Modifiable and Non-modifiable Factors Associated with Low Awareness of Hypertension Treatment in Indonesia: A Cross-Sectional Population-Based National Survey

Introduction: Low awareness about hypertension treatment is recognized as a significant cause of treatment failure. Therefore, identifying its underlying factors is essential for developing effective intervention strategies. This study aims to identify the modifiable and non-modifiable factors associated with low awareness about hypertension treatment. Method: This national, cross-sectional, population-based survey used publicly available data from the Indonesian Family Life Survey (IFLS-5) for 2014 among respondents with hypertension aged ≥15 years. Depression and insomnia, as modifiable factors, were assessed using the Centre for Epidemiologic Studies—Depression (CES-D) and the Patient-Recorded Outcomes Measurement Information System (PROMIS) questionnaire, respectively. Non-modifiable factors, such as sociodemographic information, were obtained from self-reported data. Logistic regression analysis was used to assess the association between these factors and low awareness about hypertension treatment. Odds ratios (ORs) with 95% confidence intervals (CIs) were reported. Result: The study recruited 7,920 respondents, the majority of whom were female (53.8%) and aged <60 years (71.1%). The prevalence of low awareness of hypertension treatment was 87.1% (51.8% in women and 48.2% in men). Being an elderly (OR: 1.60, 95%CI 1.36–1.88), being irregularly blood pressure control (OR: 4.40, 95% CI 3.78–5.13), having depressive symptoms (OR: 1.35, 95% CI 1.12–1.62), having insomnia (OR: 1.31, 95% CI 1.11–1.53), and having low satisfaction with health care (OR: 1.28, 95% CI 1.08–1.51) were associated with low awareness of hypertension treatment. Surprisingly, respondents with strong religiosity (OR: 1.62; 95% CI 1.25–2.09) were more likely to display low awareness of hypertension treatment. Conclusion: The main factors associated with low awareness of hypertension treatment are modifiable. Thus, health care professionals should integrate more patient-specific factors when designing tailored interventions.


INTRODUCTION
Hypertension is an increasing part of the global burden of disease, which is more predominant in low-and middle-income countries (LMICs) [1]. The current global trends in hypertension has increased from 650 million to 1.28 billion over the last three decades [2]. A population-based study on more than 90 countries revealed that the burden of hypertension has increased by 7.7% in LMICs but decreased by 2.6% in high-income countries (HICs) [3]. Globally, 8.5 million deaths were associated with uncontrolled high blood pressure due to unawareness of this condition [4]. In fact, awareness of hypertension treatment is reportedly lower in LMICs compared with other countries [5]. In 2017, Indonesia reported 91.3 million people with hypertension, which was responsible for 35% of all deaths [6,7].
Uncontrolled blood pressure is predominantly frequent in LMICs, where health awareness is the weakest and mortality rates due to cardiovascular disease are the highest [8]. According to Indonesia Basic Health Research 2018, only 37% of Indonesians with hypertension currently have obtained control of their blood pressures, whereas 45.6% exhibited low adherence to hypertension treatment [7]. Low awareness of medication is reported to be associated with medication non-adherence [9,10], which may lead to poorly controlled hypertension, high risks of complication, and increased health care costs [11][12][13]. Therefore, low awareness of hypertension treatment should be addressed due to the lack of symptoms of the disease.
Previous studies have reported several factors that may be associated with awareness of hypertension treatment [14][15][16][17]. However, such studies have mainly explored non-modifiable factors. A previous study conducted in LMICs revealed that insurance status and income were significantly associated with awareness of treatment of hypertension [18]. Moreover, low levels of education were associated with low rates of awareness about hypertension treatment in LMICs but not in HICs [5]. However, the association of modifiable factors, such as satisfaction with health care, blood pressure control status, depressive symptoms, and insomnia, to awareness of hypertension treatment in Indonesia remains unclear.
The Indonesian Family Life Survey (IFLS) is a longitudinal socioeconomic and health survey in Indonesia, which represents approximately 83% of the Indonesian population [19]. Several studies that utilized IFLS-4 or IFLS-5 demonstrated the association of non-modifiable factors, such as socioeconomic status [20] and psychosocial risk [17], to hypertension in Indonesia. However, a clear evidence of which factors or foci are required to improve medication awareness of hypertension treatment in Indonesia remains lacking. Thus, identifying its underlying factors is essential to developing effective strategies for intervention. This study aims to identify modifiable and non-modifiable factors associated with low awareness of hypertension treatment in Indonesia.

STUDY DESIGN
This study used the national longitudinal and cross-sectional data of the fifth wave of the IFLS-5, which were collected from 2014 to 2015 and were publicly available since 2016. IFLS-5 used a multistage stratified sampling design that represents 83% of the Indonesian population conducted in 13 provinces, which are four provinces on Sumatra (North Sumatra, West Sumatra, South Sumatra, and Lampung), all five of the Javanese provinces (DKI Jakarta, West Java, Central Java, DI Yogyakarta, and East Java), and four provinces covering the remaining major island groups (Bali, West Nusa Tenggara, South Kalimantan, and South Sulawesi) with a response rate of more than 90% [19]. The IFLS collected extensive measures of basic sociodemographic, certain economic characteristics of all household members, and health status, including self-reported measures of general health status, symptoms, pain, and biomarker measurements of health. The ethical review boards have approved the IFLS of Research and Development (RAND) Corporation and Gadjah Mada University. Written informed consent was obtained from all respondents before data collection [19].

STUDY POPULATION
The total number of IFLS-5 subjects was 34,464 (aged 0 to >80 years). Data were obtained from subjects who were at least 15 years of age at the completion of the survey. Subjects with available data on blood pressure, depressive symptoms, and sleeping quality were included. Subjects with missing values for outcome variables were excluded.

POTENTIAL FACTORS ASSOCIATED WITH LOW AWARENESS OF HYPERTENSION TREATMENT
The potential factors associated with low awareness of hypertension treatment were classified as modifiable and non-modifiable factors. Depressive symptoms, having insomnia, blood pressure control status, happiness status, having visited health care services, and satisfaction with health care were categorized into modifiable factors. Alternatively, non-modifiable factors consisted of gender, age, marital status, level of education, residency, coverage of health insurance, economic status, and religiosity.
A self-reported Center for Epidemiologic Studies Depression (CES-D) scale was used to measure depressive symptoms [19]. CES-D contains 10 questions to assess the feelings of subjects, which are highly correlated with depressive symptoms [25]. Eight questions were focused on negative symptoms (e.g., I was bothered by things that usually do not bother me), whereas the two other questions assessed the positive symptoms of depression (e.g., I felt hopeful about the future). The subjects indicated the frequency that each item applied to them in the past week using a four-point Likert-type scale (0 = rarely or none of the time; 1 = some or little of the time; 2 = moderately or much of the time; 3 = most or almost all the time). The final score is calculated by summing all items after reversing the positive mood items. A subject with a total score equal to or above 10 is considered to have symptoms of depression [26]. The CES-D questionnaire was translated into Indonesian (forward translation), where two independently hired translators re-translated it into English (back translation) [19].
Insomnia was assessed using 10 questions from PROMIS [19]. Five items each were used to assess sleep quality and sleep impairment in the past week [27]. Each item was rated using a five-point Likert-type scale (0 = never/not at all; 1 = a little bit; 2 = somewhat; 3 = quite a bit; 4 = always/very much). Insomnia was defined as total scores of ≥21-40 [28]. The PROMIS questionnaire was translated into Indonesian (forward translation), where two independently hired translators re-translated it into English (back translation) [19].
Blood pressure control status was categorized as regular and irregular dependent on the subject's response to the question, How regularly do you have your blood pressure checked? Happiness status was assessed using the question, Taken together, how you would say things are these days? Would you say you are very happy, happy, unhappy, or very unhappy? Those who answered very happy or happy were categorized as happy, whereas those who answered unhappy or very unhappy were categorized as unhappy [29]. Satisfaction with health care was assessed using the question, Concerning your health care, which of the following is true: it is less than adequate for my needs; it is just adequate for my needs; it is more than adequate for my needs? Low satisfaction with health care is equivalent to the response it is just less than adequate for my needs, whereas high satisfaction with health care is indicated by it is just adequate for my needs or it is more than adequate for my needs [29].
Sociodemographic factors were gender (men, women), age at the completion of the survey (less than 60 years for the non-elderly and more than 60 years for the elderly), marital status (currently unmarried and currently married), level of education (no education, elementary school, junior high school, senior high school, and university), residency (rural and urban), coverage of health insurance (no and yes), economic status (poor and non-poor), and religiosity (religious, non-religious). The total income of the family was used to measure economic status over the past 12 months in rupiah divided by family size (per capita income). Capita income was categorized per quintile. Religiosity was assessed using the question How religious are you? Those who answered very religious or religious were categorized as religious, whereas those who answered somewhat religious or not religious were considered as non-religious [29]. The subject was asked whether they visited any health care services one month prior to the survey to assess health care visit status (yes and no).
To determine whether subjects have comorbid disease status, subjects were asked the question, Has a doctor/ paramedic/nurse/midwife ever told you that you had the following chronic conditions of diseases? with response options: hypertension, diabetes, tuberculosis, asthma and other chronic lung diseases, cardiac diseases(heart attack/coronary heart disease/angina or other heart diseases), liver diseases, stroke, cancer or malignancies, arthritis/rheumatism, uric acid/gout, depression, vision and hearing abnormalities [29]. Subjects who answered only hypertension were categorized as having no comorbidity, while subjects who answered hypertension and one to three other chronic disease were then categorized as having one to three comorbidities, and subjects who had more than four comorbidities were categorized as having more than four comorbidities.

STATISTICAL ANALYSIS
Descriptive statistics were used to summarize the characteristics of the subjects. Awareness of hypertension treatment was estimated for each age group and gender. A chi-square test was performed to assess the univariate association between the characteristics of the subjects and the outcome. We conducted complete case analyses because the missing data were small. The potential factors found associated with the outcome at a significance level of p < 0.25 in the univariate analyses were included in the initial multivariate model. Multivariate logistic regression was performed to obtain the odds ratio (OR) with a 95% confidence interval (95% CI) with manual backward elimination. The p-values were set at 0.05 for factors included in the final model. Subgroup analyses by number of comorbidities was performed in no comorbid, having one to three comorbidities, and having more than four comorbidities groups. The goodness-of-fit statistic was assessed using the Hosmer-Lemeshow test. Pseudo R-squared value, which is the default value reported by Stata, is obtained as an equivalent to the R-squared reported in the linear regression. R-squared is a number that ranges from 0 to 1, which indicates the extent to which the combination of independent variables simultaneously influences the value of the dependent variable [30]. All statistical analyses were performed using Stata software version 14.0 for Windows.

RESULTS
A total sample of 7,920 subjects without missing data on awareness of hypertension treatment was included in this study. The majority were female (53.8%) and aged <60 years (71.1%; Table 1 The prevalence of low awareness of hypertension treatment was 87.1% (51.8% in women and 48.2% in men). Level of education, marital status, economic status, and happiness status were negatively associated with low awareness of hypertension treatment (   Table 3). The goodness-of-fit p-value of the model was 0.271 with a pseudo-R-squared value of 8.74%. Overall, modifiable and non-modifiable factors that were associated with low awareness of hypertension treatment were relatively similar in subgroup analysis, except that insomnia and depressive symptom status were not seen as risk factors for low awareness of hypertension treatment among subjects with more than four comorbidities (Table 3).

DISCUSSION
Among the 7,920 respondents, 87.1% reported low awareness of hypertension treatment. Age and religiosity are non-modifiable factors associated with low awareness of hypertension treatment. Meanwhile, the modifiable factors are blood pressure control status, satisfaction with health care, depressive symptoms, and insomnia. Medication awareness is critical in providing accurate information about the treatment undertaken to improve medication adherence [31]. Patients with low awareness of treatment tend to be non-adherent to treatment for chronic diseases [32]. Medication awareness is one of the five broad categories of treatment-related problems in a patient perspective that have potential implications for therapeutic success as they create barriers to medication adherence [33]. However, low adherence to treatment in chronic disease is associated with poor disease control, increased costs, increased hospitalization rates, decreased quality of life, and increased mortality [34][35][36][37].
We observed that the elderly subjects were associated with low awareness of hypertension treatment, which can be partially explained by several possible barriers, such as frailty, multiple comorbidities, polypharmacy, and cognitive impairment [38]. Surprisingly, religiosity indicated a significant association with low awareness of hypertension treatment. A possible explanation may be that people foresee and surrender healing outcomes from God. Thus, they would risk not being aware of taking medications [39,40]. In addition, previous studies reported that religiosity could increase the likelihood of using complementary and alternative medical (CAM) practices, such as herbal medicines [41,42]. The high prevalence of CAM among patients in Ethiopia with hypertension is associated with low awareness of health care, which potentially leads to ineffective hypertension management [43]. Another finding from a previous study reported decreased adherence to prescribed medication due to CAM use in patients with type 2 diabetes [44].
We further observed that blood pressure control status, satisfaction with health care, depressive symptoms, and insomnia are associated with low hypertension awareness. Subjects who irregularly control their blood pressure have a risk of low awareness of hypertension treatment. A possible explanation for this finding may be that the inadequate understanding of therapeutic goals influences their awareness of hypertension treatment [45]. Affordability and unreliable devices are the two main reasons why patients do not check their blood pressure regularly [46]. Individual medical education is the most critical aspect of blood pressure regulation [47]. The fundamental method for hypertension control is to make people more aware of it, treat it, and maintain it under control [48]. The problem of achievement of target blood pressure consists of patient-related barrier, physician-related barrier, and health care system-related barrier, so that a team-based approach to blood pressure control that involves nurses, pharmacists, and physician assistants should be emphasized [49]. People with high blood pressure can benefit from antihypertensive medication, while complications and aging make it less probable to achieve the desired blood pressure [47].
We also observed that patients with low satisfaction with health care were likely to display low awareness of hypertension treatment. Previous studies revealed that satisfaction with health care was associated with adherence to hypertension treatment [50,51]. Subjects with better satisfaction with health care may increase their awareness of hypertension treatment. Furthermore, having a depressive status was associated with low awareness of hypertension treatment, which can be partially explained by the deterioration of high-level cognitive functions, specifically attention, awareness, and memory, among subjects with depressive symptoms [52]. At the same time, having depressive disorders has been associated with hypertension [53][54][55]. A systematic review and meta-analysis demonstrated that the risk for cardiovascular diseases in subjects with depressive symptoms is approximately two times higher than those in non-depressed people [56]. Similar results were observed in subjects with insomnia. A possible explanation is that the severe decrease in sleep hours is related to cognitive function, which leads to the lack of awareness of treatment. Prior studies have reported a significant relationship between sleep duration and cognitive function [57,58]. This finding indicates the need to screen for sleep impairment and to maintain regular sleep patterns.
Hypertension frequently coexists with diabetes, hyperlipidemia, or other metabolic syndromes [59]. Patients with these comorbidities are more likely to require combination therapy, which needs more medications. The result of subgroup analysis showed that both subgroups had 9 Khoiry et al. Global Heart DOI: 10.5334/gh.1143 relatively similar modifiable and non-modifiable factors that were associated with low awareness of hypertension treatment. Therefore, health care providers need to pay more attention to patients with low awareness regardless of their number of comorbidities.
In summary, the majority of sociodemographics as non-modifiable factors were not associated with low awareness of hypertension treatment. Sociodemographics such as marital status may be extremely general for predicting the medication awareness of subjects. No association was found between gender and low awareness of hypertension, as reported in a previous study [60]. In general, gender is seemingly a less meaningful factor associated with low awareness of hypertension treatment. Moreover, the level of education is not associated with awareness, which is similar to the findings of previous studies [61,62]. Thus, health literacy may be more important than level of education [62]. By contrast, another study indicated that low levels of education were associated with low awareness of hypertension in LMICs [5]. We observed that residency was not associated with low awareness of hypertension treatment. This finding is in contrast with that of a previous study [63], which found an association between people living in urban areas and high awareness of hypertension treatment. Population in urban areas has exhibited better treatment-seeking behavior and has easy access to health care [63]. Notably, the current study observed no association between visits to health care facilities with low awareness of hypertension treatment. In contrast, a study using IFLS-4 indicated that the number of visits to any health care facility helped to increase awareness about hypertension [64]. Furthermore, no association was found between happiness status and low awareness of hypertension treatment. This result is consistent with data obtained in a cohort study with a 31-year follow-up in London, which revealed the lack of robust associations between life satisfaction and incident hypertension [65]. One unexpected finding is that health insurance coverage was not associated with low awareness of hypertension treatment. However, one could expect that the limited subjects had health coverage in 2014 since the Indonesian National Health Insurance System began in 2014. Moreover, other studies reported on the positive effect of health insurance coverage due to the reduced cost of receiving health care services [66,67].
The current finding suggests that the majority of non-modifiable factors may be insufficient to address the low awareness of hypertension treatment. Thus, health care professionals should consider modifiable factors in designing tailored and personalized interventions by implementing open communication to educate and to understand patients' views regarding patients' and family members' involvement in the treatment to improve patients' health outcomes [68]. Education and counseling are effective in terms of behavioral changes in patients to clarify misperceptions and improve the awareness of hypertension treatment [69]. Furthermore, health care professionals should to regularly check blood pressure and consider patient satisfaction by providing correct treatment and creating a respectful and caring relationship with patients by sharing the decision-making process [70]. The current findings may serve as reference for health professionals focusing on factors related to low awareness of hypertension treatment that should be addressed (Figure 1).
The strength of this study is that we used IFLS data, which represents 83% of the Indonesian population with a small attrition rate (6%). However, we acknowledge certain limitations related to methodological issues. First, the study was unable to provide causal inferences regarding the association between modifiable and non-modifiable factors related to low awareness of hypertension treatment due to the cross-sectional design. Second, we used complete case analysis, which may reduce statistical power due to the sample size, which can produce biased estimates, leading to overestimating or underestimating conclusions. Third, in this analysis, we included only subjects with a blood pressure of more than 140/90 mmHg, which may lead to overestimating low awareness of hypertension treatment since subjects with controlled blood pressure (blood pressure <140/90 mmHg) are not included. Fourth, the overall association of our model was relatively low, which indicated the presence of other unmeasured factors that may influence the low awareness of hypertension treatment, such as education about hypertension, origins of subjects, local tradition, number of medication, medication beliefs [71], comorbid diseases [72], or healthy lifestyle [73].