Retention and use of newborn resuscitation skills following a series of helping babies breathe trainings for midwives in rural Ghana

Background: The Helping Babies Breathe (HBB) program teaches basic newborn resuscitation techniques to birth attendants in low-resource settings. Previous studies have demonstrated a decrease in mortality following training, mostly in large hospitals. However, low-volume clinics in rural regions with no physician immediately available likely experience a greater relative burden of newborn mortality. This study aimed to determine the impact of HBB trainings provided to rural Ghanaian midwives on their skills retention and on first 24 hour mortality of the newborns they serve. Methods: American Acadamy of Paediatrics (AAP)-trained Master Trainers conducted two 2-day HBB trainings and 2-day refresher courses one year later for 48 midwives from Ghanaian rural health clinics. Trainee skills were evaluated by Objective Structured Clinical Examination (OSCE) at three time points: immediately after training, four months after training, and four months after the refresher. Midwives recorded the single highest level of resuscitation performed on each newborn delivered for one year. Results: 48 midwives attended the two trainings, 32 recorded data from 2,383 deliveries, and 13 completed OSCE simulations at all three time points. The midwives' OSCE scores decreased from immediately after training (94.9%) to four months later (81.2%, p < 0.00001). However, four months following the refresher course, scores improved to the same high level attained initially (92.7%, p = 0.0013). 5.0% of neonates required bag-mask ventilation and 0.71% did not survive, compared with a nationwide first 24 hour mortality estimate of 1.7%. Conclusions: The midwives' performance on the simulation exercise indicates that an in-depth refresher course provided one year after the initial training likely slows the decay in skills that occurs after initial training. Our finding that 5.0% of newborns required bag-mask ventilation is consistent with global estimates. Our observed first 24 hour mortality rate of 0.71% is lower than nationwide estimates, indicating the training likely prevented deaths due to birth asphyxia.