A systematic review identifying common data items in neonatal trials and assessing their completeness in routinely...

We aimed to test whether a common set of key data items reported across high-impact neonatal clinical trials could be identified, and to quantify their completeness in routinely recorded United Kingdom neonatal data held in the National Neonatal Research Database (NNRD).

We systematically reviewed neonatal clinical trials published in four high-impact medical journals over 10 years (2006-2015) and extracted baseline characteristics, stratification items and potential confounders used to adjust primary outcomes. Completeness was examined using data held in the NNRD for identified data items, for infants admitted to neonatal units in 2015. The NNRD is a repository of routinely recorded data extracted from neonatal Electronic Patient Records (EPR) of all admissions to National Health Service (NHS) Neonatal Units in England, Wales and Scotland. We defined missing data as an empty field or an implausible value. We reported common data items as frequencies and percentages alongside percentages of completeness.

We identified 44 studies involving 32,095 infants and 126 data items. Fourteen data items were reported by more than 20% of studies. Gestational age (95%), sex (93%) and birth weight (91%) were the most common baseline data items. The completeness of data in the NNRD was high for these data with greater than 90% completeness found for 9 of the 14 most common items.

High-impact neonatal clinical trials share common data items. In the United Kingdom, these items can be obtained at a high level of completeness from routinely recorded data held in the NNRD. The feasibility and efficiency using routinely recorded EPR data, such as that held in the NNRD, for clinical trials, rather than collecting these items anew, should be examined.

PROSPERO registration number CRD42016046138. Registered prospectively on 17 August 2016.