The Effect of Newborn Care Training Among Health Care Workers on Feeding Practices in Very Low Birth Weight Neonates in Kenyatta National Hospital and Pumwani Maternity Hospital: a Before and After Study

Abstract

Background: Optimal nutrition in Very Low Birth Weight (VLBW) neonates is essential in decreasing mortality and long-term morbidities such as postnatal growth restriction and poor neurodevelopmental outcomes. Given appropriate nutrition, postnatal growth in preterm neonates should parallel that during the intrauterine period. Unfortunately, research done in the newborn units in Kenya showed that only about half of all admitted neonates receive appropriate feed volumes on their first day of life. Less than two-thirds had the right volume for Intravenous fluids, highlighting the need for strategies to remedy this. One of the strategies to address this challenge is ensuring that health workers have adequate knowledge in appropriate neonatal feeding practices. Main objective: To determine the effect of the Newborn Care Training among health care workers on feeding practices in VLBW neonates in Kenyatta National Hospital (KNH) and Pumwani maternity hospital (PMH). Study Design: This was a before and after intervention study. Intervention: A 5-day Newborn Care Training of staff was conducted in the newborn units of KNH and PMH on 1st – 5th March 2021 and 22nd - 26th February 2021 for KNH and PMH respectively as part of ongoing national quality improvement programme. In addition, half day trainings with a focus on feeds were conducted to ensure all health care providers in the new-born units in both hospitals were trained. Methodology: The study was conducted in the Newborn Unit (NBU) of Kenyatta National Hospital (KNH) and Pumwani Maternity Hospital (PMH); Data was extracted from REDCap data platform in both hospitals, KNH and PMH. This study had 16 time points – eight before the intervention (July 2020 – February 2021) and eight after the intervention (June 2021 – January 2022). Random sampling was used to select neonates at the individual time points, up to a sample size of 36 at each time point from both sites. We abstracted data on feed prescriptions made by health care workers for all VLBW neonates on their first day in the different time points in both pre- and post- interventional periods. Analysis used a balanced design in which the number of pre – training and post- training time points was equivalent. At each time point we calculated the proportion of correct feed prescriptions as a fraction of all the prescriptions at that time point. These proportions were then modelled using segmented beta regression. Results: A total of 435 records met the inclusion criteria for the study. 181 in the pre-interventional period and 254 in the post interventional period. There were 38.1% (69/181) and 45% (115/254) correct prescriptions in the pre-intervention and post intervention period respectively (p value = 0.14). The intervention resulted in an improvement of the composite indicators for correctness of feed prescription +7.2% (-2.7 ,17.0), however, the difference between the two periods wasn’t statistically different (p = 0.14). There was a significant decrease in the proportion of stable VLBW neonates on intravenous fluids from 72.6% to 54.5% (p = 0.05, effect size = -18.1%). In the segmental analysis, there was a decreasing trend of correctness of prescriptions before the training and immediately after the training there was step increase in the number of correct prescriptions. However, with longer post intervention time, the correctness of prescriptions plateaued from around the 2nd month. Two factors were significantly associated with correct feed prescription i.e., stability and the referral status of the child (p-values < 0.001 and 0.001 respectively). Conclusions: There was no statistical effect of the newborn care training among heath care workers on feeding practices in VLBW neonates in KNH and PMH. There was a step increase in number of correct prescriptions, however the increase levels off at about 2 months after the intervention and remains at closely the same level after that. Neonates who were unstable and those who were referred from other health care centers were more likely to have a correct feed prescription compared to the rest of the VLBW neonates Recommendations: Regular short focused in-service training on feeding of VLBW neonates should be conducted to improve correct feed prescriptions per the national and WHO guidelines in order to improve neonatal outcomes in newborn units.