Nathan Hodson

BACKGROUND Competition for UK Internal Medicine Training (IMT) has risen sharply, from 1.70 applications per post in 2015 to 5.27 in 2026. Selection is by multi-station interview, which tests structured clinical and professional reasoning delivered under time pressure rather than knowledge alone. One common way candidates prepare is peer role-play, but it is not equally available: some lack a suitable practice partner or train remotely or overseas, and others opt out because they are reluctant to rehearse in front of colleagues. OBJECTIVE To describe MedPrep, an iOS application that provides on-demand simulated interview practice with automated feedback, and to report early user engagement during the IMT 2026 selection window as an initial feasibility and acceptability evaluation. METHODS In MedPrep, candidates give a timed spoken response to an interview station prompt and receive automated structured feedback and a score from a multimodal large language model (LLM). We retrospectively extracted backend usage data on 22 May 2026 for all users who registered during the observation window (1 October 2025 to 13 February 2026). Developer and test accounts, and users who completed no stations, were excluded. Metrics are reported descriptively. AI-generated scores are reported as platform outputs and were not validated against human examiners. RESULTS Of 149 registered users, 37 (24.8%) completed at least one station (real users) and 26 (17.4%) completed more than one (active users). All selected the IMT track and were recruited through social media and word of mouth. Active users completed a median of 6 stations (IQR 2-25; maximum 71) over a median of 42 app sessions (IQR 13-100; maximum 272). Real users completed 472 stations across all five IMT activities. Per available station, practice was most intensive on the formulaic portfolio and presentation tasks (12.3 completions per station) and lowest on the much larger clinical bank (5.7). The most-completed station was the 2-minute presentation (25/37 users). Of 472 completions, 445 carried a stored AI score (mean 6.35/10, SD 1.35), concentrated in the developing (4-6) and competent (7-8) bands. CONCLUSIONS MedPrep is a feasible and scalable way to deliver self-directed interview practice, with preliminary acceptability indicated by sustained voluntary use and organic recruitment. The automated feedback has not yet been validated and no learning or selection outcomes were measured. The next step is a controlled evaluation with validated outcomes and examiner-referenced scoring.

Context: During 2020–2021, many states offered financial incentives for COVID-19 vaccinations. Systematic review evidence suggests that this was often effective among adults. Objectives: We aimed to investigate whether incentives increased vaccination rates among 5-11-year-olds. Design: The synthetic control method was used. Setting: Minnesota's “Kids Deserve a Shot” policy. Participants: Data were drawn from the Centers for Disease Control and Prevention and the US Census Bureau. Relevant placebo and robustness tests were conducted. Intervention: The offer of financial incentives for pediatric COVID-19 vaccinations. Main Outcome Measure: The number of fully vaccinated 5-11-year-olds in Minnesota. Results: The synthetic control was well-matched with true Minnesota during the preintervention period. There was no difference between true Minnesota and the synthetic Minnesota during the postintervention period, indicating that the incentive policy had no effect. The robustness checks and placebo tests were all consistent with this result. Conclusions: Policymakers attempted various interventions to promote vaccine uptake, including financial incentives. Although this analysis did not find any differences between true Minnesota and synthetic Minnesota during the postintervention period, further research into which factors determine incentive effectiveness will strengthen interventions and protect communities in future public health crises.

Background. The Medicare Part D program provides stand-alone prescription drug insurance to US residents aged ≥65 y and those with disabilities. Annually, they can choose to stay with their plan or to switch to one of the many alternative plans. Behavioral decision-making research has proposed various strategies for tackling such complex choices. Here, we interviewed Medicare Part D beneficiaries to understand their real-world experiences with making their plan choices and examined quotes from those who stayed with their first plan and those who had switched. Method. In July 2023, we conducted phone interviews with 20 Medicare Part D beneficiaries aged 57 to 86 y who were recruited through the US-wide Understanding America Study. Results. Eleven interviewees had stayed with their first plan, and 9 had switched. Stayers and switchers applied similar strategies, with mixed success. Using decision-support tools such as the Medicare plan-finder Web site, booklet, and mailers could be time-consuming and confusing. Recommendations were taken from insurance brokers, agents, and pharmacists, often without questioning their advice. Interviewees varied widely in the amount of information they considered, with some using heuristics to reduce their choice set and others considering only 1 recommended option. Few beneficiaries made annual plan comparisons, and while some deferred choices until necessary, others did not act even after being confronted with plan changes or negative experiences – due to knowledge and time constraints. Discussion. Medicare beneficiaries use various strategies known from the behavioral decision-making research literature for tackling complex choices between Medicare Part D stand-alone plans. Each strategy posed its own challenges. We discuss our findings in light of the behavioral decision-making literature and recommend policies for helping beneficiaries with their plan choices. Highlights: We interviewed Medicare Part D beneficiaries about how they tackled drug insurance plan choices with many options and complicated features. Interviewees reported using decision-support tools, seeking help, applying heuristics, and relying on choice deferral, but with mixed success. Qualitatively, similar experiences were described by interviewees who stayed with their initial plan and those who had switched. We recommend policies for addressing the difficulties beneficiaries had in applying these strategies.

Medicare Part D innovatively included a market in public health insurance in the United States. Proponents argued that increased competition would drive better value for beneficiaries, but others feared that beneficiaries would struggle to navigate the complex program. Understanding how Part D beneficiaries choose between plans allows us to evaluate the extent to which Medicare Part D succeeds at increasing value to beneficiaries and where there is scope to support beneficiaries. Many Part D beneficiaries are sensitive to price cues in relation to pharmacy choice and medication adherence, yet frequently overpay for their plans. Empirical literature suggests that behavioral aspects including information overload, low salience, low trust, and practical ‘sludge’ all partly contribute to failure to switch. We propose solutions to address these barriers based on behavioral insights.

Background: Parenting skills programs are the primary intervention for conduct disorders in children. The Pause app enhances these programs by providing digital microinterventions that reinforce learning between sessions and after program completion. The potential of artificial intelligence (AI) in this context remains untapped. Hackathons have proven effective for health care innovation and can facilitate collaborative development in this space. Objective: We aimed to rapidly build AI-powered features in the Pause app to enhance parenting skills. Methods: We undertook a 1-day hackathon that included an ideation phase drawing on the Design Council’s double diamond framework and a development phase using microsprints based on agile and scrum approaches. The interdisciplinary participants included medical professionals, developers, and product managers. Results: Participants identified 3 core problems: generating age-appropriate distractions, receiving feedback on parenting efforts, and effectively using the journal function. During the solution phase, a wide range of options were explored, resulting in 3 key solutions: AI-assisted idea generation, a tool for summarizing parenting interactions, and a weekly journal roundup. During the development phase, participants completed 4 microsprints. Teams focused on 3 workstreams: building a “weekly roundup” module, creating an AI-based distraction generator, and developing a summarizer for active play sessions. These prototypes were integrated into the preproduction environment, with each workstream producing a functional component. Participant feedback (n=4) was unanimously positive, with all participants rating the event as “excellent” and highlighting the value of in-person collaboration. Conclusions: This 1-day hackathon used the double diamond approach to develop AI-powered features for parenting programs. Three solutions were explored across workstreams, resulting in 2 fully functioning and 1 near-functioning app component. The rapid problem-solving approach mirrors other health technology hackathons and highlights the untapped potential of AI in digital parenting support, surpassing traditional e-learning or video-based methods. This work suggests broader applications of AI-driven coaching in fields like social care. Despite a small team, the hackathon was focused and productive, generating relevant solutions based on prior engagement with parents and practitioners. Future research will assess the impact of the app’s AI-powered features on parenting outcomes.

Objective To evaluate adherence to National Health Service (NHS) patient registration ID guidelines among General Practitioners’ (GP) practices. Design A mystery shopper study, including website reviews and phone calls. Setting Rural and urban parts of the United Kingdom’s West Midlands. Participants 85 randomly selected GP practices. Primary and secondary outcome measures In January–April 2024, GP’s websites were reviewed before phone calls in which our ‘mystery shopper’ was asked to register without photo ID and proof of address. Results Of 85 GP practices, 60 (71%) breached NHS guidance either online or over the phone, with only 25 (29%) consistently following NHS guidance. Phone calls to rural (vs urban) GP practices were more likely to yield refusal of registration without photo ID and proof of address, despite rural (vs urban) GP practices making similar statements online. During some phone calls, practices sought to negotiate a compromise by requesting less robust ‘documentation’, such as an addressed parcel. Conclusions GP practices commonly refuse registration to people without photo ID or proof of address, thus creating ‘sludge’ and undermining access to healthcare especially for poor, vulnerable patients, including immigrants. Changing GP practices’ websites would not address this problem if erroneous information is still provided over the phone. GPs and practice managers should ensure that all staff follow NHS guidance to allow registration without these documents.

Background: Long-Acting injectable (LAI) antipsychotics prevent relapse in patients with unreliable adherence to oral antipsychotics. Non-engagement with treatment is associated with relapse and involuntary admission. Financial incentives have been shown to increase engagement with LAI antipsychotics among those at highest risk of relapse. This study aimed to explore whether community mental health professionals who deliver LAI antipsychotics find financial incentives acceptable and whether empirical research has addressed their concerns. Methods: Fourteen mental health professionals who deliver LAI antipsychotics were interviewed after snowball recruitment. Each then saw a presentation outlining recent research findings into financial incentives for LAI antipsychotic engagement before interviews resumed. Interviews explored the advantages and challenges professionals anticipated with respect to financial incentives for LAIs. Thematic analysis was conducted and themes were compared with findings from the literature. Results: Median self-reported support increased from “5/10” to “6.75/10” after the information presentation. Participants responded positively to the idea that incentives could be seen as ‘rewards’. They readily drew connections between improvements in engagement, increased insight, and stronger therapeutic relationships. Practical implementation challenges were envisioned along with proposed solutions. Several participants remained concerned that financial incentives would lead to increased drug use. Conclusions: Mental health professionals who deliver LAI antipsychotics hold a range of views about the potential role of financial incentives to increase engagement. When professionals are aware of research findings they are more positive about incentives. Further research should consider patients’ and professionals’ preferences for the design of incentive regimes.

Aim To develop a consensus curriculum describing what information service users should be given prior to cognitive behavioural therapy (CBT). Design A Delphi study was undertaken. Following development work with professionals and experts by experience, 30 initial statements were prepared and survey rounds were undertaken until 90% of statements achieved 80% consensus on inclusion or exclusion. Setting NHS Talking Therapies services in England provide CBT following referral by a professional or self-referral. Out of 1.7 million referrals in 2022–2023 more than 1 million disengaged before or during therapy. Ensuring patients are adequately prepared for CBT is one promising approach to improving engagement and outcomes. Participants Participants were only included if they had either 3 years of clinical experience in NHS Talking Therapies services or a leadership role in NHS Talking Therapies services. They reported current employment at seven NHS Trusts in England as well as private practice. Results Of the 41 participants, 36 completed all three rounds. After three rounds, 27 statements were included by consensus in the curriculum, covering six domains: Emergency information (two statements), What is CBT? (six statements), Practical preparation for CBT (five statements), What does a CBT session look like? (five statements), What is CBT not? (two statements) and How can people get the most out of CBT? (seven statements). Conclusions This consensus curriculum provides a basis for ensuring patients are well-prepared for CBT within the context of NHS Talking Therapies. Further research on improving engagement and outcomes from NHS Talking Therapies services should aim to address these 27 topics.