The Study Design, period and population
An institutional-based prospective follow up study was conducted. The study was conducted in neonatal intensive care units of Debre Markos, Felege Hiwot, and Tibebe Ghion comprehensive specialized hospitals, that the former found in Debre Markos town and the latter two found in Bahir Dar town, Amhara.
Debre Markos is located 300km from Addis Ababa, the capital city of Ethiopia, and 256 km from Bahir-Dar, which is 556 km away from Addis Ababa. Debre Markos (DMCSH), Felege Hiwot (FHCSH), and Tibebe Ghion (TGCSH) Comprehensive Specialized Hospitals have 162, 230, and 151 monthly and 1845, 2479, and 1652 annual neonatal admission according to the 2012 EFY report, respectively. Of these 63, 89, and 67; and 612, 834, and 660 were monthly and annual admissions of preterm neonates, respectively. These hospitals are equipped with 28, 43, and 27 neonatal beds and 24, 35, and 18 maternal beds respectively.
All of these hospitals are the final referral choice for other health institutions around and provide tertiary level neonatal care and are organized with necessary materials, equipment and health professional mixes.
These are organized into different service areas including term, preterm, isolation, and procedure rooms with KMC and maternal waiting rooms. The major services are general neonatal care, blood and exchange transfusion, phototherapy, and ventilation support such as CPAP. This study was conducted in these hospitals from October 1 to November 30/2020. Study participants were recruited and followed for the first seven postnatal days.
The study population included all preterm neonates admitted to the neonatal intensive care unit of Debre Markos, Felege Hiwot, and Tibebe Ghion comprehensive specialized hospitals during the study period.
Inclusion Criteria
Neonates born after 28 completed weeks, but before 37 completed weeks, and admitted to the neonatal intensive care unit of these hospitals.
Exclusion Criteria
Neonates who had started direct breast milk or other option of feeding before the time of admission, pre-diagnosed stage II/III necrotizing enterocolitis, stage III perinatal asphyxia, unknown gestational age, unknown APGAR score, and birth weight were excluded from this study.
Sample Size and sampling technique
The sample size was computed by using STATA (version 14) considering this statistical assumptions; two-sided significant level (α) of 5 %, power 80 %, Za/2= Z value at 95 % confidence interval = 1.96, death rate = 50%, Hazard Ratio (HR) = 0.5, Survival probability = 0.5, the proportion of withdrawal = 0.15 (45).
The study population was preterm neonates on whom trophic feeding is more commonly practiced. Neonates that doesn’t fulfill eligibility criteria were recruited from respective study hospitals. From neonates who were twin or triplet, only one of them was included by lottery method at each spot. A total sample of 210 children were finally selected randomly.
Data Collection Procedure
Data was collected using a semi-structured pretested English version questionnaire and extraction checklist through face to face interview and chart review. The content of the questionnaire includes neonatal and maternal socio-demographic variables, neonatal and maternal related factors, and health service-related factors. The data extraction checklist and questionnaire were adapted from different related literature, books, and guidelines (13, 22, 28, 35, 36, 37, 42–44, 47, 48).
Besides the principal investigator, six nurses working at NICU, two from each respective hospital, as a data collector, and three nurses as a supervisor were participated throughout the data collection process.
Baseline data were obtained soon after admission, and the rest of the data were obtained every day in the follow-up period. Supervisors had been following the data collection process every day.
Data Processing and Analysis
The data were checked for completeness, coded, and entered into Epi-data version 3.1; and exported to Stata/SE 14.0 for data cleaning and analysis.
Continuous data were reported with a mean (standard deviation) and median (interquartile range). The data with categorical nature was described with frequency and proportion. The outcomes of study participants were dichotomized into (code ‘1’) as a failure and (code ‘0’) as a censor. Some continuous variables were categorized for ease of analysis and otherwise used as continuous. The variance inflation factor (VIF) and correlation matrix were used to assess multi-collinearity.
The Kaplan Meier survival curve was used to estimate survival time, and a log-rank test was used to compare the survival curves of categorical variables. The necessary assumption of the Cox-proportional hazard regression model was checked using the Schoenfeld residual test, the graphical methods, and the presence of a time-dependent covariate. The overall model adequacy and fineness were assessed using the Cox-Snell residuals and global fit test, respectively. Then, bi-variable Cox-regression was computed for each predictor variable, and a P-value of < 0.25 was used as a cut-off point to enter variables to multi-variable Cox-regression. The variables were selected through backward stepwise procedures. The confounding effect was minimized using proper inclusion and exclusion criteria and a multi-variable analysis.
The result of the final model was expressed in terms of adjusted hazard ratio (AHR) with 95% confidence intervals. The significant association was declared with a p-value less than 0.05 in a multivariable Cox regression model. Finally, the result of this study is presented with tables, graphs, or text narrations.
Operational Definitions and Definition of Terms
Early feeding: neonates start trophic feeding within 24 hours of birth (28, 30).
Delayed feeding: neonates start trophic feeding after 24 hours of birth(30).
Survival time: the length of time in hours followed starting from birth to the first trophic feeding
Event: the neonates who had started first trophic feeding within the follow-up period.
Censored: neonates who died left against medical advice, transferred or referred before starting trophic feeding, or not started at end follow-up.
Follow up time: time from birth to the first seven days of life.
Hemodynamic instabilities: Blood group and RH incompatibility, anemia, polycythemia, bleeding disorders, blood glucose disturbances (46).
Trophic feeding: The first minimal enteral feeding to prime the gut regardless of method or volume (30).