The study adopted a quantitative approach and a cross-sectional design to assess the knowledge and access to rehabilitation services by families of CWDs in the Upper West Region (UWR) of Ghana. This approach facilitated an objective assessment of the study participants’ knowledge and the access barriers to rehabilitation services that can be generalised to settings of similar characteristics. The cross-sectional nature of this study allowed a one-time collection of data, which is faster, cost-effective and can provide useful information on the prevalence of the condition being studied for public health planning (Setia 2016).

The study was undertaken in four selected districts: Wa West, Wa Municipal, Jirapa Municipal and Nandom located in the UWR of Ghana. The region covers a land space of 118 476 km² and shares borders with Burkina Faso to the north, Cote d’Ivoire to the west, the Upper East region of Ghana to the east and the Savana region to the south. The region has a population of 894 326 people, with 9.4% of individuals aged 5 years and older having some form of disability (defined as having difficulty in performing at least one activity) (Ghana Statistical Service 2021). Less than half (46.0%) of the population older than 6 years has received formal education or is currently in school, with 41.7% of these individuals being female (Ghana Statistical Service 2021). The region is equipped with a regional hospital, four polyclinics, three district government hospitals, five private hospitals and five maternity homes. Additionally, there are 72 health centres, 10 clinics and 319 Community-based Health Planning Service (CHPS) compounds (Ghana Health Service, 2023). These health facilities are available across the 11 administrative districts of the region. The population density in the region is 40 persons/km²; however, some districts have a population density of 13 persons/km², with communities located far apart (Ghana Health Service, 2023).

The participants for this study were family members (classified as fathers, mothers, grandparents and others – siblings, aunties, etc.). Therefore, the eligibility for recruiting a family member was that the person must be the first person to attend to the child’s healthcare needs or directly involved in attending to the healthcare needs of the child. The child must not be more than 18 years of age.

In addition, the confirmation of the presence of disability in the child was a requirement. This was done by verifying the hospital folder of each child to confirm the health provider’s diagnosis. In cases where the folder was unavailable, parents of children with visible or identifiable disabilities such as speech impairment, hearing, cerebral palsy, and so on were contacted for recruitment. The research assistants (RAs) recruited for the data collection had previous training in disability-related fields and were further trained to facilitate the identification of disabilities. The three RAs were a Registered General Nurse, a Special Education Teacher and a Disability Practitioner (holder of Bachelor of Science [BSc] Disability and Rehabilitation Studies degree).

Finally, the families were required to be aware of rehabilitation services and recognise that their child with a disability needs a rehabilitation service. This was to ensure that the participants of this study had at least an idea of rehabilitation services and could respond to the survey questionnaire relating to access and access barriers (real or perceived).

A community-based survey was conducted in some communities of the selected districts (Wa West, Wa Municipal, Jirapa Municipal and Nandom) using a network approach as the first method. In most of the rural communities, the healthcare providers (who were largely ‘community nurses’) were the first contact persons who referred the research team to families with a child with a disability, and thereafter, the network process continued. However, in communities without a health facility or where the ‘community nurse’ knew no family with a child with a disability, a house-to-house approach was adopted.

The Cochran’s sample size determination formula was used to estimate the sample size:

where z is the confidence interval of 95%; p is the prevalence of 9.4% (Ghana Statistical Service 2021).

The prevalence of 9.4% represents the proportion of individuals aged 5 years and older with disabilities – defined as having difficulty performing certain activities – within the UWR of Ghana, where this study was conducted:

Taking into account a 10% loss rate, the minimum sample size required was 144. However, efforts were made to collect data from as many families as possible during the study period. Ultimately, 206 participants were recruited after giving their consent to participate.

The questionnaire was locally developed with keen attention to the culture of the people, taking into consideration the low educational background of the study population. It covered the following thematic areas: knowledge of what rehabilitation services entail, access to and barriers to rehabilitation services, and the level of satisfaction with those services. The section measuring participants’ knowledge provided ‘Yes’ or ‘No’ response options. The aim was to determine whether participants understood what the existing rehabilitation services in their catchment area entail and which services are most suitable for the various disabling conditions listed in this study. The questionnaire used to determine access barriers and satisfaction levels included multiple response options (derived from literature and insights from the local area), allowing participants to select all applicable responses.

The questionnaire was pre-tested among a few (n = 15) families with the same cultural background in the Wa East district. Attention was paid to the translation of the questionnaire into local languages and the use of culturally appropriate terms. The pretesting aimed to ensure the accuracy of the data during the main study and the appropriateness of the questionnaire for measuring the areas investigated.

A semi-structured researcher-administered questionnaire was administered via a digital platform (Kobo Collect). Three RAs were trained for data collection, and each is proficient in the local language of the participants (i.e., Waali, Dagaare or Brifor). The translation of the questionnaire into the various local languages was emphasised during the training, with support from selected local individuals.

The data were retrieved from the Kobo Collect server and exported to STATA software for descriptive and inferential analysis. Descriptive analysis was conducted on the demographic data of study respondents, their knowledge of rehabilitation services, and access barriers faced by families of CWDs. A Chi-square test was further carried out to establish associations between the demographic characteristics of the participants and other variables of interest in the study. A significant association was determined at a 95% confidence interval. The results are presented in frequency tables and bar charts.

Ethical clearance to conduct this study was obtained from the Kwame Nkrumah University of Science and Technology Committee on Human Research Publication and Ethics (ethical approval number: CHRPE/AP/068/23). Participation in the study was entirely voluntary. Considering the high illiteracy level in the study area, verbal consent was obtained from participants after guiding each individual through the Study Information document in their local language. Participants who could read and understand the information document were provided with hard copies to review and decide whether to give consent or not.

In Table 1, 206 participants were involved in the study, the majority of whom were females (n = 145, 70.37%). The average age of the participants was 48.6 years, with a standard deviation (s.d.) of 14.8. Most participants lived in rural areas (n = 152, 73.79%). Mothers constituted 96 (46.60%) of the participants, and 91 (44.2%) lived in households with 6–10 members, making this the largest household size in the study.

Concerning the educational level of the caregivers, the majority (n = 157, 76.21%) had no formal education, while 97 (47.09%) were unemployed. Among those who were employed, 72 (34.95%) were farmers. In terms of religion, the majority (n = 139, 67.3%) identified as Christians. Nearly all participants (n = 192, 93.2%) were categorised as having low monthly incomes (see Table 1).

Out of the 206 family members involved in the study, the majority 114 (55.3%) had male children. The mean age of the children was 11.9 years, with a s.d. of 11.6 indicating a wide variety of range. The largest proportion of the children 116 (56.3%) were not attending school, and the most prevalent disability type among the children was physical disability, affecting 98 (47.57%) of them (see Table 2).

Table 3 represents the responses regarding the rehabilitation services required for the disability types categorised in this study. The majority of participants, 195 (94.66%), answered ‘Yes’ to physiotherapy being a rehabilitation service for physical disability, while 190 (92.33%), 198 (96.12%) and 119 (57.77%) answered ‘No’ for occupational therapy, speech and language therapy and assistive devices, respectively, as rehabilitation services for physical disability. Furthermore, 206 (100%) and 193 (93.69%), respectively indicated that physiotherapy and occupational therapy are not rehabilitation services for sensory disability. Nearly equal proportions, 126 (61.17%) and 139 (67.48%), indicated that speech and language therapy and assistive devices are the rehabilitation services needed for children with sensory disabilities.

For children with intellectual disabilities, most participants – 175 (84.95%), 140 (67.96%) and 171 (83.01%), respectively indicated that they do not require physiotherapy, occupational therapy and assistive devices. However, 133 (64.56%) indicated that they require speech and language therapy services.

Table 4 shows the distribution of caregivers and their corresponding levels of knowledge about rehabilitation services. A cross-tabulation shows that over half of those aged 20–39 years and those over 60 demonstrated higher knowledge than those aged 40–59 years. Less than half of both males and females exhibited high knowledge levels. Furthermore, a greater proportion of fathers (52.0%) and grandparents (49.0%) were more knowledgeable than mothers (45.0%), while nearly all participants classified as ‘other family members’ (92.0%) showed low knowledge. Quite unexpectedly, a larger proportion of participants in rural areas (53.0%) and those who attained only primary education (64.0%) were found to be more knowledgeable than their counterparts. Additionally, a higher proportion of employed individuals (69.0%) demonstrated greater knowledge than those who were unemployed (20.0%).

A test of association revealed that the caregiver’s age (p = 0.048), guardian role (p = 0.051), place of residency (p ≤ 0.001) and employment status (p ≤ 0.001) showed statistically significant differences in knowledge levels at a confidence level of 95%.

The participants were asked direct questions to determine who had tried to access rehabilitation services for their child with a disability. The results revealed that more than half, 113 families (54.85%), had at least attempted to access such services for their CWDs. However, of the 113 families that tried, the majority, 82 (72.82%), were unsuccessful in obtaining the necessary rehabilitation services for their child.

The barriers to accessing rehabilitation services, as presented in Figure 1 showed that the majority of caregivers strongly agreed that feelings of shyness (81.07%), community discouragement (70.39%) and negative attitudes of healthcare providers (70.39%) were major obstacles. Furthermore, financial problems (95.63%), long distances and transportation issues (75.24%) and the unavailability of services (47.57%) were identified as other significant challenges. However, long waiting times and infrastructural/material barriers were not hindrance to 30.58% of participants in seeking rehabilitation services for their CWDs (see Figure 1).

From Figure 2, the participants’ satisfaction level was assessed among the 31 families that successfully accessed the rehabilitation services their ward needed. The results showed that participants were largely dissatisfied with the rehabilitation services. They expressed significant dissatisfaction with both the direct costs (18 participants, 58.06%) and indirect costs (18 participants, 58.06%), as well as with the distance to the rehabilitation facility (18 participants, 58.06%), travel time (15 participants, 48.39%) and wait time to receive the service (15 participants, 48.39%). Many participants were also dissatisfied with their ability to get access to rehabilitation services when needed (14 participants, 45.16%) and with the physical accessibility of the rehabilitation centres (10 participants, 32.25%). However, nearly half of the participants were reserved about their overall satisfaction with the rehabilitation services (14 participants, 45.16%), the providers’ attitudes (14 participants, 45.16%) and the providers’ ability to explain the child’s condition to the caregiver (15 participants, 48.39%).

In Table 5, only the variables that showed statistical significance with knowledge of rehabilitation services are presented. The study found that participants living in urban areas were less likely to have knowledge of rehabilitation services for their CWD compared to those living in rural areas (odds ratio [OR] = 0.34, 95% CI: 0.12–0.96). Other family members were 4% less likely to have knowledge of rehabilitation services (crude odds ratio [COR] = 0.04, 95% CI: 0.00–0.70).

After adjusting for all other covariates, individuals living in urban areas were less likely (adjusted odds ratio [AOR] = 0.27, 95% CI: 0.12–0.58) to have knowledge of rehabilitation services compared to those living in rural areas. In addition, Christians were less likely (AOR = 0.45, 95% CI: 0.23–0.86) to have knowledge of rehabilitation services. Participants who have ever received rehabilitation services for their CWD were more likely (AOR = 2.79, 95% CI: 1.34–5.82) to have knowledge of rehabilitation services compared to those who have never received rehabilitation services for their CWD.

In Table 6, participants who strongly disagreed that community discouragement was a barrier to having access to rehabilitation services had increased odds (OR = 151.30, 95% CI: 1.90–12024.09) of accessing rehabilitation services for their CWD. Similarly, participants who strongly agreed that community discouragement were more likely (OR = 4.28, 95% CI: 1.20–15.33) to access rehabilitation services for their CWD compared to those who were undecided. Nonetheless, caregiver age, monthly income, sex, residence, employment status, financial problems, long-distance and transportation issues, infrastructural barriers and feelings of shyness were not statistically significant in relation to access to rehabilitation services.

After adjusting for other variables, families who had no formal education were less likely to have access to rehabilitation services for their CWD (AOR = 0.22, 95% CI: 0.07–0.59). In addition, participants who strongly disagreed that long waiting times were a barrier were also less likely (AOR = 0.30, 95% CI: 0.13–0.73) to access rehabilitation services for their CWD. Furthermore, participants who agreed or strongly disagreed with health personnel’s negative attitudes were less likely to have access to rehabilitation services for their CWD (AOR = 0.08, 95% CI: 0.00–1.01) and (AOR = 0.17, 95% CI: 0.05–0.62), respectively. Caregivers who strongly agreed that community discouragement was a barrier were more likely to access rehabilitation services for their CWD (AOR = 2.79, 95% CI: 1.25–6.19) compared to those who were undecided.