2018SeatonSEPhD.pdf (21.39 MB)
Download file

Modelling Neonatal Care Pathways for Babies Born Very Preterm

Download (21.39 MB)
thesis
posted on 20.02.2018, 10:55 by Sarah Emma Seaton
Predicting length of stay in neonatal care is important for resource planning and the counselling of parents. However, it has received limited attention and two issues are: 1. Babies who die in neonatal care are not included appropriately and research should consider all babies simultaneously, irrespective of whether they live or die 2. The different levels of neonatal care (intensive, high dependency and special care) and how they contribute towards overall length of stay have not been considered This thesis contains four inter-connected studies to investigate how statistical approaches can help to address these issues. Firstly, a systematic review was conducted to identify factors commonly used to predict length of stay and mortality. Factors measurable at or around birth, such as gestational age and birthweight, were found to be important. Secondly, competing risks methods were used to predict median length of stay in neonatal care for two competing events: babies who survive to discharge and babies who die before discharge. These estimates can be used by clinicians, with their clinical judgement, to counsel parents about the risk of mortality and about potential length of stay. The third study develops this approach to account for the different levels of care received by the baby, using multistate modelling as a natural extension of the more limited competing risks approach. Mean lengths of stay at each level of care were estimated in order to facilitate commissioning of neonatal services. Finally, the differences in length of stay between Operational Delivery Networks, (groups of neonatal units that work together) were investigated to determine if differences existed. These were examined to understand whether differences were due to varying levels of intensity of specific levels of care within a network or a difference in total length of stay.

History

Supervisor(s)

Manktelow, Brad; Draper, Elizabeth; Abrams, Keith

Date of award

15/02/2018

Author affiliation

Department of Health Sciences

Awarding institution

University of Leicester

Qualification level

Doctoral

Qualification name

PhD

Language

en

Usage metrics

Categories

Keywords

Exports