Available 2007 hospital records revealed neonatal sepsis/pneumonia and mortality rates of 5.7% and 2.0%, respectively. Many interventions, such as drying, weighing, examining, providing eye prophylaxis and supplying vitamin K, were observed. Timely birth doses of hepatitis B vaccine tripled from baseline levels (25
). Unfortunately, these interventions were performed in sequences that did not allow the newborns to benefit from all
of their mothers’ natural protection in the first hour of life, i.e. provision of warmth, blood transfusion from the placenta, protection from infection via skin-to-skin contact and completion of colostrum feeding (8
). Cohorting newborns in ‘nurseries’ for mandatory observation periods and clearance exposed them to hospital-acquired flora.
Hospital practices permitted the overwhelming majority of newborns to be exposed to cold, similar to practices in other countries (27
). Only one of 26 newborns with apnoea was dried. Most newborns with primary apnoea will start breathing from stimulation during drying.
This evaluation found that 68.2% were put to the breast, similar to national survey results (3
). However, this process was not optimal because the infants’ mouths were pried open, positioned onto the areola and their cheeks were stroked to trigger rooting only 10 min after birth, a time before newborns are ready to breastfeed (14
). They were allowed only two minutes of this forced attachment.
The Academy of Breastfeeding Medicine (23
) states that, for a healthy newborn, procedures should be delayed to allow early parent–newborn interactions and the first breastfeed. Newborns were typically separated for weighing, examination, eye prophylaxis and injections at only 12 min after delivery. The majority (88%) were reunited with their mothers at 155 min, a time when newborns younger than 24 h are usually asleep. When newborns eventually initiate breastfeeding, the risk for infection-related death is doubled or tripled (21
The need for basic life-saving interventions and for beneficial parent–newborn interactions indicates that procedures carried out immediately after delivery should be standardized in time and order. Unnecessary procedures, such as routine suctioning, early bathing and separation of newborns from their mothers, should be discontinued. Aside from potential for harm, these procedures burden already overworked hospital staff.
These findings should not be surprising. Pre- and in-service training in medical and allied schools do not address key WHO guidelines in newborn care (28
). WHO guidelines have not emphasized the importance of the timing of early interventions for newborns. Even with formal training, optimal outcomes may not be realized (29
). This is particularly true when the physical and policy environments do not enable appropriate newborn care practices or disable outdated and inappropriate practices.
This evaluation was limited to the largest hospitals in only nine of the 17 regions. However, the deliveries in the 51 hospitals accounted for more than 10% of the deliveries nationwide. It did not include either home deliveries or those in smaller centres. We focused on the large hospitals because changing practices here will affect training of obstetric/paediatric residents and midwifery/nursing/medical students.
Extraction of data from hospital records was limited by problems inherent in hospital recording. Many hospitals had missing data for key outcome variables. The sensitivity analysis revealed that, regardless of missing data, a high burden of disease exists in these large hospitals. We could not validate reports of zero deaths or sepsis. Accurate reporting of sepsis cases was limited by variation in interpretation of clinical presentations and laboratory results. While gestational ageing techniques were not uniformly performed in the hospitals, birth weight categories were available.
Despite the Hawthorne effect implicit to the observational methodology, serious hospital practice issues were evident. The problems are likely to be more severe than our evaluation uncovered. Recording bias is another potential limitation for data on rooming-in and breastfeeding after the 2-h observation window.
A secondary analysis of a nationwide survey revealed death rates of neonates born to women delivering in a healthcare facility to be statistically similar to those born at home attended by a non-health professional (OR 1.0; 95% CI, 0.63–1.57) (26
). This comparative data together with the findings of the present evaluation set off a series of responses to address inappropriateness and lack of standardization of immediate newborn care practices. The Department of Health convened a technical working group to review current evidence and draft evidence-based recommendations. The resulting Essential Newborn Care protocol then underwent expert and stakeholder panel review in a guideline development process. It specifically defined the time for each intervention and made explicit statements to stop unnecessary interventions. It has now become an official publication (30
). The DOH is spearheading strategies to (i) jumpstart the hospital reform agenda as the next phase of health sector reform; (ii) develop model hospitals and networks of excellence in each of 17 regions; (iii) update the pre-service and in-service medical, nursing and midwifery curricula and (iv) conduct a nationwide social marketing campaign. Preliminary results show that the one hospital that implemented the protocol has seen historically low neonatal deaths.
Globally, 450 newborns die every hour (5
). Their limited reserves and defences make newborns vulnerable to attendant practices. The scope and seriousness of threats, even with skilled attendants, were not clear until embarking on this direct observational assessment. We believe this is the first evaluation of its kind that enables quantification of the timing and performance of various interventions.
Although standards for immediate newborn care exist in high- and low-income countries alike, direct observational studies may uncover substandard practices. Even in developed countries, the sequence and timing of critical interventions may still require changes or standardization.