Aims Despite the progress made in reducing under five mortality in line with Millennium Development Goal four, the mortality rate remains high worldwide. The most vulnerable time is the neonatal period with 44% of deaths of children under 5 in 2013 occurring within the first 28 days of life. The aim of this analysis is to compare neonatal care in three Kenyan government hospitals to uncover common themes that may be addressed to improve care and potentially reduce Kenyan neonatal mortality. Methods A retrospective review of case notes of newborn unit (NBU) admissions at three government hospitals was conducted over a two month period. Data was collected on demographics, diagnoses, observations and prescriptions. This was done by RCPCH Global Links Programme volunteers and local partners.

Results A total of 310 neonates were admitted with 221 case notes reviewed (88 excluded due to unavailability of notes and 1 as they were admitted at 2 weeks of age). Performance in each area was varied across hospitals. Overall, 29% of neonates were hypothermic on admission, 4% had blood glucose documented on admission and of those, 22% were hypoglycaemic. The accuracy of prescribing first line antibiotics varied with 0.8%–18% of gentamicin prescriptions being incorrect. Documentation of vitamin K and tetracycline eye ointment administration ranged from 21.6% to 69.7%. Overall mortality rate for the period studied was 14% (range 11%–24%).

Conclusions This analysis of clinical practice identified both areas of success, and those with potential for improvement. There is a need for ongoing training and education in addition to improvement of the standardised national neonatal admission proforma. Particular areas of strength and weakness in care varied between the hospitals, therefore it may be postulated that by NBU’s sharing experiences, neonatal care could be improved countrywide.