Description of the study area
The study was conducted in Addis Ababa, the capital city of Ethiopia. The total area of the town is 527 km2. The city has a longitude and latitude of 38° 45' 48.9996'' E and 9° 0' 19.4436'' N respectively with an elevation of 2,355 m above sea level. During the study period there were three COVID-19 treatment centers; Yeka general hospital, St. Paul’s hospital millennium medical college and St. Peter’s specialized hospital.
Study design and period
A facility based unmatched case control study design was conducted from March 30/2022 to June 30/2022.
Population
Source population
All overweight/obese COVID-19 patients admitted at Addis Ababa COVID-19 centers were the source population.
Study population
The case population were defined as overweight/obese COVID-19 patients those admitted to the intensive care unit (ICU); who had one of the following; respiratory rate > 30 breath/min; oxygen saturation < 93% on room air; acute respiratory distress syndrome (ARDS); or required mechanical ventilator.
The control population were defined as overweight/obese COVID-19 patients those who did not have severity symptoms.
Study population for cases and controls
All selected overweight/obese COVID-19 patients those who have severity symptoms; or required mechanical ventilator and admitted to COVID-19 centers during data collection period were the study population for cases and all selected overweight/obese COVID-19 patients who do not have severity sign and symptoms were study population for control group
Eligibility criteria
Inclusion criteria for case and controls
All overweight/obese COVID-19 patients with severer symptom and admitted to Addis Ababa COVID centers. All overweight/obese COVID-19 patients who did not have severer symptoms are included in the control groups.
Exclusion criteria for case and control
Those patients critically ill for COVID-19 and unable to be weighed and lack weight record during the current admission are excluded from the study. Those patients unable to be weighed due to other diseases they were admitted for and lack recorded weight during the current admission were excluded from the study.
Sample size determination
The sample size was determined by Epi-info version 7.2 by assuming 95% of confidence interval, 5% margin of error and 10% non-response rate, power of 80%, percent of control exposed 25.6%, percent of case exposed 45.5% of poor sleep quality from study done in Madrid, Spain (16). The total sample size is 252 with 10% of non-response rate and 5% margin of error. Ratio of control to case is 2:1 case 84 and control 168.
Sampling technique and procedure
All overweight/obese patients who met the inclusion criteria for both cases and controls were enrolled in the study sequentially until the sample size was reached. Since the study participants (overweight/obese COVID-19 patients) were uncommon, the number of COVID-19 patient declined during the study period the source population was not larger than study population we enrolled all overweight/obese COVID-19 patients
available in the three Hospitals who met inclusion criteria to the study. Total of 150 (50 cases and 100 controls) study participants came from Yeka general hospital, 36 participants (12 cases and 24 controls) from St. Peter’s speciality hospital, and 60 participants (20 cases and 40 controls) from St. Paul’s hospital millennium medical college. Proportional allocation was not employed since the study population is rare all overweight/obese COVID-19 patients were included to the study. And, there is no previous BMI record of COVID-19 patients which might help the allocation. Patterns of COVID-19 disease are unknown. In addition, each hospital has an equal COVID-19 treatment clinic: if proportional allocation was employed it may require much longer time till each gate is full and might be unfeasible to the current study.
Study variables
Dependent variables
The dependent variable of the study is determinants of COVID-19 severe.
Independent variables
Sociodemographic factors: - Age, Sex, Ethnicity, Occupational status
Clinical factors: - Cardiac diseases, Hypertension, DM, HIV, Renal disease, Malaria and Hypercholesterolemia
Immunologic factors: - BCG, POV, COVID-19 vaccine, Measles and steroid use
Psychological factors: - stress and sleep quality
Behavioural factor: - Smoking and Alcohol drinking
Operational definition
Overweight: patients with body mass index greater than or equal to 25 kg/m2 to 29.9kg/m2 (9).
Obese patients: patients with body mass index greater than or equal to 30kg/m2 are categorised as obese patients (9).
High stress:-those participants who score ≥ 50% by perceived stress were considered as high perceived stress (17).
Low stress:-those participants score ˂ 50% by perceived stress ware considered as having low perceived stress (17).
Severe COVID-19 disease: severe patients were those admitted to the intensive care unit (ICU), who had one of the following respiratory rate breath/min > 30; oxygen saturation < 93% on room air; acute respiratory distress syndrome (ARDS); or required mechanical ventilator (18–20).
Non severe COVID-19 disease: non severe patients without the above symptoms (18–20).
Never smoker: A person who never smokes a part or all of cigarettes in lifetime (21).
Ever smoker: A person who smokes a part or all of cigarettes in lifetime (21).
Current smoker: A person who smokes a part or all of cigarettes in the past 30 days and/or 100 cigarette lifetimes (21).
No risk for alcohol drink: those who score 0 out of 40 for the AUDIT alcohol drink questions (22).
Low risky drinker: those who score 1–7 out of 40 for the AUDIT alcohol drink questions (22).
Risky drinkers: those score 8–12 out of 40 for the AUDIT alcohol drink questions (22).
High risk drinkers: those who score greater than 13 out of 40 for the AUDIT alcohol drink questions (22).
Good sleep quality: those score 0 for Pittsburgh sleep quality index (23).
Fairly good sleep quality: those score 1 for Pittsburgh sleep quality index (23).
Fairly Bad sleep quality: those score 2 for Pittsburgh sleep quality index (23).
Bad sleep quality: those score 3 for Pittsburgh sleep quality index (23)
Data collection procedure and measurements
Data collection procedure
The data was collected using pre-tested semi structured questionnaires by interviewing the patients or guardians for critically ill patients and a medical record review was conducted for all participants. Most participants give consent and information by themselves. Consent and information of three patients from the case and three patients from controls were obtained from the guardians. Data collectors were professionals who have been working on COVID-19 clinics; data was collected by wearing appropriate personal protective equipment. During the data collection period, the target study populations were included based on the sample size of the study which is a total of 252 patients (84 severe/critical COVID-19 overweight/obese cases and168 non-severe overweight/obese COVID-19 control admitted at COVID centers for COVID or other medical problems). All patients admitted to COVID ward and other wards like medical, surgical, gynaecology and emergency, follow up clinics were assessed and categorized as case and control based on the categorization criteria and inclusion criteria. All overweight/obese COVID-19 patients available during data collection period were included sequentially till sample size gate reached.
Data collection measurements
Temperature measurement: infrared thermometer was used to assess the temperature of the patients.
Body mass index: Height and weight was measured according to WHO guidelines. Height is measured by the stadiometer by making knees straight, head in the Frankfurt plane, heels, buttocks and shoulder blades touching the vertical line of the stadiometer. Weight was measured using beam balance. Since weight measurement of critical patients was impossible we took weight measured in the current admission if they had and length was measured since height measurement in standing position was difficult. Weight measurement of the 6 participants (3 cases and 3 controls) was taken from the medical record of current admission though length was measured instead of height for these specific participants. Calibration was made in between every patient and after moving instruments by measuring a known standardised material. BMI was calculated by dividing weight by height in metre squares.
Blood examination: hypercholesterolemia, DM, HIV was collected from medical records and for those who have no recent medical record laboratory test was done by chemistry machine, glucometer and stat pack respectively. Calibration of the glucometer with a chemistry machine was done every morning of the data collection period.
Information on renal disease: was collected from medical record review and for those who have no data about renal disease; blood urea nitrogen (BUN) and serum creatinine test was done by a chemistry machine.
Malaria was examined by the malaria rapid examining test (RDT) test.
Blood pressure was assessed using a manual Sphygmomanometer.
Information about cardiac disease: was obtained from medical record and for those who lacks data on cardiac problem was assessed by history and physical examination (chest pain, palpitation, dyspnoea, body swelling, cough, pulse rate abnormalities, heart murmurs) then confirmed by Echocardiography.
Asthma: was assessed both by previous history of known asthma and/or by cough, wheezing and chest tightness. History was taken to obtain information about vaccination and smoking. Information on stress, alcohol and sleep quality: was taken based on Pittsburgh sleep quality tool to obtain information on sleep quality, for alcohol consumption history was taken based on alcohol audit screening tool and stress related information was taken based on stress scale tool .
Data processing and analysis
Data was entered into EPI data version 3.1 and then exported to Statistical Package for Social Sciences (SPSS) version 25 for cleaning and analysis. Descriptive statistics were done to describe variables and data was described by texts, frequency tables, graphs, mean with standard deviation and percentages. The goodness-of-fit of the model was measured by Hosmer and Lemeshow's test with P-value.
Bivariable and multivariable analysis was done. Variables having P ≤ 0.25 with COVID-19 severity among adult overweight/obese COVID-19 on bivariable analysis were checked for multicollinearity before the final model. Since the entire candidate had no multicollinearity (all variables had variance inflation factor less than 1.7 and tolerance greater than 0.87), all variables with P ≤ 0.25 in bivariable analyses were interred to multivariable analysis. The model fitness was checked using the Hosmer-Lemeshow test and it has Chi-square of 10.117 with p value of 0.357. Adjusted odds ratio (AOR) with 95% Confidence interval was used to determine the strength of association. P ≤ 0.05 was considered to declare statistical significance.
Data quality control
Initially, questionnaires were prepared and modified by the principal investigator based on the objectives of the study in English. To ensure consistency, it was first translated into Amharic and then back into English. Data collectors and supervisors received three days of intensive training on the purpose, study tools, and overall data collection procedures. Furthermore, pre-testing of the study tools was done in 5% of the total sample size or 12 consecutively selected overweight/obese COVID-19 patients admitted to Bishoftu hospital to identify potential problems in data collection tools. This pre tested data was not considered as part of the main data (excluded from the main data).
The principal investigator provided regular supervision and support to the data collectors. A field manual for data collection mentioning standardized techniques was developed by the principal investigator to collect data in a similarly standardized way. Calibration was also carried out for beam balance and stadiometer. The supervisor and principal investigator checked each collected data for completeness and clarity every day. To verify the accuracy of data entries; minimize errors, a pre-arranged coding sheet was prepared before entering the questionnaire results into the Epi Data. After the completion of data entry, randomly selected 10% of the questionnaires were checked for errors and inconsistencies. Using SPSS Version 25.0, missing values and outliers were properly checked and managed. Frequency distribution and cross tabulation were performed for every variable before any other statistical analysis. Intra and inter observer reliability was checked by: Pearson correlation coefficient (r) for both was 1.