Implementation of basic training in neonatal resuscitation in the outskirts of Conakry, Guinea: evaluation of neonatal mortality by 'before and after intervention' design.

High neonatal mortality remains a major health problem in Guinea (32 deaths /1,000 live births). This represents 15 000 deaths annually, without improvement over the past decade. We evaluated the impact of 2 days of neonatal resuscitation training of health professionals working in the disadvantaged outskirts of Conakry. Non-randomised interventional study with pre-and post-interventional analysis of the very early neonatal mortality with data collection over two 6-month periods, one before and one after intervention. Intervention: Theoretical and practical training given to health professionals working in private obstetric centres within a defined area. After training, all centres were equipped with basic resuscitation devices. We concentrated on the private sector, dominated by informal facilities scarcely equipped and run by often poorly trained paramedical staff. Outcome measures were very early neonatal mortality (6 hours) and the need for referral to higher equipped structures. Theoretical knowledge was assessed by a questionnaire, pre-training, post-training and 6 months later. 27 nurses, midwives and doctors participated, working in 13 health facilities. They performed 589 deliveries during the two periods analysed. The 6-hour neonatal mortality rate decreased (31.8‰ to 5.7‰, p=0.031), need for neonatal transfer dropped from 27.3% to 11.3% (p=0,19), whereas the stillbirth rate remained high and unchanged. There was a sustained improvement in theoretical knowledge (mean of correct answers: 59.3% before, 82.0% after training, (p<0.001) and 85.9% 6 months later). A 2-day training course for health workers in private facilities and provision of basic neonatal resuscitation equipment significantly improved neonatal outcome.