A virtual solution to disparate clinical information systems and consequent data silos
File(s)
Author(s)
Ramzan, Farzan
Type
Thesis
Abstract
Background: NHS healthcare services in England are fragmented in how they deliver care, and boundaries between primary and secondary care result in the fragmentation of health information. Thus, most healthcare data become trapped in silos within multiple disparate clinical information systems that lack interoperability. Viewing the issue of information fragmentation as an unintended consequence of interactions between health information technologies, this research adopts a sociotechnical stance and identifies informatics techniques that enable the linkage and integration of heterogeneous data silos within an emergency care setting.
Aim: The aim of this research is to investigate and adopt informatics techniques that address health information fragmentation and data silos within an NHS Hospital Emergency Care setting.
Method: This research employed a descriptive observational quantitative single case study design using routinely collected administrative data. The literature identified informatics techniques, incorporating ontologies, clinical data warehouse architectures, data quality indicators and record linkage strategies. A conceptual framework was developed to facilitate the sequential application of these techniques and enable the linkage and integration of multiple heterogeneous datasets to create a linked longitudinal dataset.
Results: Domain understanding via process modelling permitted identification of the extent of fragmentation. Patient ontologies, standardisation approaches and clinical data warehousing enabled interoperability. Probabilistic linkage strategies outperformed deterministic strategies. The creation of a linked dataset facilitated the analysis of departmental transfer times, four-hour
standard and interrupted time series analysis comparing emergency admission rates pre/post intervention.
Conclusion: Each integration scenario presents a specific challenge dependent upon the environment in which information fragmentation has occurred. Data integration is not just one problem, and the most important challenge lies in the correct identification and combination of informatics tools, technologies and processes to perform integration. Thus, disparate data can be linked to answer research questions and create new knowledge not previously possible using
unlinked data.
Aim: The aim of this research is to investigate and adopt informatics techniques that address health information fragmentation and data silos within an NHS Hospital Emergency Care setting.
Method: This research employed a descriptive observational quantitative single case study design using routinely collected administrative data. The literature identified informatics techniques, incorporating ontologies, clinical data warehouse architectures, data quality indicators and record linkage strategies. A conceptual framework was developed to facilitate the sequential application of these techniques and enable the linkage and integration of multiple heterogeneous datasets to create a linked longitudinal dataset.
Results: Domain understanding via process modelling permitted identification of the extent of fragmentation. Patient ontologies, standardisation approaches and clinical data warehousing enabled interoperability. Probabilistic linkage strategies outperformed deterministic strategies. The creation of a linked dataset facilitated the analysis of departmental transfer times, four-hour
standard and interrupted time series analysis comparing emergency admission rates pre/post intervention.
Conclusion: Each integration scenario presents a specific challenge dependent upon the environment in which information fragmentation has occurred. Data integration is not just one problem, and the most important challenge lies in the correct identification and combination of informatics tools, technologies and processes to perform integration. Thus, disparate data can be linked to answer research questions and create new knowledge not previously possible using
unlinked data.
Version
Open Access
Date Issued
2020-03
Date Awarded
2020-07
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Gnani, Shamini
Bottle, Robert Alex
Publisher Department
School of Public Health
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)