Study Design, Setting and Period
An institutional-based unmatched case-control study design was used to assess the determinants of pneumonia in children aged between 2 to 59 months at FGH March 2018. Fitche General Hospital is a 102 bed Hospital located approximately 114 km from Addis Ababa in the Fitche District of the North Shoa Zone of Oromiya. It is one of the government hospitals in the Oromiya Regional State. It serves more than 1.6 million people and is a referral center for 57 Health Centers and 297 Health Posts from North Shoa Zone.
Study population
The study population consisted of children aged between 2 to 59 months and their mothers or caregivers at FGH with pneumonia (case) and those who come without pneumonia (control). The study enrolled mothers or caregivers of children aged between 2 to 59 months that are cases and controls.
Cases were defined as a child aged 2 months to 59 months who received a positive diagnosis of pneumonia by trained pediatricians according to the World Health Organization (WHO) Integrated Management of Childhood Illness (IMNCI) guideline adopted by Ethiopian government.23-25
Controls were defined as a child aged 2 months to 59 months who brought to FGH with other than respiratory complaints (who came for immunization and growth monitoring service). Cases and Controls were recruited within the same facilities, period, and age group. The study excluded children belonging to the same household either cases or controls and mothers or caretakers who have hearing impairments.
Sample size calculation and sampling procedure
Sample size was determined using Epi Info version 7.2 software Stat Calc menu for unmatched case control study. By considering a 95% confidence level, 80% of Power, 1:1 ratio of controls to cases, 23.8% percent of controls exposed, 2.25 odds ratio, and 41.3% percent of cases with exposure. Through reviewing previous studies, the overcrowding is an exposure variable for pneumonia that gave the highest sample size 123 of cases and controls.16The total estimated number of pneumonia patients (cases) for six months was 792 from the pediatric outpatient department (OPD) and the emergency department. Taking the average number of pneumonia cases per month (792/6 = 132), all cases were selected consecutively during data collection until the required number of cases was obtained. Conversely, 5,178 controls from growth monitoring and expanded program of immunization (EPI) units were visited in the last six months based on the registration book. The average number of controls for one month (5,178/6 = 863) and the sampling fraction of control was (863/123=7). A systematic random sampling technique was used to select the study control every seven intervals. The first participant was selected by the lottery method from 1st to 7th and continued every other control until we finalized the calculated sample size.
Data collection tools and procedures
Data were collected by four trained professional nurses using face-to-face interviews with a pre-tested and standardized questionnaire. The questionnaires were developed objectives based on a different literature review.16, 17, 18 It contains socio-demographic factors (age of the child, sex of child, residence, marital status of mother or caregiver, educational status of the mother or caregiver, religion, monthly income), and house factors (radio or TV in the house, number of house room and number of participants who live in the same house, the type of floor, and type of roof). It also contains childhood illnesses (history of vaccination, history of measles infection, history of HIV, and previous history of acute respiratory tract infections were reviewed from the child document). Additionally, the questionnaires contain indoor pollution factors (type of fuel energy used, cigarette smoker in the house, presence of kitchen, kitchen detached from the main house), nutritional status (exclusive breastfeeding and malnourished), and place of delivery were assessed. The questionnaire adopted by the English language and translated to the local language (Affan Oromo and Amharic), and back translation was done to see the consistency of the questionnaire. Pretest was done among a sample of 24 (12 cases and 12 controls) children age between 2 to 59 months at Kuyu General Hospital in the North Shoa Zone of Oromia region. The collected data were analyzed, reviewed and correction made for the final version of the questionnaire. Anthropometric measurement was done using a standardized and calibrated measuring tools for weight (recorded using analog weight scale to the nearest 0.1 grams) and height (to the nearest 0.1 centimeters), and mid-upper arm circumference (MUAC) was also measured using a MUAC tape-record to the nearest 0.1 centimeters. The outcome was measured according to the definition of WHO for pneumonia, age between 2 to 59 months with symptoms of cough or difficult breathing and fast breathing and/or chest in-drawing.20
Data processing and analysis
The collected data were cleaned and checked for completeness; then it was entered, compiled, and analyzed by using the Epi Info for data entry and SPSS software for analysis. Percentages and simple frequencies of the given data were calculated for each variable to describe the findings. Additionally, tables were used to assist data presentation. Bivarible logistic regression analysis was carried out to select variables for multivariable binary logistic regression. Variables with P value ≤ 0.2 in the bivariable analysis were included in a multivariable logistic regression analysis to control the confounding effect among the variables. In multivariable analysis, a p-value of less than 0.05 was considered statistically significant, and adjusted odds ratios with 95% CI were calculated to determine the association.