The development of an outcome measure to assess community reintegration after stroke for patients living in poor socioeconomic urban and rural areas of South Africa
Maleka, Morake Elias Douglas
The prevalence of stroke is high in South Africa (Connor et al., 2007). However, in-hospital patient rehabilitation following stroke (Hale and Wallner., 1996; Rhoda and Henry., 2006; Mudzi., 2009) is limited, subsequently patients who have had a stroke have limited functional independence at discharge from hospital (Mamabolo et al., 2009). In addition community based rehabilitation services in South Africa are poorly developed and inadequate (Hale and Wallner., 1996; Rhoda and Henry., 2006; Mudzi., 2009). Therefore, patients are sent home without rehabilitation and they do not receive any once they are at home. Therefore, it is not known how well people reintegrate back into their communities following stroke. One reason for this dearth of knowledge is that there are no appropriate outcome measures to measure community reintegration of patients back into their communities. This is particularly so for patients living in poor rural and urban communities. Outcome measures are an essential part of clinical quality management in rehabilitation, but need to take the context in which patients live into consideration. All outcome measures that assess community reintegration for patients with stroke, have been designed in developed countries, such as the United Kingdom, and are therefore not contextual for the type of patients seen in poor socioeconomic urban and rural communities in developing countries, such as South Africa. The definition and components of community reintegration vary and differ depending on the setting and target population. Although there are similarities amongst the different outcome measures, differences occur in the definition and components of community reintegration based on contextual factors. Except for the Participation Scale, all the outcome measures reviewed were formulated in more affluent and developed countries. Some scales were considered by the author to be too long for use in a largely illiterate population where questionnaires are better when interviewer administered (the Craig Handicap Assessment and Reporting Technique, the Stroke Impact Scale, Participation Scale, the Stroke-Adapted Sickness Impact Scale Profile). Many scales were scored based on the visual analogue scale system and some use five or more points Likert scale, which does not lend itself easily to translation (the Reintegration to Normal Living Index, the London Handicap Scale, the Stroke Impact Scale, the Community Integration Measure, the Stroke Specific Quality of Life, and the Subjective Index of Physical and Social Outcome). Only six of these scales had been validated in a stroke population (the Reintegration to Normal Living Index, the London Handicap Scale, the Stroke Specific Quality of Life, the Subjective Index of Physical and Social Outcome, the Stroke Impact Scale and the Stroke-Adapted Sickness Impact Scale Profile). Many scales were not specific to community reintegration and included very few items under the participatory domain (the Reintegration to Normal Living Index, the London Handicap Scale, the Stroke Specific Quality of Life, the Stroke Impact Scale, the Nottingham Health Profile, the EurolQol Quality of life Scale, the Soweto Stroke Questionnaire, the Medical Outcomes Study Short Form 36 and the Stroke-Adapted Sickness Impact Scale Profile). Based on this review of the tools developed to measure community reintegration, there did not appear to be a tool that would be appropriate to measure community reintegration following stroke in a black South African community; a measure that takes into account contextual, cultural, multi-lingual and illiteracy factors. As a result, the researcher set out to develop an outcome measure of community reintegration that would take into account all the environmental and personal factors of patients with a stroke living in poor socioeconomic rural and urban areas of South Africa. With that in mind, the overall aim of the thesis was to develop, validate and test the reliability of an interview-administered outcome measure to assess community reintegration after stroke for patients living in poor socioeconomic rural and urban communities of South Africa. There are two parts to this thesis. The specific objectives of each of the studies are listed below: Study 1 1. To conceptualise community reintegration from the perspective of individuals who have had a stroke and their caregivers in order to develop and construct the outcome measure. 2. To develop and construct the items of the outcome measure using the information gained from the interviews and a review of the literature. Study 2 1. To validate the outcome measure using neurological and community based rehabilitation experts and patients who were interviewed in study 1. 2. To establish the reliability and factor structure of the outcome measure using factor analysis and internal consistency statistics. 3. To establish construct validity by comparing this newly developed outcome measure to another existing tool. Study settings: A community setting in the Soweto primary health care clinics around Johannesburg-Gauteng province (urban) and primary health care clinics in Elim, Siloam-Limpopo Province (rural), South Africa. The studies were carried out as follows: study 1 was qualitative in nature using semi-structured face to face interviews. Thirty two interviews were conducted with patients who had had a stroke and their caregivers, nineteen from the rural setting and thirteen from the urban setting. Interviews were recorded using an audiotape, transcribed word for word and the content analysed. Thematic content analysis was used to extract the statements concerning community reintegration, concepts were identified and grouped into themes. The document developed from this study was the new preliminary outcome measure used in study 2. Study 2 was a combination of qualitative and quantitative studies conducted in order to validate the newly developed community reintegration outcome measure. The first validation phase used three Delphi technique rounds with local neurological and community based rehabilitation experts, a statistician, psychologist and two meetings of the nominal group technique with patients as well as caregivers from study 1 to establish face and content validity of this newly developed outcome measure. The second validation phase was a quantitative, cross sectional study which was two pronged. Firstly, homogeneity (internal consistency) of items contained in this newly developed outcome measure were assessed and secondly using factor analysis items were further reduced and the construct of the outcome measure was confirmed. One hundred and twelve patients and 104 patients from the urban and rural setting respectively were included in the study. The last validation phase was a quantitative, cross sectional study which compared this newly developed outcome measure, with another existing measure the Subjective Index of Physical and Social Outcome (SIPSO), to assess construct validity. Eighty patients were used for this purpose (40 for each setting). From the analysis of the results of study 1, community reintegration was conceptualised from the perspective of patients as well as their caregivers. The conceptualisation phrased in a positive manner, incorporates the ability to move around in one‟s home and community, of not being isolated without having roles reversed and identity loss. The person should be able to work to sustain his/her life and not lose hope. Themes from the interviews gave rise to sixty seven items which were generated based on the interviews conducted and these items were categorized under 11 domains by the researcher. This document was used in study 2. The results of study 2 phase one were: the 67 items that were generated in study 1 were reduced to 44 categorised under eight domains after three rounds of Delphi technique and two nominal group meetings with patients. Study 2 phase two, factor analysis and internal consistency statistics results: As a result of the new grouping of the items, the two settings ended with a different outcome measure each. In the rural setting 12 items were removed, leaving 34 items but only six items were removed from the urban setting outcome measure, leaving 40 items. The internal consistency of these two newly formed outcome measures was very good (for both settings, the alpha coefficient was 0.95). The new outcome measure was named the Maleka Stroke Community Reintegration Measure (MSCRIM), the rural and urban setting versions. Study 2 phase three, compared the MSCRIM to SIPSO. High correlations (urban r = 0.88, p = 0.0001, 2-tailed and rural setting, r =.95, p = 0.000, 2-tailed) were found between the MSCRIM and the SIPSO. However, MSCRIM (both versions) contain items that are context specific to patients with a stroke living in poor socioeconomic urban and rural areas in South Africa. The MSCRIM is therefore a valid and reliable measure to assess community reintegration following stroke for patients living in poor socioeconomic rural and urban communities of South Africa. This outcome measure is interview-administered to either patients who have had a stroke or their caregivers. Key words: Stroke, development of outcome measures, community reintegration, poor socioeconomic, rural and urban areas.