No abstract

Background: Understanding adolescents' mental health during lockdown and identifying those most at risk is an urgent public health challenge. This study surveyed school pupils across Southern England during the first COVID-19 school lockdown to investigate situational factors associated with mental health difficulties and how they relate to pupils' access to in-school educational provision. Methods: A total of 11,765 pupils in years 8-13 completed a survey in June-July 2020, including questions on mental health, risk indicators and access to school provision. Pupils at home were compared to those accessing in-school provision on risk and contextual factors and mental health outcomes. Multilevel logistic regression analyses compared the effect of eight risk and contextual factors, including access to in-school provision, on depression, anxiety and self-reported deterioration in mental wellbeing.Results: Females, pupils who had experienced food poverty and those who had previously accessed mental health support were at greatest risk of depression, anxiety and a deterioration in wellbeing. Pupils whose parents were going out to work and those preparing for national examinations in the subsequent school year were also at increased risk. Pupils accessing in-school provision had poorer mental health, but this was accounted for by the background risk and contextual factors assessed, in line with the allocation of in-school places to more vulnerable pupils. Conclusions:Although the strongest associations with poor mental health during school closures were established risk factors, further contextual factors of particular relevance during lockdown had negative impacts on wellbeing. Identifying those pupils at greatest risk for poor outcomes is critical for ensuring that appropriate educational and social support can be given to pupils either at home or inschool during subsequent lockdowns.

BackgroundUtilisation 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.ObjectiveOur 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.MethodsWe 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.FindingsResults 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.ConclusionsThis 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.Clinical implicationsReal-world, individual patient data are an invaluable source of information, which can be used to better personalise treatment.

Age-related decline may not be as pronounced in complex activities as it is in basic cognitive processes, but ability deterioration with age is difficult to deny. However, studies disagree on whether age is kinder to more able people than it is to their less able peers. In this article, we investigated the "age is kinder to the more able" hypothesis by using a chess database that contains activity records for both beginners and world-class players. The descriptive data suggested that the skill function across age captures the 3 phases as described in Simonton's model of career trajectories: initial rise to the peak of performance, postpeak decline, and eventual stabilization of decline. We therefore modeled the data with a linear mixed-effect model using the cubic function that captures 3 phases. The results show that age may be kind to the more able in a subtler manner than has previously been assumed. After reaching the peak at around 38 years, the more able players deteriorated more quickly. Their decline, however, started to slow down at around 52 years, earlier than for less able players (57 years). Both the decline and its stabilization were significantly influenced by activity. The more players engaged in playing tournaments, the less they declined and the earlier they started to stabilize. The best experts may not be immune to aging, but their previously acquired expertise and current activity enable them to maintain high levels of skill even at an advanced age.

Background: Oxford 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. Methods: In 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 freetext documents from the Clinical Record Interactive Search database to train a named entity recognition model, a machine learning technique which recognizes concepts in freetext. 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). Results: In 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