Study design, area, and period

An institution-based unmatched case-control study was conducted from January 1, 2021, to March 15, 2021.

The study was conducted at AHMC, Adama, Ethiopia. Adama is a city located in the Oromia region, 100 km southeast of Addis Ababa, in the Great Rift Valley of East Africa. It is one of the major cities in Ethiopia, having an area of 29.86 square kilometers and a population of more than half a million. The city has four hospitals, eight public health centers, and more than one hundred thirty private health care facilities. AHMC is the first and only public and referral hospital in the city. As reported by the administrative office of AHMC, the hospital serves a population of more than 6 million from five regions (Oromia, Amhara, Afar, Somali, and Dire-Dawa). The hospital has a capacity of 527 beds for admission and serves an average of 1000 patients per day in 15 OPDs in 10 different specialties. The Department of Pediatrics is one of the treatment teams with the capacity to handle 40 cases per day at the Pediatric OPD.

Population

The source population of the study were all 2–59 months old children visiting AHMC pediatric OPD for different reasons during the study period.

Cases

Cases were defined as 2–59 months old children who visited the Pediatric OPD of AHMC and were diagnosed with pneumonia within the study period. The assigned physician were the ones to make the diagnosis using the World Health Organization (WHO) Integrated Management of Childhood Illness (IMCI) guideline adopted by the Ethiopian Government.

Controls

Controls were defined as 2–59 months old children who visited the Pediatric OPD of AHMC and were diagnosed with non-pneumonia cases within the study period.

Sample size determination

Sample size calculation for two population proportions was employed using EPI info version 7.2.4.0 software. The assumption was based on a 95% confidence level, 90% power, 1:1 case-to-control ratio, history of having no separate kitchen as the main predictor of Pneumonia, the percent of controls exposed as 3.4%, and the Odds Ratio (OR) of 5.37 from a case-control study in Northeast Ethiopia [11]. The estimated sample size was 236. After adding a 5% nonresponse rate, the total sample size becomes 248 (124 cases and 124 controls).

Sampling procedure

Cases were selected using a consecutive sampling technique until the required sample size was attained. For each case of pneumonia, the next visitor or patient from the same pediatric OPD who met the inclusion criteria was taken as a control. Finally, mothers or caregivers of the cases and controls were interviewed.

Inclusion and exclusion criteria

Mothers/caretakers who had children 2–59 months of age coming to AHMC for health care services were included in the study, and respondents who were critically ill during the data collection period were excluded from the study.

VariablesDependent variable Independent variables

Sociodemographic variables (age, sex, birth order, family size, area of residence, age of mother, occupation of parents, education status of parents, marital status of parents, and average monthly income).

Environmental factors (mainly used room for cooking, the place where the child stays during cooking, type of fuel used for cooking, presence of windows, source of water, hand washing before child feeding, cigarette smoking in the house, and domestic animals in the house).

Nutrition and vaccination status (exclusive breastfeeding up to 6 months, current breastfeeding status, age when complementary feeding started, zinc supplementation, and malnutrition).

Self or family history of illness (HIV status of the child, history of diabetes, diarrhea in the past two weeks, upper respiratory tract infection (URTI) in the past two weeks, pneumonia in the family in the past two weeks, history of measles illness, history of parental bronchial asthma).

Operational definitions Pneumonia

A child with cough or difficult breathing who has fast breathing and no general danger signs, no chest indrawing and no stridor when calm, was considered to have pneumonia.

Malnutrition

Malnutrition was considered when children had a mid-upper arm circumference (MUAC) < 125 mm or weight for length/height < -2 z-score (WHO standard).

Vaccination status

Children who had an age-inappropriate vaccination according to the Expanded Program on Immunization schedule, and those who had never been vaccinated were considered to have inadequate vaccination status. At the same time, those who had an age-appropriate vaccination and those who had completed vaccination were considered to have adequate vaccination status.

Data collection procedure

The data were collected using a pretested, structured interviewer-administered questionnaire. The questionnaire included sociodemographic characteristics, environmental factors, nutritional factors, comorbid illnesses, and related care practices. The questionnaire was adapted from previous studies with minimal modification. After each interview, a record review was done to collect information on the MUAC, height, weight, and HIV status of the child.

Data quality control

To assure the quality of data, the questionnaire was pretested with 5% of the targeted sample size at AHMC before running the actual data. Necessary modifications of the questionnaire were carried out based on the pretest feedback. The reliability of the questionnaires was checked. The data were collected under regular supervision after giving a one-day training for three data collectors and a supervisor before the initiation of the data collection process. The data collectors and the supervisor were nurses with prior training in IMCI and not an employee of AHMC. All questionnaires were reviewed and checked for completeness by the supervisors and principal investigator daily and the necessary feedback was being provided to data collectors.

Data processing and analysis

The data were checked for completeness and consistency, and then cleaned and entered using Epi info version 7.2 0.4.0 and exported to SPSS version 26 for analysis. The descriptive analyses were presented by frequency tables and percentages. A multiple logistic regression model was fitted to find factors associated with pneumonia. Initially, all variables were screened by carrying out binary logistic regression analyses. Then, variables that had a p-value of less than 0.2 were taken into the multivariable logistic regression model to identify confounders and the independent effect of different factors on the occurrence of pneumonia. A p-value less than 0.05 and an adjusted odds ratio with 95% CI were used to determine the presence and degree of association.