Presented at the Neonatal Society 2010 Summer Meeting.
Statnikov Y1, Santhakumaran S1, Manktelow B1,2, Modi N1, on behalf of the Neonatal Data Analysis Unit*
1 Section of Neonatal Medicine, Department of Medicine, Chelsea & Westminster Campus, Imperial College London, UK
2 Department of Health Sciences, University of Leicester, UK
* Members of the NDAU Steering Board: J Abbott (BLISS), P Brocklehurst, K Costeloe, E Draper, J Kemp, A Majeed, N Modi, A Wilkinson
Background: Necrotising enterocolitis (NEC) is serious inflammatory intestinal condition occurring in the neonatal period and associated with considerable mortality, and short and long-term morbidity. As a percentage of live births, mortality from NEC has increased in the UK even though total infant mortality has decreased (1). Strategies for NEC prevention remain elusive and population based information is urgently needed.
Aim: To evaluate the potential for surveillance of NEC using data captured in the course of clinical care onto the neonatal.net electronic record.
Methods: Data entered onto neonatal.net is submitted in anonymised format to the Neonatal Data Analysis Unit (NDAU) with NHS Trust Caldicott Guardian approval, from neonatal units across England. These electronic records were analysed for all babies admitted to contributing units. In addition to demographic information, data for this analysis was extracted from two different data item types: daily, updated for each day of care provided, and diagnosis fields at admission and discharge. The daily items in this analysis include a field “NEC Treatment”, which has the options “Medical” or “Surgical”, and a field “Enteral Feed”, which includes options ” Expressed Breast Milk” (mother’s or donor), “Formula” and “Other”. The fields used from the diagnosis items included NEC, based on ICD-10, the episode and point (admission, stay or discharge) at which the diagnosis was made (no field is available for actual date of diagnosis). These data sources were merged together based on the anonymous patient ID code. Each baby was included only once in the analysis.
Results: A complete record from birth was available for 48,250 babies in 94 neonatal units between 1 October 2007 and 30 June 2009. Of 1310 babies recorded as treated for NEC, the final outcome at discharge from the last episode in a neonatal unit was recorded in 1182, of whom 180 (14%) died. There were 136 (12%) who died with a reported discharge diagnosis of NEC. Of the 204 babies recorded as having received surgical NEC treatment, the final outcome at discharge was recorded for 178 babies, of whom 67 (38%) died. Of these 178 babies, 54 (30%) died with a reported discharge diagnosis of NEC. Daily item records were available from birth for 196 babies who had surgical treatment for NEC (Table 2). Some enteral feed information was recorded for 78 (60%) babies in the days from birth leading up to NEC treatment. For 5 (6%) babies the enteral feed entries consisted of either breast milk (mother or donor) or no data at all and another 11 (14%) babies had either a record of receiving formula milk or no record at all.
Table 1. Mortality in babies recorded as having any treatment for NEC, by gestational age; 1 October
2007 to 30 June 2009.
Table 2. Enteral feeding from day of birth for babies recorded as having surgical treatment for NEC, by
gestational age; 1 October 2007 to 30 June 2009.
Conclusion: These data are drawn from approximately half the neonatal unit admissions in England but should be viewed with caution. A definitive, widely accepted case definition for NEC is lacking and even when cases come to surgery, there may be ambiguity in the distinction between NEC and closely related emergencies such as spontaneous intestinal perforation. Reliable surveillance of NEC to examine between centre variation, antecedents and co-morbidities, follow temporal trends, assess long-term outcomes and underpin clinical trials, requires a nationally accepted uniform case-definition. The structure of the data capture for NEC requires amendment as information entry is inconsistent and can be reported in three locations on neonatal.net (daily records, diagnoses, and NEC screening), making analysis difficult. Enteral feeding data reveals very poor completeness. Nonetheless, resolving these issues will permit reliable population based surveillance throughout the UK.
Acknowledgements: Y Statnikov and S Santhakumaran are funded by NIHR programme grant, Medicine for Neonates.
1. Rees CM et al. (2008) Archs Dis Child Fetal Neonatal Ed. 93 (1):F395-F396
2. McGuire W. (2003) Archs Dis Child Fetal Neonatal Ed. 88 (3):F11-F14