Household Need and Unmet Need for Caregivers of Older Persons in Thailand

ABSTRACT

The need for caregivers is a crucial issue in Thailand. This research examined levels and trends of household needs and unmet needs for caregivers of older persons and explored potential factors associated with these needs. The analysis utilized data from the Survey of Older Persons in Thailand 2007, 2011, 2014, and 2017. The household need for a caregiver of older persons was defined as a household with one or more older people who needed a caregiver to help them perform basic activities of daily living. The unmet need for a caregiver referred to households where at least one older person in the household needed care but did not receive it. Findings illustrated the increasing levels and trends of household needs as well as unmet needs over time. In terms of potential determinants, older person households in Bangkok and households with higher socioeconomic status were more likely to be the household need for caregivers. In contrast, those households in the Northeastern, the poorest region, were more likely to be the unmet need household. These findings are indicative of the rising demand for long-term care services in Thailand. However, it is vital to consider unmet household needs, especially in the worse-off area, when designing national policies.

KEYWORDS:

• Caregiver
• older people
• unmet need
• household
• Thailand

Introduction

With an aging population, in which older people become increasingly restricted in their activity and their disabilities become more debilitating over time, the role of caregivers has expanded and accordingly been recognized as an important factor affecting the security and safety of older people (Faes et al., 2010; Mamani et al., 2019). Global aging and the need for caregivers as a phenomenon has to be interpreted within a specific societal context as the changing population ages at different rates and has different social needs globally (Brijnath, 2009; Karlin & Weil, 2017; Muramatsu & Yin, 2017; Bruggencate et al., 2018).

Thailand started to seriously address the changing health-care needs of the population starting with the 8^th National Economic and Social Development Plan (1997–2001). The government introduced a number of policies to address population aging and health care. This included providing subsidies for health-care expenditure (especially inpatient care) in recognition of the fact that there was a severe shortage of hospital beds for older people with chronic health problems and insufficient geriatric specialists. However, long-term institutional care is considered as a last resort by the Thai government (and most families) in terms of the provision of services for older people. As in most Asian countries, caring for older people in Thailand is shaped by the social norm which prescribes that care is the primary responsibility of family members (Jitapunkul & Wivatvanit, 2008; Kim et al., 2015; Knodel et al., 2015, 2018; Slote & De Vos, 1998). Therefore, at present, care for older persons focuses on home-based and community-based care. The 2^nd National Plan for older Persons of Thailand, 2001–2021 was launched in 2002, and one of the key domains of activities of the plan is empowering families and communities to support and provide care to older persons (Jitapunkul & Wivatvanit, 2008).

Critically, the decline of the extended family has been evident in Thailand, due to falling and the migration of adult children (Knodel et al., 2015). As a result, an emerging issue for documentation and analysis is the extent of the unmet need for caregivers of older persons, particularly where older people’s households are at risk of unmet needs. Rising trends of household need and unmet need for caregivers of older people under the demographic and socioeconomic transitions must increase stress or burden to caregivers. The potential for stress and burden from caregiving are factors that can affect the termination of a caregiver’s role or, in the worse cases, elderly abuse (Kasper et al., 1994).

In health care, the term of need for caregivers is linked to a need for caregiver assistance. At the time of this writing, there was no standard measurement of household need for caregivers. A study found that the need for a caregiver of older people depends on the number of activities of daily living (Knodel, 2015). This study defined the household need for a caregiver of older persons as having one or more older people in the household who need a caregiver for five basic ADLs (i.e., eating, dressing, bathing, toileting, brushing teeth, and washing face). The households, where at least one older person in the household who needed care but did not receive it, were defined as having an unmet household need for a caregiver.

The potential determinants of the household need and unmet need for caregivers were older people’s demographic characteristics and their health status (e.g., functional disability or cognitive impairment), caregiver characteristics, living arrangement, healthcare cost, and transportation system (Meemon & Paek, 2019; Nihtilä & Martikainen, 2007). On the societal level, geographical and household, socioeconomic factors have been mainly considered.

Changes in demographic structures have led to changes in economic behavior. Culture, defined as people’s attitudes, behavior, or way of life, varies depending on the area of residence and region. Therefore, geographical differences and cultural bias may be manifest in different patterns of health-seeking behavior, and health-related preferences, as well as perceptions of older people (Biswas et al., 2006). A comparison of household needs for caregivers of older persons between urban and rural areas or among regions should be clearly explored. Regional terms (Bangkok, Central [excluding Bangkok], Northern, Northeastern, and Southern) used by the Thai government reflect the country’s cultural diversity.

The strong link between wealth and health brings concern about inequalities in social support (Braveman et al., 2000). Thus, differences in the household need for caregivers by household socioeconomic status should be considered to ascertain the equity of health programs and other provided services. A previous study reported the prevalence of need for care of older people in a wealthier socioeconomic status household was lower than poorer ones (Melzer et al., 2000). Household wealth represents a permanent economic status the household measured by household’s ownership of various assets, whereas, an income is a theoretical indicator and lesser permanent measure of the socioeconomic status. In Thailand, households with older people face higher levels of annual health-care expenditure compared to households without older people. In this study, the household income refers to the annual income of the household head, and the education is the highest completed level of education of the head of household at the time of the survey.

Given the fact that more than half of the Thai population has no retirement pension, the household socioeconomic characteristics of older people are crucial in predicting their long-term care situation. Findings from the existing literature are still unclear whether household’s socioeconomic factors affect the household need for caregivers of older people (Knodel et al., 2018; Meemon & Paek, 2019). Also, the results of our previous study showed a fluctuating trend of the household need for caregivers (Phetsitong et al., 2019). The present study, therefore, extends the previous research by adding the recent Survey of Older Persons in Thailand in 2017. In addition, to the knowledge of the authors, profiles of unmet needs for caregiving at home for older people in Thailand have been relatively limited in the past, and there have been no studies assessing the levels and trends of the unmet need at the household level. Thus, this study aimed to analyze levels and trends of household need and unmet household need for caregivers of older persons and explore the impact of geographical and household socioeconomic determinants of these needs in Thailand from 2007 to 2017. This national trend analysis would provide a better understanding of the household need for caregivers and add to research on aging and the Long-term Care (LTC) policy for Thai older people on the unmet household need for the caregiver of older people.

Methods

Data source and sample

This study uses data from the nationally representative Survey of Older Persons in Thailand (SOPT) 2007, 2011, 2014, and 2017 conducted by the National Statistical Office (NSO). In total, the numbers of private households in the sample were 79,524 in 2007, 79,560 in 2011, 83,880 in 2014, and 83,880 in 2017. The surveys collected demographic backgrounds and socioeconomic information on all sample households. For eligible households, individuals aged 50 years and above were interviewed for the older person questions (e.g., health status, ability to perform activities of daily living autonomously, health behavior, and preparation for old age). It should be noted that only eligible older people households are available in the SOPT for 2007, while the dataset of all sample households in the SOPT for 2011, 2014, and 2017 were accessible for public use.

Our analysis included only households with older individuals aged 60 and over. The total representative samples in 2007, 2011, 2014 and 2017 were 22,688 households, 25,077 households, 27,990 households, and 25,740 households, respectively.

Household need for a caregiver of older persons

Measuring the need for a caregiver is often assessed by the older person’s difficulties with activities of daily living (ADLs). The literature found that the most basic ADLs that can identify a dependent older person and need for care are eating, dressing, bathing/using the toilet, and grooming difficulties (Katz et al., 1963; Kingston et al., 2012). In this study, the older person in the surveys was asked “Can you perform the daily living activities (eating, dressing, bathing, toileting, brushing your teeth, and washing your face)?”

Met and unmet household need for a caregiver of older persons

Households, where all older persons who needed a caregiver for assistance with basic ADLs and received it, were defined as having a met household need for a caregiver. In contrast, the remaining households, where at least one older person in the household who needed care but did not receive it, were treated as having an unmet household need for a caregiver. To measure whether or not the older person had a caregiver for assistance with ADLs, in the SOPTs for 2007, 2011, 2014 and 2017, older persons were asked whether or not they have a caregiver to provide care for the activities of daily living.

Determinants of household need and unmet household need for a caregiver of older persons

This study assessed geographic and household socioeconomic characteristics including areas of residence, region, relative household wealth, educational attainment of household head, and income of the household head as the potential determinants of household need and unmet household need for a caregiver of older persons.

The present study used relative household wealth, educational attainment, and the income of the household head as proxy indicators of socioeconomic status of the household, and this might be one of the key factors associated with unmet household need for caregivers of older persons. The wealth index is measured based on a set of household characteristics and assets ownership. These attributes have been converted into a wealth index by principal component analysis (McKenzie, 2005) and converted into relative household wealth quintiles. Higher levels of education may cause a greater lack of availability of caregivers for older members, generally (Logue, 1990). In this study, educational attainment of the household head refers to the highest completed level of education of the primary caregiver at the time of the survey. The income of the household head refers to their income during the past year of household head (in Thai baht).

Data analysis

Household sampling weight was applied to make the sample nationally representative. Since the surveys employed a stratified two-stage sampling for data collection, a chi-squared test for analysis of categorical data from a complex survey was used. The binary logistic regression analysis of pooled data was performed to explore the impact of potential determinants of the household need and unmet need for caregivers of older persons. All statistical analysis at the household level was performed using STATA/SE 14.0 (StataCorp, 2015).

This study was approved by the ethics committee of the Institute for Population and Social Research-Institutional Review Board (IPSR-IRB) on 30 January 2020 (COE. No. 2020/01-028).

Results

Levels and trends of older people household need for caregivers in Thailand

Overall, the increasing need for caregivers of older people was presented as a prominent trend. The percentage of the households need for caregivers of older people varied from the year 2007 to 2017. In 2007, 5.0% (95% confidence interval (CI) 4.6–5.5) of all older people households were in need for caregivers, this percentage was slightly lower (4.6%; 95% CI 4.2–5.0) in 2011. The percentage was, however, higher in 2014 (6.6%; 95% CI 6.1–7.0), and then marginally decreased to 6.3% (95% CI 5.9–6.7) in 2017 (Table 1).

Table 1. Percentage of older people’s household needs for caregivers by area of residence, region, household wealth, educational attainment of the household head, and income of the household head in 2007, 2011, 2014, and 2017.

Characteristics	2007 (n = 22,395)		2011 (n = 25,059)		2014 (n = 27,980)		2017 (n = 25,635)
	Estimate	95% CI	Estimate	95% CI	Estimate	95% CI	Estimate	95% CI
Overall	5.0	(4.6, 5.5)	4.6	(4.2, 5.0)	6.6	(6.1, 7.0)	6.3	(5.9, 6.7)
Area of residence
Urban	6.8	(5.9, 7.8)	5.8	(5.2, 6.4)	6.8	(6.1, 7.6)	6.9	(6.3, 7.5)
Rural	4.3	(3.8, 4.8)	4.0	(3.6, 4.5)	6.4	(5.8, 7.0)	5.9	(5.4, 6.5)
Design-based test (adjusted F)	24.699***		21.421***		0.697		4.278*
Region
Bangkok	8.4	(6.4, 10.8)	7.8	(6.4, 9.1)	9.8	(9.8, 9.8)	8.6	(7.2, 10.2)
Central (excluding Bangkok)	5.9	(5.0, 6.9)	5.9	(5.2, 6.7)	7.4	(6.6, 8.2)	7.5	(6.8, 8.4)
Northern	4.1	(3.4, 5.0)	4.2	(3.6, 5.0)	4.2	(3.5, 5.2)	5.0	(4.3, 5.9)
Northeastern	4.0	(3.2, 4.6)	3.0	(2.4, 3.6)	6.2	(5.3, 7.3)	5.5	(5.0, 6.1)
Southern	5.5	(4.3, 7.1)	4.9	(4.0, 6.0)	7.2	(6.2, 8.4)	6.0	(4.8, 7.6)
Design-based test (adjusted F)	8.641***		16.748***		14.538***		1.762
Relative household wealth
Quintile 1, poorest	na.	na.	4.1	(3.5, 4.8)	6.1	(5.4, 6.9)	6.0	(5.4, 6.8)
Quintile 2, poor	na.	na.	3.7	(2.9, 4.6)	5.9	(5.2, 6.8)	5.7	(5.0, 6.5)
Quintile 3, medium	na.	na.	4.5	(3.7, 5.5)	7.0	(6.0, 8.0)	6.7	(5.7, 7.9)
Quintile 4, rich	na.	na.	4.7	(3.8, 5.7)	7.2	(6.3, 8.3)	7.3	(6.2, 8.6)
Quintile 5, richest	na.	na.	6.3	(5.4, 7.3)	7.2	(6.4, 8.0)	6.4	(5.5, 7.5)
Design-based test (adjusted F)			5.456***		2.269		1.944
Educational attainment of household head
No education	na.	na.	6.8	(5.5, 8.3)	8.1	(6.8, 9.6)	11.1	(9.6, 12.8)
Preschool-primary	na.	na.	4.1	(3.7, 4.5)	6.3	(5.7, 6.8)	6.0	(5.5, 6.4)
Secondary	na.	na.	5.0	(4.1, 6.1)	6.5	(5.6, 7.5)	5.3	(4.0, 6.8)
Higher than secondary	na.	na.	7.0	(5.0, 9.7)	7.6	(6.4, 8.9)	5.6	(4.1, 7.7)
Design-based test (adjusted F)			9.232***		3.679*		11.470***
Income in the past year of household head (baht)
<10,000	na.	na.	6.6	(5.5, 7.9)	7.1	(5.8, 8.8)	11.0	(9.4, 12.8)
10,000–19,999	na.	na.	4.7	(3.8, 5.8)	9.0	(7.6, 10.6)	8.2	(7.2, 9.4)
20,000–29,000	na.	na.	4.6	(3.7, 5.8)	6.6	(5.6, 7.7)	7.0	(5.8, 8.3)
30,000–49,999	na.	na.	3.1	(2.4, 3.9)	6.0	(5.2, 6.9)	5.5	(4.8, 6.3)
50,000–99,999	na.	na.	3.2	(2.5, 4.1)	5.1	(4.4, 5.8)	5.1	(4.5, 5.8)
100,000+	na.	na.	4.6	(3.8, 5.5)	6.0	(5.3, 6.8)	4.6	(3.7, 5.7)
Design-based test (adjusted F)			8.553***		7.787***		11.450***

calculated from weighted samples under complex samples analysis; *** P < 0.001; **P < 0.01; and *P < 0.05; na. (not available)

With respect to an area of residence, the findings clearly showed that the level of household need for caregivers of older persons was higher in urban areas. However, a sharp increase in the need for caregivers in older-person households in rural areas was clear in 2014, and the rural-urban differential disappeared then. Yet, in 2017, the need for caregivers was significantly different between urban and rural (Table 1). When differentiated by region of Thailand, Bangkok had the highest percentage of household need for caregivers of older people in 2007, 2011, 2014, and 2017 (8.4%, 7.8%, 9.8%, and 8.6%, respectively). The need for caregivers of older persons in the Central region has obviously increased (5.9% in 2007 and 2011, 7.4% in 2014, and 8.6% in 2017), while a gradually increasing trend was revealed in the North region (4.1% in 2007, 4.2% in 2011 and 2014, and 5.0% in 2017; Table 1).

When taking household socioeconomic characteristics (household wealth, educational attainment and income of household head) into consideration, the findings showed that the prevalence of the household need for caregivers was significantly highest in the richest households (6.3%) in 2011. Conversely, the poor households showed the lowest percentage of the need for caregivers, accounting for 3.7%, 5.9%, and 5.7% in 2011, 2014, and 2017, respectively. However, the difference across the wealth index was not statistically significant in 2014 and 2017 (Table 1). Importantly, there was a significant difference in the need for caregivers across the indices of educational attainment and annual income of the household head. The highest levels and increasing trends of the household need for caregivers were shown in households where the household head had no education (6.8% in 2011, 8.1% in 2014, and 11.1% in 2017), and where the household head had an income lower than 10,000 baht (6.6%, 7.1%, and 11.0%; Table 1). It should be noted that the prevalence of household needs for caregivers of older persons by household socioeconomic characteristics in 2007 could not be calculated since the data on total households were not available (Table 1).

Met and unmet household need for caregivers of older persons

Levels and trends in 2007, 2011, 2014, and 2017

Among the samples of older people household need for caregivers in 2007 (n = 1,225), 2011 (n = 1,274), 2014 (n = 1,739), and 2017 (n = 1,579), 85.4% (95% CI 82.0–88.2), 82.5% (95% CI 78.5–85.9), 73.5% (95% CI 70.3–76.4), and 77.1% (95% CI = 72.7, 81.0) were the households with met need for a caregiver of an older person(s) in 2007, 2011, 2014, and 2017, respectively. Conversely, there was a rising trend of households with an unmet need for caregivers, from 14.6% (95% CI = 11.8, 18.0) in 2007, 17.5% (95% CI 14.1–21.5) in 2011 to 26.5% (95% CI 23.6–29.7) in 2014, which then slightly declined to 22.9% (95% CI 19.0–27.3) in 2017.

Socio-economic characteristics of the households with unmet need

As mentioned previously, for the 2007 survey, only data from households with persons age 50 years or over are available. Therefore, the household socioeconomic variables could not be tabulated for 2007.

Table 2 shows the equality profile of the household need for a caregiver of an older person(s) by household socioeconomic characteristics in 2011 and 2017. Overall, in 2011, the umnet household need for caregivers of older people was highest among the households with the highest socioeconomic (though the differences were not statistically significant). By contrast, in 2014 and 2017, the highest prevalence of unmet household need for caregivers of older persons was found in the poorest households (33%; 95% CI 27.3–29.9), and this was statistically significant in 2014. In 2011, the educational attainment of the household head was positively correlated with unmet needs for a caregiver of an older person in the household. However, this relationship was reversed in 2014, when the educational level of the household head was significantly inversely correlated with unmet need. The prevalence was highest in older people’s households where the household head had no education (32.7%; 95% CI 23.9–43.0). Yet, in 2017, the higher percentage of the unmet need was more likely to be found in households where the head had higher educational attainment.

Table 2. Percentage distribution of unmet household needs for caregivers of older persons by household socio-economic characteristics in 2011, 2014, and 2017.

Household socio-economic characteristics	Unmet household needs for caregivers
	2011 (n = 1,274)		2014 (n = 1,739)		2017 (n = 1,579)
	Estimate	95% CI	Estimate	95% CI	Estimate	95% CI
Overall	17.5	(14.1, 21.5)	26.5	(23.6, 29.7)	22.9	(19.0, 27.3)
Relative household wealth
Quintile 1, poorest	15.5	(10.4, 22.6)	33.3	(27.3, 29.9)	26.5	(26.1, 26.9)
Quintile 2, poor	20.7	(11.5, 34.5)	30.5	(24.5, 37.3)	25.6	(21.4, 30.3)
Quintile 3, medium	12.3	(6.9, 21.2)	26.0	(19.2, 34.1)	22.5	(21.2, 23.7)
Quintile 4, rich	18.4	(12.1, 26.9)	25.5	(19.4, 32.8)	17.6	(9.7, 29.9)
Quintile 5, richest	21.8	(14.2, 32.0)	15.1	(10.0, 22.2)	19.5	(15.2, 24.7)
Design-based test (adjusted F)	0.913		4.370**		3.795
Education level of household head
No education	12.3	(7.1, 20.5)	32.7	(23.9, 43.0)	15.6	(14.5, 16.8)
Preschool-primary	17.7	(13.6, 22.6)	28.9	(25.2, 32.9)	24.3	(21.6, 27.2)
Secondary	20.1	(10.3, 35.6)	20.8	(14.0, 29.3)	24.0	(14.2, 37.6)
Higher than secondary	28.3	(15.0, 46.8)	11.4	(6.3, 19.7)	23.5	(7.2 55.0)
Design-based test (adjusted F)	1.518		5.459***		1.785
Income in the past year of household head (baht)
<10,000	14.6	(9.4, 22.0)	28.4	(19.4, 39.4)	18.8	(18.2, 19.4)
10,000–19,999	14.5	(8.4, 23.7)	34.1	(26.4, 42.8)	21.0	(19.0, 23.1)
20,000–29,000	9.0	(4.4, 17.6)	27.5	(19.5, 37.3)	22.3	(20.8, 23.9)
30,000–49,999	25.3	(15.4, 38.7)	26.9	(20.9, 34.0)	24.8	(18.7, 32.0)
50,000–99,999	19.2	(11.5, 30.2)	26.8	(20.9, 33.8)	26.1	(17.6, 36.8)
100,000+	26.6	(16.0, 40.9)	27.4	(20.1, 36.1)	24.2	(15.3, 36.1)
Design-based test (adjusted F)	2.025		0.513		2.185

calculated from weighted samples under complex samples analysis; *** P < 0.001; **P < 0.01; and *P < 0.05

There was an inconsistent pattern of unmet household need for caregivers of older people in these 3 years in relation to the annual income of the household head (Table 2).

Potential determinants of household need and unmet need for caregivers of Thai older people

Binary logistic regression analysis of pooled data, 101,091 older people’s households, indicates a statistically significant difference in the household need for caregivers over time. In the unadjusted model, the levels of household need for caregivers of older people in the years 2007, 2014, and 2017 were significantly higher than in the year 2011 by 9%, 24%, and 23%, respectively (Table 3).

Table 3. Odds ratio (OR) and 95% confidence interval (CI) from binary logistic models predicting household need and unmet need for caregivers of older persons, pooled data from the SOPT 2007 to 2017.

	Household need for caregivers							Unmet household need for caregivers
	Unadjusted			Model 1 (n = 101,091)		Model 2 (n = 73,690)		Unadjusted			Model 1 (n = 5,817)		Model 2 (n = 4,166)
	n	OR	95% CI	OR	95% CI	OR	95% CI	n	OR	95% CI	OR	95% CI	OR	95% CI
Year of survey (ref: 2011)	101,091							5,817
2007		1.09*	1.01, 1.18	1.09*	1.01, 1.18				1.18	0.94, 1.49	1.16	0.92, 1.46
2014		1.24***	1.15, 1.33	1.26***	1.17, 1.36	1.43***	1.31, 1.55		2.32***	1.91, 2.84	2.19***	1.80, 2.68	2.03***	1.64, 2.51
2017		1.23***	1.14, 1.32	1.25***	1.16, 1.35	1.42***	1.31, 1.54		2.15***	1.75, 2.63	2.00***	1.63, 2.46	1.76***	1.42, 2.18
Area of residence (ref: Urban)	101,091							5,817
Rural		0.89***	0.85, 0.93	0.90***	0.85, 0.95	0.92*	0.86, 0.98		1.31***	1.15, 1.49	1.26**	1.10, 1.44	1.18*	1.01, 1.38
Region (ref: Bangkok)	101,091							5,817
Central (excluding Bangkok)		0.74***	0.66, 0.84	0.78***	0.69, 0.88	0.74***	0.64, 0.86		0.89	0.66, 1.21	0.82	0.60, 1.12	0.79	0.55, 1.13
Northern		0.54***	0.47, 0.61	0.56***	0.49, 0.63	0.48***	0.41, 0.56		1.11	0.81, 1.51	1.03	0.75, 1.42	1.02	0.71, 1.47
Northeastern		0.55***	0.49, 0.63	0.57***	0.51, 0.65	0.54***	0.46, 0.62		1.81***	1.34, 2.45	1.57**	1.15, 2.15	1.46*	1.01, 2.10
Southern		0.72***	0.64, 0.82	0.75***	0.66, 0.86	0.74***	0.63, 0.86		1.16	0.84, 1.61	1.03	0.74, 1.43	0.87	0.60, 1.28
Household socioeconomic characteristics
Household wealth (ref: Quintile 1, poorest)	78,630							4,588
Quintile 2		0.92	0.84, 1.01			0.99	0.90, 1.10		0.86	0.70, 1.06			0.85	0.69, 1.05
Quintile 3		1.01	0.92, 1.11			1.13*	1.02, 1.24		0.66***	0.52, 0.82			0.72**	0.57, 0.91
Quintile 4		1.14**	1.04, 1.25			1.27***	1.15, 1.41		0.74**	0.60, 0.92			0.82	0.65, 1.04
Quintile 5, richest		1.13**	1.04, 1.24			1.26***	1.12, 1.40		0.54***	0.43, 0.68			0.65**	0.49, 0.85
Educational attainment of household head (ref. No education)	78,646							4,591
Preschool-primary		0.65***	0.59, 0.71			0.69***	0.63, 0.76		1.23	0.98, 1.53			1.18	0.94, 1.49
Secondary		0.73***	0.65, 0.83			0.73***	0.63, 0.85		0.78	0.56, 1.08			0.89	0.61, 1.28
Higher than secondary		0.74***	0.65, 0.85			0.74***	0.62, 0.87		0.91	0.65, 1.27			1.10	0.74, 1.64
Income in past year of household head (baht) (ref.<10,000)	73,791							4,170
10,000–19,999		0.92	0.83, 1.03			0.90	0.80, 1.01		1.06	0.81, 1.37			0.91	0.69, 1.19
20,000–29,000		0.76***	0.67, 0.85			0.73***	0.64, 0.82		0.95	0.71, 1.26			0.87	0.65, 1.17
30,000–49,999		0.60***	0.53, 0.66			0.56***	0.50, 0.62		1.08	0.83, 1.40			1.02	0.78, 1.33
50,000–99,999		0.52***	0.47, 0.58			0.47***	0.42, 0.52		1.23	0.96, 1.58			1.26	0.97, 1.64
100,000+		0.58***	0.52, 0.65			0.45***	0.40, 0.52		1.17	0.90, 1.52			1.38*	1.02, 1.86
Log likelihood				-2215.7360		-15712.1610					−2787.4267		−2127.6502
LR chi^2				238.56		610.31					178.53		146.21
Prob>chi^2				0.0000		0.0000					0.0000		0.0000
Pseudo R2				0.0054		0.0191					0.0310		0.0332

*** P < 0.001; **P < 0.01; and *P < 0.05

With adjustments for potential determinants, Model 1 adjusted for years of the survey, area of residence, and region. The findings showed that these three variables were a significant determinant of the household need for caregivers of older people. The older people’s households in the year 2007, 2014, and 2017 were more likely to need caregivers for older people than those in 2011 by 1.09 times, 1.26 times, and 1.25 times, respectively. The odds ratio (OR) indicated that older people’s households in rural areas were less likely to report the need for caregivers than those in urban areas by 10%. With regard to region, older people’s households in Central, Northern, Northeastern, and Southern areas had a lower need for caregivers than those in Bangkok.

The analysis of Model 2, which also included household socioeconomic determinants, found that household wealth, educational attainment, and the income of the household head significantly predicted the risk for a household need for caregivers of older people. The third, fourth, and fifth quintile of older people’s households reported a higher need for caregivers than those in the first quintile, in other words, the poorest households, as the adjusted ORs were 1.13 (95% CI 1.02–1.24), 1.27 (95% CI 1.15–1.41), and 1.26 (95% CI 1.12–1.40), respectively. In contrast, the higher the income of the household head, the less likely was the need for caregivers. The adjusted ORs for income ranged 20,000–29,000 baht; 30,000–49,999 baht; 50,000–99,999 baht; and 100,000 baht or higher were 0.73 (95% CI 0.64–0.82), 0.56 (95% CI 0.50–0.62), 0.47 (95% CI 0.42–0.52), and 0.45 (95% CI 0.40–0.52), respectively. Moreover, older people’s households with educated household heads had a lower need for caregivers. The year of surveys and region remained strongly significant, while the area of residence became a predictor at a significant level at 0.05 (Table 3).

The model analysis of the unmet household need for caregivers

Out of the 101,091 samples of older person’s households, 5,817 households with older household members who needed caregivers were included in the unmet household need analysis. Findings of the unadjusted and adjusted models demonstrated that older people households in the years 2014 and 2017 experienced two times higher unmet need compared to those in the year 2011 (Table 3).

In Model 1, older people’s households in rural areas were more likely to respond to unmet need for caregivers than those in urban areas (OR = 1.26%; 95% CI 1.10–1.44). Similarly, the households in the Northeastern significantly explained the higher unmet need for caregivers of older people (OR = 1.57; 95% CI 1.15–2.15).

Model 2 showed that there were only older people’s households within the third and fifth quintile predicting the lower unmet household need for caregivers. The educational attainment of the household head was not a significant determinant of unmet household need, while households, where the head had an income of 100,000 baht and higher, was a significant predictor. It is notable that older people’s households in rural areas, and in the Northeastern of Thailand, were still statistically significant, and the size of the OR was decreased.

Discussion

By adding the recent dataset from 2017, the present study extended our previous research on the state of the household need for caregivers of older persons. At the same time, we further analyzed the unmet household need for caregivers of older persons from 2007 to 2017.

Based on the basic ADLs performance of older people as an indicator of the household need for caregivers, this study found that levels of the household need for caregivers of older persons were slightly decreased, from 5.0% in 2007 to 4.6% in 2011. This figure increased to 6.6% in 2014, and then marginally declined to 6.3% in 2017. The findings from this multivariate confirmed a statistically significant difference in levels of the household need for caregivers of older persons during the years 2007, 2011, 2014 to 2017. However, rising trends of household need for caregivers of older persons were revealed. The findings make evident the rising demand for LTC services in Thailand. With the growth of the older population (United Nations, Department of Economic and Social Affairs, Population Division, 2019), we believe that the levels of need for caregivers of older people’s households will continue to increase.

Unmet household need for caregivers of older people was determined by whether older household members had difficulty in performing ADLs by themselves and lacked someone to help them. Although the majority of these in-need households could arrange a caregiver for the older person(s), not all did. The levels of unmet household need in older people’s households in 2014 and 2017 were significantly higher (almost double) compared to those in the year 2011. The study highlighted that there was a rising trend of households with unmet need for caregivers from 2007 to 2017. The gradually declining pattern of co-residence, or numbers of household members, has brought new concerns about caregiving and support in many Asian countries (Khan et al., 2017). Thailand is undergoing a demographic and socioeconomic transition. The need for care of older people who are living longer than ever before is occurring in the context of decreasing availability of family assistance (Peek et al., 2015), leading to an inevitable increase in unmet household need for caregivers of older persons. Based on the analysis of background data of older person’s households of the surveys in this study, the average number of members per household of the older people’s households gradually declined from 3.6 in 2007 to 3 persons in 2017 (Appendix). A gap between the demand for – and supply of – caregivers for older people thus emerges at the household level. It is very likely that this trend will continue if the proportion of households with older people continues to grow along with the decreasing size of households.

In terms of the resulting determinants of household need and unmet need for caregivers of older Thai people, this study found that areas of residence, region, and household socioeconomic characteristics significantly predicted the household need and unmet needs for a caregiver. With respect to areas of residence, the findings highlighted older people households in urban areas significantly predicted the risk of household need for caregivers of older people, but the older people’s households in rural areas reported significantly higher unmet need for caregivers. It has been argued that modernization has led to lower co-residence and higher invisible venerable older people (Logue, 1990). The influence of social norms on care for older people in a rural area stipulates that household members provide support to older members in the form of co-residence. It was evident that co-resident households have a positive impact on the health status of the older people (Velkoff, 2001). Older Thai people from rural areas showed a preference to live in their own homes that included other family members rather than on their own (Knodel & Chayovan, 2012). However, given all the contemporary circumstances, the population of rural areas is aging because of a demographic transition from high fertility and mortality to relatively low fertility and moderate mortality rates, along with complex modern economic development in terms of diverse patterns of living arrangement. As result of this study, Thailand should be aware of the increasing trend of household needs for caregivers of older people, which will continue in rural areas throughout the coming decade.

Significant regional inequalities in household needs for caregivers were found in 2007, 2011, and 2014. Bangkok had the highest percentage of household need for caregivers of older people, while the household need for caregivers of older persons was obviously increasing in the Central region of Thailand. We should also be aware of increasing trends in the household need for caregivers of older people who live in Northern Thailand as the high level of function health problem of the older people in remote areas has been reported (Weiangkham et al., 2014). Nevertheless, it is important to note that the older people in the Northeastern of Thailand were more likely to experience unmet need for caregivers. This is probably because a higher internal migration rate across regions of younger population during the time of surveys was to be found in the Northeastern (National Statistical Office, 2020). Therefore, caregivers were less available to provide caregiving to the older people.

We also find that household wealth positively predicted a higher household need for caregivers. This finding is opposite of Melzer (Melzer et al., 2000). However, the analysis of the unmet household needs for caregivers highlighted an important issue concerning long-term care, and the extent to which households with low socioeconomic status can properly care for older members. The results of this study indicated that there were significant disadvantages in the poorest quintile households. In general, socioeconomically poor households could afford less paid care or other kinds of supplementary care. Thus, caregiving for their older household members was still mainly provided by younger household members or social networks.

Significantly, the need for care of older people in households where the head has a low level of education and income has become increasingly important. The highest level and a sharp increase of household need for caregivers were shown in households where the household head had no education. A lack of education of the household head affects both qualities of care for the older person and the quality of life of the household members. Poor education must lessen the ability to deal with difficult conditions and stress in providing care to dependent older people (Schulz et al., 2016).

Limitations

There were some limitations to this study. The first limitation is related to the nature of the secondary source; there were no standard measures of household need or unmet need for caregivers of older people at the time of the study. As a result, the physical ADL abilities of the older person were used as an indicator for the needs in this research. The second limitation is that the institutionalized people were not included in the analysis as the surveys cover only private households. Thus, an underestimate of levels of the needs could have happened. Lastly, this study was limited by the cross-sectional design. Therefore, longitudinal population-based studies are recommended to analyze the prevalence and true predictors of household need and unmet need for a caregiver of an older person. In addition, beyond geographic and socioeconomic factors of households, caregiver characteristics can impact the unmet household need for a caregiver. However, the surveys had no available data to analyze such details.

Conclusion

This study is a contribution to filling the gap in essential new knowledge on the change in need and unmet need for caregivers of older people at the household level. It has important policy implications for the Long-term Care (LTC) of older persons in community. The LTC policy must take geographical and socioeconomic differences into consideration. In views of the findings relating to unmet need, the national and local government should not only be aware of the challenges posed by the expanding need for long-term caregivers of older persons, but also should be actively intervening to ease the household burden of long-term care for older persons, in particular, for households with low socioeconomic levels.

Key points

• Rising trends of households both with need and unmet needs for caregivers were revealed.

• The long-term care policy must take geographical differences into consideration.

• The national and local government should not only be aware of the challenges posed by the expanding need for long-term caregivers of older persons but also should be actively intervening to ease the household burden of long-term care for older persons, in particular, for households with low socioeconomic levels.

Acknowledgments

We would like to thank the National Statistical Office (NSO), Ministry of Digital Economy and Society of Thailand for providing us 2007, 2011, 2014, and 2017 Survey of Older Persons in Thailand for this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research project is supported by Mahidol University under Grant [A32/2563].

Appendix

The older person’s households in 2007, 2011, 2014, and 2017

Table

Characteristics	2007	2011	2014	2017
Number of sample households with older person	22,688	25,077	27,990	25,740
Percentage of households with older persons*	na.	37.94	37.35	34.32
Mean number of older persons per household* (SD)	1.34 (0.50)	1.37 (0.52)	1.38 (0.52)	1.43 (0.53)
Mean number of members per household* (SD)	3.62 (1.87)	3.51 (1.84)	3.44 (1.85)	3.01 (1.67)

*calculated from weighted samples, na. (not available).