Burden of fatigue among adults living with HIV/AIDS attending antiretroviral therapy in Ethiopia

This study was conducted in public and private health facilities of Mekelle city, Tigray, North Ethiopia. The city has seven local administrative district (kifleketema) and one hundred sub-districts (ketena). Mekelle zone has an estimated total population of 256,000 [34]. Among the health facilities in the city, all hospitals and eight health centers provide ART service. According to the report of Tigray regional state health bureau, a total of 9662 adult ART users are accessing the service from 14 public. An institutional based cross-sectional study was conducted from April 1 to 30, 2019.

The source population was all adults (Age ≥ 18) living with HIV/AIDS and those who attended ART clinic at the health facilities. Patients diagnosed with HIV/AIDS at least 30 days prior to enrollment to the study were included. Patients having a diagnosed bipolar disorders, schizophrenia, or dementia, women with known pregnancy, and patients with chronic illness condition marked by fatigue such as renal disease, cancer, multiple sclerosis, or known cardiac patients were excluded from the study. The sample size was calculated by using single population proportion formula. Since there was no similar study in the area or countries having the same status with Ethiopia, the proportion of 50% was taken to have maximum sample size. This gave a sample of 369 and by considering 10% non-response rate and 1.5 design effect; the final calculated sample size was 609. According to Tigray health bureau report, at the end of 2018, the total patients attending ART clinic in Mekelle city were 9662 with an estimated number of 3220 patient flow per month. A multistage sampling procedure was used to select study subjects and the total numbers of health facilities were selected by using a simple random sampling technique. Accordingly, 50% of the health facilities, i.e. two public hospitals, one private hospital, and four health centers were selected. The number of study units to be sampled from each facility was determined by using a proportional allocation and systematic random sampling was employed to select each study subjects. By using ART user card number, the first participant was selected by simple random sampling technique and every K^th interval was followed until the allocated number of study subjects for each facility was reached. The sampling fraction (K^th) was stated as (the ART case load /sample size), every 5th ART user was recruited for each health facilities.

Burden is defined as the prevalence of fatigue. By using the Fatigue Severity Scale, a patient who scored greater than or equal to four average score was considered as having clinically significant fatigue [35]. A patient was considered as physically active if the total minutes reported by participants was greater than 150 min per week [19]. Co-morbidity was defined as the presence of one or more additional confirmed diseases or disorders concurrently occurring with HIV/AIDS (hypertension, epilepsy, opportunistic infections including tuberculosis) [36]. A participant was considered as physically disabled on self-report by the presence or absence of physical disability that had significant negative impacts on one or more major activities of daily living. Participants who scored ≥5 on ‘patient’s health questionnaire-9’ were considered as having depression. Participants whose score was > 7 on ‘Insomnia Severity Scale’ were considered to have insomnia. Body mass index is the ratio of weight to height in meters squared, less than 18.5 being underweight, 18.5–24.9 normal weight, 25–29.9 overweight, and > 30 obesity.

The tool was adopted from Fatigue Severity Scale (FSS) for measuring fatigue intensity, and Insomnia Severity Index for measuring insomnia. Data on socio-demographic, disease and treatment related characteristics, psychosocial and personal characteristics were collected with interviewer administered structured questionnaire. Data abstraction format was also used to collect data from patients’ medical record. Depression was measured by using patient’s questionnaire-9, and there was some modification in question number 9. Question number 9 ‘Thoughts that you would be better off dead or of hurting yourself in some way’ was modified to ‘Thoughts of hurting yourself in some way, if yes, how far’ then according to their response they were asked to grade it. Data were collected by seven bachelor of science degree holder nurses and supervised by two physiotherapists. Study participants who had severe fatigue were advised to visit physiotherapy department as soon as possible.

In order to maintain the quality of data, the questionnaire was translated to local language (Tigrigna) and back to English to ensure its consistency. Data collectors and supervisors were trained for 1 day on the purpose of the study, details of the data collection tool, interviewing techniques, the importance of privacy and ensuring confidentiality of the respondents prior to the actual data collection. Before the data was officially collected, the questionnaire was pre-tested on 5% (31) of the total sample size among the ART users in Wukro general hospital. Based on the findings of the pre-test, changes were made, including removing the questions around viral load and some modifications in language to the insomnia and depression scales. The supervisors made routine checkups for completeness and consistency of the data. The questionnaire was reviewed and checked for completeness, accuracy and consistency by the supervisors and investigators to take timely corrective measures. The data were coded and stored in a secure, confidential area.

Data were cleaned and entered into Epi Info version 7 and exported to Statistical Package for Social Sciences (SPSS) version 23 software for further process and analysis. Descriptive statistics was presented in the form of frequency, percentage, mean, median, range and standard deviation. Binary logistic regression model was fitted to identify factors that were significantly associated with the outcome (fatigue). Variables having P-value < 0.25 in the bivariate logistic regression analyses were considered as potential candidates in the final multivariable logistic regression analysis. P-value < 0.05 was used to declare statistical significance in the multivariable model. Multi-collinearity was checked using Variance Inflation Factor (VIF) and the overall goodness of fit of the model was checked by using Hosmer-Lemeshow goodness of fit test. Finally, the adjusted odds ratio (AOR) with its 95% confidence interval (CI) was used to determine statistically significance and depict the results.

Limitations of this study include the lack of information captured about study participant use of antidepressant and sleeping pills, which could have influenced the scores of ‘Patients health questionaire-9’ and ‘Insomnia Severity Scale’. Moreover, as most of the previous research was done in developed countries, it was difficult to compare the result of this study with other studies in the Sub-Saharan region. The use of secondary source can also be mentioned as a limitation of this study. Hence, further longitudinal studies in the area are needed to be conducted.