Factors associated with the use of rehabilitation services among people with disabilities in Chile: A population-based study

Study design

This analytical cross-sectional study analyzed secondary data from Chile’s 2015 Second National Disability Study (ENDISC II).¹⁰ The study population consisted of residents of urban and rural areas of the 135 communes of the 15 regions of Chile. ENDISC II was conducted between June 30 and September 5, 2015.¹¹

Setting

ENDISC II is a population-based survey that aims to “determine the prevalence and characterize disability at the national level, identifying the main gaps in access to persons with disabilities”.¹⁰ It was conducted jointly by the Ministry of Social Development (MDS) and the National Statistics Institute (INE) and supervised by the National Disability Service (SENADIS) of Chile.¹¹

ENDISC II used a probabilistic sampling in two phases. The first phase used a list of households from a previous survey (Case 2013), and the second phase systematically selected households within the strata (communes and rural–urban areas). The final selection unit was the dwelling. ENDISC II’s design allows its results to be nationally and regionally representative because it covers 80% of the total number of households in Chile. Data collection was conducted through personal interviews with paper questionnaires. The surveys were applied to 12,015 dwellings, which included 12,265 people aged ≥18 years. Details of the sample design can be reviewed in their methodology book.¹⁰

Participants

For this study, we included persons with disabilities aged >18 years of both sexes and excluded records with missing or inconsistent data on the variables of interest.

Measurement of disability

To measure the prevalence of disability, ENDISC II applied a methodology based on the Model Disability Survey (MDS) used by the WHO in the II World Report on Disability, which measures three aspects, namely, functioning, health status, and environmental factors.¹⁰ The MDS was developed after the review of 179 health and disability surveys from around the world, following the implementation of the ICF in 2001.¹² The application of this methodology in the Americas is novel because it is different from surveys that use few questions to identify persons with disabilities, such as those of the Washington group.¹³ This methodology is based on the item response theory and uses a metric scale that places different questions associated with greater or lesser disability. Thus, it is possible to identify the degrees of difficulty and adequately estimate the total prevalence of disability.¹³

Variables

The outcome variable was rehabilitation services in the previous year for persons with disabilities. This variable is of a nominal qualitative type and was measured with the question, “In the last 12 months, did you receive any rehabilitation services?” The response categories were yes and no.

The exposure variables were sex (male and female), age group (18–29, 30–44, 55–59, and 60+), educational level (no education, elementary, middle, and high school), marital status (married/cohabiting, widowed/separated/divorced, and single), origin (urban and rural), indigenous identification (no and yes), current job (no and yes), health insurance system (FONASA, FFAA, and Order, ISAPRE, none or private, and other/does not know), chronic diseases (which included diabetes, hypertension, arthritis/arthrosis, heart diseases, respiratory diseases, migraine, and AIDS/HIV diagnosed by a doctor; “Has a doctor ever told you that you have [name of disease/health condition]?” no and yes; mental illness (which included persons with physician-diagnosed anxiety and depression; “Have you ever been told by a doctor that you have [name of disease/health condition]?” no and yes), recreational activities (“During the past 6 months, did you engage in or attend the following activities or places?” no and yes), physical activity (“In the past month, did you play sports or engage in physical activity outside of your work schedule, for 30 min or more?” no and yes), caregiver assistance (“Because of your health, do you have someone to help you at home or outside your home, including family and friends, to perform the following activities?” no and yes), and perceived discrimination “In the past 12 months, have you felt discriminated against (i.e., been prevented from doing something, bothered, or made to feel inferior?” no and yes).

Bias

Self-reporting bias may increase the prevalence of disability.¹⁴ However, surveys and population censuses have widely used self-reported questions to measure this condition. For example, the Washington Group questions measure disability quickly, effectively, and economically and have been used in more than 69 national censuses worldwide.¹⁵ ENDISC II also uses self-report questions through an interview conducted by trained enumerators and has already been applied in 10 countries, Chile being the first in Latin America.¹²

Statistical processing and analysis

The data were downloaded from the SENADIS web page and analyzed with STATA version 16 (Stata Corporation, College Station, TX, USA). To select the subgroup of people with disabilities aged >18 years, the variable disc_adult was used through the sub pop command. Frequency distribution tables and weighted percentages with their 95% confidence intervals (95% CI) were prepared. In the bivariate analysis, to evaluate differences in the characteristics of people with disabilities who use or do not use rehabilitation services, the chi-square test was corrected with the F statistic for survey design. In the multivariate analysis, Poisson regression was used to estimate prevalence ratios (PR) with their 95% CI. A crude model was developed, and variables having a p-value <0.25 were included in the adjusted model. All results were weighted according to the complex sample design of the ENDISC II. A p-value of p <0.05 was assumed to be statistically significant. The possible multicollinearity among the variables of the adjusted model was also checked through the manual calculation of the variance inflation factor (VIF).¹⁶ A VIF ≥10 was an indication of collinearity. We did not perform any stratification according to sex, as previous studies have not shown that this variable is associated with the use of rehabilitation services in people with disabilities.¹⁷

Ethical aspects

The ENDISC II database is publicly accessible and available on the SENADIS website. Data are coded and do not allow the identification of participants. The Human Medicine career approved the present secondary analysis of the Universidad Científica del Sur. It was exempted from review by the institutional ethics committee according to resolution N° 407-DACMH-DAFCS-U. CIENTIFICA-2022, issued on August 1, 2022.

Prevalence of the use of rehabilitation services in people with disabilities

Approximately 19.1% of people with disabilities in Chile used some rehabilitation services in the previous year. A systematic review that included 77 studies, of which 28 measured access to rehabilitation services in persons with disabilities, found low coverage of rehabilitation in India (5%), Turkey (5%), and Bangladesh (7%) and high coverage in Brazil (80%), Malaysia (76%), Philippines (70%), and South Africa (71%).¹⁸ The present study described similar results. A population-based analysis of 13,659 individuals with disabilities in southern Brazil found that only 9.2% of people used rehabilitation services.¹⁹ A study in a Ugandan community of 318 randomly selected persons with physical disabilities reported a prevalence of 26.4% of rehabilitation service utilization.¹⁷ Another study in Uganda of 284 persons with physical disabilities receiving CBR services found a prevalence of access to rehabilitation services of 41.1%, although only 6.8% had access to physical therapy.²⁰

This variability in the prevalence of rehabilitation services could be explained by the diversity of methodological designs, different instruments used to measure disability, or differences in access to rehabilitation services. As long as methodologies for measuring disability are not standardized, these differences will remain unexplained. Thus, caution should be exercised when comparing similar studies.

Factors associated with the use of rehabilitation services

People with disabilities with lower educational levels were less likely to use rehabilitation services than those with higher education levels. These results are consistent with studies conducted in Uganda,¹⁷ Brazil,^21–23 Poland,²⁴ and Peru,²⁵ which indicates that low educational level decreases the use of rehabilitation services. A study in Brazil also found that older adults with low educational level were more likely to have a functional disability.²² In this regard, several population-based studies have shown that educational disparities are associated with a higher incidence of disability^26,27; this would be explained by the physical and intellectual demand that is necessary to achieve a higher level of education, and because of their condition, people with disabilities would not reach this level. Moreover, limited economic resources would increase these inequalities.²⁸ The more significant disability in people with lower educational levels would explain our study’s greater need for rehabilitation services.

People with disabilities and mental illnesses such as depression or anxiety were more likely to use rehabilitation services. Depression and disability have a reciprocal association. Several factors, such as social stereotyping, abuse, poverty, environmental barriers, or lack of access to health services, predispose people with disabilities to suffer up to depressive symptoms three times more than their non-disabled peers; moreover, depression can lead to disability.²⁹ This association is clearly described in several studies.^30,31 However, there is limited information on depression and its association with the use of rehabilitation services in people with disabilities. Some studies have shown results similar to ours. In 3,568 adults with disabilities in Korea, those who received rehabilitation services had a higher risk of depressive symptoms than those who did not (OR 1.23; 95% CI 1.01–1.50).³² A population-based study of 195,033 Americans found a higher prevalence of depressive symptoms in adults with disabilities who used assistive technologies (30.4%) than those who did not (7.4%).³³ Another study in Germany involving 4,020 patients with cancer reported a higher prevalence of depression in those patients coming from cancer rehabilitation centers (28%) compared with inpatients (24%) or outpatients (21%), increasing this probability up to six times.³⁴

According to our findings and the evidence reviewed, there is an increase in depressive symptoms in people who used rehabilitation services. This may be due to situations inherent to the rehabilitation process, such as pain, embarrassment, or frustrations experienced by patients when their rehabilitation takes longer than expected or when the disability is permanent. Information on this topic is limited, but some studies have described this explanation.³²

Finally, people with disabilities who have caregiver assistance were more likely to use rehabilitation services. One of the consequences of disability is dependence on performing basic activities, which leads to the need for a caregiver.³⁵ A population-based study in Peru found that 40.5% of people with disabilities were dependent on a caregiver; this function was mainly assumed by female family members.²⁵ The role of the caregiver mainly falls on a family member, either the children or the partner.³⁶ A study in China conducted on informal caregivers of people with intellectual disabilities found that 85% of the caregivers were the parents of the patients.³⁷ Therefore, family members would have an essential role in the decision to access rehabilitation services, especially in those with severe limitations and almost absolute dependence on the caregiver. This explanation has been postulated by Medeiros et al. to justify the high prevalence of rehabilitation services in minors with disabilities.¹⁹ The role of the caregiver in the success of rehabilitation is fundamental because the caregiver often becomes the channel of communication between the physician and the patient. In addition, health professionals rely on caregivers to extend the rehabilitation process at home.³⁸

Limitations

This study has some limitations. First, social desirability and recall biases could alter the prevalence of rehabilitation services in persons with disabilities. Second, certain variables related to the use of rehabilitation services, such as transportation, architectural barriers, or quality of medical care, were not available in the ENDISC II. Third, stratification according to age groups was not performed; in older adults, the associated factors may vary concerning other age groups. Fourth, the use of rehabilitation services only included the last 12 months. Their use before that time is unknown, especially in patients with chronic disabilities. Fifth, some categories of the exposure variables may not have an adequate sample size to show an association with the use of rehabilitation services. Sixth, causality between the main variables cannot be affirmed because of the study’s cross-sectional design. As a strength, ENDISC II is a population-based survey; thus, the results of this study would be representative of people with disabilities in Chile.