The impact of hospital-based and community based models of cerebral palsy rehabilitation: a quasi-experimental study

Cerebral palsy (CP) is the most common paediatric neurological condition [1] and the principal cause of disability in children globally [2]. It is defined as “a group of disorders of the development of movement and posture, causing activity limitation, that are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain. The motor disorders of cerebral palsy are often accompanied by disturbances of sensation, cognition, communication, perception, and/or behaviour, and/or by a seizure disorder” [3]. CP is a universal problem [2] with a global incidence of 2 to 3 cases per 1 000 births [4]. The exact prevalence in Zimbabwe is unknown: however, from extrapolated data, the incidence is similar and estimated at 1.55/1000 in rural areas and 3.3/1000 in urban areas [5].

Children with CP face multiple bio-psychosocial challenges [6],[7]. This coupled to the fact that CP is a lifetime condition [2],[3],[8], may result in a considerable burden on caregivers of children with severe impairments, affecting their health and health related quality of life [9],[10]. Rehabilitation treatment is an essential component [1],[11] of the multi-disciplinary approach required to address the problems of children with CP and their families [8],[12]. Researchers have not yet identified the most effective method of service delivery in terms of optimising the child’s potential and providing support to the caregiver, especially in low-income countries such as Zimbabwe. Issues such as accessibility and acceptability of services, compliance with training and efficacy of the intervention need to be considered when implementing any model of service delivery.

Different models of rehabilitation service delivery have been proposed in an attempt to provide affordable and appropriate support to people living with disabilities and these can be broadly classified as either institution–based (IB) or community based rehabilitation (CBR)/outreach-based (OR) approaches [13],[14]. The roots of CBR can be traced back to the Declaration of Alma-Atta which led to the adoption of the global primary health care strategy by the World Health Organization (WHO). The aim of CBR was to provide primary health care and rehabilitation services to people with disabilities within their communities [15]. CBR has been in existence for more than 3 decades [15]-[18] yet little is known about its efficacy, effectiveness, relevance, appropriateness and sustainability as a service delivery model and public health strategy [19]-[21].

Zimbabwe utilizes a hybrid model of provision of rehabilitation services that is a blend of hospital-based and community-based approaches which are provided at district, provincial and central hospitals [22]. Unfortunately, a decade of socio-economic meltdown has resulted in deterioration of the health care delivery system [23]. At present, organization of rehabilitation services varies from institution to institution and is mainly governed by resource availability. Most institutions are now offering hospital-based services only. For instance, out of the six state central hospitals in Zimbabwe, only Harare Central Hospital (HCH) is at present running a consistent outreach program through its Children Rehabilitation Unit (CRU) [22]. The CRU is a specialized paediatric rehabilitation centre, which, for more than twenty years, has run a peri-urban, community-based outreach programme based on the WHO CBR model. Children and their care-givers (mostly mothers) gather in groups twice a month in community centres. Children receive some individual treatment from therapists and/or rehabilitation technicians (who have undergone two years of training). In addition, there are group activities and education sessions. In contrast, children in another high-density area of Harare, which is serviced by a different hospital, receive regular physiotherapy within an institutional out-patient setting. As the outreach programme relies on a certain amount of donor funding, it is somewhat more expensive to run [22].

There was a clear need to compare the two models of service delivery, not only to inform the on-going re-structuring of rehabilitation services in Zimbabwe [23], but to provide empirical evidence of the relative impact of CBR/outreach services as compared to institutionally-based rehabilitation [11].

The objectives of the study were therefore to compare the impact of the outreach (OR) and the institutionally based (IB) programs in terms of their impact on the children’s functioning, the strain on their caregivers, compliance with scheduled appointments and the overall satisfaction with the services received. It was anticipated that there would be little difference in the functional change between the two groups. The greater group interaction and support were expected to result in a greater decrease in the strain of the care-givers attending the outreach group. The satisfaction with services was expected to be greater in the outreach group as the service was brought to them and they did not need to travel far to get support for their children.

A quasi-experimental design was used as it was a pragmatic trial and it was not possible to randomly assign children to one group or the other. The geographical location determines a child’s program allocation as the two areas are some distance apart, children and caregivers were thus obliged to attend one or other programme depending on their place of residence. A sample of convenience was drawn from the children treated under the OR program and IB CP clinics who attended the clinics during the first four weeks of the study.

The children had to have received a diagnosis of CP according to their patient notes. They were to be between 0.5 to 12 years of age as the Gross Motor Function Measure (GMFM) has good content and face validity for children in the age range 0.5 to 13 years [24] and the discharge age for the CRU Outreach program is 12 years. No participant was recruited if they were scheduled for surgery or if they had any other significant medical and nutritional problems or other clinical factors that might have biased the rehabilitation program [25],[26]. Children who had co-morbid neurological conditions e.g. Spinal Bifida or who were receiving other forms of therapeutic interventions such as private physiotherapy were similarly excluded.

The burden of care as measured by the Caregiver Strain Index (CSI) [27] was one of the major variables under scrutiny. Assuming mean CSI scores of 7 and 9 (SD = 2) for both groups [28] at the conclusion of the study period, the expected minimal number of cases per group was 16 (alpha = .05, power = 95%). Oversampling was done to counteract effects of attrition due to e.g. sickness and non-compliance.

The objectives of the study were to compare the impact of the outreach (OR) and the institutionally based (IB) programs in terms of their impact on the children’s functioning, the strain on their caregivers, compliance with scheduled appointments and the overall satisfaction with the services received. It should be noted that the “entire package”, which encompassed the location (community or institutional based), the increased training and experience of the OR personnel, the structure of the therapy sessions and the provision of refreshments during the OR sessions were compared. It was not possible, using this research design, to isolate which components of the programmes resulted in the differences seen. The results of the study indicate that in several respects the OR programme was superior to the IB programme. The sample appeared to be representative of children with CP in that the majority had spastic type CP (80%) which is the most common variant of CP as reportedly accounts for 80–83% of cases [11],[35]-[37]. The spread across the different GMFCS levels was similar to a large scale study in Canada, which reported 42% of the children were severely affected (Levels IV and V) compared to 44% in this study [31]. The predominance of males in the sample has also been reported in other samples of children with CP [6],[38],[39]. It would therefor appear that the children in this study were representative of most samples of children with CP.

A problem with quasi-experimental studies is that there may be confounding variables that may bias the results of the study. In this case, there were no differences found between the participants in the OR group and the IB group in terms of demography or nature of their impairment. The differences that were noted, that the OR children were older and more were severely affected (although not statistically so), would have introduced bias into a randomised trial. However, in this pragmatic trial it was an indication of the strength of the OR intervention in that older, more severely affected children were still being brought in for treatment. It has been reported that older and more severely affected children might respond less to interventions, [12]. It was thus necessary to control for these factors by doing regression analysis. Children in the OR group showed greater improvement and several factors can account for this difference. Firstly, some of the rehabilitation workers in the OR group are based in a specialist unit and have developed skills in child treatment whereas the IB rehabilitation professionals are responsible for treatment across a wider spectrum of conditions and ages. Secondly, the lower child to therapist ratio in the OR group ensures ample time for treatment and demonstration of techniques to caregivers. Thirdly, continuity of care in the OR group,and the inherent good therapist-child relationship may have led to increase in-treatment adherence and this may have enhanced treatment efficacy [40].

The situation in Zimbabwe is typical of a resource constrained country in that children with severe CP are not necessarily catered for within institutions or special schools. The response to therapy might be different in children who have had on-going intensive rehabilitation within specialised centres. In addition, parents who have had access to sophisticated services may not demonstrate the same degree of satisfaction with the type of service provided. The results of the study may therefore only be of relevance to low and middle income countries.

As the children in the community based group were older and higher proportions were in GMFCS Levels IV and V, their caregivers might be expected to report greater strain. This was not the case, which might indicate that the outreach based intervention mitigated the impact of severity and chronicity of care to a certain extent. This hypothesis however, needs to be empirically tested. Alternatively, it may be that parents of younger children are in earlier phases of ‘grief’ in response to having a child with a disability which may dissipate over time to some extent [41]. It is clear that the care-givers are in need of additional support, particularly financial and emotional as there are no disability grants in Zimbabwe.

Caregivers in the OR group seemed to be more satisfied with services and were more compliant as compared to recipients of IB services. It is essential to evaluate patient satisfaction with services delivery as satisfaction is related to treatment compliance and outcomes [33],[42]. Services in the OR group were provided every fortnight and this could have enhanced satisfaction and compliance with services. Furthermore, consistent booking schedules have been demonstrated to affect the levels of compliance and satisfaction with services [43]-[45]. Additionally, provision of services within the recipients’ communities, a more natural environment, negates the need for transportation costs and adapted transportation (which may not be available in low resourced settings). It also results in an increased amount of social support. All of which have been identified cited as enhancers to satisfaction and compliance [46]-[50].

Lack of knowledge of the child’s impairment can lead to non-compliance [51]. As caregivers in the OR group would have attended CP workshops prior to joining the outreach group, it was expected that they would have been more knowledgeable about CP. This might have enhanced compliance and satisfaction with the explanations given by therapists on CP and its treatment. This hypothesis was not tested however and a weakness of the study was that the amount of information that the caregivers had regarding CP was not compared between the groups.

The IB group had a higher patient to therapist ratio; this inherently leads to time pressure during treatment sessions. Time pressure may lead to decreased in-treatment adherence, less satisfaction with explanations and therapy given. This may have accounted for the lower compliance and satisfaction in the IB group. The nature of the patient/practitioner relationship also affects the extent of compliance and satisfaction [51]. Further, the absence of continuity of care in the IB group, might also have accounted for lower rate of compliance and satisfaction.

Research on the effect of treatment frequency has yielded inconclusive evidence [12],[40],[43]. However, a study by Christiansen & Lange, [43] suggests that intermittent frequency is equally efficacious when compared to continuous dosage. Therefore, evidence from our findings suggests that a two week gap may be tolerable for caregivers and may result in equal gains in functional outcome.

Results from this study need to be interpreted with caution as children and caregivers had had interventions for varying lengths of time and changes in the outcome measures might have taken place prior to the study. Secondly, methodological weakness of the study design and the lack of randomisation and blinding of the assessor may limit the generalizability of our findings. Thirdly, there was a difference in expertise and clinical experience for therapists in the comparison groups with those in the OR arm more experienced and this could have introduced bias.

In conclusion, long term caregiving leads to strain in caregivers and there is a need to design individualized interventions to alleviate the burden on caregivers as it may ultimately affect the child’s functional prognosis and health outcomes. The study was a pragmatic, quasi-experimental study, which by its nature cannot lead to causal inference. However findings from this study suggest that the provision of care within a community setting as part of a well-structured OR programme may be preferable. It was associated with a greater improvement in functioning, greater satisfaction with services and better compliance. In addition, care-givers continued to bring in older children for therapy, which was encouraging. It is therefore suggested that may be the preferred method of service delivery. Further research is needed however, to cost the methods of service delivery in order to determine the cost of transferring the management of children with CP from institutions to the community.