Optimizing Helping Babies Breathe implementation in a resource limited setting to improve perinatal outcome

Abstract

Globally, deaths around the time of birth are unacceptably high of which there are about 2.7 million neonatal deaths and 2.6 million stillborn annually. Perinatal mortality accounts for deaths after 28 weeks of gestation to seven days after birth. Perinatal mortality is a public health concern with a huge impact on the health, social and economic well being affecting both family and society. The burden of perinatal mortality is obvious in the low and middle-income countries, and more on the countries south of sub-Sahara Africa. East Africa being among the sub-Sahara countries is also experiencing a high rate of perinatal mortally, with Tanzania taking the lead. Almost half of stillborns are alive at the start of labor offering an opportunity for prevention. Likewise, 44% of the neonatal deaths occur on the first day of life and are predominantly the result of intrapartum events. Events during labor, including birth asphyxia (interruption of placental blood flow) account for one-quarter of the global newborn deaths. These deaths can be substantially reduced by improving quality of care around the time of labor and childbirth. The Helping Babies Breathe (HBB) curriculum, involves training to improve knowledge and skills of midwives and other birth attendants, to provide improved basic care to new-borns including timely breathing support and as a consequence to improve newborn survival as needed. It became apparent the course did not alter outcome following one day training i.e. improve survival. This pointed to the need for more frequent training. Thus the concept of frequent brief onsite simulation (FBOS) HBB training was introduced at Haydom Lutheran Hospital as part of continuous quality improvement (CQI), in an effort to reduce perinatal mortality. This became the central thrust of this thesis as described below. Aim: The aim of this thesis is to evaluate the process and impact for optimizing implementation of HBB project to improve early perinatal outcome at Haydom Lutheran Hospital, a rural referral hospital in north-central in Tanzania. Method: We conducted three studies between February 2011 and January 2017 to evaluate the process and impact of HBB project to improve early perinatal outcome. The study site was labor ward and operating theatre at Haydom Lutheran hospital. The study intervention involved implementation of FBOS training using a low fidelity manikin with the ability to provide bag/mask ventilation and feel a pulse on different simulation scenarios and also having repeated feedback. Study I was a one-year project from February 2011 through January 2012 that involved FBOS. This was a before-after prospective education intervention study in a cohort of midwives (birth attendants), pregnant women attending to give birth and their newborns. The labor management process and outcomes of birth in the first 24hrs were evaluated. The outcome of pregnancy (n=4814) was compared to a baseline period (n=4894), which was also a one-year period between February 2010 through January 2011. Secondary outcomes included care provider change in behavior i.e. frequency in resuscitation practice, labor management which involved, fetal heart rate monitoring, mode of delivery and resuscitation practice. Study II was a five years follow-up from February 2011 through January 2016. Perinatal outcome during the study period was compared to the baseline period as in study I (Feb. 2010 through Jan 2011). The study involved continuous observation to trace and document perinatal outcomes over time and evaluate the implementation process. The cohort involved 22,176 newborns and compared the outcome to the baseline (n=4894). Factors included in the analysis involved those with potential co-relationship with perinatal outcomes as interventions, administrative events and facility process. Study III was also a continuous observation to trace and document perinatal outcomes as in study II. The cohort involved a total of 31122 newborns of which intervention period was for six years from Feb 2011 through Jan 2017 with 26220 newborns and one year of baseline period (Feb 2010 through Jan 2011). Logistic regression modeling was used to construct risk-adjusted variable-life adjusted display (VLAD) and cumulative sum (CUSUM) plots to monitor changes in perinatal survival (primary outcome). Plots of unadjusted changes in perinatal risks were compared to risks adjusted plots. Results: In Study I, There was a significant reduction in early neonatal mortality rate (eNMR) from 11.1/1000 during baseline to 7.2/1000 (p0.040) after implementation of FBOS HBB training. During the period, the proportion of resuscitation through stimulation increased from 14.5% to 16.3% (p 0.016), and suction increased from 13.0% to 15.8% (p ≤ 0.0005) while the proportion receiving bag-mask ventilation (BMV) decreased from 7.3% to 5.9% (p ≤ 0.005) in Cohort 1 versus Cohort 2, respectively. In study II, the CUSUM plot in most of the period was lower than the baseline level of 2.7% with slight variation on ePMR months indicating reduction after implementation of FBOS HBB training. In the VLAD plot there was a continuous upward trend on cumulative monthly number of lives saved compared to baseline, with few fluctuations indicating that the outcome (perinatal survival) was better than in the baseline. The trend indicated continuous improvement in perinatal outcome during the five years follow-up period. The trend of outcomes had some variations in some point, which could be linked with different interventions and events of which improvement in survival linked refresher HBB training and reduced survival linked trained midwifes leaving the hospital and new recruited who have not attended FBOS HBB training. The VLAD plot showed an overall positive trend, reflecting more than 120 extra lives saved over the 5-year period. In study III, Persistent and steady increase in perinatal survival was observed following implementation of FBOS HBB training. Six years follow-up revealed 150 extra lives saved according to VLAD plot. After adjusting for the risk factors VLAD plot indicated that an estimated 250 extra lives were saved which indicate that survival was maintained even when the cohort included high risks cases indicating a further improvement in survival compared to when the risks were not considered. Conclusion: This PhD project show that optimizing the implementation of FBOS simulation training is associated with improvement on clinical practice and neonatal survival. This is the first published report that documented the important association of FBOS and reduce neonatal mortality. During the CQI, continuous evaluation in the SPC revealed that the improvement in perinatal outcome matched with the activities related to FBOS training. Additionally, the reduction on perinatal mortality was even more evident when adjusting for risks in the cohort. To conclude, optimizing implementation of HBB training has the potential to improve perinatal outcome.