Bloed, zweet en tranen...

Lieve Ilse,

Bloed, zweet en tranen... Dat heeft het ons absoluut gekost om af te studeren! Maar ik ben onzettend trots op het resultaat en ik had deze opdracht met niemand anders willen volbrengen!
Het wordt gelukkig nog gepubliceerd in het boek van Adri, maar ik vind dat het ook hier niet in mag ontbreken...

Zie hier het eindresultaat:

Differential Effects of Deprivation on Functional Development of Children with a Developmental Disability in Rural South Africa.

I. Kroon, Department of Clinical Psychology, Utrecht University, Utrecht, the Netherlands.

N.K. Sinnaeve, Department of Health Psychology, Utrecht University, Utrecht, the Netherlands.


Abstract
The present study aimed to investigate the differential effects of deprivation on functional development of children with disabilities in rural South Africa. Some participants (N=17) were living in a residential facility, Sizanani Children’s Home (institutional group), and others (N=20) were living in their community environment (community group) after they were discharged, because of a period of turbulence in that Home in December 2006. A modified version of the HOME-Inventory was used to assess the quality of environment. The total score on the HOME was used to split the community group into two different groups, i.e. participants living in severe hardship (N=12) and participants living in relatively better life circumstances (N=8). The functional development was assessed with the Pediatric Evaluation of Disability Inventory (PEDI). Significant results showed that the consequences of the period of turbulence at Sizanani Children’s Home indirectly negatively effected the functional development of all participants. Although, the participants in the community group showed a greater decline in their functional development compared to the participants in the institutional group, this was not significant. In contrast to the expectation, within the community group the participants living in relatively better life circumstances had a greater decline in their functional development in comparison with the participants living in severe hardship. These results were not significant. Altogether, these results might suggest that the level of deprivation could influence the functional development of children with a disability.

Keywords Children, Deprivation, Disability, Functional Development, Post-institutionalization, Rural South Africa.


Introduction
The Individuals with Disabilities Education Act (IDEA) (2000) states that children with a developmental disability are permanently, or for an extended period, experiencing delays in cognitive, physical, communicative, social or emotional, or adaptive development or they have a diagnosed physical or mental condition that has a high probability of resulting in developmental delay (in Malone, McKinsey, Thyer & Straka, 2000). They need a high level of constant care to maintain comfort, sustain life, or attend to a bodily function that the children cannot manage themselves. The prevalence of developmental disabilities and associated epidemiological factors has been well documented in industrialized countries. Conversely, only limited information is currently available from developing nations. Statistics South Africa has published a document in 2005, where it is estimated that in South Africa 2.3 million people were living with a certain form of disability. This number constituted 5% of the total sample. The prevalence of sight disability was the highest (32%), followed by physical disability (30%), hearing disability (20%), emotional disability (16%), intellectual disability (12%) and communication disability (7%) (Statistics South Africa, 2005).

It is likely that the extent of the disability and additional problems associated with it contributes to children’s vulnerability. A range of definitions has been used for describing vulnerability in children across a number of African countries. Smart (2003) formulated a definition of vulnerable children from South Africa. He included those children who are neglected, destitute or abandoned, those living with terminally ill parents, those born to single mothers, those living with unemployed caretakers and those who are abused or ill-treated by caretakers (in Skinner et al., 2006). Having a disability on top of this will exacerbate their vulnerability.

In South Africa children with disabilities are slowly being viewed as respected human beings, who should be able to participate in mainstream activities on an equal ground. The Republic of South Africa (1996) has enfranchised the rights of people with disabilities in the country’s Constitution (in Christianson et al., 2002). Still a person with a disability is someone who has little or no access to basic needs or rights. But people with disabilities need to learn to cope with and to be creative in any situation in order to find a solution to a problem and to enhance their independence, so that they would not remain dependent and part of the burden on the household. Turmusani (2003) acknowledged that in reality children with a disability often get abandoned by their families and need to be taken care of in residential facilities (in Magyarszeky & Mbethe, 2005).

In order to take care of disabled and vulnerable children, in 1989 an Austrian Priest, Father Charles Kuppelwieser, founded a centre for children with neurological disorders in Bronkhorstspruit, South Africa. Sizanani (‘help each other’) Village opened its doors in March 1993. Over the years Sizanani Village expanded to a HIV Community Centre, a school for learners with special educational needs, a home for physically and mentally abused women and their children, farming projects, a conference centre and a home for children with disabilities, Sizanani Children’s Home (Broek, van den & Janssen, 2007; Magyarszeky & Mbethe, 2005; Tamboer & Groenhuijzen 2003; Vermeer, Wijnroks, Magyarszeky & Mbethe, 2005).
Since the establishment of Sizanani Children’s Home its core function has been nursing care, because most of the parents (or other caregivers) of children with a disability were not able to take care of their child in a financial way as well as they suffered from mental pressure. The childcare workers were especially recruited for taking care of the children 24 hours a day, including feeding, bathing, dressing, changing diapers and any other activities related to the physical wellbeing of the children.
From around 2001 the Utrecht University in the Netherlands started with several research projects at Sizanani Children’s Home. The study by Tamboer and Groenhuijzen (2003) concluded that the introduction of a developmental programme would be very beneficial for the entire organization. Therefore the management of Sizanani Children’s Home reconsidered their policy, which resulted in a shift from a nursing approach towards a more developmental approach, and introduced the Conductive Education (CE) programme (Magyarszeky & Mbethe, 2005; Tamboer & Groenhuijzen, 2003). Conductive Education is a developmental stimulation programme for children with motor disorders resulting from cerebral palsy and other brain disorders.
After the introduction of a pilot study to examine the effects of CE it was chosen to implement the basic principles of CE throughout Sizanani Children’s Home, adapting it in accordance with the varying needs of the children (Vermeer, Wijnroks, Magyarszeky & Mbethe, 2006).
The implementation of CE in all the units of Sizanani Children’s Home commenced in 2005 and the effects of CE were assessed again in a longitudinal study in a larger sample, using a multiple baseline design. At the end, five baseline measurements were conducted.
Around the end of 2006, however, a turbulent crisis at Sizanani Village emerged and the research on the effects of CE was stopped. More than hundred children, including thirty-two children of the representative sample under study, were discharged in January 2007 and send back to their families living in rural community settings. Moreover, the Trust that operated the Home was dissolved. Only fifty-six orphaned, abandoned and severely neglected children who had nowhere else to go stayed in Sizanani Children’s Home, of which eighteen were from the original research group. When labour relations deteriorated and strikes broke out, CE was terminated in the entire Home. The following months had been spent stabilizing the organization by ensuring that it continued to provide quality care and attempting to find a permanent future for the Home and the children under its care.
Besides, a new project to evaluate the need of an outreach programme on the children who are living in rural community settings is planned to be established. This programme will provide follow-up care of the discharged children of Sizanani Children’s Home.

Due to the period of turbulence at Sizanani Children’s Home in December 2006, an unique situation emerged for investigating the possible influence of changing environmental circumstances on the development of children who are disabled. In the first place, CE was terminated for all children. In addition, for children who had been placed into the communities, their life situation changed from living in an institution with sufficient nutrition, medical support and adequate stimulation of their development by CE, into a situation of living in the communities by their family where the majority will experience no adequate life support. That is, the environment for these children would change from a caring surrounding, perhaps a relatively enriched environment, to an environment characterized by possible severe deprivation, at least in most of the cases. Moreover, many children were placed in Sizanani Children’s Home because their caregivers were not able to take care for them. It is likely that changes in the functional development of children who were discharged from Sizanani Children’s Home, will depend on the quality of the home environment. In addition, the differences in functional development within the children who were discharged from Sizanani Children’s Home, are likely to be related to the degree of severity of the community environment. Higgs (2007) conducted a study on the everyday quality of life in South Africa, and the author regarded a score of 50 as the poverty line in the distribution of poverty scores. At a score of 50, access to basic amenities falls precipitously, basic household durables are lacking and proper nutrition cannot be afforded. Amongst black people, 49% lie above the partition of 50, 32% score 60 or more and 10% above 80, suffering hardship and deprivation of the most severe kind. According to Higgs people living in rural areas fare the worst.

Because of the lack of therapeutic stimulation as well as healthy nutrition and other aspects of quality of life that poverty provokes, a possible factor which could influence the functional development of the discharged children would be a poor quality of the home environment. Christianson et al. (2002) have observed an association between typical socio-economically deprived rural areas in South Africa and a high rate of mild intellectual disability. Factors observed in this study included poor living conditions, malnutrition and limited intellectual stimulation of infants and children. It was proposed that for many of the children with mild intellectual disabilities with an undefined aetiology, their level of functioning could have been the result, or aggravated by their environment.
Furthermore, it can be assumed that providing a high level of care to a child with long-term functional limitations can become burdensome and may impact upon both the physical and psychological health of the caregiver and the child. Evidence suggests that there is considerable variation in how parents adapt to this care giving demand (McKinney & Peterson, 1987; Sloper & Turner, 1993; Wallander, Varni, Babani, DeHaan, Wilcox & Banis, 1989). Some adapt well to challenges of caring for a child with a disability, while others do not. Cadman, Rosenbaum, Boyle & Offord (1991) found that parents of children with disabilities were more likely to experience depression and distress than parents of children without disabilities. Also the setting in which care giving takes place and the level of disability are likely to be correlated with exposure to care related stressors. According to Eicher and Batshaw (1992), socioeconomic resources may help contain the extent to which a caregiver’s condition become burdensome.
To date, many studies have only examined the effects of early deprivation on the development of adopted children who have been in impoverished institutional care in Romania during the first years of their life and due to adoption by families living in developed countries their situation changed to much better life circumstances. Romanian institutions were characterized, according to Wilson (2003), by a lack of adequate nutrition, minimal interpersonal contact and little to no sensory stimulation. It has been generally noted in this study that children adopted before six months develop normally in social and cognitive domains. However, children who stayed longer in institutional care had less certain outcomes. Another study by O'Connor, Rutter and Michael (2000) examined a large, systematically ascertained sample of children exposed to severe early deprivation in Romania before being adopted into the United Kingdom. They found evidence to suggest that the duration of institutionalized care accounted for the attachment dirsorder with those children. It remains to be seen whether the results reported by Wilson also apply to children with disabilities. Furthermore, it should be noted that the studies of O’Connor et al. and Wilson were conducted in Europe, thus it remains unclear what the effects are of post-institutionalization from qualitative high standard institutional care to an impoverished environmental surrounding in rural South Africa for children with disabilities.
The aim of the present study was to investigate the possible changes in functional development of children with disabilities within the temporary period after the crisis in Sizanani Children’s Home. This study is part of a broader intervention study in which the effects of the centre-based CE-programme of the Home will be conducted.
The current level of functioning of all children will be compared with the level of functioning in 2005. During this period in 2005 and the most of 2006 the children were stimulated with CE, but was then terminated in Sizanani Children’s Home. So, when the functional development of the children in Sizanani Children’s Home would have been improved over this time period, it could imply the effectiveness of CE. Besides, the measurements in the communities would imply crucial information to formulate goals of the outreach programme, for which implementation is planned at the beginning of 2008. Moreover, the outreach programme is aimed to establish own community-based centers and develop effective cooperation with existing community-based organizations (Magyarszeky, 2007).

In this study the following general research question was posed:

Did the period of turbulence in December 2006 at Sizanani Children’s Home, and discharging some children into their community environment, negatively effect the functional development of the children?

In line with this general research question the following additional hypotheses were:

1) Between December 2006 and one year later, the children will show a decline in their functional development.

2) The children who were send back to their community environment will show a greater decline in their functional development in comparison with the children who stayed in the institutional environment.

3) Within the community group, the decline in functional development will be greatest for children living in severe hardship in comparison with children living in relatively better life circumstances.


Methods
Population and sample
At the start of the implementation of Conductive Education in Sizanani Children’s Home in 2005, fifty-five participants were randomly selected from the total population of hundred-and-seventy children and young adults to participate in a longitudinal study to assess the effects of CE. In January 2007 more than hundred children, including thirty-two children of the representative sample, were discharged from Sizanani Children’s Home and the longitudinal study was stopped, as was the implementation of CE.
The sample of the present study was part of the representative sample of fifty-five participants of the longitudinal study. Eighteen participants were still living at Sizanani Children’s Home and thirty-two participants were living in surrounding communities within a 300 miles circumference of Sizanani Village. The other five children of the original research group passed away or were excluded due to fractures or other injuries. Accordingly, two groups were formed: i.e. (i) the children living in Sizanani Children’s Home (institutional group) and (ii) the children living in their communities (community group). Only twenty participants of the thirty-two who were living in their communities participated in the present study. The missing twelve participants were excluded for different reasons: one child passed away, two families were too angry at Sizanani Children’s Home and refused to co-operate, four children could not be traced because any contact details were missing, three children were currently living in another residential facility and two children were living too far away from Sizanani Children’s Home.

So, the final sample of the present study consisted of thirty-eight children and young adults with disabilities, of which nineteen were male (50%). The institutional group consisted of ten male participants (56%) and eight female participants (44%) and there were nine male (45%) and eleven female participants (55%) in the community group. As presented in Table 1, statistical comparisons revealed that the research groups were comparable in terms of gender (2(1)=-.106, p=.74).
The age of the participants of the total sample ranged from 7 years to 28.6 years, with an average of 17.97 years (SD=15.32). Although the mean age of the male participants (M=19.64, SD=5.47) was higher than the mean age of the female participants (M=16.30, SD=4.73), this difference tended towards significance (t(38)=2.017, p=.051). As presented in Table 1, the institutional group had a mean age of 17.75 years (SD=6.42) and the average age of the participants in the community group was 18.15 years (SD=4.40), which was not significant (t(38)=-.228, p=.821).

The most common disabilities among the participants were cerebral palsy 50%, cerebral palsy with spasm 15.8%, 13.2% hydrocephalus, 7.9% cerebral palsy with hydrocephalus, 13.2% did not fit into one of above mentioned classes of disability. Additionally, the total sample included 5 participants with a mild/moderate disability (13.5%), nineteen participants with a severe disability (51.4%) and thirteen participants with a profound disability (35.1%). The institutional group contained one participant with a mild disability (5.9%), seven participants with a severe disability (41.2%) and nine participants with a profound disability (52.9%). The community group consisted of four participants with a mild disability (20%), twelve participants with a severe disability (60%) and four participants with a profound disability (20%). Pearson’s chi-square test revealed that the research groups were not comparable in terms of level of disability (2(4)=9.48, p=.050), as presented in Table 1.


Design
This part of the planned longitudinal research consisted originally of five baseline measurements and one intervention measurement within one single group. Because of a delay in the implementation of CE and a lack of a control group, instead of three planned baseline measurements there were five baseline measurements conducted. Due to the period of turbulence at Sizanani Children’s Home, the design of the present study changed into a quasi-experimental design and the single group has been split into two different groups: i.e. (i) participants living in Sizanani Children’s Home (institutional group) and (ii) participants living in the communities (community group). For that reason, an unique situation emerged for investigating the influence of changing environmental circumstances from a caring surrounding (institutional group) to possible deprivation (community group), as presented in Figure 1. The participants in the community group run a risk of getting into a situation of deprivation.
In the previous study on the implementation of CE (Vermeer et al., 2005), a representative sample of fifty-five children was assessed by means of five baseline measurements. Taking the completeness of the data-set into consideration, data from the fifth baseline measurement (T5, May 2006) was used as the pre-test measurement for the present study.
It is theoretically interesting to investigate the influence of the quality of environment and its effects of possible deprivation on functional development. “The types of experiences children are likely to have, the manner in which these experiences are likely to be perceived by children, and the impact they are likely to have on various aspects of development are partly determined by cultural, economic, community, and familial factors” (Bradley, Caldwell, Brisby, Magee & Whiteside, 1992, pp. 314). Moreover, children are active both in producing their own environments and in eliciting things from the environment (Bradley et al., 1992). In order to ascribe an effect of the possible deprivation, participants in the community group should show a decline in their functional development in comparison with the data of the fifth baseline measurement. In addition, within the community group the participants who are living in severe hardship should have a greater decline in their functional development compared with the participants living in relatively better life circumstances.

Instruments
Pediatric Evaluation of Disability Inventory
The level of functional abilities can be assessed by using the Pediatric Evaluation of Disability Inventory (PEDI). The PEDI can be used with parents or a caregiver who knows the child well, for evaluation and discrimination of the level of functional ability of children with an age range from 6 months to 7.5 years, or older children with severe disabilities in daily functioning (Vermeer et al. 2005). According to the PEDI, ‘disability’ can be seen as an increased amount of help the child needs from the caregiver during daily activities compared to other children of the same age. Therefore, also the amount of assistance needed from the caregiver can be measured using the PEDI (Custers, Net, van der & Helders, 1997).
The PEDI is a reliable and valid structured interview measuring the level of functional abilities of the child (Custers et al., 1997; Vermeer et al., 2005). The inventory is developed to confirm a child’s functional abilities concerning three scales; (1) functional skills scale (the level of functional skills), (2) caregiver assistance scale (the physical assistance received from the caregiver) and (3) modification section (adaptations or use of objects, such as braces or wheelchairs). Each scale is divided into three domains; i.e. self-care, mobility and social functioning and all questions can be answered with ‘able’ or ‘unable’ (Ketelaar, Vermeer & Helders, 1998). The first scale, the ‘functional skills’ scale consists of questions concerning the typical functional skill level of the child. For example, questions about the self-care within the ‘functional skills’ scale are: “Child opens mouth for teeth to be brushed” and “Child obtains soap when washing hands”. Questions about mobility within the ‘functional skills’ scale are: “Child raises to sitting position in bed or crib” and “Child gets in and out of own bed, not needing own arms”. “Child tells about own feelings and thoughts” and “Child shows awareness around hot objects” are examples of questions of social functioning within the ‘functional skills’ scale.
The second scale, ‘caregiver assistance’ scale, consists of keywords concerning the physical assistance the children need to carry out their functional activities. These questions can be answered on a six-point scale, ranging from ‘independent’ to ‘total assistance’. In the self-care domain within the ‘caregiver assistance’ scale an example of keywords are “brushing teeth and caring for nose” and “clothes, toilet management or external device use, and hygiene; do not include toilet transfers, monitoring schedule, or cleaning up after accidents”. “Child's wheelchair, adult sized chair, adult sized toilet” and “getting in and out and changing positions in child's own bed” are examples of keywords of the mobility domain within the ‘caregiver assistance’ scale. The social functioning domain of this scale consists of the following questions: “understanding of request and instructions” and “ability to provide information about own activities and make own needs known, include clarity of articulation”.
The third scale has not been used in this study because of a lack of specialized equipment in Sizanani Children’s Home (Broek, van den, et al., 2007; Custers, 2001). This scale could be left out without affecting the other scales, because each PEDI scale is self-contained. Further, some items were deleted because they weren’t applicable for Sizanani Children’s Home. For example, items on the domain of hair brushing were deleted because all children had very short hair. Unfortunately, new measurements to determine the reliability weren’t accomplished again (Broek, van den et al., 2007). The raw scores of the ‘functional skills’ scale and the ‘caregiver assistance’ scale are frequency counts of the functional abilities the child has mastered. They can be converted into scaled scores. These scaled scores are distributed along the scale from zero to 100. Increasing numbers on the ‘functional skills’ scale represents increasing degrees of functional performance. They provide an indication of the performance of the child along the continuum from relatively easy to relatively difficult items in a particular domain on the PEDI. The scaled scores of the ‘caregiver assistance’ scale reflect low capability near zero and high capability for scores near 100 (Haley, Coster, Ludlow, Haltiwanger & Andrellos, 1992). A higher scaled score represented a higher level of functional performance and respectively a higher level of independence (Vermeer et al., 2005).
The present study only applied scaled scores, because they are not adjusted for age and can therefore be used to describe the functional status of children and young adults with disabled conditions (Custers, 2001). As the PEDI was used as an evaluative instrument, the scores of the participants were not compared with norm groups, only compared over time (Custers, 2001; Custers et al., 1997).

Home Observation for Measurement of the Environment Inventory
The Home Observation for Measurement of the Environment Inventory (HOME) was used to measure the quality and quantity of stimulation and support available to a child in the home environment (Bradley et al., 1992). The HOME was in its current form not suitable for the present study. For this reason, only four domains were used: i.e. (1) emotional climate, (2) family participation, (3) opportunities for variety and (4) quality of the environment. Each domain consists of four questions. “Caregiver spontaneously vocalized to the child twice during the visit” is an example of a question within the ‘emotional climate’ domain, whereas “Caregiver has a fairly and predictable daily schedule for the child (e.g.: meals, bedtime)” is an example of a question in the ‘family participation’ domain. One of the questions to measure ‘opportunities for variety’ was “Child gets out of house at least two times per week” and “There is running water available in the house” was one of the questions within the ‘quality of environment’ domain.
The questions in the first three domains were mainly based on the original questions from different versions of the HOME. Small adaptations were fulfilled to make the instrument more appropriate for children with a disability in rural South Africa. The questions in the fourth domain were based on the study by Higgs (2007) on the quality of life in South Africa.
The procedure of this instrument is a low-key semi-structured observation and interview, in which a binary-choice (‘yes’ or ‘no’) format is used. The total score on this instrument will be used to determine which children are living in severe hardship and who are living in better life circumstances.
Statistical analyses revealed that some inter-item-correlations were weak or very low, sometimes even negative. After examining the mutual variances, six questions were deleted. The reliability in terms of internal consistency (a=.58) of the revised HOME was considered to be acceptable.

Procedure
All measurements of this study were conducted over a period of three months, from September until December 2007. The first two weeks of this period were scheduled to learn more about the everyday life in Sizanani Children’s Home, this included working in the units and getting acquainted with the childcare workers and the children. The researchers helped in the different units during the daily activities as bathing and feeding and spent extra time with the children in their leisure time.
Since this research was part of the longitudinal study on the effects of CE, also other assessments were done. These were the Functional Motor Assessment Scale for measuring the functional motor abilities of children with cerebral palsy and the Levels of Intervention Observation Instrument for assessing the interactions between the caregiver and the child.
All assessments were done in English, sometimes translated from Zulu or Sotho by a childcare worker. The measurements started first with a couple of trial assessments in the different units of Sizanani Children’s Home. Then the measurements for the other children started. Every instrument was assessed once per child.

Measurements in Sizanani Children’s Home (institutional group)
During each measurement within the units of Sizanani Children’s Home the same set of instruments has been executed.
The first measurements were conducted in the unit where the most profound male participants were living. During the break of the unit manager/ childcare workers in the morning and the afternoon the PEDI was administrated. Since the assessment of the PEDI is very time consuming, the maximum amount of interviews was two per day per childcare worker. In this way, the reliability of the answers of the childcare worker was protected.
The HOME was administered with the unit manager. All questions applied to the participants within the unit, except the questions in the ‘opportunities for variety’ domain.
The same procedure was followed in the other operating units.

Measurements in the Communities (community group)
The assessments in the communities took either place when there was a physical appointment arranged with the family, or when no measurements in Sizanani Children’s Home could be assessed. At first the social worker of Sizanani Children’s Home tried to contact the caregivers of the children who were discharged from the Home. When she was able to contact any of those caregivers, she organized an appointment on a fixed date. Then the researchers tried to combine the set date with other community visits within the same area. During each community visit at least one childcare worker joined the researchers for translating the different assessments as well as safety guidance. Before the assessments were conducted, informed consent with the caregiver of the participant was acquired. The caregiver with whom the assessments were conducted was the one that took daily care of the participant. During each measurement within the communities the same set of instruments as in Sizanani Children’s Home has been executed.
At the first few community visits, the different instruments were successively strictly assessed. As there was more routine in the assessments, the interviews went naturally and lost its strict element. Since the caregivers were more comfortable they contributed valuable information, which would not be the case otherwise.

Data analysis
The Statistical Package for Social Sciences (SPSS, version 14.0) was used for computing descriptive statistics, correlations and carrying out t-test. The alpha was set at 5% for all analyses.

Before starting the actual assessments the inter-observant reliability for the PEDI and the HOME-Inventory was administered. The assessments were done in synchrony by the two researchers. The inter-observer reliability was calculated by using Pearson’s R. The level of agreement between the researchers on the assessments has proven to be good (.96 < r < .99).
Since this research was part of a longitudinal study, missing values were common due to changes in the research group. In the analyses of this research the missing data were excluded pair wise, the cases were only excluded if they were missing the data required for the specific analysis.

A score for the quality of the home environment was constructed by calculating the sum score of each child on the revised HOME-Inventory. The median of this sum score was used to split the quality of environment into two groups. As a consequence, quality of environment was dichotomized into severe hardship and better life circumstances in subsequent analyses. A subgroup was discerned with 12 participants (60%) living in severe hardship and 8 participants (40%) living in relatively better life circumstances.
Also a sum score of the PEDI was constructed and was used to measure the functional development over time, this score was called ‘PEDItotalT’. However, the PEDI was not assessed for participants who functioned at an extremely low level. In that case they received a zero sum score on the PEDI. Eight participants in Sizanani Children’s Home and three participants in the communities received zero as score on the PEDI.
A Kolmogorov-Smirnov (K-S) One-Sample Test was conducted, to analyze whether the distribution of the data of the present study was normal and homegeneous. The results showed that the data satisfied this requirement (Z=.420, p=.994). Also, the data of the present study were measured on interval level and had equal variances. Besides, a Kolmogorov-Smirnov (K-S) One-Sample Test was again conducted, to analyze whether the distribution of the dependent variable within the community group was normal and homegeneous. The results showed that the dependent variable satisfied this requirement (Z=.78, p=.577).

To examine the first hypothesis whether all children will show a decline in their functional development between December 2006 and one year later, a paired t-test was carried out.
The expectation, as formulated in the second hypothesis, was that the participants living in the community group will show a greater decline in functional development in comparison with the participants who stayed in the institutional group. First, differences in the level of functioning between the two groups at the pre-test measurement (T5) were analyzed with a t-test for independent samples. A t-test for independent samples was also used to analyze changes (mean subgroup difference on the scores between T5 and T7) in functional development of the participants in the two groups over time. This score of difference was also used in subsequent analyses.
In order to examine the third hypothesis that within the community group the decline in level of functioning of the children who were living in severe hardship will be greater than the decline in level of functioning of the children who are living in relatively better life circumstances, another series of independent t-tests were carried out. Differences in the level of functioning between the two groups within the community group at the pre-test measurement (T5) were analyzed with a t-test for independent samples. Then again, with the score of difference a t-test for independent samples was used to analyze the changes in functional development of the participants in the two groups over time.


Results
In order to investigate the first hypothesis that all participants of the present study showed a decline in their functional development over time, a paired t-test for all participants with their level of functioning at T5 and T7 was conducted. Statistical comparisons revealed that at T5 the mean score of level of functioning of all participants was 204.98 (SD=180.14) and at T7 191.60 (SD=166.0). The mean scores of level of functioning showed a decline over time, the difference between these mean scores was significant (t(35)=1.705, p=.048, one-tailed). This means that this hypothesis was confirmed.

The second hypothesis, that the decline in functional development was greater for the participants in the community group than for the participants in the institutional group, was examined by analyzing the difference in the mean of functional development between T5 and T7 for both groups. First, it was necessary to measure if the groups were already significantly different on their level of functioning at the pre-test measurement (T5). At T5, the mean level of functioning of the participants in the community group and the mean level of functioning of the participants in Sizanani Children’s Home were respectively 255.10 (SD=176.16) and 142.34 (SD=169.91). The mean difference between the two groups was 112.76 and the 95% confidence interval for the estimated sample mean difference was between 230.98 and 5.46. An independent t-test showed that the difference between the two groups was not significant (t(34)=1.94, p=.061, two-tailed). In subsequent analyses the score of difference in level of functioning between T7 and T5 was computed and used. Although, the community group as expected and as presented in Table 2, demonstrated a mean score of decline in functional development of -18.98 (SD=56.62) and the participants in the institutional group had a mean score of decline of -6.39 (SD=31.98), an independent t-test showed that the difference between the two groups was not significant (t(34)=.79, p=.21, one-tailed).

In order to examine the third hypothesis whether within the community group the decline in functional development of the participants who were living in severe hardship was greater than the decline in functional development of the participants in relatively better life circumstances, independent t-tests were conducted. As presented in Table 3, results showed that in comparison with children living in severe hardship (M=207.85, SD=153.95), children who were living in better life circumstances (M=351.94, SD=163.88) displayed higher scores on their level of functioning at T5. Surprisingly, in contrast to the expectation, the mean decline in functional development of the participants living in better life circumstances was -45.75 (SD=63.60) and the mean decline in functional development of the participants living in severe hardship was -1.23 (SD=48.01). However, the difference was not statistically significant (t(17)=1.74, p=.099, two-tailed).


Discussion
After a temporary period of turbulence in Sizanani Children’s Home, the present study aimed to investigate the possible changes in functional development of children with disabilities living there and those living in their community. Results showed that the consequences of this period of turbulence in December 2006 had a negative effect on the functional development of the children. The overall decline in functional development showed a trend towards significance. The difference between the decline in functional development of the participants in the community group and the decline in functional development of the participants who stayed behind in Sizanani Children’s Home was in the expected direction and the statistical analysis revealed that this difference was significant. In contrast to the expectation, within the community group the participants living in better life circumstances showed a greater decline in their functional development in comparison with the participants living in severe hardship. This difference tended towards significance.

With regard to the first hypothesis that all children showed a decline in their functional development between December 2006 and one year later, the statistical analyses revealed that the results were in the expected direction. It is certain that there was a decline in functional development of the children from Sizanani Children’s Home. The question remains whether the consequences of the period of turbulence at Sizanani Children’s Home in December 2006 accounted for this decline in functional development. In order to investigate this, three groups were formed (institutional group, relatively better life circumstances community group and severe hardship community group) and the following specific hypotheses were formulated.

With respect to the second hypothesis in which the functional development of the participants living in the community group had a greater decline in comparison with the functional development of the participants who stayed in Sizanani Children’s Home, no significant results were found. However, considering the mean scores of difference between the pre-test measurement and the current measurements (T7-T5), these were in line with the formulated expectation. In addition, in comparison with the situation in Sizanani Children’s Home, living in the community environment means for most participants living in deprived situations. This finding is partly in line with findings reported by Christianson et al. (2002). That is, both Christianson et al. as well as the present study found the suggestion that a poor physical home environment could be associated with a low level of functioning of the participants living there. However, as Christianson et al. demonstrated that a poor physical home environment accounted for a unique proportion of the variance in the level of functioning in children with a disability, the current data seem to indicate that the poor physical home environment plays a less prominent role in accounting for the decline in functional development of the participants. One explanation that may account for this difference is that the design of the study of Christianson et al. was not longitudinal and therefore it did not measure a developmental process over time. Also, the caregivers in the study by Christianson et al. were the only ones who were constantly taking care of their child with a disability from birth on, while in the present study the caregivers were in a totally different situation. The children of the caregivers within the community group of the present study were institutionalized in Sizanani Children’s Home for a couple of years. But due to the period of turbulence at Sizanani Children’s Home in December 2006, the caregivers were suddenly forced to be responsible for taking care of their child again. McKinney and Peterson (1987), Sloper and Turner (1993) and Wallander et al. (1989) found considerable variation in how parents adapt to this care giving demand. For various reasons, some adapt well to challenges of caring for a child with a disability, while others do not. Also Love, Schochet and Meckstroth (1996) reported that child care quality was positively correlated with child development (in Blau, 1999). In this light, it might be interesting to investigate care giving quality on functional development more. Specifically, the impact and the burden on the caregivers may facilitate a decline in functional development of their child with a disability. Therefore, by developing the revised HOME-Inventory, future research should take emotional bonding and attachment between the child and its caregiver into consideration.

However, it should be kept in mind that before the period of turbulence at Sizanani Children’s Home, the level of functioning between the two groups already differed on the pre-test measurement (T5). This difference showed a trend towards significance. In addition, statistical comparisons revealed that the mean level of functioning of the children, before they were discharged from Sizanani Children’s Home, was higher than the mean level of functioning from those who stayed in the Home. Nevertheless, there are at least two possible explanations that may account for the fact that no significant differences were found in functional development between the two groups over time (T5-T7).
First, the twelve children in the community group who dropped out of the study may have had a relatively low level of functioning. When included, they could have lowered the mean level of functional development of the community group. Moreover, since one child passed away and three children were hospitalized in another residential facility, it can also be assumed that those children had the lowest level of functioning in the community group. Although speculative, it is likely that if those children were included in the present study, the level of functioning between the two groups on T5 would not have differed in the first place.
Secondly, it is likely that the functional development of the participants in Sizanani Children’s Home could not decline more because there might be a bottom effect for their level of functioning. Additionally, the level of functioning of the participants in the community group was relatively higher than those in Sizanani Children’s Home, therefore their level of functioning could drop more easily. In this light, the decline in functional development of the participants in Sizanani Children’s Home, in comparison with the participants in the communities, should be interpreted carefully.

With regard to the third hypothesis, whether within the community group the functional development of the children who are living in severe hardship would show a greater decline than the decline in functional development of the children who are living in relatively better life circumstances, no significant differences were found. Importantly, in contrast to the expectation, the participants who were living in better life circumstances showed a greater decline in their functional development, compared to the participants living in severe hardship. First of all, it is likely that the functional development of the participants who were living in relatively better life circumstances could decline more because there might be a bottom effect for the level of functioning of the participants living in severe hardship. Whereas, at the pre-test measurement the level of functioning of the participants living in better life circumstances was relatively higher than the level of functioning of those living in severe hardship, their level of functioning could more easily drop. Therefore, the decline in functional development of the participants living in relatively better life circumstances should be interpreted carefully. But on the other hand, factors as emotional attachment to the caregiver, bonding to relationships within the community environment and the children’s drive to cope with and manage their disability, could controvert this possible explanation. It can be supposed that the caregivers living in relatively better life circumstances could be away from home more often in comparison with the caregivers living in severe hardship. Because they need to earn more money to maintain their relatively good living conditions, and to maintain the principle necessities of the family in their everyday life. For this reason, it might be a plausible explanation that those caregivers are less frequently present to stimulate and give attention and affection to their child.
Further, another interesting note for interpreting the differences in functional development between the two research groups should be reported. The quality of environment, as measured with the fourth scale of the modified version of the HOME-Inventory only consisted of four questions. Maybe this was not sufficient enough to assess the quality of the environment. Also, it could be possible that the subdivision, as created by the median score of the sum score on this instrument, did not match with the definitions of poverty as described in Higgs (2007). He suggested that people living in severe hardship need to collect water from far away, they do not have electricity available, they cannot afford to eat the correct kinds of food and mostly live in traditional shacks. People living in relatively better circumstances generally have access to municipal water in or outside their home and they usually have electricity available in their small brick houses. To a larger extent they can also afford to eat the correct kinds of food. Therefore, it might be possible that the subdivision within the community group was wrongfully formulated. Instead of dichotomizing the living conditions into severe hardship and relatively better life circumstances, it might be more reliable to split the community group into ‘poor living circumstances’ and ‘severe hardship’. Also, when applying the definitions of poverty as formulated by Higgs, it might be more reliable to the reality.
In this light, a limitation of the current study should be mentioned. As the present study for the first time conducted the assessments in the communities, including the quality of environment and its level of deprivation, data from this study should be handled carefully. Because the assessments within the community group took place in only about one and a half hour, the observations and registrations of that moment are questionable. In addition, the quality of environment of Sizanani Children’s Home is only examined after the period of turbulence in December 2006. For that reason, the results on the quality of environment of the current study cannot be compared to previous data. Besides, due to the small sample size within the two conditions in the community group, the observed power was very low and may have accounted for the fact that no significant differences were found. Therefore, it would be necessary to conduct further measurements to investigate the influence of the quality of environment and its level of deprivation on functional development.

It should be acknowledged that the present study suffers from various difficulties. First of all, the problems at Sizanani Children’s Home during and after the period of turbulence in December 2006 perhaps could have influenced the results of this study. In Sizanani Children’s Home, there was a lack of trust between the childcare workers and the management. It is possible that this lack of confidence has persuaded the childcare workers’ attitude and behaviour, concerning the quality of care as well as their answers during the assessments. Whereas it is just a conjecture, the general lack of some childcare workers has possibly affected the results of this study. Another difficulty was the fact that Conductive Education was ended in Sizanani Children’s Home. It is likely that this lack of stimulation has generally influenced the results on the functional development of the children in this study.
Second, the consequences of discharging more than hundred children in December 2006 could also possibly affected the caregivers of those children in the communities. The majority of caregivers could not understand and appreciate why their child was discharged from Sizanani Children’s Home. That provoked some anger and sometimes a defensive attitude towards the researchers. Additionally, two families did not want to meet with the researchers because they were too angry to invite them. Further, during and after most of the assessments in the communities the caregivers were often desperately asking whether their child could be taken back to Sizanani Children’s Home. Caregivers were regularly explaining that taking care of their child was extremely intractable with the responsibilities to maintain the principle necessities of the family in their everyday life. This is in line with findings that have been reported earlier in pediatric psychology literature (e.g. Eicher & Batshaw, 1993). It has been suggested that the settings in which care giving for a child with a disability in the communities takes place, would aggravate care related stressors within the caregiver and the care giving demand could become burdensome. Sjobu (1994) has shown that also the severity of the disability is related to parental stress. Since the majority of the caregivers in the communities experienced tremendous stress because they were not able to cope with the care giving demands, it is possible that they sometimes exaggerated their situation in order to imply a lower level of functioning of their child.
From this point of view, it might be important that caregivers of children with a disability receive some assistance in developing stress management techniques and coping skills. It would be relevant to start a day care centre in the communities, in which professional caregivers will take daily care of the children. Taking this relevance into consideration, it would imply the need for an outreach programme in which Sizanani Children’s Home can provide follow-up care to the discharged children. Additionally, the caregivers will have the opportunity to maintain their jobs and being responsible for the principle necessities of the family. Besides, a day care centre can be a place where caregivers can set up a support system. According to a study by Walton (1993) one of the most successful ways of regenerating energy, developing skills and acquiring knowledge is to become involved within a support group. It concluded that a support group allows the caregivers to become aware that they are not alone. They find out that there are others who must deal with the same frustrations and stressors that they do, and they will begin to learn not only how to accept these stressors, but how to become more efficient in managing some of the stressors. This study by Walton on the evaluation of counselling parents of children with a disability on stress management, also showed that there is a tendency for caregivers of children with a disability to become isolated socially. Because of the time restraints upon the caregivers as well as the inappropriate reactions of society towards the child, the caregivers may irrationally think that it is best not to get involved socially. It would be a suggestion to appoint a counselor who can teach the caregivers how to cope with the stressors that they face in the process of stress management.
The third essential issue concerns the inter-observer reliability of this research. The measurements were done by different researchers in different time-spans. This could have lead to differences in administrating the assessments. This possible lack of continuity between the researchers can violate the inter-observer reliability and may have affected the results. However, the level of agreement between the researchers on the assessments in the present study has proven to be good.
Finally, during the assessments in the communities, it was sometimes necessary to appeal to the help of the child care worker, who served as a translator, because most of the caregivers could not properly speak English. Despite given instructions to literally translate the information the caregiver was responding, it could be possible that some important details were lost.

Despite these difficulties, the current data slightly support the notion that quality of environment might contribute to the level of functioning of children with a disability. Considering the mean scores, the period of turbulence at Sizanani Children’s Home in December 2006 has clearly negatively effected the functional development of the participants within all conditions. It can be concluded that after the period of turbulence at Sizanani Children’s Home, the participants because of their vulnerability sensitively reacted in their functional development.