The study was conducted at the University of Gondar Comprehensive Specialized Hospital Northwest Ethiopia from February 1 to March 30, 2020. The University of Gondar Comprehensive Specialized Hospital is one of the biggest teaching hospitals in the Amhara region, serving over five million people requiring tertiary care. It is situated 728 km north of Ethiopia’s capital city, Addis Ababa. The hospital provides both comprehensive and referral-level treatment. There are 938 full-time registered healthcare workers. This research’s study population included all healthcare professionals working in the hospital.

An institution-based cross-sectional study was carried out among all healthcare workers who were permanently employed at the University of Gondar Comprehensive Specialized Hospital and were available during data collection. During the data collection time, healthcare professionals who were unable to communicate due to a serious illness were excluded.

The sample size for this study was determined using a single population proportion formula by assuming the prevalence of poor sleep quality among healthcare professionals (p = 50%), 95% confidence interval, and 5% margin of error. After adding a non-response rate of 10%, the final sample size became 423. There were a total of 876 healthcare professionals. Therefore, a stratified random sampling technique was applied based on their profession. First, a proportional allocation was used after identifying the number of healthcare workers in each profession. The study subjects were then chosen to form a total of 423 healthcare professionals using a computer-generated simple random sampling method from those strata.

Data were gathered from February 1 to March 30, 2020, using a self-administered questionnaire that included socio-demographic variables, work-related factors, substance use, depression (as assessed by the Beck’s Depression Inventory), and the Pittsburgh Sleep Quality Index (PSQI). The PSQI is a validated tool used to assess adult sleep quality and patterns (37). Subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, medication use, and daily dysfunction are the seven groups of the instrument. The seven domains are then added to create a total score, with a score greater than 5 indicating “poor” sleep quality and a score of less than or equal to 5 indicating “good” sleep quality (38).

Poor sleep quality was the dependent variable in this research. Socio-demographic variables (age, gender, marital status, having children, and profession), substance use variables (alcohol consumption, cigarette smoking, drinking coffee, and khat chewing), work-related variables (working shift, working hour, and work experience), and health status variables (depression, history of chronic illness, exercise, and self-rated health) are independent variables. To assess depression, we used the Beck Depression Inventory (BDI), which is a 21-item self-report rating assessment that assesses depression-related attitudes and symptoms (39). The Beck Depression Inventory, second edition (BDI-II) is one of the most widely used tools in research and application to assess the existence and degree of depression in the previous 2 weeks (40). The BDI-II assesses 21 symptoms and attitudes, which include Mood, Pessimism, Sense of failure, Lack of satisfaction, Guilt feelings, Sense of punishment, Self-dislike, Self-accusation, Suicidal wishes, Crying, Irritability, Social withdrawal, Indecisiveness, Distortion of body image, Work inhibition, Sleep disturbance, Fatigability, Loss of appetite, Weight loss, Somatic preoccupation, and Loss of libido (41, 42). Participants with BDI-II scores ranging from 0 to 13 were deemed normal. Participants with BDI-II scores of 14–19, 20–28, or 29–63 were classified as having mild, moderate, or severe depression, respectively (39). In terms of substance use, participants were considered current substance users if they had used cigarettes, khat, alcohol, or coffee at least once in the month preceding the survey (43–45). Khat (Catha edulis) is an herbal trade-related cultivated plant growing in most parts of the world, particularly in Eastern Africa and Arab countries (46, 47). It contains psychoactive substances cathinone and cathine, which can cross the blood–brain barrier to enter and stimulate the brain (47–49). A semi-structured question “Did you have a previous history of chronic illness?” was used to evaluate the history of chronic illness. A “yes” answer was considered as having a history of chronic illness. The questionnaire assesses shift work by asking if they have day and night shift work; if they answer positively with “yes”, they were classified as having shift work. Shift work entails switching between day and night shifts. A healthcare shift worker is a healthcare professional who often works in a healthcare facility switching between day and night (50). Regular exercise is described as a subset of physical activity that is planned, structured, and repetitive, with the ultimate or intermediate goal of improving or maintaining physical health (51). According to WHO guidelines, 150–300 minutes of moderate-intensity physical exercise or 75–150 minutes of vigorous-intensity physical activity is recommended (52).

After checking for the completeness and consistency of the collected data, the data were entered into Epi-data version 3.02 and exported to SPSS version 25 for analysis. To express descriptive results, frequency with percent and mean with standard deviation were computed. The study fulfilled all chi-square assumptions, which assume that both factors are categorical, all observations are independent, and cells in the contingency table are mutually exclusive; there should be no zero cell values, and the expected value of cells should be 5 or higher in at least 80% of cells. A binary logistic regression was performed to determine the crude association between each independent variable and poor sleep quality. Variables in the bi-variable analysis with a p-value <0.25 were candidates for multivariable binary logistic regression analysis. The crude odds ratio (COR) and adjusted odds ratio (AOR) with 95% CI were calculated. For the multivariable analysis, variables with a p-value <0.05 were considered statistically significant.

To ensure the data quality, 1-day training was given for four BSc nursing data collectors. The questionnaire was also pretested, and regular supervision of the data collection process was also made to increase the completeness, accuracy, and consistency of the data. Incomplete questionnaires were discarded from the analysis. Cronbach’s alpha was also calculated to test the reliability of the PSQI tool, and it was 0.72.

Ethical clearance was obtained from the Ethical Review Board of the University of Gondar with ethical clearance number 1840/2012. To ensure the confidentiality of respondents, informed written consent was obtained from each study participant, and their names were not written on the questionnaire, which reduced the authors’ access to information that could identify individual participants during or after data collection.

This survey had 418 participants. Initially, the plan was to collect data from 423 professionals, but due to incomplete and non-returned questionnaires, the response rate became 98.8%. Two hundred thirteen (51%) were female. The subjects’ mean age was 30.7 years (SD 6.3). One-third (33.3%) of the individuals were married. More than half of the participants in the research were nurses, 226 (54.1%), with midwives accounting for the remaining 60 (14.4%) ( Table 1 ).

Three hundred eighty-five (92.1%) of the participants were on duty. Four hundred sixty (97.8%) of them worked fewer than 56 hours per week. In terms of substance use, 45 (10.8%), 260 (62.2%), and 35 (8.4%) of the subjects were current khat chewers, alcoholic drinkers, and cigarette smokers, respectively. Thirty-five (8.4%) of those who participated had a history of chronic disease. The majority of subjects (82.3%) reported moderate stress ( Table 2 ).

The overall prevalence of poor sleep quality among healthcare professionals in this study was 58.9% [95% CI (54.2, 63.6%)]. Participants’ average night sleep duration and sleep latency were 6.5 hours (SD 1.27) and 18.6 minutes (SD 16.6), respectively. Three hundred five (73%) of the participants had less than 7 hours of sleep per day. Thirty-three (7.9%) and 73 (17.5%) had low habitual sleep efficiency (<65%) and used sleep medication in the past month, respectively ( Table 3 ).

In the logistic regression analysis, variables with p-value <0.25 were included in the multivariable logistic regression model. Accordingly, sex, shift work, current khat chewing, depression, and not performing exercise were significantly associated with poor sleep quality (p < 0.05).

Female participants were two times [AOR = 1.9, 95% CI (1.2, 2.9)] more likely to experience poor sleep quality compared to male participants. Participants who worked in shifts were six times [AOR = 5.7, 95% CI (2.3, 14.3)] more likely to have poor sleep quality than their counterparts. Participants who did not have regular exercise were two times more likely to have poor sleep quality than those who had regular exercise. Current khat chewers were three times [AOR = 3.1, 95% CI (1.2, 7.6)] more likely to experience poor sleep quality than those who did not chew khat. Participants with moderately severe depression were three times more likely to have poor sleep quality than those without depression [AOR = 2.6, 95% CI (1.3, 6.8)] ( Table 4 ).

Our study is not free from limitations. Because the study is cross-sectional, we are unable to show a cause–effect relationship between independent variables and poor sleep quality. Additionally, we could not show the dose–response relationship between substance use and sleep quality. Another limitation of this research could be recall bias.