Clinical Record Interactive Search Research Articles

Using de-identified electronic health records to research mental health supported housing services: A feasibility study.

Mental health supported housing services are a key component in the rehabilitation of people with severe and complex needs. They are implemented widely in the UK and other deinstitutionalised countries but there have been few empirical studies of their effectiveness due to the logistic challenges and costs of standard research methods. The Clinical Record Interactive Search (CRIS) tool, developed to de-identify and interrogate routinely recorded electronic health records, may provide an alternative to evaluate supported housing services. The feasibility of using the Camden and Islington NHS Foundation Trust CRIS database to identify a sample of users of mental health supported accommodation services. Two approaches to data interrogation and case identification were compared; using structured fields indicating individual's accommodation status, and iterative development of free text searches of clinical notes referencing supported housing. The data used were recorded over a 10-year-period (01-January-2008 to 31-December-2017). Both approaches were carried out by one full-time researcher over four weeks (150 hours). Two structured fields indicating accommodation status were found, 2,140 individuals had a value in at least one of the fields representative of supported accommodation. The free text search of clinical notes returned 21,103 records pertaining to 1,105 individuals. A manual review of 10% of the notes indicated an estimated 733 of these individuals had used a supported housing service, a positive predictive value of 66.4%. Over two-thirds of the individuals returned in the free text search (768/1,105, 69.5%) were identified via the structured fields approach. Although the estimated positive predictive value was relatively high, a substantial proportion of the individuals appearing only in the free text search (337/1,105, 30.5%) are likely to be false positives. It is feasible and requires minimal resources to use de-identified electronic health record search tools to identify large samples of users of mental health supported housing using structured and free text fields. Further work is needed to establish the availability and completion of variables relevant to specific clinical research questions in order to fully assess the utility of electronic health records in evaluating the effectiveness of these services.

Real-world effectiveness of admissions to a tertiary treatment-resistant psychosis service: 2-year mirror-image study.

Treatment-resistant schizophrenia is a major disabling illness which often proves challenging to manage in a secondary care setting. The National Psychosis Unit (NPU) is a specialised tertiary in-patient facility that provides evidence-based, personalised, multidisciplinary interventions for complex treatment-resistant psychosis, in order to reduce the risk of readmission and long-term care costs. This study aimed to assess the long-term effectiveness of treatment at the NPU by considering naturalistic outcome measures. Using a mirror image design, we compared the numbers of psychiatric and general hospital admissions, in-patient days, acuity of placement, number of psychotropic medications and dose of antipsychotic medication prescribed before and following NPU admission. Data were obtained from the Clinical Records Interactive Search system, an anonymised database sourced from the South London and Maudsley NHS Trust electronic records, and by means of anonymous linkage to the Hospital Episode Statistics system. Compared with the 2 years before NPU admission, patients had fewer mental health admissions (1.65 ± 1.44 v. 0.87 ± 0.99, z = 5.594, P < 0.0001) and less mental health bed usage (335.31 ± 272.67 v. 199.42 ± 261.96, z = 5.195 P < 0.0001) after NPU admission. Total in-patient days in physical health hospitals and total number of in-patient days were also significantly reduced (16.51 ± 85.77 v. 2.83 ± 17.38, z = 2.046, P = 0.0408; 351.82 ± 269.09 v. 202.25 ± 261.05, z = 5.621, P < 0.0001). The reduction in level of support required after treatment at the NPU was statistically significant (z = -8.099, P < 0.0001). This study demonstrates the long-term effectiveness of a tertiary service specialising in treatment-resistant psychosis.

Dementia in People from Ethnic Minority Backgrounds: Disability, Functioning, and Pharmacotherapy at the Time of Diagnosis

ObjectivesIncreasingly, older populations in the United Kingdom and other well-resourced settings are ethnically diverse. Despite a concern that the prevalence of dementia is expected to rise, very little is known about the association of ethnicity and dementia among aging older adults. The current study aimed to compare ethnic group differences in symptom profile, functioning and pharmacotherapy at dementia diagnosis. DesignCross-sectional study of patient characteristics at the point of dementia diagnosis. Setting and ParticipantsIn total, 12,154 patients aged 65 years or older diagnosed with dementia in Southeast London between 2007 and 2015. MethodsData were extracted from the Clinical Record Interactive Search system, which provides anonymized access to the electronic health records of a large mental healthcare provider in Southeast London. Patients from ethnic minority backgrounds were compared with white British individuals on mental and physical well-being, functional scales and medications prescribed at dementia diagnosis, as well as subtype of dementia documented anywhere in the record. ResultsCompared with white British patients, Black African and Black Caribbean patients were more likely to present with psychotic symptoms and were less likely to have an antidepressant prescribed; white Irish patients had higher rates of substance/alcohol use and depressive symptoms were more prevalent in South Asian patients; all ethnic minority groups had higher odds of polypharmacy; and vascular dementia diagnoses were more common in Black and Irish ethnic minority groups. Conclusions and ImplicationsAt dementia diagnosis, there are substantial differences in noncognitive mental health symptoms and pharmacotherapy across ethnic minority groups and compared with the white British majority population. Some of these differences might reflect access/treatment inequalities or implicit unconscious bias related to ethnicity, influencing both. They need to be taken into consideration to optimize pathways into care and personalize assessment and management.

Clozapine treatment and risk of COVID-19 infection: retrospective cohort study.

Clozapine, an antipsychotic with unique efficacy in treatment-resistant psychosis, is associated with increased susceptibility to infection, including pneumonia. To investigate associations between clozapine treatment and increased risk of COVID-19 infection in patients with schizophrenia-spectrum disorders who are receiving antipsychotic medications in a geographically defined population in London, UK. Using information from South London and Maudsley NHS Foundation Trust (SLAM) clinical records, via the Clinical Record Interactive Search system, we identified 6309 individuals who had an ICD-10 diagnosis of schizophrenia-spectrum disorders and were taking antipsychotics at the time of the COVID-19 pandemic onset in the UK. People who were on clozapine treatment were compared with those on any other antipsychotic treatment for risk of contracting COVID-19 between 1 March and 18 May 2020. We tested associations between clozapine treatment and COVID-19 infection, adjusting for gender, age, ethnicity, body mass index (BMI), smoking status and SLAM service use. Of 6309 participants, 102 tested positive for COVID-19. Individuals who were on clozapine had increased risk of COVID-19 infection compared with those who were on other antipsychotic medication (unadjusted hazard ratio HR = 2.62, 95% CI 1.73-3.96), which was attenuated after adjusting for potential confounders, including clinical contact (adjusted HR = 1.76, 95% CI 1.14-2.72). These findings provide support for the hypothesis that clozapine treatment is associated with an increased risk of COVID-19 infection. Further research will be needed in other samples to confirm this association. Potential clinical implications are discussed.

Clozapine Treatment and Risk of COVID-19

Background: COVID-19 is a novel infection that leads to pneumonia in severe cases. Clozapine, an antipsychotic with unique efficacy in treatment resistant psychosis, is known to be associated with increasing risk of pneumonia. Aims: To investigate possible associations between clozapine treatment and increased risk of COVID-19 in patients with schizophrenia-spectrum disorders who are receiving antipsychotic medications, using electronic health records data, in a geographically defined population in London. Method: We identified 6304 individuals who had an ICD-10 diagnosis of schizophrenia-spectrum disorders and were taking antipsychotic during the 3 months prior the follow-up period. People who were on clozapine treatment were designated as the exposed group. Those on any type or combination of antipsychotic treatment that did not include clozapine were designated unexposed. We tested associations between clozapine treatment and COVID-19, adjusting for potential confounders. Results: 70 patients developed COVID-19. Individuals who were on clozapine had a nearly threefold increased risk of COVID-19 compared with those who were on other antipsychotic medication (hazard ratio HR=2.70, 95% CI 1.62 - 4.51). This association was attenuated but remained after adjustment for potential confounders. Conclusion: These findings provide support for the hypothesis that clozapine treatment is associated with increased risk of COVID-19. Further research will be needed in other samples to confirm this association. Potential clinical implications are discussed. Funding Statement: This work was supported by the Clinical Records Interactive Search (CRIS) system funded and developed by the National Institute for Health Research (NIHR) Mental Health Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London and a joint infrastructure grant from Guy’s and St Thomas’ Charity and the Maudsley Charity (grant number BRC-2011-10035). RDH DF and JHM receive salary support from the National Institute for Health Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. Declaration of Interests: RDH, received research funding from Roche, Pfizer, Janssen and Lundbeck, DF has received research funding from Janssen and Lundbeck. JHM has received research funding from Lundbeck. All other authors have no conflict of interest. Ethics Approval Statement: The research was conducted under ethical approval reference 18/SC/0372 from Oxfordshire Research Ethics Committee C.

The association between neighbourhood characteristics and physical victimisation in men and women with mental disorders.

How neighbourhood characteristics affect the physical safety of people with mental illness is unclear. To examine neighbourhood effects on physical victimisation towards people using mental health services. We developed and evaluated a machine-learning-derived free-text-based natural language processing (NLP) algorithm to ascertain clinical text referring to physical victimisation. This was applied to records on all patients attending National Health Service mental health services in Southeast London. Sociodemographic and clinical data, and diagnostic information on use of acute hospital care (from Hospital Episode Statistics, linked to Clinical Record Interactive Search), were collected in this group, defined as 'cases' and concurrently sampled controls. Multilevel logistic regression models estimated associations (odds ratios, ORs) between neighbourhood-level fragmentation, crime, income deprivation, and population density and physical victimisation. Based on a human-rated gold standard, the NLP algorithm had a positive predictive value of 0.92 and sensitivity of 0.98 for (clinically recorded) physical victimisation. A 1 s.d. increase in neighbourhood crime was accompanied by a 7% increase in odds of physical victimisation in women and an 13% increase in men (adjusted OR (aOR) for women: 1.07, 95% CI 1.01-1.14, aOR for men: 1.13, 95% CI 1.06-1.21, P for gender interaction, 0.218). Although small, adjusted associations for neighbourhood fragmentation appeared greater in magnitude for women (aOR = 1.05, 95% CI 1.01-1.11) than men, where this association was not statistically significant (aOR = 1.00, 95% CI 0.95-1.04, P for gender interaction, 0.096). Neighbourhood income deprivation was associated with victimisation in men and women with similar magnitudes of association. Neighbourhood factors influencing safety, as well as individual characteristics including gender, may be relevant to understanding pathways to physical victimisation towards people with mental illness.

Impact of matching error on linked mortality outcome in a data linkage of secondary mental health data with Hospital Episode Statistics (HES) and mortality records in South East London: a cross-sectional study

ObjectivesLinkage of electronic health records (EHRs) to Hospital Episode Statistics (HES)-Office for National Statistics (ONS) mortality data has provided compelling evidence for lower life expectancy in people with severe mental...

The Development of the Military Service Identification Tool: Identifying Military Veterans in a Clinical Research Database Using Natural Language Processing and Machine Learning.

BackgroundElectronic health care records (EHRs) are a rich source of health-related information, with potential for secondary research use. In the United Kingdom, there is no national marker for identifying those who have previously served in the Armed Forces, making analysis of the health and well-being of veterans using EHRs difficult.ObjectiveThis study aimed to develop a tool to identify veterans from free-text clinical documents recorded in a psychiatric EHR database.MethodsVeterans were manually identified using the South London and Maudsley (SLaM) Biomedical Research Centre Clinical Record Interactive Search—a database holding secondary mental health care electronic records for the SLaM National Health Service Foundation Trust. An iterative approach was taken; first, a structured query language (SQL) method was developed, which was then refined using natural language processing and machine learning to create the Military Service Identification Tool (MSIT) to identify if a patient was a civilian or veteran. Performance, defined as correct classification of veterans compared with incorrect classification, was measured using positive predictive value, negative predictive value, sensitivity, F1 score, and accuracy (otherwise termed Youden Index).ResultsA gold standard dataset of 6672 free-text clinical documents was manually annotated by human coders. Of these documents, 66.00% (4470/6672) were then used to train the SQL and MSIT approaches and 34.00% (2202/6672) were used for testing the approaches. To develop the MSIT, an iterative 2-stage approach was undertaken. In the first stage, an SQL method was developed to identify veterans using a keyword rule–based approach. This approach obtained an accuracy of 0.93 in correctly predicting civilians and veterans, a positive predictive value of 0.81, a sensitivity of 0.75, and a negative predictive value of 0.95. This method informed the second stage, which was the development of the MSIT using machine learning, which, when tested, obtained an accuracy of 0.97, a positive predictive value of 0.90, a sensitivity of 0.91, and a negative predictive value of 0.98.ConclusionsThe MSIT has the potential to be used in identifying veterans in the United Kingdom from free-text clinical documents, providing new and unique insights into the health and well-being of this population and their use of mental health care services.

The anticholinergic effect on cognition (AEC) scale-Associations with mortality, hospitalisation and cognitive decline following dementia diagnosis.

To investigate associations between central anticholinergic burden (determined through the anticholinergic effect on cognition [AEC] scale) and mortality, hospitalisation and cognitive decline in patients with dementia. The South London and Maudsley NHS Foundation Trust (SLaM) Clinical Records Interactive Search (CRIS) application was used to identify patients with a first diagnosis of dementia. Medication exposure was extracted through a natural language processing algorithm, allowing for calculations and comparisons of AEC scores. Data were linked to national mortality and hospitalisation data sources, and serially recorded Mini-Mental State Examination (MMSE) scores were used to investigate cognitive decline. We identified 14 093 patients with dementia, 60.7% were female and the mean age at diagnosis was 79.8 years. Patients for whom a review of their medication was indicated (AEC score ≥ 2 for any individual drug or total AEC score ≥ 3) had an increased risk of mortality (hazard ratio 1.07; 95% confidence interval [CI]: 1.01-1.15) and emergency hospitalisation (1.10; 95% CI: 1.04-1.17), but there were no associations with duration of hospitalisation. Cognitive trajectory analyses showed that this exposure group had lower MMSE scores at diagnosis and a sharper increase in MMSE scores over the subsequent 6 months, but similar slopes for the 6 to 36 months period compared to the remainder of the sample. Patients with dementia receiving medication with high central anticholinergic activity appear to have worse prognosis in terms of mortality and hospitalisation risk, but have, primarily, acutely impaired cognitive function, rather than longer-term differences in cognitive decline. J Am Geriatr Soc 68:-, 2020.

M128. THE ASSOCIATION BETWEEN CHILDHOOD SEXUAL ABUSE AND OUTCOMES IN PATIENTS WITH SEVERE MENTAL ILLNESS

BackgroundConsistent evidence suggests that there is an association between childhood trauma and psychotic disorders. Exposure to childhood trauma among people with severe mental illness (SMI) is associated with a range of negative outcomes, including exacerbation of symptoms, higher rates of readmission and relapse, poorer levels of social and vocational functioning, substance use and self-harm (Grubaugh et al., 2011). We aimed to identify people with SMI and a history of CSA as recorded in electronic health records in a large mental healthcare provider in the UK, describe the clinical and functional characteristics these patients and compare their clinical outcomes (admissions, medication) to those of patients with no recorded history of CSA.MethodsData for this study were obtained from Camden & Islington NHS Foundation Trust using the Clinical Record Interactive Search (CRIS) tool. CRIS is an application developed to enable routinely collected electronic health records to be used in research. C&I NHS FT is a large mental health provider serving a geographic catchment area of two inner-city London boroughs, and approximately 470,000 residents. For purposes of this study we identified patients with an ICD-10 diagnosis of F20-F29 or F30-31 with at least one year of follow-up. We searched text records (clinical notes and documents) for keywords associated with CSA, using an iterative process of inclusion and exclusion rules. We then tested the positive predictive value of this search. We compared the demographic, clinical and functional characteristics of patients with and without a history of CSA using chi square tests. Next, we used logistic regression models to examine the outcomes of inpatient admission and receipt of antipsychotic medication among the study population.ResultsWe identified 7,000 patients with SMI, of which 619 (8.8%) hade a recorded history of CSA. Patients with a recorded history of CSA were more likely to be female (63.8% vs. 43.3%), single (76.9% vs. 66.5%), of white ethnic origin (66.4% vs. 54.3%) than their counterparts without CSA. There was no difference in the distribution of social deprivation between the groups. Major depressive disorder, posttraumatic stress disorder and personality disorders were all more prevalent in patients with CSA (13.4% vs. 7.6%, 4.7% vs. 1.4%, and 22.0% vs. 5.8%, respectively). Higher rates of moderate-severe psychotic symptoms, depressed mood, self-harm, substance use and aggression were also evident in this group, as were problems with relationships and accommodation. There was a 2-fold increase in the odds of inpatient admissions in patients with a history of CSA than in those without (adjusted OR=1.95, 95% CI: 1.64–2.33), and they were more likely to have spent at least 10 days a year as inpatients (adjusted OR=1.32, 95% CI: 1.07–1.62). Similarly, patients with a history of CSA were more likely to be prescribed antipsychotic medication (adjusted OR=2.48, 95% CI: 1.69–3.66), and more likely to be given over 75% of the maximum recommended dose (adjusted OR=1.72, 95% CI: 1.44–2.04).DiscussionThe current study used routinely recorded data from electronic health records to identify a history of childhood sexual abuse in patients with SMI. Consistent with previous findings, the study demonstrated that exposure to childhood trauma is associated with negative clinical and functional outcomes in these patients. Clinicians working with patients with SMI should be trained and skilled at assessing childhood adversity, and addressing such adversity in the clinical setting.

Maximizing the use of social and behavioural information from secondary care mental health electronic health records.

The contribution of social and behavioural factors in the development of mental health conditions and treatment effectiveness is widely supported, yet there are weak population level data sources on social and behavioural determinants of mental health. Enriching these data gaps will be crucial to accelerating precision medicine. Some have suggested the broader use of electronic health records (EHR) as a source of non-clinical determinants, although social and behavioural information are not systematically collected metrics in EHRs, internationally. In this commentary, we highlight the nature and quality of key available structured and unstructured social and behavioural data using a case example of value counts from secondary mental health data available in the UK from the UK Clinical Record Interactive Search (CRIS) database; highlight the methodological challenges in the use of such data; and possible solutions and opportunities involving the use of natural language processing (NLP) of unstructured EHR text. Most structured non-clinical data fields within secondary care mental health EHR data have too much missing data for adequate use. The utility of other non-clinical fields reported semi-consistently (e.g., ethnicity and marital status) is entirely dependent on treating them appropriately in analyses, quantifying the many reasons behind missingness in consideration of selection biases. Advancements in NLP offer new opportunities in the exploitation of unstructured text from secondary care EHR data particularly given that clinical notes and attachments are available in large volumes of patients and are more routinely completed by clinicians. Tackling ways to re-use, harmonize, and improve our existing and future secondary care mental health data, leveraging advanced analytics such as NLP is worth the effort in an attempt to fill the data gap on social and behavioural contributors to mental health conditions and will be necessary to fulfill all of the domains needed to inform personalized interventions.

Clinical factors associated with progression to dementia in people with late-life depression: a cohort study of patients in secondary care

ObjectivesDepression can be a prodromal feature or a risk factor for dementia. We aimed to investigate which clinical factors in patients with late-life depression are associated with a higher risk...

Characteristics and risk of repeat suicidal ideation and self-harm in patients who present to emergency departments with suicidal ideation or self-harm: A prospective cohort study

BackgroundCharacteristics and outcomes of patients presenting to Emergency Departments (EDs) have been under-examined. This paper describes the characteristics and risk of repeat suicidality amongst patients presenting to EDs with (1) suicidal ideation and (2) self-harm, compared to (3) controls in mental health crisis. MethodsThe Clinical Record Interactive Search tool identified 2211 patients who presented to three London EDs with suicidal ideation or self-harm, and 1108 control patients. All patients received a full psychosocial assessment. Chi-squared tests examined group characteristics. Cox regression models assessed the risk of re-presentation with suicidal ideation or self-harm within one year. ResultsThere were a higher proportion of females and individuals under the age of 25 in the self-harm group. Patients presenting with suicidal ideation or self-harm were more likely to be white, live in more deprived areas, and less likely to receive a psychiatric diagnosis within one year compared to controls. Risk of repeat suicidality within one year was 3–4 times higher in those with baseline suicidal ideation (adjusted HR = 3.66, 95% CI 2.44–5.48) or self-harm (HR = 3.53 95% CI 2.47–5.04) compared to controls. LimitationsTo be included patients needed to have a full psychosocial assessment. Incomplete records meant 21.4% of the sample was excluded. This will have introduced bias which might confound observed associations. ConclusionIndividuals presenting with either suicidal ideation or self-harm have similar risk for re-presentation within one year. Both groups would benefit from personalised risk management plans and active follow-up to reduce the risk of repeat suicidal behaviour.

Identifying Predictors of Suicide in Severe Mental Illness: A Feasibility Study of a Clinical Prediction Rule (Oxford Mental Illness and Suicide Tool or OxMIS).

BackgroundOxford Mental Illness and Suicide tool (OxMIS) is a brief, scalable, freely available, structured risk assessment tool to assess suicide risk in patients with severe mental illness (schizophrenia-spectrum disorders or bipolar disorder). OxMIS requires further external validation, but a lack of large-scale cohorts with relevant variables makes this challenging. Electronic health records provide possible data sources for external validation of risk prediction tools. However, they contain large amounts of information within free-text that is not readily extractable. In this study, we examined the feasibility of identifying suicide predictors needed to validate OxMIS in routinely collected electronic health records.MethodsIn study 1, we manually reviewed electronic health records of 57 patients with severe mental illness to calculate OxMIS risk scores. In study 2, we examined the feasibility of using natural language processing to scale up this process. We used anonymized free-text documents from the Clinical Record Interactive Search database to train a named entity recognition model, a machine learning technique which recognizes concepts in free-text. The model identified eight concepts relevant for suicide risk assessment: medication (antidepressant/antipsychotic treatment), violence, education, self-harm, benefits receipt, drug/alcohol use disorder, suicide, and psychiatric admission. We assessed model performance in terms of precision (similar to positive predictive value), recall (similar to sensitivity) and F1 statistic (an overall performance measure).ResultsIn study 1, we estimated suicide risk for all patients using the OxMIS calculator, giving a range of 12 month risk estimates from 0.1-3.4%. For 13 out of 17 predictors, there was no missing information in electronic health records. For the remaining 4 predictors missingness ranged from 7-26%; to account for these missing variables, it was possible for OxMIS to estimate suicide risk using a range of scores. In study 2, the named entity recognition model had an overall precision of 0.77, recall of 0.90 and F1 score of 0.83. The concept with the best precision and recall was medication (precision 0.84, recall 0.96) and the weakest were suicide (precision 0.37), and drug/alcohol use disorder (recall 0.61).ConclusionsIt is feasible to estimate suicide risk with the OxMIS tool using predictors identified in routine clinical records. Predictors could be extracted using natural language processing. However, electronic health records differ from other data sources, particularly for family history variables, which creates methodological challenges.

Life Expectancy, Mortality Risks and Cause of Death in Patients with Serious Mental Illness: A Comparison between 2008-2012 and 2013-2017

Background: People with serious mental illness have a significantly shorter life expectancy than the general population. This study investigates whether the mortality rate in this group has changed over the last decade. Method: Using Clinical Record Interactive Search (CRIS) software, we extracted data from a large electronic database of patients in South East London. All patients who had received a diagnosis of schizophrenia, schizoaffective disorder or bipolar affective disorder from 2008-2012 and/or 2013-2017 were included. Estimates of life expectancy at birth, standardized mortality ratios and causes of death were obtained for each cohort according to diagnosis and gender. Comparisons were made between cohorts and with the general population using data obtained from UK Office of National Statistics. Findings: 26,005 patients with serious mental illness were included. In men, life expectancy was greater in 2013-2017 (64.9 years; 95% CI: 63.6-66.3) than in 2008-2012 (63.2 years; 95% CI: 61.5-64.9). Similarly, in women, life expectancy was greater in 2013-2017 (69.1 years; 95% CI: 67.5-70.7) than in 2008-2012 (68.1 years; 95% CI: 66.2-69.9). The difference with general population life expectancy fell by 0.9 years between cohorts in men, and by 0.5 years in women. In women with SMI, cancer was the leading cause of death in 2013-2017, whereas the leading cause in 2008-2012 was cardiovascular diseases. In men with SMI there were more deaths due to cancer in 2013-2017 than in 2008-2012. Cancer overtook cardiovascular disease as the leading cause of death in women. Interpretations: Relative to the general population, life expectancy for people with serious mental illness it is still much worse though appears to be improving. The increased cancer-related mortality in these patients suggests that physical health monitoring should consider including cancer as well as cardiovascular and metabolic disease. Funding Statement: EC is funded by a National Institute for Health Research Doctoral Research Fellowship (NIHR300273). CK was funded by Taipei City Government. RP is funded by National Institute for Health Research. Academy of Medical Sciences, The Wellcome Trust, MRC, British Heart Foundation, Arthritis Research UK, the Royal College of Physicians and Diabetes UK. Declaration of Interests: All authors have completed the ICMJE uniform disclosure form and declared no interest of conflicts. Ethics Approval Statement: The SLaM BRC Case Register and CRIS have received ethical approval as an anonymised data set for secondary analyses from the Oxfordshire Research Ethics Committee C (18/SC/0372). The Clinical Data Linkage Service was used for data protection, linkage and extraction.

Community treatment orders and associations with readmission rates and duration of psychiatric hospital admission: a controlled electronic case register study

ObjectivesLimited evidence is available regarding the effect of community treatment orders (CTOs) on mortality and readmission to psychiatric hospital. We compared clinical outcomes between patients placed on CTOs to a...

‘Real time’ monitoring of antipsychotic prescribing in patients with dementia: a study using the Clinical Record Interactive Search (CRIS) platform to enhance safer prescribing

BackgroundThe use of antipsychotic drugs in dementia has been reported to be associated with increased risk of cerebrovascular events and mortality. There is an international drive to reduce the use...

Synthetic cannabinoid use in psychiatric patients and relationship to hospitalisation: A retrospective electronic case register study.

Introduction and objectives:Cannabis use has been associated with psychosis and with poor outcome in patients with mental illness. Synthetic cannabinoids (SCs) have been suggested to pose an even greater risk to mental health, but the effect on clinical outcome has not been directly measured. In this study, we aimed to investigate the demographics and hospitalisation of psychiatric patients who were SC users.Methods:We searched the Biomedical Research Centre Clinical Record Interactive Search register for SC users and age- and sex-matched SC non-users who had been psychiatric patients under the South London and Maudsley NHS Trust. We recorded diagnosis, homelessness, cannabis use and the total number of days admitted as an inpatient to secondary and tertiary mental-health services.Results:We identified 635 SC users and 635 age- and sex-matched SC non-users. SC users were significantly more likely to be homeless (χ2=138.0; p<0.0001) and to use cannabis (χ2=257.3; p<0.0001) than SC non-users. SC users had significantly more inpatient days after their first recorded use of SCs than controls (M (SD)=85.5 (199.7) vs. 25.4 (92.32); p<0.0001). Post hoc tests revealed that SC non-users who used cannabis had fewer inpatient days than SC users (p<0.0001), and that non-users of both SC and cannabis had fewer inpatient days than SC non-using cannabis users (p=0.02).Conclusions:SC use may lead to an increase in the number of days spent in hospital in patients with psychiatric illness. This highlights the need for clinicians to ask specifically about SC use.

Suicide attempts requiring hospitalization in patients with eating disorders: A retrospective cohort study.

Suicide attempts requiring hospitalization are known to be common in patients who are diagnosed with eating disorders. Attempting suicide is a major indicator for those at risk of completed suicide. Both the specific eating disorder diagnosis and the influence of psychiatric comorbidities on suicide attempts requiring hospitalization were investigated, with demographic and socioeconomic variables as confounders, over a 10-year observation period from January 2007 to March 2017. Anonymized health-record data from the South London and Maudsley NHS Foundation Trust (SLaM) were retrieved through the Clinical Record Interactive Search (CRIS) data resource; this is linked to national Hospital Episode Statistics (HES) data. These data include all diagnoses for inpatient admissions. Hazard ratios, with 95% confidence intervals (CIs), were calculated from cox regression analyses and the effects of a number of confounders were estimated by performing multivariable analyses. In total, 4,895 patients were diagnosed with anorexia nervosa (AN), bulimia nervosa (BN), or eating disorder otherwise not specified (EDNOS). Of these, 331 (6.7%) had attempted suicide requiring hospitalization and 21 (0.04%) completed suicide. The eating disorder category associated with the highest risk of a suicide attempt was AN (HR: 1.43, 95%CI: 1.08-1.89, p = .01). The risk was significantly increased further if the patient had a comorbid diagnosis of personality disorder, depression, bipolar affective disorder, and substance misuse. Suicide attempts requiring hospitalization have a high incidence rate among patients with eating disorders, and the risk is significantly increased in AN. Comorbid psychiatric illness and suicidal ideation should be carefully assessed in all eating disorder patients.

Natural language processing for structuring clinical text data on depression using UK-CRIS.

Utilisation of routinely collected electronic health records from secondary care offers unprecedented possibilities for medical science research but can also present difficulties. One key issue is that medical information is presented as free-form text and, therefore, requires time commitment from clinicians to manually extract salient information. Natural language processing (NLP) methods can be used to automatically extract clinically relevant information. Our aim is to use natural language processing (NLP) to capture real-world data on individuals with depression from the Clinical Record Interactive Search (CRIS) clinical text to foster the use of electronic healthcare data in mental health research. We used a combination of methods to extract salient information from electronic health records. First, clinical experts define the information of interest and subsequently build the training and testing corpora for statistical models. Second, we built and fine-tuned the statistical models using active learning procedures. Results show a high degree of accuracy in the extraction of drug-related information. Contrastingly, a much lower degree of accuracy is demonstrated in relation to auxiliary variables. In combination with state-of-the-art active learning paradigms, the performance of the model increases considerably. This study illustrates the feasibility of using the natural language processing models and proposes a research pipeline to be used for accurately extracting information from electronic health records. Real-world, individual patient data are an invaluable source of information, which can be used to better personalise treatment.