Applied Clinical Informatics Lead, SLaM Biomedical Research Center, South London & Maudsley Foundation NHS Trust, UK
Mizanur Khondoker
Senior Lecturer in Medical Statistics, University of East Anglia, Norwich Medical School, UK
Rashmi Patel
MRC UKRI Health Data Research UK Fellow, Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, Kings College London; and South London & Maudsley Foundation NHS Trust, UK
Marcella Fok
Visiting Researcher, Psychological Medicine, Institute of Psychiatry, Psychology and Neuroscience, Kings College London; and Central and North West London NHS Foundation Trust, UK
Robert Harland
Clinical Director of Psychosis, Psychosis CAG, South London & Maudsley Foundation NHS Trust, UK
Matthew Broadbent
Informatics Lead, SLaM Biomedical Research Center, South London & Maudsley Foundation NHS Trust, UK
Paul McCrone
Professor of Health Economics, School of Health Science, Institute of Psychiatry, Psychology and Neuroscience, Kings College London, UK
Robert Stewart
Professor of Psychiatric Epidemiology and Clinical Informatics, Psychological Medicine, Institute of Psychiatry, Psychology and Neuroscience, Kings College London; and South London & Maudsley Foundation NHS Trust, UK
BackgroundThe density of information in digital health records offers new potential opportunities for automated prediction of cost-relevant outcomes.AimsWe investigated the extent to which routinely recorded data held in the electronic health record (EHR) predict priority service outcomes and whether natural language processing tools enhance the predictions. We evaluated three high priority outcomes: in-patient duration, readmission following in-patient care and high service cost after first presentation.MethodWe used data obtained from a clinical database derived from the EHR of a large mental healthcare provider within the UK. We combined structured data with text-derived data relating to diagnosis statements, medication and psychiatric symptomatology. Predictors of the three different clinical outcomes were modelled using logistic regression with performance evaluated against a validation set to derive areas under receiver operating characteristic curves.ResultsIn validation samples, the full models (using all available data) achieved areas under receiver operating characteristic curves between 0.59 and 0.85 (in-patient duration 0.63, readmission 0.59, high service use 0.85). Adding natural language processing-derived data to the models increased the variance explained across all clinical scenarios (observed increase in r2 = 12–46%).ConclusionsEHR data offer the potential to improve routine clinical predictions by utilising previously inaccessible data. Of our scenarios, prediction of high service use after initial presentation achieved the highest performance.
