Introduction
Disability has many destructive individual and social effects. According to the “International Classification of Functioning, Disability, and Health”, disability is a condition that limits and disrupts a person’s participation and activities [1]. The World Health Organization (WHO) estimated that more than one billion people, or 15% of the world’s population, live with some disabilities, 80% of whom are in low- and middle-income countries [2]. 2017 WHO report showed that about 183 million people have severe disabilities, indicating an increase of 23% compared to 2005 [3]. People with disabilities, in addition to the need for care from healthy people (such as vaccinations and other prevention and treatment services), need rehabilitation services [4]. Rehabilitation services include a wide range of services that seek to optimize the performance of people with disabilities [2]. The 2015 Global Burden of Diseases study showed that 74% of all years lived with disabilities (YLDs) in the world was due to disabilities, indicating the need for rehabilitation interventions [3].
Although the demand for rehabilitation services is growing, its supply resources in many parts of the world do not meet the current needs since rehabilitation services have not been considered seriously by many governments [5]. This neglecting has led to improper access to these services and no coordination between the demand and supply [2]. Some studies in South Africa show that only 26% of the population receives the required rehabilitation services [6,7,8,9]. The WHO report in 2014 showed that about 76%-85% of people with disabilities in developing countries do not receive any rehabilitation services [10]. A set of factors influence the utilization of rehabilitation services. To date, various studies have examined the factors affecting the utilization and non-utilization of rehabilitation services [11]. Patel et al. showed that the utilization of rehabilitation services by people with disabilities depends not only on socioeconomic factors but also on cultural factors, residence, literacy status, gender, etc. [12]. Borker et al. showed that the non-utilization of rehabilitation services by 75.8% of disabled people in a rural community in India [13]. Another study in 2017 showed that 76.3% of stroke patients in Ghana did not use rehabilitation services due to economic problems. Some studies have reported low utilization of rehabilitation services by people with disabilities in Iran [14, 15]. Given the importance and necessity of rehabilitation services, the governments should take the necessary measures to facilitate these services for people in need. To reduce the existing information gap, the present study examines the inequality in utilizing rehabilitation services among Iranian households.

Materials and Methods
The present study is a descriptive-analytical study with a cross-sectional design using household, expenditure, income survey data from the Statistics Center of Iran in 2018. The relevant data were extracted and categorized by province, urban and rural areas. The initial analysis of raw data was performed. In this regard, the provinces of Sistan and Baluchestan, Ardabil, Kohgiluyeh and Boyer-Ahmad, Kurdistan, and West Azerbaijan were excluded from the study due to lack of data on the use of rehabilitation services (audiometry, optometry, speech therapy, and physiotherapy). Households that declared zero essential expenses (e.g. in food expenses) were also excluded from the study. Finally, the relevant data of 15929 households in rural areas, 17467 households in urban areas, and 38958 households in the whole country were included in the study.
The Chi-square test was used to investigate the relationship between the utilization and non-utilization of rehabilitation services with the study variables. In the next step, inequality in the utilization of rehabilitation services was measured using the concentration index (CI). To calculate this index, household income was used as a ranking variable to measure inequality. Households were divided into five categories based on income level from the first quintile (with the lowest income) to the fifth quintile (with the highest income). The CI was calculated as Fourmula 1:


where µ represents the average rate of the dependent variable (percentage of households using rehabilitation services), ri refers to the ranking of each household according to the income quintile, and yi shows the utilization of rehabilitation services by household i. The numerical value of the CI is between -1 and +1. The positive value indicates that the use of rehabilitation services is higher among households with higher economic status. The negative value indicates that it is higher among households with lower economic status. In this equation, the dependent variable value is 0 or 1 and is not bounded within the range of -1 and +1. The normalization of the concentration index was performed by multiplying the value of CI by 1/1-µ, according to Wagstaff [16].
Access and Excel applications were used for data extraction and STATA v.14.1 software for data analysis. To better describe each province’s situation in the utilization of rehabilitation services, the rate of use of these services by urban/rural areas and the whole country was also displayed on the map using ArcGIS Map v.10 software.

Results
Of the total study households, about 258 (0.77%) had used rehabilitation services in 2018. Among the households that used rehabilitation services, 226 (87%) had a male head, and 32 (13%) had a female head. Also, 52 (21%) had 1-2 members, 156 (60%) 3-4 members, and 50 (19%) had ≥5 members. Besides, 239 (92%) had insurance coverage, and 19 (8%) had no insurance coverage, and this index had a significant effect on the use of rehabilitation services (P<0.05). Heads of 173 households (67%) who used rehabilitation services were employed. Moreover, the households in the fifth quintile of income (36%) used rehabilitation services more than other quintiles (Table 1).

In terms of the utilization at the provincial level, the lowest household rate using rehabilitation services was related to Bushehr Province (0.27%), followed by Lorestan, Semnan, Isfahan, and Kerman. The highest rate was related to Qom, Mazandaran, East Azerbaijan, Golestan, and Yazd provinces (>1%) (Table 2).

According to the utilization mapping (Figure 1), Zanjan, Qazvin, Khuzestan, Isfahan, Lorestan, Bushehr, and Semnan provinces had the lowest use of rehabilitation services in urban and rural areas and the whole country (shown in red on the map).

In urban areas, only the situation in Qom province was reported to be high. In contrast, a high utilization rate was reported in rural areas of East Azerbaijan, Mazandaran, Golestan, Yazd, Fars, and Hormozgan provinces. In the whole country, only for the provinces of East Azerbaijan, Mazandaran, and Qom, a high level of utilization was reported (shown in green on the map).
The study of inequality in the utilization of rehabilitation services in the whole country showed a concentration index of CI=0.24 (95% CI: 0.17-0.30), indicating inequality in favor of the rich. This condition suggests that the rich use more of these services than the poor. The value of this index for rural and urban areas was 0.27 and 0.19, respectively (P<0.001). At the provincial level, the CI values for the provinces of Kermanshah, Kerman, Isfahan, and Ilam were negative, indicating inequality in favor of the poor. However, these values were not significantly different from zero (equality in service utilization). In 11 out of 26 provinces, inequality was statistically significant (P<0.001). Among these, the highest level of inequality was related to Bushehr (CI=0.74), Lorestan (CI=0.70), and North Khorasan (CI=0.59) provinces. The lowest CI value was related to Khorasan provinces (CI=0.04), but it was not statistically significant. The lowest inequality that was statistically significant was related to Qom province (CI=0.26) (Table 3).

Discussion and conclusion
The purpose of this study was to investigate the distribution of the utilization of rehabilitation services among Iranian households. The results showed that about 0.77% of the Iranian households in 2018 had used rehabilitation services. Considering the prevalence of disability in Iran in 2011, which was reported to be about 1.35%, this finding can be justified [17]. The type of population can also affect the use of services. A study by Fullard et al. in the United States found that about 14% of Parkinson patients in 2007 used services such as physiotherapy, occupational therapy, and speech therapy [18].
The value of the concentration index in this study was 0.24 for the whole country, and it was higher in households with higher income. A study by Ahmadi et al. showed that, for specialized medical and dental services, inequality was in favor of the rich, while for the general medical, family physician, and primary health care services, inequality was in favor of the poor [19]. The study by Rezapour et al. in Kerman also showed that inequality in the use of outpatient and inpatient services was in favor of poorer groups [20].
Results reported that household size and health insurance coverage had a significant relationship with the utilization of rehabilitation services. With the increase of the household size, the use of rehabilitation services increased. Lack of insurance coverage for some rehabilitation services (speech therapy, occupational therapy, and technical orthopedics), long duration of use of these services, and high deductible for services such as physiotherapy led to the higher utilization of these services by the households with higher income. In other words, the use of rehabilitation services had disproportionately been concentrated on high-income households. These results highlight the need to review and modify the basic health insurance package in Iran. Some studies believe that misunderstandings of the provisions of the Public Health Insurance Act by policymakers in the past have led to the exclusion of rehabilitation services from basic health insurance packages [21]. Since studies have shown a higher prevalence of disability among poorer groups in society, it is necessary to make improvements in the regulations of health insurance to increase the benefit of these people from rehabilitation services so that they can receive the rehabilitation services according to their needs and without financial pressure [22,23,24].
In the present study, Bushehr and Qom provinces of Iran had the highest and lowest inequality in the use of rehabilitation services, respectively, indicating that the income gap of households in Bushehr Province had caused more inequality in the use of rehabilitation services compared to Qom Province. According to the results, inequality in the utilization of rehabilitation services was higher in rural households than in urban households. This condition probably indicates that, due to the higher income gap between households, unfair distribution of rehabilitation services, and low geographical access to these services, lower-income households in less developed areas are less likely to use rehabilitation services. Chavehpour et al. showed that the development rate is directly related to the concentration of health resources such that 70.6% of hospital beds in Isfahan and Tehran provinces were located in areas with higher social and economic status [25].
Among the provinces, Qom and Bushehr had the highest and lowest rate of using rehabilitation services, respectively. This condition indicates that the share of the public health centers in providing rehabilitation services in Bushehr Province is less than that of the private health centers. Since the cost of these services in private health centers is higher, lower-income groups are less tended to use these services. Other reasons can be the insufficient supply of rehabilitation services or less awareness of these services in Bushehr Province. Vamaghi et al. in a study in Tehran, showed that one of the reasons for not using speech therapy services was the lack of awareness of the parents about the existence of such services [26]. Studies by Rais Dana et al. [53], Soltani et al. [54], and Abdi et al. [7] showed that the existence of cultural factors such as wrong perceptions and attitudes towards people with disabilities could be one of the barriers in using these services in Iran.
The main limitation of the present study was the assumption that households would spend on rehabilitation services. In other words, only households that stated that they had paid for rehabilitation services were considered as households using rehabilitation services. Accordingly, the households that might have received free rehabilitation services for having supplementary insurance were considered as households with no utilization of rehabilitation services. However, studies in Tehran have shown that rehabilitation services covered by insurance are less popular than other services [29], which may be due to the limited options and low quality of available services [30].
Income inequality in the use of rehabilitation services significantly reduces the access of low-income people to these services, including rehabilitation services in the primary health insurance package with appropriate price and population coverage can increase equity in access to rehabilitation services. Moreover, fair distribution of rehabilitation services in accordance with the needs of both public and private rehabilitation centers can play an essential role in increasing the use of these services by households. The committed participation of institutions and organizations such as the Ministry of Health and Medical Education, health insurance organizations, welfare organizations, and non-governmental organizations (NGOs) related to people with disabilities in health policy processes can play an essential role in fulfilling the expectations of this group of people about receiving rehabilitation services.

Ethical Considerations
Compliance with ethical guidelines
This study was approved by the Ethics Committee of Kermanshah University of Medical Sciences (Code: IR.KUMS.REC.1398.516).

Funding
This research was funded by Kermanshah University of Medical Sciences.

Authors' contributions
Study design: Jafar Yahyavi Dizaj, Farogh Nomani, Ali Kazemi Karyani; Methods of study and statistical analysis: Manijeh Soleimanifar, Mohsen Fateh, Shahin Soltani; Data analysis and interpretation: Amir Massoud Arab, Jafar Yahyavi Dizaj, Ali Kazemi, Shahin Soltani; Consulting, editing and final writing of the article: Manijeh Soleimanifar, Farogh Nomani. All authors approve the content article.

Conflict of interest
there is no Conflict of interests.

Acknowledgments
We would like to thank the Statistics Center of Iran for making the data available, as well as the professors who helped guide this study. Kermanshah University of Medical Sciences is also appreciated for its financial support of this research.
References