2025 年 3 巻 2 号 p. 42-57
Non-communicable diseases (NCDs) represent a significant global health challenge, resulting in 41 million deaths annually, 86% of which are in low- and middle-income countries. In Indonesia, the prevalence of NCDs has increased in recent years. Health-seeking behavior (HSB) varies among individuals and is influenced by personal perspectives on wellness and illness. The factors influencing HSB among NCDs patients in urban and rural communities in Indonesia, remain unknown. This study examined factors influencing HSB among patients with NCDs in both urban and rural communities in Java Island, Indonesia. A random sample of a total of 332 patients with NCDs was selected and a structured questionnaire was utilized for data collection by direct interviews. A quantitative method design with cross-sectional multivariate logistic regression analysis was conducted. The behavioral economics-judgment (p = 0.01, odds ratio [OR] = 2.484); evaluated need (p = 0.016, OR = 1.800); and labor status-informal sector (p = 0.031, OR = 0.489) were significantly associated with HSBs in patients with NCDs. These results emphasize the need for improving the healthcare system in Indonesia by focusing on education, diagnosis, referrals, and patient-required information. Additionally, the findings highlight the necessity for implementing specific methods for health education to engage with informal workers for improving healthcare system. Furthermore, these findings may be insightful for other countries facing similar healthcare challenges.
Non-communicable diseases (NCDs) are chronic conditions that result from a complex interaction between genetic, physiological, environmental, and behavioral factors. NCDs represent a significant global health challenge, accounting for 86% of cases occur in low- and middle-income countries1). Indonesia is a developing country with the fourth largest population in the world2). It continues to face a multitude of health challenges, and both infectious diseases and NCDs represent a significant health burden. The Indonesian government has identified NCDs as a priority area due to their high prevalence and increasing incidences. According to data from the World Health Organization, 78% of deaths from NCDs in Indonesia are projected to occur in 20243). The prevalence of NCDs has increased in recent years; these diseases include cancer (from 1.4% to 1.8%), stroke (from 7% to 10.9%), chronic kidney disease (from 2% to 3.8%), diabetes mellitus (from 6.9% to 8.5%), and hypertension (from 25.8% to 34.1%)4). The failure to effectively manage NCDs can potentially result in considerable economic loss for the country.
Health-seeking behavior (HSB) is closely related to the health status of a country with economic development. HSB is defined as action taken based on a personal decision to promote wellness or recover from an illness or disease5). Experts in the fields of health interventions and policy are recognizing the importance of human behavioral factors in providing quality healthcare services6). In health behavior science, HSB can be an important variable to consider when developing a personalized treatment plan, as well as when considering prognosis7). Prior research has shown that multiple factors influence HSBs. Individuals’ sex has been consistently correlated with HSB8–11). Education level and labor status also play significant roles in shaping HSBs8–10). Additionally, age11–13) the patient’s distance from healthcare facilities13,14) and place of residence11,15) impact HSBs. Insurance status is another key factor influencing HSB9,16), along with socioeconomics, service quality, family support, income status, and knowledge of healthcare9,13,17,18). An investigation of HSB is essential to gain a deeper understanding of the factors that influence individual to seek healthcare facilities19). The previous research explains that urban communities have more differences in HSBs than rural communities in Malaysia20). In addition, rural communities perceive access as a significant barrier to healthcare services. In contrast, it is not a significant barrier in urban communities, with the health financing factor posing a greater challenge21). Therefore, conducting HSB research in both communities is essential to ensure representativeness, as differences between the two regions are evident.
The present study is grounded in Andersen’s Behavioral Model of Health Services Use, Green’s PRECEDE-PROCEED Model, and the Behavioral Economics framework22–24). According to Andersen’s model, the determinants of health service utilization are classified into three categories: predisposing factors (age, sex, residence, marital status, education status, labor status, knowledge, and attitudes), enabling factors (access to health facilities), and need factors (perceived need and evaluated need)22). Green’s model, in addition, incorporates behavioral and reinforcing factors that influence individual decision-making23,24). This theoretical framework is instrumental in the integration of family and peer support as reinforcing factors in the present study. Furthermore, behavioral economics constructs, such as judgment, decision-making, and choice architecture, are integrated to explain the potentially irrational or non-linear reasoning that individuals may use when deciding to use health care services25). Taken together, the behavioral economics is important for the individual in influencing decisions related to health issues26,27). The integration of these diverse models provides a comprehensive framework for understanding the interplay between rational and non-rational influences on HSB, particularly in both urban and rural contexts as representation.
Several studies have examined HSB in Indonesia that related to pregnant women, children, adolescents, and individuals with communicable diseases (e.g., tuberculosis, malaria, acute respiratory infection). It has focused on social, cultural, and stigma-related factors and access to healthcare, including the impact of universal health coverage in Indonesia28–37). However, research related to HSB among patients with NCDs and factors related to behavioral economic in Indonesia, remains to be explored. This study aimed to identify associated factors influencing HSB among patients with NCDs in Indonesia. This study was conducted on Java Island, the most populated region in Indonesia, containing 56.7% of the Indonesian population38). This study was conducted in South Jakarta and Bogor Regency, representing urban and rural areas, respectively. It provided an understanding of the rationale behind the HSBs of individuals residing in both communities. This study provides insights useful for improving the quality of healthcare among patients with NCDs and may serve as a model for other diseases worldwide.
Due to this study conducted in Indonesia, there are terms in the health care system in Indonesia that may not commonly familiar. To clarify and better understand the paper, the following is the definition of specific terms in this study:
Puskesmas is a public primary health care facility that organizes and coordinates promotive, preventive, curative, rehabilitative, and/or palliative health services in its working area39). Posbindu is a platform for the early detection, monitoring, and intervention of NCD risk factors, conducted continuously through Puskesmas supervision40). Posyandu is a community-based service that supports local governments in planning and delivering public services at the village or subdistrict level41).
Study designThis study employed a quantitative research utilized a cross-sectional design. Cross-sectional is an observational method that analyzes data from a population at a single timepoint that is often used to measure the prevalence of health outcomes, understand determinants of health, and describe features of a population42). This research was approved by the institutional review board at the Syarif Hidayatullah State Islamic University of Jakarta following the Ministry of Health, Republic of Indonesia (approval number: Un.01/F.10/KP.01.1/KE. SP/03.08.027/2024). The study protocol conformed to the ethical guidelines of the 2016 declaration of Helsinki or Council for International Organizations of Medical Sciences (CIOMS). In this study, the informed consent was obtained from all participants.
Definition of HSB and conceptual frameworkThe term ‘HSBs’ among patients with NCDs refers to the actions taken by respondents to obtain treatment at healthcare facilities within the six months before the study’s implementation. The following section presents a conceptual framework, accompanied by an explanation of the dependent and independent variables utilized in this study. The respected theories in the behavior field were used, including Anderson and Green theory and Purnell’s study22–25), and adjusted to the objections in this study (Fig. 1).
This study was conducted from March to July of 2024 in Puskesmas Cilandak, South Jakarta (urban communities), and Puskesmas Sentul, Bogor Regency (rural communities). According to the 2018 Indonesia Basic Health Survey, South Jakarta and Bogor Regency have high risk factors for NCDs. The prevalence of smoking is 13.45% in South Jakarta and 27.84% in Bogor, both higher than the national standard. South Jakarta also has a higher prevalence of obesity at 15.08% and the highest consumption of high-salt foods at 29.45%. In Bogor, 51.40% regularly consumed fatty or fried foods. Hypertension rates are also elevated in both areas: 10.92% in South Jakarta and 10.41% in Bogor Regency43,44).
Areas were defined according to the World Bank’s definition of cities, where an urban area was defined by at least 50,000 inhabitants in contiguous dense grid cells (>1,500 inhabitants/km2), with rural areas primarily consisting of low-density grid cells (<300 inhabitants/km2)45). Puskesmas Cilandak is located in South Jakarta has a middle to upper class population, a population density of 11,780 people/km2 (classified as rural area by the World Bank), and is one of the largest primary health care systems in South Jakarta, representing a potential collaborative opportunity to enhance data accuracy. The center oversees five satellite facilities and has seen a notable increase in NCD-related visits (from 5,489 patients in 2022 to 5,939 patients in 2023). Therefore, Puskesmas Cilandak was selected as the sample urban community.
In addition, the poverty rate in Bogor Regency remained at 7.27% indicating ongoing socioeconomic challenges46). Puskesmas Sentul was one the Puskesmas under the Bogor Regency, is classified as a rural area45). Puskesmas Sentul has a population density of 226 people/km2 (classified as rural by World Bank). In addition, the prevalence of NCDs such as hypertension and diabetes remain a significant concern at Puskesmas Sentul. The proportion of hypertension cases increased from 2.85% in 2020 to 4.38% in 2021. In addition, diabetes mellitus is among the ten most commonly treated conditions in 202147). In comparison to other Puskesmas in Bogor Regency, data from Sentul was more readily accessible; therefore, Puskesmas Sentul was selected as the sample rural community for this study.
The population comprised Indonesian patients with NCDs, aged 20–59 years, diagnosed with hypertension, heart disease, asthma, diabetes, or cancer, and residing in Jakarta or Bogor District. Individuals within this age range are more likely to experience NCDs at an early stage and actively seek treatment48). Individuals who are unwilling to participate, those with severe conditions such as strokes, and patients outside the specified age range were excluded. The total population of patients with NCDs who visited Puskesmas Sentul and Puskesmas Cilandak every month was 576 and 580, respectively.
Sample size and samplingThe total sample size for the quantitative survey was 332 patients with NCDs distributed in both urban and rural communities. The required sample size was calculated using the sample size software by Lun & Chiam (NUS, Singapore)49) based on a hypothesis test formula for the difference between two proportions50) with a 95% confidence level and 95% power for each independent variable. The simple random sampling method was used in this study (Fig. 2). The sampling method was random by selecting participants from the attendance list using a random number generator (calculatorsoup.com) to randomly pick the selected participants. The sampling method involved coordinating with healthcare personnel to confirm whether the sample came from a patient with an NCD during community visits as part of the NCD screening program. Screening for the 20–59-year-old age group was typically conducted through general health campaigns such as Posbindu and Posyandu, which involve broader community health efforts. In this study, the informed consent was obtained from all participants. Patients who met the inclusion criteria were asked about their treatment-seeking behavior using pretested and structured questionnaires, with direct interviews conducted. The duration of each interview ranged between 15–20 minutes. Data collection was facilitated by five enumerators who had degrees in public health. The enumerators were provided with an explanation regarding technical details and interview procedures.
Quantitative data were analyzed using SPSS Version 29 (IBM, Armonk, New York, USA)51) with univariate and bivariate analysis performed using the Chi-square test and multivariate analysis performed using multivariable logistic regression analysis. A p value of <0.05 was considered significant. Before data collection, validity and reliability tests were conducted on the questionnaire to ascertain its capacity to consistently and accurately measure the intended variables. However, before proceeding with multivariable analysis, confirmatory factor analysis (CFA) was conducted to assess the construct validity and internal consistency of variables related to attitudes and behavioral economics (judgment, decision making, choice architecture). The variables with strong loading factors (score ≥0.5) and acceptable model were selected for further multivariate analysis. These variables were then transformed into composite variables, which were subsequently employed as independent variables in the final model. The CFA is based on established theories and concepts; it allows for the determination of the number of factors to be created and the variables to be included in each factor. The objective was guiding the identification of the desired factors and the grouping of variables52).
Table 1 presents the distribution frequency of variables (HSB; predisposing, enabling, and reinforcing, needs; and behavioral economics). The majority sample of this study were female (89.8%), with 84.4% of those being married and residing in urban communities and 84.8% residing in rural communities. The distribution of age groups was similar in both rural and urban communities, with a particularly high proportion of individuals aged 40–59 years (73.2%). The majority of respondents were unemployed (78.9%) and received income from their spouses. In general, both communities were classified as having low-incomes (52.1% of urban participants and 77.6% of rural participants). Participants in urban communities exhibited a higher proportion of middle- and high-incomes (47.9%). Over half of respondents (64.7%) in urban communities completed a tertiary education; and a larger proportion of respondents in rural communities only completed basic education (68.5%), with 19.4% having received no formal education.
| Variable | Frequency, urban & rural; n (%/332) | Frequency, urban; n (%/167) | Frequency, rural; n (%/165) |
|---|---|---|---|
| Health seeking behavior of patients with NCDs* | |||
| No | 110 (33.1) | 57 (34.1) | 53 (32.1) |
| Yes | 222 (66.9) | 110 (65.9) | 112 (67.9) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Location health seeking behavior | |||
| Puskesmas | 93 (28.0) | 31 (18.6) | 62 (37.6) |
| Hospitals | 13 (3.9) | 3 (1.8) | 10 (6.1) |
| Puskesmas and hospitals | 116 (34.9) | 76 (45.5) | 40 (24.2) |
| Other | 110 (33.1) | 57 (34.1) | 53 (32.1) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Age | |||
| 20–39 years | 89 (26.8) | 39 (23.4) | 50 (30.3) |
| 40–59 years | 243 (73.2) | 128 (76.6) | 115 (69.7) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Sex | |||
| Male | 34 (10.2) | 24 (14.4) | 10 (6.1) |
| Female | 298 (89.8) | 143 (85.6) | 155 (93.9) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Marital status | |||
| Not married | 51 (15.4) | 26 (15.6) | 25 (15.2) |
| Married | 281 (84.6) | 141 (84.4) | 140 (84.8) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Education status | |||
| No schooling | 39 (11.7) | 7 (4.2) | 32 (19.4) |
| Basic (Graduated elementary and middle school) |
165 (49.7) | 52 (31.1) | 113 (68.5) |
| Tertiary (Completed high school, bachelor’s, master’s, doctoral/PhD) |
128 (38.6) | 108 (64.7) | 20 (12.1) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Labor status | |||
| Unemployed | 262 (78.9) | 124 (74.3) | 138 (83.6) |
| Informal sector | 46 (13.9) | 23 (13.8) | 23 (13.9) |
| Formal sector | 24 (7.2) | 20 (12.0) | 4 (2.4) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Insurance status | |||
| No insurance | 66 (19.9) | 9 (5.4) | 57 (34.5) |
| Insurance | 266 (80.1) | 158 (94.6) | 108 (65.5) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Income status | |||
| Low (<Rp.3.000.000/$194.69) |
215 (64.8) | 87 (52.1) | 128 (77.6) |
| Middle (Rp.3000.000/$194.69–Rp.6000.000/$389.38) |
109 (32.8) | 73 (43.7) | 36 (21.8) |
| High (≥Rp.6000.000/$389.38) |
8 (2.4) | 7 (4.2) | 1 (0.6) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Duration sick | |||
| <5 years | 198 (59.6) | 104 (62.3) | 94 (57.0) |
| ≥5 years | 134 (40.4) | 63 (37.7) | 71 (43.0) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Multimorbidity | |||
| More than one NCD | 123 (37.0) | 73 (43.7) | 50 (30.3) |
| Only one NCD | 209 (63.0) | 94 (56.3) | 115 (69.7) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Family history of NCDs | |||
| Yes | 184 (55.4) | 103 (61.7) | 81 (49.1) |
| No | 148 (44.6) | 64 (38.3) | 84 (50.9) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Knowledge of NCDs | |||
| Poor | 153 (46.1) | 56 (33.5) | 97 (58.8) |
| Good | 179 (53.9) | 111 (66.5) | 68 (41.2) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Attitude surrounding NCDs | |||
| Negative | 199 (59.9) | 86 (51.5) | 113 (68.5) |
| Positive | 133 (40.1) | 81 (48.5) | 52 (31.5) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Access to healthcare facilities | |||
| ≥5 km, far | 38 (11.4) | 12 (7.2) | 26 (15.8) |
| <5 km, close | 294 (88.6) | 155 (92.8) | 139 (84.2) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Family support | |||
| Not supported | 69 (20.8) | 43 (25.7) | 26 (15.8) |
| Supported | 263 (79.2) | 124 (74.3) | 139 (84.2) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Peer support | |||
| Not supported | 147 (44.3) | 67 (40.1) | 80 (48.5) |
| Supported | 185 (55.7) | 100 (59.9) | 85 (51.5) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Perceived need | |||
| Not necessary | 151 (45.5) | 72 (43.1) | 79 (47.9) |
| Required | 181 (54.5) | 95 (56.9) | 86 (52.1) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Evaluated need | |||
| Not necessary | 139 (41.9) | 80 (47.9) | 59 (35.8) |
| Required | 193 (58.1) | 87 (52.1) | 106 (64.2) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Behavioral economics (judgement) | |||
| Negative | 271 (81.6) | 131 (78.4) | 140 (84.8) |
| Positive | 61 (18.4) | 36 (21.6) | 25 (15.2) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Behavioral economics (decision making) | |||
| Negative | 184 (55.4) | 86 (51.5) | 98 (59.4) |
| Positive | 148 (44.6) | 81 (48.5) | 67 (40.6) |
| Total | 332 (100) | 167 (100) | 165 (100) |
| Behavioral economics (choice architecture) | |||
| Negative | 281 (84.6) | 135 (80.8) | 146 (88.5) |
| Positive | 51 (15.4) | 32 (19.2) | 19 (11.5) |
| Total | 332 (100) | 167 (100) | 165 (100) |
*NCDs: Noncommunicable diseases
HSB among patients with NCDs in urban and rural communities was assessed by interviewing patients on specific variables of HSB (Table 1). In total, 66.9% of respondents indicated that they had visited healthcare facilities within the past six months, while 33.1% indicated that they had not. The most frequently visited health service facilities were the puskesmas and hospitals (34.9%); 33.1% of respondents visited other health service facilities, including private healthcare, indicating respondents visited multiple types of health service facilities.
Differences were observed in the HSB activities of urban and rural communities. The respondents in urban communities indicated a high healthcare preference for both puskesmas and hospitals (45.5%), whereas those in rural communities preferred puskesmas alone (37.6%). Furthermore, a total of 66.5% of respondents in urban communities demonstrated a good knowledge on NCDs, whereas in rural communities, 58.8% exhibited poor knowledge on NCDs. An analysis of the data revealed that 61.7% of respondents in urban communities reported a family history of NCDs. Conversely, 50.9% of respondents in rural communities did not report this information. However, it is possible that some families in rural communities may have undiagnosed NCDs without being aware of them.
Factors associated with the HSB of patients with NCDsTo investigate factors associated with the HSB of patients with NCDs, bivariate and multivariate analyses were conducted. In the bivariate Chi-square analysis (Table 2), three variables were identified as having a statistically significant correlation with healthcare service utilization among patients with NCDs (p < 0.05): labor status (informal sector), evaluated need (the interpretation provided by doctors to respondents regarding their health condition), and behavioral economics judgment (respondents’ assessment on the availability of health services by health workers).
| Variable | Health-seeking behavior of patients with NCDs | Total n (%) 332 | p | COR* (95% CI) | |
|---|---|---|---|---|---|
| No n (%) | Yes n (%) | ||||
| Age | |||||
| 20–39 years | 27 (30.3) | 62 (69.7) | 89 (100) | 0.599 | 0.839 (0.497–1.418) |
| 40–59 years | 83 (34.2) | 160 (65.8) | 243 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Sex | |||||
| Male | 11 (32.4) | 23 (67.6) | 34 (100) | 0.919 | 0.961 (0.451–2.051) |
| Female | 99 (33.2) | 199 (66.8) | 298 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Residence | |||||
| Rural | 53 (32.1) | 112 (67.9) | 165 (100) | 0.727 | 0.913 (0.578–1.443) |
| Urban | 57 (34.1) | 110 (65.9) | 167 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Marital status | |||||
| Not married | 16 (31.4) | 35 (68.6) | 51 (100) | 0.872 | 0.909 (0.479–1.727) |
| Married | 94 (33.5) | 187 (66.5) | 281 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Education status | |||||
| No schooling | 15 (38.5) | 24 (61.5) | 39 (100) | 0.202 | |
| Basic | 47 (28.5) | 118 (71.5) | 165 (100) | 0.225 | 1.569 (0.757–3.251) |
| Tertiary | 48 (37.5) | 80 (62.5) | 128 (100) | 0.914 | 1.042 (0.498–2.178 |
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Labor status | |||||
| Unemployed | 83 (31.7) | 179 (68.3) | 262 (100) | 0.046 | |
| Informal sector | 22 (47.8) | 24 (52.2) | 46 (100) | 0.035* | 0.506 (0.268–0.954) |
| Formal sector | 5 (20.8) | 19 (79.2) | 24 (100) | 0.76 | 1.762 (0.636–4.881) |
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Insurance status | |||||
| No insurance | 18 (27.3) | 48 (72.7) | 66 (100) | 0.307 | 0.709 (0.390–1.289) |
| Have national insurance | 92 (34.6) | 174 (65.4) | 266 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Income status | |||||
| Low (<Rp.3.000.000/$194.69) |
67 (31.2) | 148 (68.8) | 215 (100) | 0.586 | |
| Middle (Rp.3000.000/$194.69 –Rp.6000.000/$389.38) |
40 (36.7) | 69 (63.3) | 109 (100) | 0.317 | 0.781 (0.481–1.268) |
| High (≥Rp.6000.000/$389.38) |
3 (37.5) | 5 (62.5) | 8 (100) | 0.705 | 0.755 (0.175–3.249) |
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Duration of sickness | |||||
| <5 years | 59 (29.8) | 139 (70.2) | 198 (100) | 0.124 | 0.691 (0.435–1.098) |
| ≥5 years | 51 (38.1) | 83 (61.9) | 134 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Multimorbidity | |||||
| More than one NCD | 42 (34.1) | 81 (65.9) | 123 (100) | 0.81 | 1.075 (0.671–1.723) |
| Only one NCD | 68 (32.5) | 141 (67.5) | 209 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Family history of NCDs | |||||
| Yes | 60 (32.6) | 124 (67.4) | 184 (100) | 0.907 | 0.948 (0.599–1.501) |
| No | 50 (33.8) | 98 (66.2) | 148 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Knowledge of NCDs | |||||
| Poor | 56 (36.6) | 97 (63.4) | 153 (100) | 0.243 | 1.336 (0.845–2.113) |
| Good | 54 (30.2) | 125 (69.8) | 179 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Attitudes | |||||
| Negative | 72 (36.2) | 127 (63.8) | 199 (100) | 0.155 | 1.417 (0.882–2.278) |
| Positive | 38 (28.6) | 95 (71.4) | 133 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Access to healthcare facilities | |||||
| Far (>5 km) | 12 (31.6) | 26 (68.4) | 38 (100) | 1 | 0.923 (0.447–1.907) |
| Close (<5 km) | 98 (33.3) | 196 (66.7) | 294 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Family support | |||||
| Not supported | 20 (29.0) | 49 (71.0) | 69 (100) | 0.473 | 0.785 (0.440–1.400) |
| Supported | 90 (34.2) | 173 (65.8) | 263 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Peer support | |||||
| Not supported | 47 (32.0) | 100 (68.0) | 147 (100) | 0.726 | 0.910 (0.574–1.443) |
| Supported | 63 (34.1) | 122 (65.9) | 185 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Perceived need | |||||
| Not necessary | 52 (34.4) | 99 (65.6) | 151 (100) | 0.726 | 1.114 (0.704–1.762) |
| Required | 58 (32.0) | 123 (68.0) | 181 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Evaluated need | |||||
| Not necessary | 55 (39.6) | 84 (60.4) | 139 (100) | 0.044* | 1.643 (1.035–2.607) |
| Required | 55 (28.5) | 138 (71.5) | 193 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Behavioral economics (judgement) | |||||
| Negative | 98 (36.2) | 173 (63.8) | 271 (100) | 0.016* | 2.313 (1.174–4.557) |
| Positive | 12 (19.7) | 49 (80.3) | 61 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Behavioral economics (decision making) | |||||
| Negative | 57 (31.0) | 127 (69.0) | 184 (100) | 0.412 | 0.804 (0.508–1.273) |
| Positive | 53 (35.8) | 95 (64.2) | 148 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
| Behavioral economics (choice architecture) | |||||
| Negative | 98 (34.9) | 183 (65.1) | 281 (100) | 0.145 | 1.740 (0.871–3.477) |
| Positive | 12 (23.5) | 39 (76.5) | 51 (100) | ||
| Total | 110 (33.1) | 222 (66.9) | 332 (100) | ||
*COR: Crude odd ratio
In the multivariate analysis, all independent variables were included in the model to determine which were most significantly associated with HSBs in patients with NCDs (Table 3). Following the initial adjusted model, 19 additional models were obtained by removing one independent variable at a time (p > 0.05). The variables were removed in the following order: residence, access to healthcare facilities, behavior economics (choice architecture), perceived need, marital status, multimorbidity, family history of NCDs, peer support, insurance status, income status, duration of sickness, sex, knowledge of NCDs, age, family support, education status, attitudes, and behavior economics (decision making).
| Variable | B | p | AOR* (95% CI) |
|---|---|---|---|
| Age | -0.385 | 0.219 | 0.681 (0.369–1.256) |
| Sex | -0.596 | 0.227 | 0.551 (0.210–1.448) |
| Residence | -0.002 | 0.994 | 0.998 (0.515–1.932) |
| Marital status | 0.041 | 0.909 | 1.042 (0.511–2.127) |
| Education status (no schooling) | 0.365 | 1.00 (reference) | |
| Education status 1 (Basic/Graduated elementary and middle school) | 0.29 | 0.485 | 1.337 (0.592–3.020) |
| Education status 2 (Tertiary/Completed high school, bachelor, master, doctor/PhD) | -0.141 | 0.769 | 0.868 (0.338–2.232) |
| Labor status (unemployed) | 0.082 | 1.00 (reference) | |
| Labor status 1 (informal sector) | -0.77 | 0.052 | 0.463 (0.213–1.006) |
| Labor status 2 (formal sector) | 0.374 | 0.534 | 1.454 (0.448–4.721) |
| Insurance status | -0.246 | 0.508 | 0.782 (0.378–1.620) |
| Income status (low <Rp.3000.00,00) | 0.492 | 1.00 (reference) | |
| Income status 1 (middle Rp.3000.000,00–Rp.6000.000,00) | -0.279 | 0.345 | 0.757 (0.424–1.350) |
| Income status 2 (high ≥Rp.6000.000,00) | -0.75 | 0.389 | 0.472 (0.086–2.601) |
| Duration of sickness | -0.254 | 0.331 | 0.776 (0.465–1.295) |
| Multimorbidity | 0.078 | 0.781 | 1.081 (0.626–1.867) |
| Family history of NCDs | -0.082 | 0.753 | 0.922 (0.555–1.531) |
| Knowledge of NCDs | 0.388 | 0.171 | 1.474 (0.845–2.571) |
| Attitudes | 0.482 | 0.115 | 1.619 (0.890–2.946) |
| Access to healthcare facilities | 0.031 | 0.94 | 1.032 (0.457–2.327) |
| Family support | -0.396 | 0.25 | 0.673 (0.343–1.321) |
| Peer support | 0.174 | 0.537 | 1.19 (0.685–2.066) |
| Perceived need | -0.021 | 0.941 | 0.979 (0.564–1.699) |
| Evaluated need | 0.749 | 0.012 | 2.114 (1.175–3.804) |
| Behavioral economics (judgement) | 0.949 | 0.043 | 2.583 (1.029–6.487) |
| Behavioral economics (choice architecture) | -0.427 | 0.146 | 0.653 (0.367–1.161) |
| Behavioral economics (decision making) | 0.039 | 0.937 | 1.04 (0.388–2.786) |
*AOR: Adjusted odds ratio.
The final model was designated as Model 19 (Table 4). The results indicated that factors associated with HSBs were behavioral economics-judgment (p = 0.01, odds ratio [OR] = 2.484); evaluated need (p = 0.016, OR = 1.800); and labor status-informal sector (p = 0.031, OR = 0.489). Respondents employed in the informal sector had a 51.1% lower likelihood of having HSBs than those who were unemployed, indicating unemployed were more likely to engage in HSB. In the formal sector, respondents demonstrated a 1.7-fold increase in HSBs compared to those who were unemployed, although this result was statistically insignificant (p > 0.05). Evaluated need had a significant increase in the odds of association with HSB, evaluations provided by doctors increased HSB participation by 1.8-fold greater (p < 0.05). Behavioral economics (judgment) demonstrated a significant association with HSBs, with an odds ratio of 2.5 (p < 0.05). Overall, the results indicated a significant correlation between labor status (informal sector), evaluated need, and behavioral economics (judgment) with HSB.
| Variable | B | p | AOR* (95% CI) |
|---|---|---|---|
| Labor status (unemployed) | — | 0.045 | 1.00 (reference) |
| Labor status 1 (informal sector) | -0.716 | 0.031* | 0.489 (0.255–0.936) |
| Labor status 2 (formal sector) | 0.534 | 0.313 | 1.706 (0.604–4.818) |
| Evaluated need | 0.588 | 0.016* | 1.800 (1.118–2.899) |
| Behavioral economics (judgment) | 0.910 | 0.010* | 2.484 (1.242–4.968) |
*AOR: Adjusted odds ratio
This study provided evidence of original research investigated the associations between HSB and factors influencing the utilization of healthcare facilities by patients with NCDs in both urban and rural communities in Indonesia. The concept of behavioral economics provided a distinctive opportunity to understand how economic and psychological elements impact HSB. This approach has not been widely explored in previous studies, especially in relation to the HSB developed in this research.
The variables dominant in influencing HSB were behavioral economics (judgment), evaluated needs, and labor status (informal sector). Behavioral economics is the study of understanding the reason people engage in making economic decisions, including decisions related to healthcare53). The majority of respondents have a negative judgment of the information provided by healthcare workers, a significant factor influencing their HSB.
This finding can be explained by behavioral economics concepts, specifically the affect heuristic and the availability bias. The affect heuristic suggests that emotional factors, such as fear of illness or pressure, can drive decisions even when quality is low54,55) or driven by necessity behavior56) when the availability of healthcare access is limited. In addition, the availability bias may lead individuals to act based on how easily instances come to mind57,58). In this study, the judgment of respondents could be affected by, for example, any neighbor getting sick or dying due to illness (dramatic event) or high cases of disease vividly reported in news media. These cognitive shortcuts can strongly influence HSB, especially in the absence of reliable health information, which affects individual emotional dimensions59). Judgments are often a combination of several emotional states, indicating that they arise spontaneously, quickly, and without deep consideration60,61). The majority of respondents in this study were female and housewives. Consistent with previous reports, females have more positive attitudes for seeking physiological help62). However, this study was unable to conclude regarding sex-based decisions due to respondents’ characteristics. In general, respondents may assume negative judgments about healthcare workers based on quick and spontaneous assessments. Despite negative judgments and dissatisfaction with healthcare services, most respondents continued to engage in HSB due to the limited options available and sought to access the healthcare, including self-medication treatment. This finding reveals the association between behavioral economics and HSB in influencing the healthcare system in Indonesia.
Evaluated needs variable was associated with HSB in patients with NCDs. Respondents who received an evaluation or recommendation from a doctor to visit a health facility were more willing to do health seeking than patients who had not. These findings are consistent with previous reports indicating that the quality of health services provided by health workers has an impact on HSB in Kenya63,64). Additionally, the HSB of patients is shaped by the information provided by health workers, such as the evaluation of treatment outcomes and the role of health workers as a trusted source of information or consultation65,66). This implies that HSB is affected by health services provided by health workers, including evaluation and recommendation. Consequently, the quality of interaction between health workers and patients can significantly influence the healthcare system.
Labor status was another factor that significantly impacted the HSB of patients with NCDs, though this association was only seen in the informal sector. Respondents employed in the informal sector were less willing to seek healthcare services than those who were unemployed. As most of the respondents who were unemployed in this study were housewives, they may receive support from family or the government (for low income family) and also have more available time than those employed in the informal sector to seek healthcare. Whereas, formal sector workers have support for health insurance from their office67–69). In addition, informal sector workers have to pay for national health insurance themselves, which implies a financial problem. This finding was in line with a previous study, the type of occupation has a significant impact on the HSB of individuals both in rural and urban communities in Nigeria10). However, monthly income was not significantly associated with HSB in this study. This is dissimilar with previous studies that revealed a higher monthly income was associated with greater HSBs in Ethiopia, Georgia, and Mongolia15,70,71). The lack of distinction in income levels criteria among respondents may have contributed to this. In addition, Indonesia’s population has a high number of informal workers, reaching 56%72). Informal sector workers in Indonesia face health problems due to a lack of health insurance, low incomes, and changing needs73). Whereas, other countries like Thailand provide subsidies to informal workers through general taxation, or Vietnam subsidizes 50 percent of their insurance premiums74). Overall, this research provides a basis for important findings that reveal an association between the informal sector and health problem issues.
This study had several limitations common to survey research. First is, the sampling method employed a cross-sectional design, is its inability to establish causal relationships, which may have limited the generalizability of the findings, as it only allows observation of the relationship between variables. Second, the sampling may be biased due to the overrepresentation of women and housewives, as it was done at Posbindu or Posyandu during work hours. To mitigate potential bias, inclusion and exclusion criteria were carefully established, and all sub-districts or villages within the operational scope of Puskesmas Cilandak and Sentul were included in the study. Future research should explore additional outreach strategies to include more varied participants. Moreover, further research is recommended to include urban slum communities that were not included in this study.
This study identified significant associations between HSBs and several factors, including labor status (informal sector), evaluated needs, and behavioral economics (judgement). These findings suggest possible behavioral and structural differences in how patients with NCDs engage with health services. Patients who were employed in the informal sector were less likely to utilize health services. In addition, patients who received a recommendation from a medical professional were more likely to seek treatment. Interestingly, patients with a negative judgment of medical professionals were also found to significantly engage in HSB. Overall, this study highlights the importance of improving the healthcare system in Indonesia by implementing healthcare promotion, screening, and health education targeting the informal sector workers. This study revealed that communication between patients and healthcare workers is a crucial factor in patients to seek healthcare services.
As recommendations based on this study, (i) the health education program can be implemented and expanded through screening activities in Posyandu or Posbindu under the supervision of Puskesmas. (ii) educational outreach to informal worker communities can be strengthened in Indonesia, (iii) the registration process for the National Health Insurance can be simple and straightforward, and (iv) the registration system can be improved to ensure accurate identification of informal workers for potential subsidy allocation and automatic enrollment in the National Health Insurance. Therefore, the implementation of a systemic and continuous training program is recommended to bridge the gap between patients and healthcare workers. Taken together, this study could serve as a model for HSB research in other countries or regarding other diseases.
I would like to express my gratitude to the Department of Health in Jakarta, the Department of Health in Bogor District, Puskesmas Cilandak, and Puskesmas Sentul for their invaluable assistance in facilitating research permits. I also extend thanks to Dr. Kiyoshi Hamashima and Dr. Shinji Teraji for the helpful discussion and support throughout this process. I would like to thank Editage (www.editage.jp) for English language editing. This research was conducted without any external grant or funding support.
Hasanah IJ contributed to all aspects of the study, including conceptualization, data collection, analysis, interpretation of results, drafting the original manuscript, as well as reviewing and editing the final version.
There are no conflicts of interest to disclose.
