Machine learning (ML) models show promise in predicting post-traumatic stress disorder (PTSD) treatment outcomes, but it is unknown how their predictions compare to those of clinicians. This study directly compared the accuracy of clinicians’ predictions of patient treatment outcomes with those of three ML models.

Twenty clinicians providing cognitive processing therapy repeatedly predicted outcomes for 194 veterans. We compared their accuracy against three ML models on two key endpoints: clinically meaningful symptom reduction (≥10-point PCL-5 decrease) and posttreatment severity (final PCL-5 < 33). Clinician predictions were compared against a recurrent neural network, a mixed-effects random forest, and a generalized linear mixed-effects model. We analyzed prediction accuracy and the association between clinician confidence and accuracy using logistic mixed-effects models.

ML models were significantly more accurate than clinicians at predicting whether a patient’s posttreatment PCL-5 score would be below 33 (p < .001). However, no significant difference in accuracy was found for predicting a ≥10-point symptom reduction (p = .734). Clinician confidence increased throughout treatment and was significantly associated with greater prediction accuracy for both outcomes (ORs = 1.06, ps < .001).

ML models can outperform clinicians in predicting posttreatment symptom severity, particularly early in treatment, suggesting they could be a useful tool for identifying patients at risk for suboptimal outcomes. However, ML models were not superior in predicting symptom reduction, where clinicians also performed at a high level. Findings support the selective use of ML to enhance, rather than replace, clinical judgment in PTSD treatment.

There is a significant mortality gap between the general population and people with psychosis. Completion rates of regular physical health assessments for cardiovascular risk in this group are suboptimal. Point-of-care testing (POCT) for diabetes and hyperlipidaemia – providing an immediate result from a finger-prick – could improve these rates.

To evaluate the impact on patient–clinician encounters and on physical health check completion rates of implementing POCT for cardiovascular risk markers in early intervention in psychosis (EIP) services in South East England.

A mixed-methods, real-world evaluation study was performed, with 40 POCT machines introduced across EIP teams in all eight mental health trusts in South East England from March to May 2021. Clinician training and support was provided. Numbers of completed physical health checks, HbA1c and lipid panel blood tests completed 6 and 12 months before and 6 months after introduction of POCT were collected for individual patients. Data were compared with those from the South West region, which acted as a control. Clinician questionnaires were administered at 2 and 8 months, capturing device usability and impacts on patient interactions.

Post-POCT, South East England saw significant increases in HbA1c testing (odds ratio 2.02, 95% CI 1.17–3.49), lipid testing (odds ratio 2.38, 95% CI 1.43–3.97) and total completed health checks (odds ratio 3.61, 95% CI 1.94–7.94). These increases were not seen in the South West. Questionnaires revealed improved patient engagement, clinician empowerment and patients’ preference for POCT over traditional blood tests.

POCT is associated with improvements in the completion and quality of physical health checks, and thus could be a tool to enhance holistic care for individuals with psychosis.

Accurate diagnosis of bipolar disorder (BPD) is difficult in clinical practice, with an average delay between symptom onset and diagnosis of about 7 years. A depressive episode often precedes the first manic episode, making it difficult to distinguish BPD from unipolar major depressive disorder (MDD).

We use genome-wide association analyses (GWAS) to identify differential genetic factors and to develop predictors based on polygenic risk scores (PRS) that may aid early differential diagnosis.

Based on individual genotypes from case–control cohorts of BPD and MDD shared through the Psychiatric Genomics Consortium, we compile case–case–control cohorts, applying a careful quality control procedure. In a resulting cohort of 51 149 individuals (15 532 BPD patients, 12 920 MDD patients and 22 697 controls), we perform a variety of GWAS and PRS analyses.

Although our GWAS is not well powered to identify genome-wide significant loci, we find significant chip heritability and demonstrate the ability of the resulting PRS to distinguish BPD from MDD, including BPD cases with depressive onset (BPD-D). We replicate our PRS findings in an independent Danish cohort (iPSYCH 2015, N = 25 966). We observe strong genetic correlation between our case–case GWAS and that of case–control BPD.

We find that MDD and BPD, including BPD-D are genetically distinct. Our findings support that controls, MDD and BPD patients primarily lie on a continuum of genetic risk. Future studies with larger and richer samples will likely yield a better understanding of these findings and enable the development of better genetic predictors distinguishing BPD and, importantly, BPD-D from MDD.

Despite the global expansion of electronic medical record (EMR) systems and their increased integration with artificial intelligence (AI), their utilization in disaster settings remains limited, and few studies have evaluated their implementation. We aimed to evaluate Fast Electronic Medical Record (fEMR), a novel, mobile EMR designed for resource-limited settings, based on user feedback.

We examined usage data through October 2022 to categorize the nature of its use for disaster response and determine the number of patients served. We conducted interviews with stakeholders and gathered input from clinicians who had experience using fEMR.

Over eight years, fEMR was employed 60 times in 11 countries across four continents by 14 organizations (universities, non-profits, and disaster response teams). This involved 37,500+ patient encounters in diverse settings including migrant camps at the US-Mexico and Poland-Ukraine borders, mobile health clinics in Kenya and Guatemala, and post-earthquake relief in Haiti. User feedback highlighted adaptability, but suggested hardware and workflow improvements.

EMR systems have the potential to enhance healthcare delivery in humanitarian responses, offer valuable data for planning and preparedness, and support measurement of effectiveness. As a simple, versatile EMR system, fEMR has been deployed to numerous disaster response and low-income settings.

Human immunodeficiency virus (HIV)-associated neurocognitive disorder (HAND) prevalence is expected to increase in East Africa as treatment coverage increases, survival improves, and this population ages. This study aimed to better understand the current cognitive phenotype of this newly emergent population of older combination antiretroviral therapy (cART)-treated people living with HIV (PLWH), in which current screening measures lack accuracy. This will facilitate the refinement of HAND cognitive screening tools for this setting.

This is a secondary analysis of 253 PLWH aged ≥50 years receiving standard government HIV clinic follow-up in Kilimanjaro, Tanzania. They were evaluated with a detailed locally normed low-literacy neuropsychological battery annually on three occasions and a consensus panel diagnosis of HAND by Frascati criteria based on clinical evaluation and collateral history.

Tests of verbal learning and memory, categorical verbal fluency, visual memory, and visuoconstruction had an area under the receiver operating characteristic curve >0.7 for symptomatic HAND (s-HAND) (0.70–0.72; p < 0.001 for all tests). Tests of visual memory, verbal learning with delayed recall and recognition memory, psychomotor speed, language comprehension, and categorical verbal fluency were independently associated with s-HAND in a logistic mixed effects model (p < 0.01 for all). Neuropsychological impairments varied by educational background.

A broad range of cognitive domains are affected in older, well-controlled, East African PLWH, including those not captured in widely used screening measures. It is possible that educational background affects the observed cognitive impairments in this setting. Future screening measures for similar populations should consider assessment of visual memory, verbal learning, language comprehension, and executive and motor function.

There is a paucity of data guiding treatment duration of oral vancomycin for Clostridiodes difficile infection (CDI) in patients requiring concomitant systemic antibiotics.

To evaluate prescribing practices of vancomycin for CDI in patients that required concurrent systemic antibiotics and to determine whether a prolonged duration of vancomycin (>14 days), compared to a standard duration (10–14 days), decreased CDI recurrence.

In this retrospective cohort study, we evaluated adult hospitalized patients with an initial episode of CDI who were treated with vancomycin and who received overlapping systemic antibiotics for >72 hours. Outcomes of interest included CDI recurrence and isolation of vancomycin-resistant Enterococcus (VRE).

Among the 218 patients included, 36% received a standard duration and 64% received a prolonged duration of treatment for a median of 13 days (11–14) and 20 days (16–26), respectively. Patients who received a prolonged duration had a longer median duration of systemic antibiotic overlap with vancomycin (11 vs 8 days; P < .001) and significantly more carbapenem use and infectious disease consultation. Recurrence at 8 weeks (12% standard duration vs 8% prolonged duration; P = .367), recurrence at 6 months (15% standard duration vs 10% prolonged duration; P = .240), and VRE isolation (3% standard duration vs 9% prolonged duration; P = .083) were not significantly different between groups. Discontinuation of vancomycin prior to completion of antibiotics was an independent predictor of 8-week recurrence on multivariable logistic regression (OR, 4.8; 95% CI, 1.3–18.1).

Oral vancomycin prescribing relative to the systemic antibiotic end date may affect CDI recurrence to a greater extent than total vancomycin duration alone. Further studies are needed to confirm these findings.

The absence of clinical information in the aftermath of disasters in resource-constrained environments costs lives. fEMR– fast Electronic Medical Records–is a medical records system designed for mobile clinics and has proven useful in post-disaster settings. While the original version of the system was developed for areas without access to the Internet, a new version of this system was developed in 2019 to accommodate regions with connectivity.

We reviewed the design, implementation, and usage of fEMR from June 2014 to October 2022. We used logged data of the number of users, patient encounters, and the circumstances of each deployment. We compared usage between the original fEMR system and fEMR-on-chain.

The original fEMR system was created in an iterative process by students in Computer Science classes at three different American universities. The system creates a closed intranet signal to which clinicians connect their own device to access the software. The hardware is transported to the medical team in a carry-on suitcase prior to deployment. All data are stored on a laptop that acts as a server. The online version, fEMR On-Chain, was developed under a grant, but is sustained in development through academic partnerships. Both versions are designed so that the provider can complete an encounter with as few clicks as possible and with as little input as necessary to identify patients.The original fEMR system has been deployed to mobile clinics worldwide since 2014. The system has about 14,181 patients and 16,021 clinical encounters from 12 different countries. fEMR On-Chain has been deployed to refugee and migrant settings since 2019, containing about 18,000 patients and 22,000 encounters in two different countries.

Successive versions of the fEMR system have been used in a variety of conditions and settings, with usage accelerating since 2019 in refugee and migrant health centers.

The psychosis continuum implies that subclinical psychotic experiences (PEs) can be differentiated from clinically relevant expressions since they are not accompanied by a ‘need for care’.

Using data from Wave 2 of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC; N = 34 653), the current study examined variation in functioning, symptomology and aetiological risk across the psychosis phenotype [i.e. variation from (i) no PEs, ‘No PEs’ to (ii) non-distressing PEs, ‘PE-Experienced Only’ to (iii) distressing PEs, ‘PE-Impaired’ to (iv) clinically defined psychotic disorder, ‘Diagnosed’].

A graded trend was present such that, compared to those with no PEs, the Diagnosed group had the poorest functioning, followed by the PE-Impaired then PE-Experienced Only groups. In relation to symptom expression, the PE-Impaired group were more likely than the PE-Experienced Only and the Diagnosed groups to endorse most PEs. Predictors of group membership tended to vary quantitatively rather than qualitatively. Trauma, current mental health diagnoses (anxiety and depression) and drug use variables differentiated between all levels of the continuum, with the exception of the extreme end (PE-Impaired v. Diagnosed). Only a few variables distinguished groups at the upper end of the continuum: female sex, older age, unemployment, parental mental health hospitalisation and lower likelihood of having experienced physical assault.

The findings highlight the importance of continuum-based interpretations of the psychosis phenotype and afford valuable opportunities to consider if and how impairment, symptom expression and risk change along the continuum.