Use this URL to cite or link to this record in EThOS:
Title: Can increased integration of spatial analysis improve the understanding of how hospital design may affect healthcare associated infection dynamics and patient safety?
Author: Davis, Grahame
ISNI:       0000 0004 9350 4143
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Background: Healthcare associated infections (HCAIs) are considered to be the most frequent adverse event that threatens patients’ safety worldwide. A great deal of work has been carried out looking to improve surveillance, control and prevention of HCAIs within the NHS and health systems worldwide. Like in other research settings, an understanding of the localised environment is critical to understand its potential effect on disease dynamics. Numerous techniques, not yet exploited within healthcare settings, have been developed to quantify this environmental impact. Aim: To examine how integration of spatial epidemiology and modelling could improve the understanding of how hospital design may affect healthcare associated infection dynamics and patient safety, allowing improvement of HCAI investigations within the NHS. Methods: This thesis made use of data from Imperial College Healthcare NHS Trust, including hospital floorplans, information on patient transfers, laboratory tests and traditional infection surveillance data. A literature review was carried out to identify limitations in the application of spatial epidemiology within healthcare settings. Geographical Information Systems (GIS), graph-mining and network modelling were applied to analyse existing datasets, leading to the production of ward maps and networks. Results: The key findings of the work were numerous. Digitalisation of ward maps allows for wider dissemination of this data to infection control teams and other researchers to aid in investigations. Certain ward characteristics appear associated with increased numbers of infections and the physical layout of the ward showed relations between ward subgraphs and infection risk in both positive and negative manners. Network modelling allowed visualisation and analysis of patient movements within and between hospitals. Allowed identification of how wards are linked and specifically which wards merge departments/specialities and link patient populations. Identifiable control points allowed for greater understanding of the ward network and this was shown to help with response and planning for extreme or emergency situations. Conclusion: These techniques, whilst commonly used in other areas, have been underutilised within healthcare and have within this thesis been shown to extract greater information from pre-existing data. The work identified numerous ways existing datasets can be further explored to help researchers; whilst also providing outputs that can help day to day work, such as ward maps.
Supervisor: Drumright, Lydia ; Sevdalis, Nick Sponsor: UK Clinical Research Collaboration
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral