Prevalence of mental health disorders and medication use in Cambridge and Peterborough National Health Service foundation trust service users

Abstract:

Background: Electronic health records provide a valuable resource for health-related research. The data contained in the electronic mental health records could inform us of the prevalence of the mental health conditions in the popula- tion, service use, and treatment. However, in the United Kingdom, mental health data held by the National Health Services (NHS) has rarely been used in epidemiological re- search, for designing surveys and audits, and for recruiting participants to suitable research studies and clinical trials. These problems stem largely from information security is- sues, as clinical records contain highly sensitive and confi- dential information that cannot be shared outside the clin- ical care-giving team without explicit consent, other than in exceptional circumstances [1–3]. In order to facilitate the use of NHS data, we have created a software tool that de-identifies and anonymises clinical records, applies natu- ral language processing (NLP) system, and simplifies partic- ipant recruitment into research studies [4]. The software is currently being used for Cambridge and Peterborough NHS Foundation Trust (CPFT) mental health data. Here we will provide an overview of the prevalence of the mental health disorders and medication use within the CPFT Re- search Database. Methods: CPFT Research Database implements the Clin- ical Records Anonymisation and Text Extraction (CRATE) software, which allows for: 1) anonymisation and de- identification of structured data and free text (using NLP) from clinical records; 2) searching and the use of these clin- ical records for epidemiological research; 3) recruitment of consenting individuals to research. Results: The CPFT Research Database system currently con- tains data from 191066 patients since 2013, of whom 32805 have a coded diagnosis recorded in the ICD-10 "F" chap- ter (at any point in time). Of these patients, 13255 have been diagnosed with organic, including symptomatic mental disorders such as dementia, followed by mood (affective) disorders (n=8731), neurotic, stress-related and somato- form disorders (n=5620), schizophrenia, schizophreniform and delusional disorder (n =3016), mental and behavioural disorders due to psychoactive substance use (n=2969), dis- orders of adult personality and behaviour (n=2765), be- havioural syndromes associated with physiological distur- bances and physical factors (n=1814), behavioural and emotional disorders with onset usually occurring in child- hood and adolescence (n=1585), disorders of psycho- logical development (n=1015) and mental retardation (n=511). Medication information was extracted from free- text records of 5395 patients via NLP and the top 10 drugs used were citalopram (n=889 patients), sertraline (n=741), mirtazapine (n=671), diazepam (n=614), zopi- clone (n =609), quetiapine (n= 568), fluoxetine (n=517), olanzapine (n=513), omeprazole (n=471), and simvastatin (n=436). Conclusion: Here we have presented the mental health dis- order prevalence data and given an overview of the med- ication use information extracted from free-text records amongst the patients using the CPFT mental health ser- vices. Such information provides a valuable resource for large-scale epidemiological studies, and have great poten- tial in advancing our understanding of mental health disor- der prevalence and provision of care. Future studies could further explore the contributing factors and health out- comes for people diagnosed with mental health disorders by linking the mental health databases with physical health and mortality data. References [1] Caldicott, F., 2013. Information: To share or not to share? The Information Governance Review. UK Department of Health Available from: https://assets.publishing.service.gov.uk/ government/uploads/system/uploads/attachment_data/file/ 192572/2900774_InfoGovernance_accv2.pdf . [2] UK Department of Health, 2003. Confidentiality: NHS Code of Practice. Available from: https://www.gov.uk/government/ publications/confidentiality- nhs- code- of- practice [Accessed: 1 April, 2019] [3] UK General Medical Council, 2013. Good practice in research and Consent to research. Available from: http://www.gmc-uk. org/Good_practice_in_research_and_consent_to_research. pdf_58834843.pdf [Accessed: 1 April, 2019] [4] Cardinal, R.N., 2017. Clinical records anonymisation and text extraction (CRATE): an open-source software system. BMC Med Inform Dec Mak. https://doi.org/10.1186/s12911- 017- 0437- 1∗ . doi: 10.1016/j.euroneuro.2019.09.684