Knowledge and awareness of and perception towards cardiovascular disease risk in sub-Saharan Africa: A systematic review

Introduction Cardiovascular diseases (CVDs) are the most common cause of non-communicable disease mortality in sub-Saharan African (SSA) countries. Gaps in knowledge of CVD conditions and their risk factors are important barriers in effective prevention and treatment. Yet, evidence on the awareness and knowledge level of CVD and associated risk factors among populations of SSA is scarce. This review aimed to synthesize available evidence of the level of knowledge of and perceptions towards CVDs and risk factors in the SSA region. Methods Five databases were searched for publications up to December 2016. Narrative synthesis was conducted for knowledge level of CVDs, knowledge of risk factors and clinical signs, factors influencing knowledge of CVDs and source of health information on CVDs. The review was registered with Prospero (CRD42016049165). Results Of 2212 titles and abstracts screened, 45 full-text papers were retrieved and reviewed and 20 were included: eighteen quantitative and two qualitative studies. Levels of knowledge and awareness for CVD and risk factors were generally low, coupled with poor perception. Most studies reported less than half of their study participants having good knowledge of CVDs and/or risk factors. Proportion of participants who were unable to identify a single risk factor and clinical symptom for CVDs ranged from 1.8% in a study among hospital staff in Nigeria to a high of 73% in a population-based survey in Uganda and 7% among University staff in Nigeria to 75.1% in a general population in Uganda respectively. High educational attainment and place of residence had a significant influence on the levels of knowledge for CVDs among SSA populations. Conclusion Low knowledge of CVDs, risk factors and clinical symptoms is strongly associated with the low levels of educational attainment and rural residency in the region. These findings provide useful information for implementers of interventions targeted at the prevention and control of CVDs, and encourages them to incorporate health promotion and awareness campaigns in order to enhance knowledge and awareness of CVDs in the region.


Introduction
Cardiovascular diseases (CVDs) are the most common cause of non-communicable disease mortality in sub-Saharan African (SSA) countries. Gaps in knowledge of CVD conditions and their risk factors are important barriers in effective prevention and treatment. Yet, evidence on the awareness and knowledge level of CVD and associated risk factors among populations of SSA is scarce. This review aimed to synthesize available evidence of the level of knowledge of and perceptions towards CVDs and risk factors in the SSA region.

Methods
Five databases were searched for publications up to December 2016. Narrative synthesis was conducted for knowledge level of CVDs, knowledge of risk factors and clinical signs, factors influencing knowledge of CVDs and source of health information on CVDs. The review was registered with Prospero (CRD42016049165).

Results
Of 2212 titles and abstracts screened, 45 full-text papers were retrieved and reviewed and 20 were included: eighteen quantitative and two qualitative studies. Levels of knowledge and awareness for CVD and risk factors were generally low, coupled with poor perception. Most studies reported less than half of their study participants having good knowledge of CVDs and/or risk factors. Proportion of participants who were unable to identify a single risk factor and clinical symptom for CVDs ranged from 1.8% in a study among hospital staff in Nigeria to a high of 73% in a population-based survey in Uganda and 7% among University staff in Nigeria to 75.1% in a general population in Uganda respectively. High educational a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Introduction aims at synthesizing existing evidence on knowledge, awareness and perception towards these conditions.

Methods
This review was conducted according to the recommendations outlined in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. [18] (S1 File). It was registered with Prospero (CRD42016049165).

Search strategy
We searched PubMed, Medline, Science Direct, Google Scholar, Africa Index Medicus (AIM), Africa Journals Online (AJOL) databases to retrieve relevant primary studies conducted in SSA, using pre-defined search (Title/Abstract) and indexing terms (MeSH/Emtree). Keywords and MeSH terms and their combinations used in the searches were "knowledge", "stroke", "heart attack", "coronary heart disease", "myocardial infarction", "congenital heart disease", "heart diseases", "vascular diseases". Reference lists of full-text papers were hand searched for additional articles and reviewed for relevance in this review. The strategy is provided as a supplementary file (S1 Text).

Inclusion criteria
We included studies that were published in SSA, in English, and in peer-reviewed journals between 2007 and 2015. Papers were from primary research of any design and methodology: quantitative and qualitative and exploring knowledge, awareness and perception of CVD and the risk factors. Studies that were carried out among SSA populations living in Western countries or only described interventions leading to increased knowledge and awareness of CVDs or risk factors and symptoms of CVDs were also excluded.

Definition of terms/concepts
CVDs include vascular diseases in general, CHD, cerebrovascular disease (e.g. stroke), myocardial infarction (MI) and congenital heart diseases. Individuals were required to correctly identify CVD conditions, risk factors and clinical symptoms from a list to gauge their knowledge. Perception was based on individuals' self-assessment of chances of developing CVDs, as well as their understanding of who was at risk to develop the condition. Perception was mostly explored in qualitative studies. The SSA region was classified based on the United Nations classification of countries. [19] Data extraction Two reviewers (DB, FW) conducted data extraction from the identified studies. Information was extracted on: authors, year of publication, study design and population, research methods, types of CVDs studied, findings on the knowledge, awareness of and perception towards CVDs and the risk factors. We extracted additional data on the factors influencing knowledge and perceptions of CVD and the reported sources of information on CVD and risk factors. The exercise was reviewed by JB and KKG, who were also consulted on the extraction process.
NIH quality assessment tool uses 13 criteria to assess and rate the quality of studies. This included the research question, study population, sample size estimation, exposure and outcome assessment, loss to follow-up and statistical analysis. General guidance is provided for determining the overall quality of the studies and to grade their level of quality as good, fair or poor.
Qualitative studies were appraised using the Critical Appraisal Skill Programme (CASP) tool.
[21] The CASP tool has 10 items that look at the relevance and clarity of research goals, appropriateness of the research design and methodology in addressing the research question, recruitment strategies, data collection, data analysis, findings, ethical consideration and value of the research. Questions attached to these items enable critical self-reflection about biases and assess the extent to which findings from the study could be transferred to other settings or groups. The quality assessment and criteria are available as a supplementary file (S2 File).

Synthesis of findings
Qualitative data synthesis of the findings on the knowledge, awareness of and perception towards CVD risk and risk factors in SSA was conducted. Findings from the quantitative papers were absorbed using the multi-source synthesis method, an analytical technique that enhances transparency when synthesizing quantitative and/or contextual data, thus providing a platform for comparison between studies.
[22] Findings from qualitative articles were integrated with those from the quantitative studies based on similar themes or topics. Due to the heterogeneity in outcomes, data were not pooled to conduct a meta-analysis.

Study characteristics
A total of 2212 titles were identified from electronic database searches. 2167 titles were excluded for being irrelevant to the review question, and 45 full-text articles were assessed for inclusion. Twenty-five articles were excluded based on reasons such as not reporting the link between risk factors to general knowledge and awareness of CVDs or reporting results of an impact of an intervention in the levels of knowledge and awareness of CVD and risk factors. In the end, 20 articles were included in the review. The assessment and inclusion criteria are reported in Fig 1. One of the 18 quantitative studies out of the final 20 studies was quasi experimental, while the rest were cross-sectional. Respondents were recruited from varied settings, including from general population samples living in urban and rural areas, and from specific samples like academic staff, hospital staff and health professionals, patients, and employees in banks and in the military. The age of the participants in the different studies ranged from 16 to 82 years. More information on characteristics of study participants is presented in Table 1.

Quality of included studies
The majority of the quantitative studies were rated to be of good or high quality (n = 10). They described in detail the design and methodology used, the process of recruiting participants, justification and methods of arriving at required sample size, study setting, clear and detailed presentation of findings. Studies that were rated to be of fair or poor quality (n = 8) were papers that failed to describe details of subject recruitment processes including inclusion criteria and sampling strategies and lacked justification of sample size and other issues that could lead to a high risk of bias and undermine generalizability of the study (S1 File).

Knowledge and awareness regarding cardiovascular diseases
Most studies in this review did not state a priori the criteria used in measuring and classifying levels of knowledge and awareness. However, most of them classified knowledge and awareness of CVD or the risk factors as poor, acceptable or good. In the study by Akintunde et al, [23] among university staff, a knowledge score of <50% was classified as low; 50-69% moderate and !70% good. Nakibuuka et al [24], in a study in Uganda classified urban and rural residents who could identify 5-10, 2-4 and <2 CVD risk factors or warning signs as having good, fair and poor knowledge respectively.
Awareness of CVDs was high among studies that reported on it; 76.2% among bankers and teachers[25] and 75.6% among military personnel [26] in Nigeria. Most people in a low-income peri-urban community in South Africa, [27] were familiar with the terminology used to describe CVDs. However, the studies reported generally low knowledge levels of CVDs with most studies reporting less than 50.0% of respondents having good knowledge. In studies conducted among workers in a Nigerian University Hospital, one reported that 19.0% had good knowledge of CVDs[23] while another showed that 53.5% knew the mechanism through which stroke occurs.
[28] Findings on the knowledge and awareness of CVDs in SSA is summarized in Table 2    Age, gender and education not associated with knowledge of CVDs.
Low knowledge of symptoms of heart disease; 24.6% Education; gender not associated with awareness of CVDs.

Knowledge of risk factors for cardiovascular diseases
To gauge knowledge of risk factors for CVDs, individuals were required to correctly identify them from a list. Just like it was the case with CVD risk, majority of the studies also reported low levels of knowledge on risk factors for CVDs. Hypertension and stress were the most known and cited risk factors in most of the studies. Participants who were unable to identify a single risk factor for CVDs ranged from as low as 1

Diabetes
The knowledge level of diabetes as a risk factor of CVD ranged from 0.3% in a study among urban adult population in Benin [36] to 47.4% among secondary school teachers in Nigeria.
[31] Two community-based studies from Ghana [35] and Uganda [24] reported less than 15% of study participants possessing any knowledge of diabetes as a risk factor for stroke. Knowledge of diabetes as a CVD risk factor among hypertension and diabetes patients at a specialist medical centre in Southern Nigeria was very low, at 7.3%. [37] Smoking Knowledge of smoking as a CVD risk factor was 70.6% among military personnel in Nigeria [26] and less than one percent among the general populations in Central Uganda.
[24] Less than 50% of respondents across all studies could identify smoking as a risk factor for CVD, with the exception of the study among Armed Forces personnel in Nigeria, 70.6%.
[26] In a study in rural Uganda, none of the respondents identified smoking as a risk factor for CVD. [32] In all, 14 studies reported on knowledge of smoking as CVD risk factor, three of which reported <5% with knowledge of smoking as a risk factor for stroke 26,42 and for CHD. [38] Physical inactivity Knowledge of physical inactivity or sedentary lifestyle as risk factors for CVD ranged from 0.6% [37] to 57%, [31] in Nigeria. Two other studies reported knowledge level of less than 10%; 1.2% in a rural Nigerian community [33] and 3.8% among hospital outpatients. [37] Heavy alcohol consumption Heavy alcohol consumption as a risk factor for CVD was reported by 4.5% in a study among patients with hypertension and/or diabetes at specialist medical outpatient clinics in Nigeria [37] [34] whereas studies among urban communities in Ghana [35] and Benin [36] reported low knowledge of stress (22% and 7.6%) respectively.

Other risk factors
Other risk factors for CVD were ageing, family history, obesity and unhealthy diet. Knowledge of these risk factors was low across studies reviewed and was least cited or known among study subjects. Ageing was identified as a risk factor for CVD by 63 [40] and as low as 1.3% in the study conducted among people living with HIV/AIDS in Cameroon. [38] Of nine studies, five that were conducted among people living with HIV, [38] hypertension and/ or diabetes outpatients, [37] rural population, [33] urban population [36] and the general population,[24] <10% identified obesity as a CVD risk factor. The biggest proportion with knowledge of obesity as a CVD risk factor, 56.1% was reported among staff of a University in Nigeria.
[29] Knowledge on diet as risk factor for CVD was 99.1% among secondary school teachers [31] and <10% among Armed Forces personnel [26] in Nigeria and the general household population in Uganda. [32] Unhealthy diet was also reported as a risk factor in two studies. [27,39] Knowledge of symptoms/ clinical signs of cardiovascular disease The proportion of respondents who could not identify a single symptom of any CVD condition ranged from 7.0% among academic staff in a University in Nigeria[29] to 75.1% among the general population in Uganda.
[24] The proportion of respondents who could identify all symptoms ranged from 3.5% among teachers [31]  more likely to agree that they were at a higher risk for CVDs. [41] In a qualitative study from South Africa,[27] participants were described as being generally unfamiliar with the concept of risk, while the two respondents who were familiar with the concept of risk could also not explain in detail what it actually meant. In a study of medical out-patients in a tertiary health institution in Nigeria, [40] majority (65.8%) of the respondents were never concerned about the possibility of developing stroke, 16.1% sometimes thought of it, 12.3% occasionally and 5.8% always had the concern. 34.1% of respondents in a population-based study from Uganda [24] perceived no chance while 14.4% perceived high chance of possible stroke in lifetime.

Factors influencing knowledge of cardiovascular diseases and risk factors among reported studies
Factors such as age and family history, type of residence and education were reported to be associated with knowledge of CVDs. The significant influence of age on knowledge of CVD was reported by three studies. [29,36,37] In two studies from Nigeria conducted among hospital outpatients [37] and university staff,[29] age <55 and <40 were a significant predictor of knowledge of CVDs. There was a significant relationship between educational attainment and knowledge of CVDs. [17,29,33,36,37,40] As reported in a study from rural South-Western Nigeria, [33] people with tertiary education were three times more likely to be knowledgeable of CVD risk factors and a study among hospital outpatients in Nigeria [37] showed that more than 12 years of education increased the odds of being knowledgeable about CVD risk factors by more than twice. A significant association between type of residence and knowledge of CVD was also described: urban residents were more knowledgeable about CVDs compared to their rural counterparts in a community study in Uganda[24] and a study among diabetic/ hypertensive outpatients. [37] No study reported a relationship between gender with knowledge of CVDs. [33,42] Sources of information on cardiovascular diseases  Table 3.

Discussion
This review identified low levels of knowledge and awareness of CVDs and associated risk factors and clinical signs or symptoms for CVDs among populations in SSA. The knowledge gap is also apparent in the low perception regarding the risk of developing and dying from CVDs in the region. [24,40] In population-based studies conducted in Uganda[24] and Benin, [36] respondents were unable to identify the organ affected by stroke, despite it being a condition with poor survival outcomes in this region. [43][44][45] Knowledge of clinical symptoms was as low as 3.5% among teachers in Nigeria, [31] while as few as 16.2% in a rural Nigerian community [33] knew that of hypertension, 0.3% for diabetes and 1% for obesity and 7.6% for stress in Urban Beninese population, [36] as risk factors or developing CVD. A systematic review of awareness of hypertension in West Africa reported overall low knowledge of hypertension. [46] Studies that explored knowledge and perceptions of obesity and sedentary lifestyles showed poor perceptions and subjective norms such as overweight being socially desirable, and a sign of beauty and riches thereby inducing unwillingness to lose weight. [47,48] African belief systems are however not static-they are complex and dynamic, tied as they are to shifting social identities. Other body of evidence suggests that contrary to the often-cited fatness equals wealth, health and beauty theory, young African women view fatness as a precursor for CVDs. [49] These women are interested in living a healthy life and are willing to reduce their body size in order to reduce the risk of obesity-related diseases despite the resistance to lose weight because of the cultural value on weight and the impact of the husband's preference. [50] These inherent perceptions and desire to lose weight should be important considerations when designing educational interventions to improve knowledge of CVDs.
Despite the rise in CVD risk factors in SSA populations, our findings indicate that the populations generally did not recognize their potential relation to the development of CVDs. In SSA, the incidence and prevalence of classical risk factors of CVDs such as smoking, [51] hypertension, [52] obesity, [53][54][55] high cholesterol, fatty diets, alcohol consumption [56][57][58] and lowered physical activity [59] are rising. This rise is linked to rapid urbanization, resulting in an epidemiological and nutrition transition, where energy-dense diets replace traditional diets and sedentary lifestyles prevail poverty. [10] As such, there is a shift in disease burden from under-nutrition and highly active lifestyle to over-nutrition-related and sedentary lifestyle related chronic diseases. Knowledge and awareness of and perception towards cardiovascular disease risk in sub-Saharan Africa Knowledge of alcohol intake as a risk factor for CVD was low in the region. Four studies [24,34,37,41] reported on this and found that <30% of study participants cited alcohol consumption as a risk factor for CVDs; in a study among medical outpatients, [40] none identified alcohol consumption as a risk factor for CVD. In most societies in SSA, use of alcohol has been defined by cultural and religious parameters, with little acceptance of the potential health effect of alcohol consumption on health. [60] This is of concern, considering the expansion of alcoholic industries commercial activities in SSA to increase sales in this region. [61,62] Adequate policies to address these challenges in SSA are however few whereas there are no developed multi-sectorial approaches, that involves the private sector, civil society, informal sector, community leaders and traditional healers. [63] Further, in countries where there are preventive interventions such as enactment of drinking and driving laws, taxation, restrictions on advertising and community information, implementation is ad hoc, informal, fragmented and often lacks adequate control and enforcement systems. [63] The relationship between alcohol consumption and CVDs is nuanced. Light to moderate drinking has been suggested to decrease the incidence of ischaemic stroke, whereas heavy drinking has been implicated as an independent risk factor for ischaemic and haemorrhagic stroke. [64][65][66] For hypertension, cardiac dysrhythmias and haemorrhagic stroke, alcohol is considered to be an independent risk factor, regardless of the drinking pattern. [67] This emphasizes the need for the development and enforcement of adequate and effective policy measures, public awareness and surveillance mechanisms in the SSA region. Without awareness of personal susceptibility and health consequences related to alcohol consumption, alcohol consumption behaviours are less likely to be modified to reduce risk of CVD.
Knowledge on stress as a risk factor of CVD was relatively high, especially among urban populations, despite the complex relationship between stress and CVDs. [68] Susceptibility to stress is influenced by type of personality, social support, coping strategies and genetic vulnerability. [68] Stress could be positive, by forcing us to adopt and thus to increase the strength of our adaptation mechanisms (eustress) or negative, when it exceeds our ability to cope, fatigues body systems and causes behavioural or physical problems (stressors). [68,69] A strong association has been observed between perceived stress and CHD [70][71][72] and current evidence shows perceived stress to be an independent risk factor for stroke. [73] The belief and perception of the influence of stress on CVDs in SSA populations could however be related to experiences of psychosocial stressors arising out of urbanization and poverty. [74,75] Experiences of chronic poverty-related stressors, such as inadequate housing, sanitation, water, overcrowding, environmental conditions, low education and unemployment, are potent predictors of poor cardiovascular health. [76][77][78] Strategies to deal with perceived psychosocial stress among these populations, include smoking and alcohol consumption, which themselves are precursors of poor cardiovascular health. [79,80] This review shows knowledge of CVDs and their risk factors to be significantly related to the type of population studied and place of residence, and the level of exposure to health information about CVDs. Studies that formally tested the association between place of residence and education on knowledge of CVDs, also reported a significant relationship. [24,32,37] There is the possibility that the differences observed in the levels of knowledge among the urban and the rural populations are driven by the fact that the urban, and mostly formally employed/ working population is more likely to be educated and more exposed to the media and other modern sources of health information, including the internet. [81,82] The rural population and uneducated on the other hand, are most likely to be poor, and less likely to be exposed to print and electronic media which have been reported as major sources of information on CVDs and risk factors. The rural populations in SSA have also been shown to utilize health services less than their urban counterparts, [83,84] and rely on information from their families. [33] Exploring the determinants of health in rural areas, such as the role of the family, is therefore important if health promotion policies and strategies are to result in significant improvements in health status.
Traditionally the major sources of information on CVD, respectively CVD risk factors have been shown to include electronic and print media (television, radio, newspaper) and health workers, [85,86]. Recent studies have quoted the internet as an important source of health information, especially among urban populations, teachers and other formally employed individuals, clearly illustrating the influence of the internet in health care. This situation presents an important consideration for public health policy and resource allocation for health promotion strategies in these settings.

Strengths and limitations
This review presents evidence regarding the knowledge and awareness of CVDs in SSA. To the best of our knowledge, this is the first systematic review of the knowledge and perceptions of CVDs in SSA. Our results are based on a systematic search of five databases, integrating both qualitative and quantitative evidence on the topic. The inclusion of qualitative studies in this review meant that research findings on perceptions towards CVDs were incorporated and contributed to our understanding of and explanation of the trends of knowledge of CVDs in this study setting. As the criteria of measurement of knowledge of CVD (risk factors) was not uniform across studies (different criteria were used for classifying knowledge into low, medium or high resulting in heterogeneity across study findings), a meta-analysis could not be conducted. As the study populations differ considerably within and between countries it is difficult to disentangle to what extent educational level or cultural or country level determine knowledge and awareness levels. Still, the qualitative synthesis of available evidence of knowledge and perceptions of and perception towards CVD risk and risk factors presented in this review should speak to the current situation as most studies were published.

Conclusions
Generally, inadequate knowledge of CVDs and the associated risk factors continues to be one of the most important factors in determining health-seeking behaviours in SSA. Knowledge levels of CVDs, risk factors and warning signs were mainly varied by type of populations and influenced by the type of employment, education levels and place of residence. Formal workers were more aware of and knowledgeable about CVD and the risk factors compared to studies conducted within rural and urban households. What this means is that education must be tailored for different groups. One-size fits all messaging is unlikely to work. Misconceptions (damaging cultural beliefs such as witchcraft and spiritual causal theories) must be addressed in ways that enhance biomedical understandings without stigmatizing cultural understandings. Adequate attention and awareness creation on the adverse implications of CVD related risk behaviours such as smoking, alcohol consumption and sedentary lifestyle on this population cannot be overemphasized. Effective policy measures, public awareness and surveillance mechanisms that takes into consideration the socio-cultural context of these behaviours need to be developed and implemented in this region. Evidence provided in this study can guide context specific interventions, aimed at mitigating CVDs by improving levels of knowledge and awareness of the conditions and risk factors among SSA populations.