Clinical Record Interactive Search Research Articles

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.

Late-life depression in people from ethnic minority backgrounds: Differences in presentation and management

BackgroundAn elevated risk of late-life depression has been suggested in older adults from minority ethnic groups, but little is known about ethnic group differences in symptom and treatment profiles. The current study aimed to compare symptoms and types of treatment between ethnic groups in patients with late-life depression. MethodsData were extracted from the Clinical Record Interactive Search (CRIS) system , which provides access to the anonymised electronic health records of a large mental health care provider in South London. In total, 5,546 individuals aged 65 years and older, and diagnosed with late-life depression between 2006 and 2017, were included. Patients from ethnic minority backgrounds were compared to White British individuals on the following features recorded at depression diagnosis: mental and physical wellbeing,functional scales, individual depressive symptoms recorded, and treatments administered. ResultsBlack Africans and Black Caribbeans more frequently presented with psychotic problems and were significantly less likely to have antidepressant treatment prescribed post diagnosis compared to White British. White Irish had higher rates of substance use and sleep disturbance. Depressive symptoms of hopelessness, guilt feelings, and suicidal thoughts were less common in Black Caribbeans, Black Africans, and South Asians compared to White British. LimitationsOnly patients with depression under a specialist mental health care provider were included in the study. ConclusionsEthnic minority elders have significantly different presentations and undertake different types of treatment both across groups and relative to their White British counterparts. These differences need to be taken into consideration to optimise pathways into care and to personalise treatment.

Identifying Veterans Using Electronic Health Records in the United Kingdom: A Feasibility Study.

There is a lack of quantitative evidence concerning UK (United Kingdom) Armed Forces (AF) veterans who access secondary mental health care services—specialist care often delivered in high intensity therapeutic clinics or hospitals—for their mental health difficulties. The current study aimed to investigate the utility and feasibility of identifying veterans accessing secondary mental health care services using National Health Service (NHS) electronic health records (EHRs) in the UK. Veterans were manually identified using the Clinical Record Interactive Search (CRIS) system—a database holding secondary mental health care EHRs for an NHS Trust in the UK. We systematically and manually searched CRIS for veterans, by applying a military-related key word search strategy to the free-text clinical notes completed by clinicians. Relevant data on veterans’ socio-demographic characteristics, mental disorder diagnoses and treatment pathways through care were extracted for analysis. This study showed that it is feasible, although time consuming, to identify veterans through CRIS. Using the military-related key word search strategy identified 1600 potential veteran records. Following manual review, 693 (43.3%) of these records were verified as “probable” veterans and used for analysis. They had a median age of 74 years (interquartile range (IQR): 53–86); the majority were male (90.8%) and lived alone (38.0%). The most common mental diagnoses overall were depressive disorders (22.9%), followed by alcohol use disorders (10.5%). Differences in care pathways were observed between pre and post national service (NS) era veterans. This feasibility study represents a first step in showing that it is possible to identify veterans through free-text clinical notes. It is also the first to compare veterans from pre and post NS era.

Depression and changes in educational attainment using administrative data: The role of socio-demographic characteristics

Background Research has shown depression to result in poorer educational attainment among children. However, comparatively little is known about the relationship between depression and attainment over time, and the role that children’s socio-demographic characteristics might play.
 Aim To describe the educational attainment trajectory profiles of children with a clinical diagnosis of depression, and to explore whether these trajectory profiles vary by gender, ethnicity, area-level deprivation and age at diagnosis.
 Methods This is a historical, longitudinal cohort study, making use of two linked administrative datasets. First, a cohort of children (<18 years) were identified who had received a first diagnosis of depression from the South London and Maudsley NHS Foundation Trust between 2007 and 2013 (n=1490). Their pseudonymised secondary mental healthcare records were extracted via the Clinical Record Interactive Search application at the Trust’s Biomedical Research Centre. These records were then linked to England’s National Pupil Database, which contains attainment data for these children at three timepoints of interest: Key Stage 1 (ages 5 to 7 years), Key Stage 2 (7 to 11 years) and Key Stage 4 (14 to 16 years). Longitudinal mixed models for repeated measures will be conducted to model changes in attainment over these three timepoints according to gender, ethnicity, area-level deprivation and age at diagnosis.
 Results Preliminary results are suggestive of attainment trajectories in this cohort varying according to their socio-demographic characteristics. Analysis will be completed August 2019.
 Conclusion The findings from this study will identify whether some groups of depressed children experience better or worse patterns of educational attainment over time. This could highlight priorities for intervention in England’s ongoing efforts to improve child mental health provision in schools.

Understanding geographical patterning of self-harm prevalence within a diverse urban population: a mixed methods spatial analysis and qualitative study

BackgroundPublic health understanding of variations in self-harm prevalence between populations largely relies on routinely collected service-use data. An unexpectedly low number of hospital inpatient admissions following self-harm has been observed in inner London despite established associations of area deprivation and social fragmentation with self-harm. We aimed to measure associations between rates of attendance at an Emergency Department following self-harm and area characteristics, and to explore the processes that might underlie the incidence of self-harm observed in inner inner-city areas. MethodFor the spatial analysis, we created a dataset of Emergency Department attendances following self-harm from 2009 to 2016 by residents of four southeast London boroughs who were older than 11 years using linked electronic patient record data and Hospital Episode Statistics. Age-standardised and sex-standardised small-area rates of self-harm attendance (SARs) were spatially smoothed using Bayesian models and mapped. Associations with area deprivation (assessed using the Index of Multiple Deprivation) and social fragmentation (Congdon Index) were analysed. For the qualitative study, we identified a case-study area from within the four boroughs studied of four contiguous census lower super output areas (LSOAs; approximately 2600 households that were both in the most deprived quintile of areas in the UK and had below average SARs for self-harm). We undertook semi-structured interviews with 14 individuals working in community organisations and two focus groups of 12 people older than 16 years and resident in the study area, recruited through community organisations. Ethical approval for data access and linkage was granted by the Oxford Research Ethics Committee C, Clinical Records Interactive Search system Oversight Committee, and the NHS Health Research Authority Confidentiality Advisory Group. The qualitative study was given ethical approval by the King's College London, Psychiatry, Nursing and Midwifery Subcommittee. FindingsWe identified 20 750 Emergency Department attendances following self-harm by 12 577 individuals. The ratio of SARs in the highest versus the lowest 5% of areas was 2·87 (95% credible interval [CrI] 2·65–3·13). Associations with self-harm were found for area deprivation (most vs least deprived quintile: rate ratio 2·07, 95% CrI 1·88–2·26) and social fragmentation (1·39, 1·23–1·56); however, we identified clusters of deprived areas with a low incidence of self-harm Emergency Department attendance. Preliminary qualitative analysis suggests that individuals in the case-study population have substantial cumulative stressors and mental health difficulties despite a low incidence of Emergency Department presentation for self-harm (SARs for the four LSOAs ranged from 0·80 to 0·94 before adjustment for deprivation and 0·68–0·81 after adjustment). Frequent experiences of violence and an emphasis on appearing tough combine with discomfort with talking about self-harm and fear of negative treatment by mental health and other services, particularly within the Caribbean and African population who make up 44% of the local population. These factors might make people more likely to respond to distress in ways not defined by researchers as self-harm (even if they are also damaging); equally, these pressures might make individuals who self-harm less likely to identify themselves to services. InterpretationRelying on service-use data for self-harm risks underestimating the extent of distress and subsequent harmful behaviours in some deprived urban populations already disadvantaged by high levels of violence and minority ethnic and migrant status. FundingNational Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London.

Adverse life experiences and triggers for self-harm in young people aged 12–26 years in inner south London: a retrospective observational study

BackgroundThe prevalence of self-harm among children and young people in the UK is increasing, with few accessing formal support. Drivers of self-harm remain poorly understood, and evidence-based opportunities to intervene are scarce. We aimed to identify factors associated with self-harm in young people to inform local action. MethodsIn this retrospective, observational study, we identified 196 residents of the south London Borough of Lambeth, aged 12–26 years, presenting to two emergency departments for injuries related to self-harm in the 2015–16 financial year (April 1, 2015, to March 31, 2016). A pragmatic sample size of 100 was selected via computer randomisation for further analysis. Electronic mental health records were retrospectively analysed using a data collection protocol that was validated with researchers before data collection. Demographics and psychiatric history were obtained from pre-coded databases linking electronic mental health records and Hospital Episode Statistics data on emergency department attendance for self-harm. Presentation details and social risk factors were derived from assessment of case notes pertaining to the 2 weeks after first presentation; we noted any risk factors that had been positively recorded. Data relating to the individual's first presentation within the study period were used; subsequent attendances were not included in the sample. We analysed the data using descriptive statistics. Ethical approval for data access and linkage was granted by the Oxford Research Ethics Committee C (18/SC/0372), Clinical Records Interactive Search system Oversight Committee (project numbers 14–026 and 19–004), and the National Health Service (NHS) Health Research Authority Confidentiality Advisory Group (reference ECC 3–04(f)/2011). This study received approval as an audit project through Southwark Operational Directorate in the South London and Maudsley Foundation NHS Trust. Findings48 (48%) of the 100 selected participants had documented experience of psychological trauma, including exposure to violence. 18 (18%) lived in unstable accommodation, and 20 (20%) were known to social services. Difficulties at home (16 [16%]) and stress relating to housing, school, or financial pressures (13 [13%]) were common documented triggers for the self-harm episode at presentation. Peer influence was rarely noted, although 27 (27%) had recorded experience of bullying. InterpretationAdverse life experiences were prevalent and align with wider population prevalence estimates. Common triggers for self-harm relate to home and school environment and wider socioeconomic stressors, supporting the link between wider determinants of health and self-harm, and contributing to the case for public health action. Risk factors and triggers identified provide potential areas for additional targeted support to prevent development or escalation of self-harm behaviours. FundingUK National Institute for Health Research Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, and King's College London.

Use of Natural Language Processing to identify Obsessive Compulsive Symptoms in patients with schizophrenia, schizoaffective disorder or bipolar disorder

Obsessive and Compulsive Symptoms (OCS) or Obsessive Compulsive Disorder (OCD) in the context of schizophrenia or related disorders are of clinical importance as these are associated with a range of adverse outcomes. Natural Language Processing (NLP) applied to Electronic Health Records (EHRs) presents an opportunity to create large datasets to facilitate research in this area. This is a challenging endeavour however, because of the wide range of ways in which these symptoms are recorded, and the overlap of terms used to describe OCS with those used to describe other conditions. We developed an NLP algorithm to extract OCS information from a large mental healthcare EHR data resource at the South London and Maudsley NHS Foundation Trust using its Clinical Record Interactive Search (CRIS) facility. We extracted documents from individuals who had received a diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder. These text documents, annotated by human coders, were used for developing and refining the NLP algorithm (600 documents) with an additional set reserved for final validation (300 documents). The developed NLP algorithm utilized a rules-based approach to identify each of symptoms associated with OCS, and then combined them to determine the overall number of instances of OCS. After its implementation, the algorithm was shown to identify OCS with a precision and recall (with 95% confidence intervals) of 0.77 (0.65–0.86) and 0.67 (0.55–0.77) respectively. The development of this application demonstrated the potential to extract complex symptomatic data from mental healthcare EHRs using NLP to facilitate further analyses of these clinical symptoms and their relevance for prognosis and intervention response.

How to measure the need for transition to adult services among young people with Attention Deficit Hyperactivity Disorder (ADHD): a comparison of surveillance versus case note review methods

BackgroundHealth services have not provided adequate support for young people with long term health conditions to transfer from child to adult services. National Institute of Health and Care (NICE) guidance on transition has been issued to address these gaps. However, data are often sparse about the number of young adults who might need to transition. Using Attention Deficit Hyperactivity Disorder (ADHD) as an exemplar, this study used an existing surveillance system and a case note review to capture the incidence of the transition process, and compared and contrasted the findings.MethodsThe Child and Adolescent Psychiatry Surveillance System (CAPSS) was used to estimate the incident transition of young people with Attention Deficit Hyperactivity Disorder (ADHD) from child to adult services. This involves consultant child and adolescent psychiatrists from the United Kingdom (UK) and Republic of Ireland (ROI) reporting relevant young people as they are seen in clinics. In parallel, a case note review was conducted using the Maudsley Biomedical Research Centre (BRC) Clinical Records Interactive Search (CRIS). The study period ran for twelve months with a nine month follow up to see how the transition proceeded.ResultsCRIS identified 76 cases in the study period, compared to 18 identified using surveillance via CAPSS. Methodological issues were experienced using both methods. Surveillance issues; eligibility criteria confusion, reporting errors, incomplete questionnaires, difficulties contacting clinicians, and surveillance systems do not cover non-doctors and psychiatrists who are not consultants. Case note review issues using CRIS included the need for researchers to interpret clinical notes, the availability and completeness of data in the notes, and data limited to the catchment of one particular mental health trust.ConclusionsBoth methods demonstrate strengths and weaknesses; the combination of both methods in the absence of strong routinely collected data, allowed a more robust estimate of the level of need for service planning and commissioning.

ETHNIC DIFFERENCES IN COGNITION AND AGE IN PEOPLE DIAGNOSED WITH DEMENTIA: A STUDY OF ELECTRONIC HEALTH RECORDS IN TWO LARGE MENTAL HEALTH CARE PROVIDERS

Qualitative studies suggest that people from UK minority ethnic groups with dementia access health services later in the illness than white UK-born elders but there are no large quantitative studies investigating this. We aimed to investigate inter-ethnic differences in cognitive scores and age at dementia diagnosis. We used the Clinical Record Interactive Search (CRIS) applied to the electronic health records of two London mental health trusts to identify patients diagnosed with dementia between 2008 and 2016. We meta-analysed mean Mini Mental State Examination (MMSE) and mean age at the time of diagnosis across Trusts for the most common ethnic groups, and used linear regression models to test these associations before and after adjustment for age, sex, Index of Multiple Deprivation and marital status. We also compared percentage of referrals for each ethnic group with catchment Census distributions. Compared to White patients (N=9380), unadjusted mean MMSE scores were lower in Asian (−1.25; 95% CI −1.79, −0.71; N=642) and Black patients (−1.82, 95% CI −2.13, −1.52; N=2008) as was mean age at diagnosis (Asian patients: −4.27 (−4.92, -3.61); Black patients −3.70 (−4.13, -3.27) years). These differences persisted after adjustment. In general, ethnic group distributions in referrals did not differ substantially from those expected in the catchments. People from Black and Asian groups were younger at dementia diagnosis and had lower MMSE scores than White referrals.

Psychiatric comorbidity as a risk factor for the mortality of people with bulimia nervosa

BackgroundBulimia nervosa (BN) is associated with increased mortality. Frequent comorbidities of BN include substance use disorders, affective disorders and personality disorders (PD). These comorbidities may add an additional risk for mortality.MethodsWe investigated the influence of these psychiatric comorbidities on all-cause mortality with demographic and socioeconomic factors considered as confounders over an observation period from January 2007 to March 2016 for 1501 people with BN using anonymised health records data from the South London and Maudsley NHS Foundation Trust (SLaM), retrieved through its Clinical Records Interactive Search (CRIS) data resource. Mortality was ascertained through monthly linkages to the nationwide tracing system administered by the Office for National Statistics (ONS). We used Cox proportional hazards regression to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). Multivariable analyses were also performed to estimate effects when controlling for confounding of age, sex, ethnicity, borough, marital status and deprivation score.ResultsA total of 18 patients with BN died during the observation period. The standardised mortality ratio (SMR) for our study cohort (against the population of England and Wales in 2012 as a standard) was 2.52 (95% CI 1.49–3.97). Cox regressions revealed significant associations of mortality with older age and male gender. Comorbid PD (HR: 3.36; 95% CI 1.05–10.73) was significantly associated with all-cause mortality, even after controlling for demographic and socioeconomic covariates.ConclusionsThese results highlight increased mortality in patients with BN and the importance of recognising and treating PDs in patients with BN.

Ethnic differences in cognition and age in people diagnosed with dementia: A study of electronic health records in two large mental healthcare providers

Qualitative studies suggest that people from UK minority ethnic groups with dementia access health services later in the illness than white UK-born elders, but there are no large quantitative studies investigating this. We aimed to investigate interethnic differences in cognitive scores and age at dementia diagnosis. We used the Clinical Record Interactive Search (CRIS) applied to the electronic health records of two London mental health trusts to identify patients diagnosed with dementia between 2008 and 2016. We meta-analysed mean Mini Mental State Examination (MMSE) and mean age at the time of diagnosis across trusts for the most common ethnic groups, and used linear regression models to test these associations before and after adjustment for age, sex, index of multiple deprivation, and marital status. We also compared percentage of referrals for each ethnic group with catchment census distributions. Compared with white patients (N=9380), unadjusted mean MMSE scores were lower in Asian (-1.25; 95% CI -1.79, -0.71; N=642) and black patients (-1.82, 95% CI -2.13, -1.52; N=2008) as was mean age at diagnosis (Asian patients: -4.27 (-4.92, -3.61); black patients -3.70 (-4.13, -3.27) years). These differences persisted after adjustment. In general, ethnic group distributions in referrals did not differ substantially from those expected in the catchments. People from black and Asian groups were younger at dementia diagnosis and had lower MMSE scores than white referrals.

An approach to linking education, social care and electronic health records for children and young people in South London: a linkage study of child and adolescent mental health service data

ObjectivesCreation of linked mental health, social and education records for research to support evidence-based practice for regional mental health services.SettingThe Clinical Record Interactive Search (CRIS) system was used to extract...

Psychiatric comorbidity as a risk factor for mortality in people with anorexia nervosa.

Anorexia nervosa (AN) is found associated with increased mortality. Frequent comorbidities of AN include substance use disorders (SUD), affective disorders (AD) and personality disorders (PD). We investigated the influence of these psychiatric comorbidities on all-cause mortality with demographic and socioeconomic factors considered as confounders in the observation window between January 2007 and March 2016 for 1970 people with AN, using data from the case register of the South London and Maudsley (SLaM) NHS Foundation Trust, an almost monopoly-secondary mental healthcare service provider in southeast London. We retrieved data from its Clinical Records Interactive Search (CRIS) system as data source. Mortality was ascertained through nationwide tracing by the UK Office for National Statistics (ONS) linked to CRIS database on a monthly basis. A total of 43 people with AN died during the observation period. Standardized Mortality Ratio (SMR) with England and Wales population in 2012 as standard population for our study cohort was 5.21 (95% CI 3.77, 7.02). In univariate analyses, the comorbidity of SUD or PD was found to significantly increase the relative risks of mortality (HRs = 3.10, 95% CI 1.21, 7.92; and 2.58, 95% CI 1.23, 5.40, respectively). After adjustment for demographic and socioeconomic covariates as confounders, moderately but not significantly elevated risks were identified for SUD (adjusted HR = 1.39, 95% CI 0.53, 3.65) and PD (adjusted HR = 1.58, 95% CI 0.70, 3.56). These results suggest an elevated mortality in people with AN, which might be, at least partially, explained by the existence of the comorbidities SUD or PD.

O4.5. INVESTIGATING GENETIC PROFILES ASSOCIATED WITH ‘REAL WORLD’ CLINICAL OUTCOMES IN PSYCHOSIS: A RETROSPECTIVE COHORT STUDY

BackgroundThe Clinical Record Interactive Search (CRIS) database provides anonymised data from the full electronic health records of all patients at the South London and Maudsley NHS Foundation Trust, a large provider of secondary mental health care. We have previously shown how the large volumes of available CRIS data pertaining to outcomes can be mined and integrated with patient data collected by historical research interview.The applications of this futuristic translational research model are yet to be fully explored. The aim of this study is to determine whether transcriptomic profiles at the onset of psychosis can discern the likely trajectory symptoms over the subsequent 5 years of illness.MethodsThe study sample consists of 200 first-episode psychosis cases (ICD-10 codes: F20-F29 or F30-F33) aged 18–65 years who presented to SLAM (South London and Maudsley NHS Trust) mental health services between the 1st of January 2010 and the 1st of January 2015. Patients were subsequently recruited to the GAP study. Patients were followed-up electronically for 5 years post recruitment using the CRIS research platform.RNA samples were collected at the baseline timepoint via PAXgene blood tubes and interrogates, using the Illumina HumanHT-12.v4 beadchip array. Samples were run at the National Institute for Health Research’s (NIHR) Biomedical Research Centre for Mental Health (BRC-MH) at the Institute of Psychiatry, Psychology and Neuroscience. A total of 4756 probes passed a stringent quality control across the 200 samples.ResultsCRIS data pertaining to the GAP cohort was interrogated for information on clinical symptoms over a 5-year period using text-mining and natural language processing apps that represent over 70 different dictionary definitions of psychotic and affective symptoms. Confirmatory factor analysis was used to reduce this to a much smaller set of orthogonal symptom dimensions which were then the subject of a genetic interrogation using gene expression data. The analysis was conducted using a statistical learning framework which combines Elastic net penalised regression methodology with K-fold cross-validation (via the GLMnet package in R). This identified gene transcripts that were predictive of longer term symptom trajectories in half of the available sample. The veracity of the model was further validated using the second withheld portion of the sample.DiscussionThe results of this discovery phase may provide a rationale for subsequent multi-modal investigations whose aims will be to further enrich the biomarker signature and to also understand the molecular mechanisms that sustain them.

F228. EFFECTIVENESS OF PALIPERIDONE PALMITATE VS OTHER LONG-ACTING INJECTABLE (LAI) ANTIPSYCHOTICS – AN ELECTRONIC CASE REGISTER STUDY

BackgroundPaliperidone palmitate is a 2nd generation long-acting injectable (LAI) antipsychotic which is increasingly prescribed for patients with chronic schizophrenia. However, it is more expensive than 1st generation LAI antipsychotics and little is known about its effectiveness in a real world clinical setting. We sought to address this issue by analyzing a large electronic case register of patients with schizophrenia treated with LAI antipsychotics.MethodsData were obtained from 1,281 patients in the South London and Maudsley NHS Foundation Trust (SLaM) Biomedical Research Centre (BRC) Case Register who were treated with an LAI antipsychotic between 1st April 2011 and 31st January 2015. The number of days spent as a psychiatric inpatient and the number of admissions to a psychiatric hospital were extracted using the Clinical Record Interactive Search tool (CRIS) and analyzed in each of the 3 years before and after LAI prescription using multivariable regression.ResultsPatients who received paliperidone palmitate (n=430; 33.6%) spent more time in hospital (β coefficient 12.3 days, 95% CI 2.3 to 19.2, p=0.001) and were admitted to hospital more frequently (IRR 1.44, 95% CI 1.29 to 1.61, p<0.001) in the year prior to treatment than those treated with other LAI antipsychotics (n=851, 66.4%). However, there were no significant differences between paliperidone and the other LAI antipsychotics in the 3 years after initiation with respect to the number of days spent in hospital (β coefficient 5.4 days, 95% CI -57.3 to 68.2, p=0.86) or frequency of hospital admissions (Incidence rate ratio 1.07, 95% CI 0.62 to 1.83, p=0.82).DiscussionPaliperidone palmitate was more likely to be prescribed in patients with more severe illness, as indicated by a history of more frequent and lengthy hospital admissions prior to initiation. The absence of differences in outcomes after initiation indicates that the effectiveness of paliperidone palmitate was similar to that of other LAI antipsychotics. However, paliperidone palmitate may be better tolerated than other 1st generation LAI antipsychotics with a lower rate of discontinuation. These findings merit consideration in relation to the high cost of paliperidone palmitate compared to other LAI antipsychotics.

Using data linkage to electronic patient records to assess the validity of selected mental health diagnoses in English Hospital Episode Statistics (HES).

BackgroundAdministrative data can be used to support research, such as in the UK Biobank. Hospital Episode Statistics (HES) are national data for England that include contain ICD-10 diagnoses for inpatient mental healthcare episodes, but the validity of these diagnoses for research purposes has not been assessed.Methods250 peoples' HES records were selected based on a HES recorded inpatient stay at the South London and Maudsley NHS Foundation Trust with a diagnosis of schizophrenia, a wider schizophrenia spectrum disorder, bipolar affective disorder or unipolar depression. A gold-standard research diagnosis was made using Clinical Records Interactive Search pseudonymised electronic patient records using, and the OPCRIT+ algorithm.ResultsPositive predictive value at the level of lifetime psychiatric disorder was 100%, and at the level of lifetime diagnosis in the four categories of schizophrenia, wider schizophrenia spectrum, bipolar or unipolar depression was 73% (68–79). Agreement varied by diagnosis, with schizophrenia having the highest PPV at 90% (80–96). Each person had an average of five psychiatric HES records. An algorithm that looked at the last recorded psychiatric diagnosis led to greatest overall agreement with the research diagnosis.DiscussionFor people who have a HES record from a psychiatric admission with a diagnosis of schizophrenia spectrum disorder, bipolar affective disorder or unipolar depression, HES records appear to be a good indicator of a mental disorder, and can provide a diagnostic category with reasonable certainty. For these diagnoses, HES records can be an effective way of ascertaining psychiatric diagnosis.

Is lithium monitoring NICE? Lithium monitoring in a UK secondary care setting.

Lithium is widely used for the treatment of bipolar disorder. Owing to its narrow therapeutic index and side-effect profile, regular monitoring of serum levels, renal and thyroid function has been recommended by all major guidelines on lithium use. We investigated whether lithium monitoring during maintenance phase treatment in clinical practice meets the latest recommendation by the National Institute for Health and Clinical excellence (i.e. lithium levels between 0.6 and 1.0 mmol/L and lithium level, thyroid and renal function tests every 6 months) in one of the largest mental health organizations in Europe, the South London and Maudsley (SLaM) NHS Foundation Trust. Retrospective data were extracted from SLaM's Clinical Record Interactive Search (CRIS) system. Adult patients with a psychiatric disorder who were on lithium at any point during the period January 2012-January 2016 and had at least one lithium level test result in the system were included in the analyses. A total of 2639 lithium level tests results were retrieved for 412 patients. Overall, the serum level was within the recommended range in 50.7% of all tests, below the range in 42.4% and above in 6.9%. Lithium level, renal and thyroid function tests were performed at the recommended frequency of 6 months (or less) in 76.2%, 72.7% and 60.2% of patients, respectively. These data demonstrate that there is a gap between the NICE 2014 recommendation and lithium monitoring practice in secondary care, with a high number of lithium level results below the therapeutic minimum. Reminder strategies for secondary care practitioners, shared care agreements or a central registry for lithium users could improve monitoring performance.

SemEHR: A general-purpose semantic search system to surface semantic data from clinical notes for tailored care, trial recruitment, and clinical research.

ObjectiveUnlocking the data contained within both structured and unstructured components of electronic health records (EHRs) has the potential to provide a step change in data available for secondary research use, generation of actionable medical insights, hospital management, and trial recruitment. To achieve this, we implemented SemEHR, an open source semantic search and analytics tool for EHRs.MethodsSemEHR implements a generic information extraction (IE) and retrieval infrastructure by identifying contextualized mentions of a wide range of biomedical concepts within EHRs. Natural language processing annotations are further assembled at the patient level and extended with EHR-specific knowledge to generate a timeline for each patient. The semantic data are serviced via ontology-based search and analytics interfaces.ResultsSemEHR has been deployed at a number of UK hospitals, including the Clinical Record Interactive Search, an anonymized replica of the EHR of the UK South London and Maudsley National Health Service Foundation Trust, one of Europe’s largest providers of mental health services. In 2 Clinical Record Interactive Search–based studies, SemEHR achieved 93% (hepatitis C) and 99% (HIV) F-measure results in identifying true positive patients. At King’s College Hospital in London, as part of the CogStack program (github.com/cogstack), SemEHR is being used to recruit patients into the UK Department of Health 100 000 Genomes Project (genomicsengland.co.uk). The validation study suggests that the tool can validate previously recruited cases and is very fast at searching phenotypes; time for recruitment criteria checking was reduced from days to minutes. Validated on open intensive care EHR data, Medical Information Mart for Intensive Care III, the vital signs extracted by SemEHR can achieve around 97% accuracy.ConclusionResults from the multiple case studies demonstrate SemEHR’s efficiency: weeks or months of work can be done within hours or minutes in some cases. SemEHR provides a more comprehensive view of patients, bringing in more and unexpected insight compared to study-oriented bespoke IE systems. SemEHR is open source, available at https://github.com/CogStack/SemEHR.

The Association Between Comorbid Autism Spectrum Disorders and Antipsychotic Treatment Failure in Early-Onset Psychosis: A Historical Cohort Study Using Electronic Health Records.

In a sample of children and adolescents with first-episode psychosis, we investigated whether multiple treatment failure (MTF, defined as the initiation of a third trial of novel antipsychotic due to nonadherence, adverse effects, or insufficient response) was associated with comorbid autism spectrum disorders. Data were from the electronic health records of 638 children (51% male) aged from 10 to 17 years with first-episode psychosis (per ICD-10 criteria) from January 1, 2008, to November 1, 2014, referred to mental health services in South London, United Kingdom; data were extracted using the Clinical Record Interactive Search (CRIS) system. The effect of autism spectrum disorder comorbidity on the development of MTF during a 5-year period was modeled using Cox regression. There were 124 cases of MTF prior to the age of 18 (19.4% of the sample). Comorbid autism spectrum disorders were significantly associated with MTF (adjusted hazard ratio = 1.99; 95% CI, 1.19-3.31; P = .008) after controlling for a range of potential confounders. Other factors significantly associated with MTF included higher age at first presentation (P = .001), black ethnicity (P = .03), and frequency of clinical contact (P < .001). No significant association between other comorbid neurodevelopmental disorders (hyperkinetic disorder or intellectual disability) and MTF was found. Children with first-episode psychosis and comorbid autism spectrum disorders at first presentation are less likely to have a beneficial response to antipsychotics.

Uncovering Capgras delusion using a large-scale medical records database.

BackgroundCapgras delusion is scientifically important but most commonly reported as single case studies. Studies analysing large clinical records databases focus on common disorders but none have investigated rare syndromes.AimsIdentify cases of Capgras delusion and associated psychopathology, demographics, cognitive function and neuropathology in light of existing models.MethodCombined computational data extraction and qualitative classification using 250 000 case records from South London and Maudsley Clinical Record Interactive Search (CRIS) database.ResultsWe identified 84 individuals and extracted diagnosis-matched comparison groups. Capgras was not ‘monothematic’ in the majority of cases. Most cases involved misidentified family members or close partners but others were misidentified in 25% of cases, contrary to dual-route face recognition models. Neuroimaging provided no evidence for predominantly right hemisphere damage. Individuals were ethnically diverse with a range of psychosis spectrum diagnoses.ConclusionsCapgras is more diverse than current models assume. Identification of rare syndromes complements existing ‘big data’ approaches in psychiatry.Declaration of interestsV.B. is supported by a Wellcome Trust Seed Award in Science (200589/Z/16/Z) and the UCLH NIHR Biomedical Research Centre. S.W. is supported by a Wellcome Trust Strategic Award (WT098455MA). Q.D. has received a grant from King’s Health Partners.Copyright and usage© The Royal College of Psychiatrists 2017. This is an open access article distributed under the terms of the Creative Commons Non-Commercial, No Derivatives (CC BY-NC-ND) license.