Unmatched case-control study was conducted from 01 January 2016 to 30 December 2016 at Adare General Hospital, Southern Ethiopia. Respondents were 18 to 50 years old that do not have a diagnosis of medical or other psychiatric disorders and mental retardation. ‘Cases’ were respondents that have a clinical diagnosis of schizophrenia. Controls were relatives, friends or care-givers of schizophrenia patients that visit the hospital during the study period. Sample size was calculated using Epi-Info/StatCalc version-7 with the following assumptions; significance level α = 0.05, power = 80%, Odds Ratio = 3, case ratio controls = 1:3 and 29.9% was the proportion of controls with substance exposure. The sample size was found to be 142 respondents; with 36 cases and 106 control. However, 10% of non-response rate was added to maximize the power and 157 (40 cases and 117 controls) was the final sample size of this study. All existing cases and their relatives, friends and caregivers who visited psychiatric clinic during the study period were invited to participate in the study.
Eligibility criteria for cases and controls
All respondents were expected to be in the age group 18 to 50 years with no mental retardation and chronic medical illness (like; hypertension, diabetes mellitus, heart diseases or others). Cases were those with clinical diagnosis of schizophrenia and without Comorbidity to other psychiatric illnesses; while controls were their relatives, friends or caregivers with no history of diagnoses to any psychiatric illnesses.
Data Collection Procedure and Tools
Trained research assistants have collected primary data on demographic and quality of life of respondents through interview. Referring different standard tools like World Health Organization STEPWISE, the researcher has developed a tool for respondent’s socio-demographic data. However, SF-36 health survey was used to collect data on quality of life (11). SF-36 health survey is standardised 36 items questionnaire that were translated, adapted and validated in Ethiopian language (13). This health survey questionnaire assesses generic health-related quality of life. It includes physical and mental component summaries and eight specific health domains. The first four domains comprise the physical component summary, while the other four comprises the mental component summary. 1) Physical Functioning is 10 items that assesses both the presence and extent of physical limitations. 2) Role Physical is 4 items that measures role limitation owing to physical problems. 3) Bodily Pain is two items domain; one item is relevant to the intensity of body pain, while the second item is relevant to the extent of interference with work or normal activities due to body pain. 4) General Health is five items; one item rates health from excellent to poor and four items address respondent’s view and expectations of their health. 5) Vitality is four-item domain and measures energy level and fatigue to capture differences in subjective well-being. 6) Social Functioning is two-item domain assesses impact of either physical or emotional problems on respondents’ social functioning and the degree to which physical or emotional problems interfere with normal social activities. 7) Role Limitation is three-item domain that assesses health-related role limitations due to emotional problems. 8) Mental health is five-item domain that includes items from the four major mental health problems (anxiety, depression, loss of emotional control and psychological wellbeing). Scores on each domain-specific scales range from zero to 100. Higher score represents better health related quality of life and lower scores reflect poor quality of life (11).
Validity and Reliability
Validity is the ability of a tool to measure what it should and reliability is the ability of a tool to produce consistent results (12). To reduce threats to content validity, SF-36 was adapted and a forward-backward translated in to local language (Amharic) following the steps set by the developers. Finally, translations were reviewed to determine whether the content of the questionnaire had appropriately addressed the research objectives. The overall reliability of the instrument (SF-36) was tested using Cronbach’s alpha during the pre-test and it was found to be 0.89, which exceeds the acceptability threshold (alpha, 0.70).
Collected data were entered to computer software SPSS v23 program and were coded, checked and cleaned for errors and prepared for analysis. Tables were used to summarise and describe variables. Median was used for the comparison of age between cases and controls and, median and percentiles were used to summarise scores of quality of life between cases and controls. While mean rank was used to compare the difference in scores of Quality of Life between cases and controls. Mann-Whitney U test and Kruskal-Wallis 1-way ANOVA were used to test score differences in quality of life between cases and controls and among different age groups and among respondents with different educational backgrounds respectively.
Ethical clearance was obtained from University of South Africa Health Studies Higher Degrees Committee and necessary approval letters were obtained from responsible bodies like, Dilla University Research and Dissemination Office, Southern Nation Nationalities and Peoples’ Regional Health Bureau and from officials of Adare General Hospital. Informed consent was signed from respondents before data collection.