projects » PhD thesis

I was employed by the EPICure Research Group based at the Institute for Women's Health at University College London from April 2011 until October 2014, during which time I submitted my PhD thesis in neonatology / epidemiology.


My viva was on the 17th February 2015. The examiners were:

  • INTERNAL: Ruth Gilbert, Professor of Paediatric Epidemiology, Institute of Child Health, UCL

  • EXTERNAL: Pierre-Yves Ancel, Médecin épidémiologiste, INSERM, Paris-Descartes University

I received a pass, subject to minor amendments. These were submitted in June 2015 and deemed acceptable by the examiners.

PhD award

My certificate was sent to me in July, although the graduation ceremony did not take place until the following year, in July 2016. This was because it is an annual event, for all graduands from the Faculties of Medical Sciences and Population Health Sciences at UCL graduating within the previous year.

Thesis abstract

TITLE: Research into the care of extremely premature babies in England from 1995 to present

This thesis, arising from the EPICure studies into extremely preterm births, seeks to explain demographic, clinical and organisational factors associated with a large increase in admissions to neonatal intensive care. Using six pre-existing data sets, along with a repeat Unit Profile Study of English perinatal centres conducted in 2011 as part of this thesis, three studies were performed:

  • Validation of the 44% increase in the number of admissions to neonatal intensive care at 22-25 weeks gestation seen between 1995 and 2006 in England was attempted using three methods of probabilistic record linkage with Hospital Episode Statistics (HES) data.

  • The effects of antenatal steroid administration, tocolysis and Caesarean delivery on perinatal outcomes in the extremely preterm population were investigated.

  • Changes in organisational characteristics - staffing and 'activity' (expressed as throughput and intensity) - that have occurred in England were examined using data from three time points.

An increase was seen in the number of extremely premature babies in HES data. Linkage with EPICure data demonstrated that routine data are insufficiently precise for use in epidemiological investigations at the margins of viability. Tocolysis was associated with improved outcomes. Antenatal steroids were associated with improved outcomes at birth following vaginal delivery. No effect was demonstrated for Caesarean delivery on birth outcomes but there was evidence of case selection at gestations below 26 weeks. Organisational data (from 1997, 2006 and 2011) demonstrated reduced numbers of cots between 1997 and 2011 with increases in both throughput and intensity of workload. Staffing levels increased, but still failed to meet recommended standards.

Current knowledge of extremely low gestational age births is inadequate for national policy or health care reorganisation. Suggestions were made for how knowledge could be improved.