Study design and setting
A facility based cross-sectional study was conducted at Hawassa University College of Medicine and Health Sciences, which is 270 km south of Addis Ababa. The college is one of the 7th colleges under Hawassa University. The college is established to teach medical, nursing, public and laboratory students. In the academic year 2013/2014 the college had enrolled a total of 2408 students. From these students 1289 of them were medical students. Medical students are expected to stay in the college for 6 years. There are four major phases medical students have to pass during their 6 years study periods: pre-medical (6 months), preclinical (nearly 2 years), clinical (two and half years) and internship (1 year). During pre-medical phase students are expected to take courses like Medical Ethics and Legal Medicine, Medical Psychology, Information Communication Technology, Medical Anthropology and Sociology, Sophomore English and Civics and Ethical Education, and Environmental Health. During their preclinical phase major basic sciences courses like Anatomy, Physiology, Biochemistry, Pharmacology, Pathology and Microbiology together with other relevant courses will be covered. During clinical years students will have both theoretical and practical session exposure for clinical major courses like Internal Medicine, Surgery, Gynecology/Obstetrics, Pediatric and Child Health together with minor clinical courses like Psychiatry, Dermatology, Ear Nose Throat (ENT), Ophthalmology, Radiology, and Dentistry. The Medical Schools use letter and grade point scale based on Hawassa University Senate Legislation . In order to be promoted to the next level and graduate successfully with Medical Doctorate Degree, a student should obtain at least 2.00 points Cumulative Grade point Average (CGPA) out of maximum score of 4.00 points.
The participants of this study were undergraduate medical students enrolled in the year 2013/2014 and attending classes in regular program under School of Medicine. Students were stratified based on the year of study. Sample size was calculated by anticipating the prevalence of mental distress as 21.6% (based on recent study conducted in Adama University ), 95% confidence level and 5% margin of error. Considering the total population (N = 1289 medical students), correction factor was used and adding 10% for non-response made our final sample size 240. To account for the relative size of students in different years of studies, proportional allocation technique was employed. Based on this, majority of students were from pre-medicine year (n = 68 students) while the least were taken from clinical year II (n = 20). The students from each class year were randomly selected. Students who formally enrolled but not attending class during study periods were excluding from the study.
Data was collected from May 12–26/2014. The periods were exam free periods. Assessments consist of an evaluation of basic socio-demographic characteristics, educational background, history of substance use, respondents’ monthly income, students’ attitude towards medical school, satisfaction by school sanitation and recreational center, students’ Cumulative Grade point Average (CGPA), relationship with their fellows, attitude about living room and their safety. Variables were compiled based on previous studies conducted in Ethiopia. Presence of mental distress was evaluated using a 20 items self-administered structured Self-Reporting Questionnaire (SRQ-20). Self-Reporting Questionnaire (SRQ-20) was originally developed by World Health Organization (WHO)  to detect presence of mental distress and validated in developing country including Ethiopia [22–24]; and found to be highly sensitive and specific. In this study a cut-off point 8 and above was taken as having mental distress based on previous finding . Since medium instruction in Ethiopian higher education is English we used the English version of the instruments.
Data were collected by final year medical students after thoroughly discussed about data collection procedures. Data was collected at the end of each class. Moreover, study participants were also informed to fill the questionnaire independently so that bias may be minimized. Data was entered into and analyzed using IBM SPSS statistics 21. Variables with a P value less than 0.05 were declared to have a statistically significant association. Each independent variable against the dependent variable was tested for having statistical significant association using binary logistic regression. Multivariable logistic regression analysis was done for those variables with P-value less than or equal to 0.25 during binary logistic regression analysis. Age was highly correlated with year of study (r = 0.8) and removed from the multivariate model.