemic heart disease, myocardial infarction, heart failure, coronary artery disease, and excluding cerebrovascular disease). We included articles that involve patient pathways, which we define as patient's interactions with different forms of care through the following phases: symptoms, diagnosis, treatment, long-term management, and any contact with secondary and tertiary care. We included papers that contained findings on two or more phases, as well as papers focused on one phase reporting data captured at two or more points in time (i.e., management over time). We excluded intervention studies, as these by definition impose a particular pathway which would not otherwise occur. We also excluded studies which reported preliminary results superseded by subsequent papers, papers reporting individuals in institutionalized care and studies published prior to 2000.

Search strategy
Medline, Embase, Global Health, International Bibliography of the Social Sciences, Scopus, ISI-Web of Science and PsycINFO were searched in February 2018, with a repeat search conducted in July 2020 to update the results. Grey literature was sought through relevant websites, Google, and hand searches of university libraries and online periodicals. No limits were placed on initial publication date or the age range of the included participants. Final search terms were informed by exploratory Medline searches conducted in September-December 2017 within the scope of the 'Responsive and Equitable Health Systems -Partnership on Non-communicable Diseases' (RESPOND) Study 5 . Search terms pertained to two areas: CVD (including signs and symptoms such as elevated blood pressure and angina) and treatment patterns (including such terms as 'patient navigation' and 'care pathway'). The detailed Medline search strategy is provided in the Extended data 15 .

Selection process
Studies were selected according to the preferred reporting items for systematic reviews and meta-analyses (see Reporting guidelines for a completed PRISMA checklist) 16 . Two reviewers (MS and GL) independently reviewed titles and abstracts against inclusion/exclusion criteria, with disagreements resolved by a third reviewer. Full texts of included articles were screened independently by MS and GL, with those fulfilling the inclusion criteria retained. Reference lists of included studies were manually searched for additional relevant articles.
Data extraction and analytic approach Location, sample characteristics, methodology, particular CVD, pathway portion(s), and barriers and facilitators along the pathway were extracted from each article to an Excel spreadsheet. A narrative synthesis was conducted rather than a meta-analysis or meta-synthesis due to study heterogeneity.
Findings are reported by condition, where possible extracting information on different segments (including symptoms, diagnosis, management, hospitalization, and post-discharge) of each pathway are described, along with associated barriers and facilitators. Barriers were extracted and organized according to a framework suggested by Khatib et al. on hypertension control 11 which drew upon earlier theoretical work by Michie et al. 17 Khatib et al. define three types of barrier to hypertension control, including capability barriers (knowledge of behaviors required to achieve control, or the capacity to conduct these behaviors), intention barriers (attitudes or motivations toward the behaviors necessary to achieve control), and health care system-related barriers (availability of resources, affordability of care, and/or acceptability of services).
Quality assessment A quality assessment of included articles was completed by NHHMK, using the RATS guidelines 18 and STROBE checklist 19 for qualitative and quantitative articles respectively. The quality assessment was not used to exclude articles, but to appraise the quality level of the evidence base.

Study selection
The results of the search are summarized in Figure 1. The main database search yielded 784 records, and a further eight were identified via complementary search strategies. After duplicates were removed, the titles and abstracts of the remaining 635 articles were independently screened by GL and MS. Of these articles, GL and MS agreed on inclusion/exclusion on 623 (IIR = 98.1%). Disagreements on 12 articles were resolved in discussion with DB, leading to a full article screen of 82 articles. In total, 64 articles were excluded due to insufficient focus on the patient pathway, leaving 18 studies selected for final review. A manual review of reference lists yielded three additional studies, leading to a total of 21 (Table 1).
Of the 21 studies, 15 focused on hypertension, of which 8 were quantitative, 6 qualitative, and 1 mixed method. Three examined secondary prevention post myocardial infarction, all quantitatively, and there were one each studying atrial fibrillation (quantitative), heart failure and acute myocardial infarction (both qualitative). The final quantitative study focused on modifiable CVD risk factors. Only three studies were set in the Philippines, compared to 18 in Malaysia.

Risk of bias and quality assessment
The quality assessment of included studies is available in the Extended data 15 . Common weaknesses in the quantitative articles included a lack of discussion of sources of bias, study limitations, and source of funding, along with a lack of validity measures. Few qualitative articles reported details on participant recruitment and informed consent procedures.

Hypertension
Initial presentation. A community-based survey in the Philippines 20 found high rates of undiagnosed hypertension (HT), and the same conclusion was reached in Malaysia by a study using a nationally-representative household health survey 21 . Four papers examined how patients found to have HT entered the health system 13,[22][23][24] . In studies of those with known HT, most patients presented at clinics with non-specific symptoms, including headaches or dizziness, causing them to seek medical advice. A minority attended after checking their own blood  Public tertiary hospital in Penang, Malaysia Prospective cohort study. Drawing on medical records, patients were followed from their admission day until discharge. Patients admitted for high BP were followed from admission to discharge at a public tertiary hospital, with BP readings taken at admission, day 1, and discharge. BP was controlled in 24% on admission and 54% upon discharge. Hypertension was poorly controlled among in-patients. BP control was better on discharge than on admission. The number of patients not on treatment decreased significantly to 1% upon discharge.
Kassab et al. Urban and rural locations, Malaysia Health systems assessment including document review and semi-structured interviews, purposive sampling. Most patients with HT were diagnosed after presenting to a clinic with symptoms. Some are diagnosed at community health promotion sessions, though these patients rarely seek out a confirmation of the diagnosis after an initial high BP reading.
Of those diagnosed at the clinic, due to patient denial of their condition, some do not return to the clinical for follow-up and end up in tertiary care after an acute event. Patients tend to visit multiple health care centres, sometimes switching between public and private, and Western and TCAM, impacting negatively upon continuity of care. Patient frequently supplement their HT medication with traditional or complementary medicine, using the latter to treat the side effects of the former. Some patients started taking TCAM when they felt that western medication was not working or they experienced side effects. Patients tend to stop coming in for follow-up once their BP is controlled, only to return later with dangerously high readings. pressure in their homes. An ethnographic study in the Philippines explored health beliefs that may delay seeking care 24 . Hypertension was viewed as "blood that had gone to the head," associated with menstruation or menopause among women, and stress, gambling losses, or difficulties at work among men. While the majority of participants recognized the role of prescribed drugs in treatment, almost as many placed their faith in the power of prayer, a balanced diet, and food such as garlic.

ACE-I = angiotensin
Treatment initiation. As noted above, most studies included patients who had attended clinics where they were diagnosed, with treatment initiated at the time or soon after. One Malaysian study reported that when HT was detected at a community outreach event, few are actually followed up 13 . Qualitative studies in Malaysia 27,35,38 reported patients who declined treatment, which the authors attributed to being in denial of or being confused about treatment, although some overcame denial when faced with adverse health implications. Studies in both countries reported patient concerns about the cost of lifelong treatment 24,29,37 and side effects 13,22,28,31,37 . In particular, some were concerned that they might become dependent on medication, or suspected that prolonged use of medication might harm them, concerns often perpetuated by family members and friends of patients 22 . This points to a failure of communication of the nature of HT and its treatment creating an intention barrier. The complexity of medication regimes 26 and reluctance to adopt a 'hypertensive' label 38 created capability and intention barriers to adherence respectively. Healthcare barriers related mainly to the doctor-patient relationship 13, and 24 , with patients expecting more information from their doctors about HT and medication use 38 .

Long term managemen.
Common themes in many of the papers include frequent discontinuation or changes in treatment 13,24,34 , self-management 24,28,38 , combined use of western medicine (WM) and traditional and complementary medicine (TCAM) 13,28,31,37,38 , and uneven adherence by health professionals to treatment guidelines 35 . Adherence to medication was variable but often poor, although it is not meaningful to calculate summary measures as the samples differed regarding inclusion criteria and measurement of adherence.
One Malaysian study examined continuity of care 34 . Most patients attended a clinic three times per year but this varied from 1 to 13. Only 13% saw the same doctor each time. A cluster of qualitative studies 24,28,38 on choices for treating HT reported widespread patient self-management, in terms of drug types and dosages.
Several Malaysian studies examined the use of traditional medicines. One, using mixed methods, including interviews with health workers and patients, concluded that use of TCAM, either to supplement or replace WM, was common 13 . This view was supported in two further studies, one comprising in-depth interviews with 10 patients in a university clinic 38 and another in a community clinic 28 . In both, many patients rejected WM, at least initially, reducing their prescribed dosage of western drugs and taking TCAM instead or in addition to the western ones, and/or taking medication (both TCAM and WM) only when symptoms were evident. They justified this practice by asserting they 'knew their body the best' or that the body needs to 'take a break' from medications 38 . However, another study, in a Malay village, found that almost all patients taking treatment for HT used WM, with fewer than 5% attending traditional practitioners 31 . Thus, it is plausible that there is considerable variation within the country.
Though HT requires long-term monitoring, attendance at followup appointments was low. Some patients in both countries visited multiple clinics, switching between public and private facilities 24,28,38 . Among patients who initially attended follow-up visits, attendance often decreased over time, with some patients eventually presenting with dangerously high blood pressure readings 13 . One Malaysian study, following patients with HT over a year, reported that only 13.6% always saw the same health professional 34 , but this was undertaken in a university clinic that would be expected to have a high turnover of junior staff.
The advent of automated sphygmomanometers has made it possible for patients to monitor their blood pressure at home. One Malaysian study found mixed effects, leading some participants to become more engaged in HT management (improving adherence), but others to became confused about their treatment 27 . Some participants used the readings to adjust their medications, stopping or decreasing medication if readings were low. While unable to find associations between patient satisfaction and clinical outcomes, one study nonetheless suggests that satisfied patients are more likely to follow physician advice 33 .
A cluster of four studies assessed professional adherence to published HT treatment guidelines in Malaysia 25,32,35,36 . One found that prescribing of antihypertensive agents in primary care clinics was largely inconsistent with current guidelines, with some treatments under-utilized. Medication costs, a shortage of storage space for medications, and a lack of time to discuss diet and lifestyle were identified as barriers to guideline adherence, with one study finding that only two-thirds of Malaysian patients with HT received guideline-compliant pharmacotherapy 36 . One survey found that the care provided in both public and private clinics in Malaysia, including prescription of antihypertensive medications, failed to meet standards set out in guidelines though adherence was worse in public clinics 25 . Barriers include the failure by health professionals to motivate patients to attend follow up, a lack of patient trust in their doctor, and long distances between patients and clinics 13,22 .

Other conditions
While most of the literature focused on HT, a few studies examined care pathways for other CVDs or conditions. Even as we recognize that it is unreasonable to draw firm conclusions from so few studies, we have included them in the interest of completeness.
Initial presentation. One qualitative study examined delays in seeking care by patients with acute myocardial infarction (AMI) in Malaysia 39 . On average, patients took 204 minutes to decide to seek care at hospital after experiencing AMI symptoms. Delays were due to uncertainty over where to seek help, whether symptoms justified medical attention, and denial, with patients persuading themselves that symptoms were controllable and non-life threatening. These factors constitute both capability (individuals were not able to recognize the severity of their condition) and intention barriers (a sense of obligations to employers or family prevented them seeking care).

Long term management.
Most studies of maintenance treatment addressed hypertension but one cohort study reported that 14% of Malaysian patients with atrial fibrillation discontinued dabigatran (an anticoagulant) by 20 months following commencement of therapy, with about a half discontinuing within the first 6 months of treatment 29 . 38% transferred to warfarin and 16% to antiplatelet agents but 46% were on no therapy. While one study reported high adherence to secondary prevention medication 30 , the others showed downward trends over time.
One of those studies reported much greater adherence to regimens that included medications (95.1%) than one that only involved relaxation techniques (22.1%) 40 .
One Malaysian study examined capability barriers to effective treatment of patients with heart failure, based on a small sample of inpatients 41 . These patients had some knowledge of how to manage their conditions, in particular the importance of medication and limiting fluid intake as important to managing heart failure, there was less awareness of advice on limiting sodium 41 . Intention barriers arose from lack of communication with health care practitioners, difficulty in remembering medications, complexity of regimens, and lack of noticeable symptoms.

Discussion
Despite the growing burden of CVD in Malaysia and the Philippines, as well in Southeast Asia as a whole, we found relatively few papers that described the pathways to effective treatment and follow up for patients with CVD (n=21), and most of them are from Malaysia. They typically examine discrete stages in the pathway (e.g. entry or start of medication), providing an incomplete picture of the patient journey. Few consider how the health system might facilitate the patient's journey or how it might take account of their social context, as well as local, cultural conceptions of illness 10 . There are few qualitative or mixed methods studies that could aid interpretation of behavior of patients and providers. Many papers are descriptive and do not seek to conceptualize the different steps on the patient pathway.
The vast majority of the research is on hypertension, with very little on pathways to care and subsequent management of acute coronary syndrome or other cardiovascular conditions. Barriers identified in both Malaysia and the Philippines involve all three of the categories described by Khatib et al., intention, capability, and health system. They challenge a simplistic view that financial constraints are the sole barrier for healthcare. They also highlight the challenges of managing conditions that are lifelong and asymptomatic, for instance, the view that they require treatment only when they have symptoms and that the 'body needs rest' from medicines (capability barriers), and forgetfulness that is inherent to diseases that require daily medication for long, if not indefinite, periods of time. More in-depth longitudinal studies with a qualitative component are needed to understand the nuances of how patients' socio-cultural contexts interact with healthcare services or healthcare system components in the pathway. The influence of TCAM, foregrounded in some of the Malaysian studies, also needs further exploration.
The lack of studies that span the entire patient pathway, or that addresses issues beyond the health system, reflect a limited conceptualization of "patient pathways." To understand why interventions for effective management of HT work, for instance, it would be useful to know how patients' wider social and material contexts contribute to success or failure to progress along these pathways. Patient pathways should be examined over time and space to truly capture their dynamic nature and should be examined alongside other aspects of their life such as routes out of employment due to illness, family dynamics after diagnosis, and local beliefs about disease. Qualitative investigation can help to capture how wider social and material dimensions facilitate or hinder HT control but it should include a longitudinal component to capture pathways in the long-term and engage with the experiences and voices of those affected by CVD. This can then be related and used to complement existing literature in the Asia Pacific region that, while likewise inadequate in characterizing patient pathways, generally reports adequate knowledge of CVDs among patients, but nonetheless identifies various patient-related motivation barriers including 'low desire' 42 , 'ambivalence' 43 , and a sense of lack of urgency 44 , as well as health care system barriers including the failure of doctors to motivate their patients 44 and the 'cynicism' among health care providers with long term management 43 .
This review has some clear limitations. It is limited to academic articles published in English and further analysis, including meta-analysis of quantitative articles, was not possible due to the heterogeneity of study topics, methods, and outcomes. Obviously, the specific findings cannot be generalized to other countries but that is not the point here as we are seeking to understand the context in each country to inform further research to understand in detail the barriers that people with cardiovascular disease face in achieving effective management of their conditions. We would encourage those in other countries, many also facing challenges in delivering proven interventions to their population, to ensure that what information does exist on the management of cardiovascular (and other) diseases in their populations are brought together to inform future research and policy. We hope that our conceptualization of the patient pathway and the identification of barriers as relating to intention or capability of patients and providers and the health system will be useful in these endeavors.

Conclusion
Cardiovascular diseases are the leading contributor to the burden of disease in South East Asia. Studies on CVD patient pathways in South East Asia are critical in order to understand and lessen the burden. The literature we have identified reveal important gaps in our understanding of the entire patient pathway. Additional research is especially needed in the Philippines, using longitudinal mixed methods approaches. One such initiative is the 'Responsive and Equitable Health Systems -Partnership on Non-communicable Diseases' (RESPOND) Study, currently underway in both countries, is doing this 5 .

Data availability
Underlying data All data underlying the results are available as part of the article and no additional source data are required.