Factors affecting access to health services by older adults in an urban community in Thailand: a cross-sectional study

Background: The impending rapid change in Thailand’s older population has many important implications for health policy, especially older adults’ health problems, which are major cause of them accessing health services. This study aimed to study factors affecting access to health services for older adults in urban communities in Thailand, as well as performing a situational survey of health service utilization. Methods: This cross-sectional study included 886 older adults from four different neighborhoods (slum, city, suburban, and community building). Data were collected using an interview questionnaire. Information about health service variables were extracted and followed the five dimensions of accessibility by Penchansky & Thomas: availability, accessibility, accommodation, affordability and acceptability. Data were analyzed using percentage mean and standard deviation (SD). Analysis of factors affecting access to health service was performed using Logistic regression. Results: The utilization of health services was high among five dimensions. There is no significant difference among other dimensions statistically from all neighborhoods. Only the acceptability dimension with notable difference (p=0.003), city neighborhood has the lowest acceptability dimension (mean 4.0, SD.=0.62), while community building has the highest one (mean 4.3, SD.=0.55). Factors affecting older adults access to health services were using health insurance rights for health care service, and concern about the necessity of health care. Conclusions: Encouraging older adults to change their health insurance rights to the nearest hospital and promoting the provision of holistic health information, which will support older adults in accessing more health services to improve their health outcomes.


Introduction
An older adult is defined by the United Nations as a person who is over 60 years of age. In 2017, the world had a population of 7.550 million people. Approximately 12.7% were older adults, meaning that our world has become an 'aging society'. In 2017, older adults made up 17.1% of Thailand's population, 1,2 and this expected to increase to 25.2% by 2030 3 and 30% by 2035. 2 This changing population has a direct impact on public health policy and countries due older adults having different physical, mental, emotional, and social health issues to other demographics. The most common issues are related to economic, social and health issues, especially chronic health issues that require dependency from other people.
Due to chronic illnesses of older adults, health policies need to ensure that health services cover patient treatment and prevention of health issues, as well as providing a well-prepared health service system. 4 The World Health Organization 5 has suggested three concepts (factors of health, participation and security factors) to ensure that people will be covered by health services regardless of income, ethnicity gender and age. Their access to health services depends on financial factors, sufficient service and the organization of the health service. In 2002, Thailand's government has implemented medical insurance rights 6 with the goal of giving people access to health care services. However, medical insurance rights do not ensure that Thai people will have access to health services completely due to the fact that their health insurance status unmatched with their service point. Even though Thailand has universal coverage policy, most individual can only access to health services at a certain medical facility where they are registered. For those who live far away from their registered medical facility have to commute to the service point, and such situation must be challenging specially for elderly. Therefore, other related factors must be considered, such as socioeconomic factors, distance from habitation to service facilities, travel expenses, and presence of reliable caretakers. 7 In 2013, Thailand's urban population was 53.6% of the total population, with an increasing trend. Previous research has shown that people in the urban areas have limited access to health services. 8 Good health service system management is significant to older adults in allowing them to access the service thoroughly. Therefore, the Ministry of Public Health's primary objective is to ensure that people have access to health services when they need them. Previous studies in Thailand have shown that older adults have access to health services using health insurance status through the Government (44%), the Universal Coverage Scheme (24%), or the Social Security Scheme (3%). It was also reported that use of health services was different dependent on the health insurance rights used and geographical region. 9 In this study, an individual's access to and use of health services is considered by three characteristics as defined by Aday & Anderson 10 : a) Predisposing Factors; b) Enabling Factors; and c) Need Factors. These three factors are independent factors, affecting elderly access to health services, which comprising of five dimensions: availability, accessibility, accommodation, affordability, and acceptability. 11 Due to the fact that there are different types of community in Bangkok; which are slum, city, suburban, housing community, community building, and housing estate, this study only focused on slum, city, suburban, and community building. While in demographic perspectives, population in Bangkok consists largely of migrant workers and latent population as well as those who were born and rise in the capital, and these different backgrounds may affect access to health services. This study aimed to study the factors affecting access to health services for older adults in urban communities, as well as to perform a situation survey of health service utilization.

Study design
This cross-sectional study was conducted among older adults from four different neighborhoods (slum, city, suburban, and community building). A purposive sampling method was used to select communities. A stratified random sampling technique was used to ensure proportional numbers of sample from slum, city, suburban, and community building urban types. Given a type I error of 0.05 and 80% power, the sample size was estimated to be 1,086.

REVISED Amendments from Version 1
For the new version, the title remains the same, the abstract was changed to show the notable difference of acceptability dimension from different neighborhoods. In the text, I followed the comment from reviewers. I used neighborhoods to represent the subgroups and used the term community only if necessary, based on the suggestions of reviewer 1. I explained clearly framework and gap of analysis. Table 2 was modified for more informative from mean score of access to health services to access to health services from different neighborhoods and deleted Table 4. Data analysis I still using logistic regression that I explained this study combines the two levels together of low and moderate level access to healthcare service as a group in order to analyze overall data. In discussion, additional discussed about four different types of neighborhoods which make no significant difference on accessing to health services. Finally, I added the limitation of this study.
Any further responses from the reviewers can be found at the end of the article Participants and setting Participants in this study were older adults who responded to the survey from September to December 2018. The inclusion criteria were a) aged >60 years; b) agree to participate in research. Those unwilling to participate or did not provide consent were excluded. Convenience sampling was used to select the participants. The setting was in Bangkok, Thailand. The numbers of participants from neighborhoods of slum, city, suburban, and community building who lived in neighborhoods were 430, 432, 108, and 116 respectively. In order to contact the participants, local community leaders and community health volunteers publicly announced the project and recruited participants in the community. Data collection took place in the community's common space.

Questionnaire
Data were collected using an interview questionnairequestions were read to the participants by the researcher, and the participants responded verbally. The questionnaire was based on the five dimension of accessibility developed by Penchansky & Thomas 11 (availability; accessibility; accommodation; affordability; and acceptability) and the three factors affecting access to health services, defined by Aday & Anderson 10 ; Predisposing Factors, Enabling Factors, and Need Factors. Predisposing Factors mean the socio-cultural characteristics of individuals that exist prior to their illness, such as age, gender, religion, education. Enabling Factors mean the logistical aspects of obtaining care, such as monthly income, health insurance status, transport to hospital, distance and time to hospital. Need Factors mean the most immediate cause of health service use, from functional and health problems that generate the need for health care services, such as self-perceived health.
The questionnaire was composed of three sections. Section 1 surveys general demographic information of the participants. Section 2 'Access to health service' consisted of 11 items including items relating to Enabling Factors and Needs Factors. Section 3 'Opinion on healthcare service utilization' contained 16 items: availability (4 items), accessibility (2 items), accommodation (4 items), affordability (3 items), and acceptability (3 items). A five-point Likert scale (1=strongly disagree, 2=disagree, 3=moderately, 4=agree, and 5= strongly agree) assess the 16 items. Interpretation of opinion on healthcare service utilization score in each item was as follows: 1.00 -2.33 indicating low level access to healthcare service, 2.34 -3.67 indicating moderate level access to healthcare service, and 3.68 -5.00 indicating high level access to healthcare service. With minimal amount of both low and moderate levels, this study combines the two levels together as a group in order to analyze overall data via Logistic regression.
A pilot test was performed with 30 older adults who had the same characteristics as the participants, which were aged>60 years and willing to answer. The reliability of the questionnaire was tested using Cronbach's alpha, with results of 0.9. As a result of the pilot study, the questionnaires had no change in the wording.
A copy of the questionnaire both in Thai and English can be found in the Extended data.

Data analysis
Data were analyzed by using SPSS Statistics for Windows, Version 28.0 (IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY: IBM Corp). Descriptive data were using percentage mean and standard deviation (SD). Analysis of factors affecting access to health service was performed using Logistic regression.
This study used the STROBE cross-sectional checklist when writing the report. 12

Ethical considerations
This study was one of 12 projects in a set of ageing research projects approved by the Institutional Review Board of the Faculty of Medicine Vajira Hospital, Navamindradhiraj University (IRB No.067/60). Information about the study was explained to the participants (objectives, data collection steps, timing, and benefits of the study). After explanation, written informed consent was obtained from all participants. Upon completing the interview, each participant received 300-baht modest amount of financial compensation for their participation.

Results
A total of 886 older adults were included in this study. Of these 70.8% were women, the average age was between 60 and 69.9 years, and most (42.3%) participants were married and lived together. In total, 58% of participants had finished primary education and most (63.8%) were currently unemployed. Almost half of participants (45.5%) had an income less than 5,000 baht.; 58% of participants stated that their income was sufficient to cover monthly bills. Most participants came from the urban neighborhoods.
In addition, 56.5% of participants reported accessing health services in the past year for an illness or health problems that required medical treatment and most of them reported that their condition was minor. More than half of respondents reported accessing health services was necessary to them with awareness that their health conditions were minimal. Most (>60%) of participants went to health facilities by themselves, and 83.6% used health services with direct health insurance. Public health facilities were the preferred health services visited.
In terms of expenses, participants reported that transportation fares were mostly below 100 baht, while 79.5% had no fees relating to medical expenses and miscellaneous fees. Most of the transport services to public health facilities were taxis, motorcycle-taxis, tricycles and vans. A total of 45% of participants visited hospitals two or three times per month. In terms of distance, most participants reported less than 5 kilometers from home to public health services, with a travel time of less than 30 minutes most reported. Most of the medical insurances were Universal Coverage Scheme. Participant demographics are shown in Table 1. Access to health services Participants from all neighborhoods reported that access to health services was high among all five dimensions. Apart from the acceptability dimension with notable difference (p=0.003), there is no significant difference among other dimensions, statistically. City neighborhood has the lowest acceptability dimension (mean 4.0, SD.=0.62), while community building has the highest one (mean 4.3, SD.=0.55) as the findings are shown in Table 2. Factor affecting access to health service In analyzing data on various factors, it was found that there is only data for Enabling Factors and Needs Factors that affect access to health services (Table 3). In terms of Enabling Factors, using health insurance rights for health care service has a great impact on access to health services; those who use the rights directly at health facilities have more than 1.9 times accessibility compared with those that use rights indirectly (p-value=0.006, 95%CI=1.214-3.094). For the Needs Factors, people who are concerned and aware about their health condition tend to access health services more than those who are not (about 1.5 times odds ratio) (p-value=0.043, 95%CI=1.013-2.340).

Discussion
This study is a part of the Health Promotion Project under the Principle for the Elderly in the Urban Area of Bangkok Metropolitan, which is influenced by a well-known national practice guideline, the Philosophy of Sufficiency Economy, initiated by King Rama IX-Bhumibol Adulyadej. Previous studies have shown that the utilization of health services by older adults in urban areas does adhere to the five dimensions as described by Penchansky & Thomas. 11 Firstly, based on previous data there is a high level of access to health service units at all levels from public to private sector, which are located across the country, and especially in the Bangkok Metropolis. In the Thailand's capital city, there are a total of 4,687 health service units, and only 141 units are able to admit patients in the facility overnight. Approximately 56% of the units are in the public sector and 44% are in the private sector. 13 This result is consistent with previous studies 14,15 which explored the factors that affect access to health services of older adults in Bangkok. However, an earlier study revealed that availability of health service at national level is relatively insufficient. 16 Secondly, for the accessibility dimension, high level access has been shown by a previous study, 14 but this result is inconsistent with some studies. 15,16 Obstacles in accessing health services have been reported as difficulties to travel to health facilities. 17 However, in our study, most participants (60.4%) stated that the distance from home to health facilities was less than 5 kilometers, and it takes 76.2% of participants less than 30 minutes to travel to health facilities. This short distance enables participants to access health services easily, which are mostly public hospitals. Therefore, the participants are able to access and utilize the service conveniently.
Thirdly, for the accommodation dimension, data from our study indicates a high level of service access. A previous study 17 revealed that the main problem in this dimension is the long duration between making an appointment and seeing general doctors or specialists. A study by Rahman et al. 18 found that the long waiting time for doctors affected the satisfaction of patients. In addition, it was found that personal factors of patients, such as experience of previous medical care or expectation, affected satisfaction as well. 19 Participants in our study had high satisfaction with appointment of doctors, as they easily made appointments. Therefore, the level of accommodation for our study is at a high level access. Fourthly, for the affordability dimension, our study revealed that there is a high level of access to services. The ability to pay a medical expense is one problem in accessing health services. 20 The constitution of the Kingdom of Thailand B.E. 2550 21 states that "Everyone has equal rights in utilization of the public health service without medical expenses and able to received public health services thoroughly and efficiently". Accordingly, the established health insurance rights for Thai people gives everyone the opportunity to access health services equally as prices are standardized by the state. 22 Our study indicated that participants were satisfied with the rights and expenses at a high level because most of them used Universal Coverage Scheme rights, which has no payment. Therefore, the ability to access the service in the affordability dimension is at a high level. Fifthly, for the acceptability dimension, our study showed a high level service. A previous study 23 found that for this dimension service characteristics, such as attitude and expectations of service recipients, can hinder the acceptance of service quality. The concept of satisfaction of service quality is significant as patients as customers should be satisfied by the service and facility, particularly when they frequently use the health services.
In comparison, the study showed that these four different types of neighborhoods individually have mean score of access to health services slightly above four, which make no significant difference on accessing to health services. Only acceptability dimension was found different because of service mind of service providers and satisfaction of patients. Elderly from city neighborhood generally have the least acceptable services due to having no option, except public hospitals. While elderly from community building neighborhood have the most acceptability rate because they are able to access health services other than public hospitals such as private hospitals. With limited number of medical staffs in public hospitals compared to number of patients, resulting in hospital administration and human resource management, as well as satisfaction of service receivers.
This study found two factors affecting health service accessibility: a) visiting health service unit they registered at, and b) the need for the service. The former factor is aligned with Enabling Factor, the latter is aligned with Need Factor. For health insurance right factor, this study statistically found that participants that visited a health service unit that they were registered at were 1.9 times more likely to access to health service, compared to those who are not registered at the unit (95%Cl=1.214-3.094). In total 83.7% of public hospital patients visited hospitals they were registered at, which is fourtimes more than those who visited health service units they are not registered. This was consistent with a previous study 15 that found that the relationship between accessibility to health services and utilization of health services was statistically significant (p<0.05). For this reason, if patients use their rights directly the health care service is free of charge for them, or only causes minimal payment. Besides, one of factors why people access public hospitals is no medical expenses. 24 It's been shown that 61.6% of older adults used Universal Coverage Scheme rights, which effects accessibility to health services. 25 In addition, use of treatment services increase with the use of health insurance rights, especially in outpatients. 26,27 The lack of ability to pay for medical expenses has been shown to be an issue with utilization of health services of older adults. 20 For the second factor, necessity of health care, our study revealed that participants who perceived that they needed to be treated in a hospital increases access to health facilities by 1.5 times compared to those that consider it unnecessary to be treated in a hospital (95%Cl=1.013-2.340). The necessity was making people be aware of the importance in medical treatment. However, it has been observed that using health insurance rights with no payment leads to a more accessible health services. These health services are mostly free of charge, especially for those who registered their health insurance rights at the hospital. Even though patients do not have to pay for the health services, free services are not what they actually want; usually they visit health service units due to their illness and awareness on health conditions and not just because these are free. This finding is supported by previous research 24 that studied the needs for access to health services from the perspective of patients and doctors. It indicated that most patients came to access services at the outpatient department; 71.8% of them thought that it was necessary to access to health services and only 0.6% thought it was unnecessary. Nonetheless the viewpoint on treatment between patients and doctors could be different; perspective is important because different perspectives lead to different decisions about necessity in treatment. A previous study 28 revealed the patients assumed that they knew about their health and were able to decide whether they needed treatment.
Lack of knowledge of health may lead them to not taking care of themself or lead them going to hospital. While doctors are likely to suggest them to initially get diagnosed by general practitioners at health service units.
Findings from our study suggested that the government should give the opportunity to get health insurance rights, so individuals can access health facilities nearest their habitation. Similarly the government should facilitate the ability to move health insurance rights to the health facility nearest their habitation. Due to the fact that receiving health services from health service units where patients do not have their health insurance causes extra payment for them. Instead, rights could be based on the use of identification card, as there are some older adults who do not have real address in urban communities as they may be renting a house etc. Medical facilities near the habitation will make it easier way to visit, with no payment, meaning that older adults could have access to more services. Both the public and private sectors should help raise awareness of basic health care, including health literacy to increase awareness of necessity for treatment. Indeed, a previous study indicated there is a correlation between health literacy and the quality of healthcare interactions. 29 Enhancing an understanding of the concept of health literacy will eventually increase patient participation in health care standard and quality in the country holistically. 30 There are a few limitations in our study, primarily the number of participants did not reach the set target due to accessibility to participants in urban communities. Also, participants were able to be contacted only via community leaders or public health volunteers for data collecting. A target number of 1,086 participants was set initially. However, this study successfully reached about 82% of the set target, 886 participants responded in the study. This study conducted on a convenience sample will have limited generalizability.

Conclusion
Among older adults in urban Bangkok, factors affecting access to health services were using health insurance rights for health care service, and concern about the necessity of health care. Encouraging older adults to change their health insurance rights to their nearest hospital and promoting the provision of public health information, will support older adults in accessing more health services to improve their health outcomes.

Data availability
Underlying data Underlying data cannot be shared as the ethical committee that approved the study stated that only aggregated data could be shared openly. In addition, this study is part of a set of research projects about ageing, which all use the same underlying dataset; sharing the dataset must be permitted by all researchers of the ageing projects.
Researchers interested in accessing the data will need to submit an official letter of request for the data to Navamindradhiraj University and will be asked to confirm that they will not violate the ethical standards of the ethical committee and protect the anonymity of the participants. Researchers can contact the corresponding author, who can facilitate this process.

Chaisiri Angkurawaranon
Department of Family Medicine, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand Thank you for the opportunity to review this paper. There are some comments and suggestions.
In the Introduction, I am a bit surprised at the low access to health services given that Thailand has universal coverage -can the authors expand just a bit more as to why it is such a low number? 1.
In the Introduction, the authors mention the five dimensions of access to health services, then three individual characteristics (by Aday & Anderson). So I'm not really sure which framework the authors will use or if they are using both, how they can connect and intersect.

2.
I think the authors should also clearly state the gap in the literature that they are trying to address. Are the authors trying to compare different communities? 3.
As the main objective (gap) is not clearly stated, the data analysis is also a bit underdeveloped. As the authors are presenting their access to care score as categorical data (low, moderate or high), it may not be appropriate to use only mean and sd. What is the binary outcome for the logistic regression? What were the potential confounders?

4.
Would Table 1 and Table 2 be more informative if the authors could also add columns detailing each type of community?

5.
For Table 3, the outcome in the logistic model has not been clearly defined. What did the models adjust for?
6. The section of the Discussion should be a summary of your main finding to answer the question. As the question was not clearly defined in the Introduction (please see comment 3), this section may need to be revised accordingly to discuss the key findings (based on the 8. objectives identified) within the context of the literature, including a more indepth discussion about access to health services within Thailand (for example, why the estimate differed from previous studies mentioned in the Introduction). A discussion or limitation about the generalisability of the results may also be required.
From reading the Discussion, it seems like the authors want to estimate access to care for Bangkok but the authors may want to consider the variation in access to care based on the types of community.

9.
In the Conclusion, the authors focus their conclusion on changing the patient's insurance rights to their nearest hospital but from my reading (based on the results and tables), most of the patients are already using their insurance right at a nearby facility so I'm not sure about this conclusion.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Partly
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Primary care and epidmiology I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Areeya Jirathananuwat
Dear Prof Chaisiri Angkurawaranon, First of all, thank you for your comments and suggestions that allowed me to greatly improve the quality of the manuscript. I agree with all your comments, and I corrected it in the new version. Additionally from my analysis of the data again, I did not find the association of types of community and access to health service. In the conclusion, the results also showed as version 1.
Thank you once again for your valuable time Best regards, Areeya percentage and provide an appropriate term. I'm not sure if this should be a 'response rate' or a 'successfully reached rate' or other terms. Table 1 -"Sufficiency of income": Provide rationale and/or reference for this measure.
Access to health services should be comparatively presented across the five domains (Table 2). For example, why was the Accommodation relatively lower than the other dimensions?
Also, it would be more interesting to compare the findings across the four neighborhoods.
The terms "medical rights" or "health insurance rights" might be confusing; consider replacing them with 'health insurance status' or more relevant terms.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Partly Are the conclusions drawn adequately supported by the results? Partly