Bipolar disorder in rural Ethiopia : community-based studies in Butajira for screening, epidemiology, follow-up, and the burden of care

Abstract: Background: The challenges of research in economically stunted countries’ settings remains a profound concern and is linked to socioeconomic development of these countries. More research is needed regarding psychiatric morbidity in rural areas of the developing and poverty stricken countries. The present studies were undertaken within the framework of a broader ongoing community-based project on the course and outcome of major psychiatric disorders in the rural Butajira district located in Ethiopia. This thesis treats the course and outcome of bipolar I disorder in the district. Objectives: Through appraising mental health and population based research in a rural Ethiopian district, to evaluate the utility of modern research instruments, and to obtain baseline information relating to bipolar I disorder in the poverty stricken rural Butajira district of Ethiopia. The specific objectives were: 1. Evaluating and comparing two different screening methods of case detection and identification for schizophrenia and bipolar I disorder in the adult population of Butajira district. 2. Assesing the prevalence and clinical characteristics of of bipolar I disorder in Butajira at the community level. 3. Evaluating short-term outcome at follow-up of bipolar I disorder in the Butajira district. 4. Determining Neurological Soft Signs in community-identified cases of bipolar I disorder in Butajira district in comparison with healthy controls. 5. Assessing the burden of care among caregivers of those affected by bipolar I disorder identified in the Butajira Study. Methods: The district’s entire adult population aged 15-49 was identified through a double-sampling design. In the first stage of screening, door-to-door interviews were conducted by lay trained high school completed individuals who knew the culture of the people. Females interviewed females whereas males interviewed males. Additionally, the key-informants method was used to identify cases that would be missed by the CIDI or otherwise. The final confirmatory diagnostic interview was conducted by clinicians using the SCAN on door-to-door basis as well. The probable cases that fulfilled the lifetime DSM-IV diagnosis of bipolar I disorder were assigned for assessment by other baseline research instruments such as Neurological Evaluation vii Scale (NES), Young Mania Rating Scale, Hamilton Rating Scale for Depression, LCSS, PANS and SANS, BISS, BII, FIS and so on. Cases so identified with bipolar I disorder were subject to a follow-up for upto 2.5 years on the average (range 1 to 4 years). Two of the main clinical outcomes assessed were relapse to a mood episode, and remission from a mood episode. Outcomes were assessed annually by the instruments, and were further assessed monthly by trained psychiatric nurses. We also did a cross-sectional study of caregivers of bipolar I disorder cases, and assessed objective burden on the caregivers as considered from social, family strain, occupational and financial domains. Results: Information provided by the key informants was better at detecting schizophrenia or chronic psychiatric disease, whereas the CIDI was better at detecting affective disorders. Of the around 100 000 adults living in Butajira, 83.3% were found by the project’s census, of which 82% (68,378 subjects) were successfully screened by the CIDI, yielding 2,161 CIDI positive. These, together with 719 cases identified by the key informants, were invited for the SCAN interview, of which 74.7% agreed. This yielded 315 SCAN positive cases for bipolar I disorder, and complete information could be collected on 295 of these. Lifetime prevalence was estimated as 0.6% for males and 0.3% for females. The mean age of onset of the manic phase was 22.0 years and that of the depressive phase was 23.4 years. For 22.7% of the cases the illness started with a depressive episode and for the remaining 77.3% it started with a manic episode. Over half of the cases (55.9%) had never sought help from modern health care sector, and only 13.2% had ever been admitted to psychiatric hospital. At follow-up, 65.9% had exprerienced a relapse and 31.1% had persistent illness, while only 5% of the patients were in remission for most of the follow-up time. The bipolar I cases, as compared with healthy controls, performed worse on several items of NES, thus having more neurological dysfunction compared to controls. Caregives were largely (80.3%) first-degree relatives and spouses. Overall, 84% of the caregivers reported difficulties in at least one of the domains of family burden. Of these, 58.7% reported a severe degree of difficulties. Caregivers reported a high level of difficulties in intrafamilial relationships and social restrictions, disruption in earning a livelihood, and financial difficulties. Conclusions: The prevalence of bipolar I disorder is comparable to the prevalences reported from other countries, and our findings support the cross-cultural validity of the concept of bipolar I disorder. Majority of the cases are not treated in contrast to that in the developed countries. The burden of care for the caregivers is substantial in the population studied.