Study design and setting
An institutional-based cross-sectional study was conducted at Wolaita Sodo University Teaching Referral Hospital in Wolaita Sodo town from March 15 to May 15, 2021. Wolaita Sodo is 313 km from Addis Ababa, Ethiopia’s capital. It is one of the ten kebeles in Ethiopia’s Southern Region that make up the zonal administrative divisions of cities. There were 4963 women between the ages of 15 and 49, 28,499 children under the age of five, and 4576 infants under the age of one year in the reproductive age group. In the town, there are two hospitals (one public and one private), three health facilities, 19 medium and lower level clinics, and 17 health posts [25, 26].
Wolaita Sodo University Tertiary Referral Hospital (WSUTRH) was founded in 2007 and now offers outpatient and inpatient medical, surgical, pediatric, psychiatric, ophthalmology, emergency, gynecological, and obstetric services. The hospital serves more than three million people in Wolaita and neighboring zones, with an average bed capacity of 416 and roughly 1155 health professionals and other technical and other personnel [27].
The pediatric emergency admission unit, the Neonatal Intensive Care Unit (NICU), the pediatric surgical admission ward, the pediatric medical admission ward, and the stabilization center (SC) unit are the four main wings of the pediatrics department, which have a total of 133 patient beds. There are 29 beds in the pediatric emergency department [27].
Participants of the study
All children who visited Wolaita Sodo University Hospital’s pediatric emergency department were included in the study’s source population. Children who visited a pediatric emergency department during the data collecting period were included in the study. Children who died shortly after arriving and patients who were critically ill were, on the other hand, omitted from the study.
Sample size determination and sampling procedure
The sample size was calculated using a single population proportion formula based on the following statistical assumptions: P = 50% (assuming a 50% proportion of pediatric emergency department length of stay because there is no published study on the study area of interest), 95% confidence level, and a 5% margin of error. The sample size was calculated using this formula, and a 10% non-response rate was applied. As a consequence, data from 422 pediatric patients who visited the pediatric emergency department was acquired. The actual study participants were chosen using a rigorous random sample technique at Wolaita Sodo University’s pediatric emergency department. The average number of pediatric patients who visited Wolaita Sodo University Teaching and Referral Hospital’s pediatric emergency department in the three months prior to the study was determined using client registration to select the desired sample of pediatric patients among Wolaita Sodo University Teaching and Referral Hospital’s attendants. The estimated client flow rate for the research period (one month) was 1,200 based on this. After that, the sample interval (k) was computed by dividing the expected number of mothers visiting the unit throughout the study period (N) by the number of respondents (n) (1200/422 = 2.8). Finally, a systematic random selection process was used to select one out of every three attendant child combinations until the required sample was reached.
Data collection tools and procedures
To obtain the necessary data, face-to-face interviews, semi-structured questionnaires, and chart reviews were employed. Tools were adapted and modified from previous studies [5, 10, 14, 22, 28, 29]. The questionnaire is divided into four sections. The first section contains 13 questions about the study participants’ socio-demographic characteristics. The second section includes five questions about time factors affecting the length of stay. The third section contains 11 questions about clinical aspects of the participant acting ED length of stay. The fourth section has 16 questions about organizational variables.
The data was collected thanks to the training of five nurses for data collection and one nurse for supervision, all of whom are Amharic and Wolaitigna native speakers. A pretest was provided to the twenty-one respondents at the Arbaminch referral hospital. The inter-observer variability was also investigated, and a correction based on the pretest results was used. The language fluency of the study participants was examined, and they were offered the choice of Wolaitigna or Amharic interviews. After confirming their willingness to engage in the study, verbal agreement and assent were obtained. Data was obtained from patients, and their parents via interviewer administered based questionnaire, and medical record inspections by qualified data collectors following their presentation, admission, and discharge from the ED.
The data gathering technique is depicted in the diagram below: First, data collectors in the triage department identified eligible patients, while resident personnel triaged patients as they came. As a result, when the patient entered the triage room, the time and other presentation factors were logged. Following that, at various treatment stages, additional information such as socio-demographic factors and other organizational-related elements was obtained through interviews and chart review. Once the patient was stabilized, diagnostic investigations and comprehensive treatment data were eventually recorded from medical records. The patient’s overall length of stay and final disposition, on the other hand, were documented as soon as the patient was discharged from the emergency department. Later, the essential time-related factors were ED arrival time and the overall length of stay between presentation and discharge from ED.
Operational definitions
Pediatrics emergency department length of stay (PEDLOS): According to the Ethiopian federal minister of health, the duration between ED arrival and ED discharge, hospitalization, or referral to another health facility should not exceed 24 h [28].
Prolonged length of stay: A patient in the emergency department for more than 24 h is characterized as [28].
Waiting time: is a period that begins when a patient arrives at the ED and ends when the patient is triaged by health care providers and should not exceed 5 min [29].
Data processing and analysis
Data were coded and put into Epi Data version 4.6 software after being checked for completeness, and then exported to SPSS version 26.0 for additional data analysis. Essential variables were provided in percent frequency tables and figures, as well as descriptive statistical analyses. Outliers and multicollinearity were checked in standardized residuals using the variance inflation factor, and variables with a variance inflation factor of more than ten were excluded. The Hosmer-Lemeshow goodness of fit test was used to assess model fitness.
To adjust for all possible confounders and identify the essential factors, independent variables with a p-value of 0.25 in the bivariable logistic regression analysis were included in the analysis. In the multivariable logistic regression adjusted odds ratios with 95% confidence intervals were used to assess the strength of the relationship between the dependent and independent variables with a p-value of < 0.05.