The new paradigm in machine learning - foundation models, large language models and beyond: a primer for physicians.

Foundation machine learning models are deep learning models capable of performing many different tasks using different data modalities such as text, audio, images and video. They represent a major shift from traditional task-specific machine learning prediction models. Large language models (LLM), brought to wide public prominence in the form of ChatGPT, are text-based foundational models that have the potential to transform medicine by enabling automation of a range of tasks, including writing discharge summaries, answering patients questions and assisting in clinical decision-making. However, such models are not without risk and can potentially cause harm if their development, evaluation and use are devoid of proper scrutiny. This narrative review describes the different types of LLM, their emerging applications and potential limitations and bias and likely future translation into clinical practice.

Developing an Adolescent and Young Adult Oncology Program in a Medium-Sized Canadian Centre: Lessons Learned.

The Adolescent and Young Adult (AYA) Program at CancerCare Manitoba (CCMB) has experienced tremendous growth since its inception. This report provides an overview of how the AYA program at CCMB was established and the crucial factors that led to its early accomplishments and continued expansion. These factors included actions and decisions made at the individual and organizational level that helped lay a strong foundation for the program's sustained success. We hope that some of these lessons learned can be adapted and implemented by other oncology agencies to improve the care outcomes and experiences of AYAs living with cancer.

EPidemiology Of Cardiogenic sHock in Scotland (EPOCHS): A multicentre, prospective observational study of the prevalence, management and outcomes of cardiogenic shock in Scotland.

Background: Despite high rates of cardiovascular disease in Scotland, the prevalence and outcomes of patients with cardiogenic shock are unknown. Methods: We undertook a prospective observational cohort study of consecutive patients with cardiogenic shock admitted to the intensive care unit (ICU) or coronary care unit at 13 hospitals in Scotland for a 6-month period. Denominator data from the Scottish Intensive Care Society Audit Group were used to estimate ICU prevalence; data for coronary care units were unavailable. We undertook multivariable logistic regression to identify factors associated with in-hospital mortality. Results: In total, 247 patients with cardiogenic shock were included. After exclusion of coronary care unit admissions, this comprised 3.0% of all ICU admissions during the study period (95% confidence interval [CI] 2.6%-3.5%). Aetiology was acute myocardial infarction (AMI) in 48%. The commonest vasoactive treatment was noradrenaline (56%) followed by adrenaline (46%) and dobutamine (40%). Mechanical circulatory support was used in 30%. Overall in-hospital mortality was 55%. After multivariable logistic regression, age (odds ratio [OR] 1.04, 95% CI 1.02-1.06), admission lactate (OR 1.10, 95% CI 1.05-1.19), Society for Cardiovascular Angiographic Intervention stage D or E at presentation (OR 2.16, 95% CI 1.10-4.29) and use of adrenaline (OR 2.73, 95% CI 1.40-5.40) were associated with mortality. Conclusions: In Scotland the prevalence of cardiogenic shock was 3% of all ICU admissions; more than half died prior to discharge. There was significant variation in treatment approaches, particularly with respect to vasoactive support strategy.

The adverse inpatient medication event and frailty (AIME-frail) risk prediction model.

BACKGROUND: Medication harm affects between 5 and 15% of hospitalised patients, with approximately half of the harm events considered preventable through timely intervention. The Adverse Inpatient Medication Event (AIME) risk prediction model was previously developed to guide a systematic approach to patient prioritisation for targeted clinician review, but frailty was not tested as a candidate predictor variable. AIM: To evaluate the predictive performance of an updated AIME model, incorporating a measure of frailty, when applied to a new multisite cohort of hospitalised adult inpatients. METHODS: A retrospective cohort study was conducted at two tertiary Australian hospitals on patients discharged between 1st January and April 31, 2020. Data were extracted from electronic medical records (EMRs) and clinical coding databases. Medication harm was identified using ICD-10 Y-codes and confirmed by senior pharmacist review of medical records. The Hospital Frailty Risk Score (HFRS) was calculated for each patient. Logistic regression analysis was used to construct a modified AIME model. Candidate variables of the original AIME model, together with new variables including HFRS were tested. Performance of the final model was reported using area under the curve (AUC) and decision curve analysis (DCA). RESULTS: A total of 4089 patient admissions were included, with a mean age ± standard deviation (SD) of 64 years (±19 years), 2050 patients (50%) were males, and mean HFRS was 6.2 (±5.9). 184 patients (4.5%) experienced one or more medication harm events during hospitalisation. The new AIME-Frail risk model incorporated 5 of the original variables: length of stay (LOS), anti-psychotics, antiarrhythmics, immunosuppressants, and INR greater than 3, as well as 5 new variables: HFRS, anticoagulants, antibiotics, insulin, and opioid use. The AUC was 0.79 (95% CI: 0.76-0.83) which was superior to the original model (AUC = 0.70, 95% CI: 0.65-0.74) with a sensitivity of 69%, specificity of 81%, positive predictive value of 0.14 (95% CI: 0.10-0.17) and negative predictive value of 0.98 (95% CI: 0.97-0.99). The DCA identified the model as having potential clinical utility between the probability thresholds of 0.05-0.4. CONCLUSION: The inclusion of a frailty measure improved the predictive performance of the AIME model. Screening inpatients using the AIME-Frail tool could identify more patients at high-risk of medication harm who warrant timely clinician review.

Reducing potentially inappropriate polypharmacy at a national and international level: the impact of deprescribing networks.

INTRODUCTION: Over the past decade, polypharmacy has increased dramatically. Measurable harms include falls, fractures, cognitive impairment, and death. The associated costs are massive and contribute substantially to low-value health care. Deprescribing is a promising solution, but there are barriers. Establishing a network to address polypharmacy can help overcome barriers by connecting individuals with an interest and expertise in deprescribing and can act as an important source of motivation and resources. AREAS COVERED: Over the past decade, several deprescribing networks were launched to help tackle polypharmacy, with evidence of individual and collective impact. A network approach has several advantages; it can spark interest, ideas and enthusiasm through information sharing, meetings and conversations with the public, providers, and other key stakeholders. In this special report, the details of how four deprescribing networks were established across the globe are detailed. EXPERT OPINION: Networks create links between people who lead existing and/or budding deprescribing practices and policy initiatives, can influence people with a shared passion for deprescribing, and facilitate sharing of intellectual capital and tools to take initiatives further and strengthen impact.This report should inspire others to establish their own deprescribing networks, a critical step in accelerating a global deprescribing movement.

How should artificial intelligence be used in Australian health care? Recommendations from a citizens' jury.

OBJECTIVE: To support a diverse sample of Australians to make recommendations about the use of artificial intelligence (AI) technology in health care. STUDY DESIGN: Citizens' jury, deliberating the question: "Under which circumstances, if any, should artificial intelligence be used in Australian health systems to detect or diagnose disease?" SETTING, PARTICIPANTS: Thirty Australian adults recruited by Sortition Foundation using random invitation and stratified selection to reflect population proportions by gender, age, ancestry, highest level of education, and residential location (state/territory; urban, regional, rural). The jury process took 18 days (16 March - 2 April 2023): fifteen days online and three days face-to-face in Sydney, where the jurors, both in small groups and together, were informed about and discussed the question, and developed recommendations with reasons. Jurors received extensive information: a printed handbook, online documents, and recorded presentations by four expert speakers. Jurors asked questions and received answers from the experts during the online period of the process, and during the first day of the face-to-face meeting. MAIN OUTCOME MEASURES: Jury recommendations, with reasons. RESULTS: The jurors recommended an overarching, independently governed charter and framework for health care AI. The other nine recommendation categories concerned balancing benefits and harms; fairness and bias; patients' rights and choices; clinical governance and training; technical governance and standards; data governance and use; open source software; AI evaluation and assessment; and education and communication. CONCLUSIONS: The deliberative process supported a nationally representative sample of citizens to construct recommendations about how AI in health care should be developed, used, and governed. Recommendations derived using such methods could guide clinicians, policy makers, AI researchers and developers, and health service users to develop approaches that ensure trustworthy and responsible use of this technology.

Monoclonal antibodies for treating early Alzheimer disease-a commentary on recent 'positive' trials.

Recent phase 3 randomised controlled trials of amyloid-targeting monoclonal antibodies in people with pre-clinical or early Alzheimer disease have reported positive results, raising hope of finally having disease-modifying drugs. Given their far-reaching implications for clinical practice, the methods and findings of these trials, and the disease causation theory underpinning the mechanism of drug action, need to be critically appraised. Key considerations are the representativeness of trial populations; balance of prognostic factors at baseline; psychometric properties and minimal clinically important differences of the primary efficacy outcome measures; level of study fidelity; consistency of subgroup analyses; replication of findings in similar trials; sponsor role and potential conflicts of interest; consistency of results with disease causation theory; cost and resource estimates; and alternative prevention and treatment strategies. In this commentary, we show shortcomings in each of these areas and conclude that monoclonal antibody treatment for early Alzheimer disease is lacking high-quality evidence of clinically meaningful impacts at an affordable cost.

Through a Tainted Lens: A Qualitatve Study of Medical Learners' Thinking About Patient 'Deservingness' of Health Advocacy.

Introduction: While health advocacy is a key component of many competency frameworks, mounting evidence suggests that learners do not see it as core to their learning and future practice. When learners do advocate for their patients, they characterize this work as 'going above and beyond' for a select few patients. When they think about advocacy in this way, learners choose who deserves their efforts. For educators and policymakers to support learners in making these decisions thoughtfully and ethically, we must first understand how they are currently thinking about patient deservingness. Methods: We conducted qualitative interviews with 29 undergraduate and postgraduate medical learners, across multiple sites and disciplines, to discuss their experiences of and decision-making about health advocacy. We then carried out a thematic analysis to understand how learners decided when and for whom to advocate. Stemming from initial inductive coding, we then developed a deductive coding framework, based in existing theory conceptualizing 'deservingness.' Results: Learners saw their patients as deserving of advocacy if they believed that the patient: was not responsible for their condition, was more in need of support than others, had a positive attitude, was working to improve their health, and shared similarities to the learner. Learners noted the tensions inherent in, and discomfort with, their own thinking about patient deservingness. Discussion: Learners' decisions about advocacy deservingness are rooted in their preconceptions about the patient. Explicit curriculum and conversations about advocacy decisions are needed to support learners in making advocacy decisions equitably.

Designer Small-Molecule Control System Based on Minocycline-Induced Disruption of Protein-Protein Interaction.

A versatile, safe, and effective small-molecule control system is highly desirable for clinical cell therapy applications. Therefore, we developed a two-component small-molecule control system based on the disruption of protein-protein interactions using minocycline, an FDA-approved antibiotic with wide availability, excellent biodistribution, and low toxicity. The system comprises an anti-minocycline single-domain antibody (sdAb) and a minocycline-displaceable cyclic peptide. Here, we show how this versatile system can be applied to OFF-switch split CAR systems (MinoCAR) and universal CAR adaptors (MinoUniCAR) with reversible, transient, and dose-dependent suppression; to a tunable T cell activation module based on MyD88/CD40 signaling; to a controllable cellular payload secretion system based on IL12 KDEL retention; and as a cell/cell inducible junction. This work represents an important step forward in the development of a remote-controlled system to precisely control the timing, intensity, and safety of therapeutic interventions.

Systematic review and longitudinal analysis of implementing Artificial Intelligence to predict clinical deterioration in adult hospitals: what is known and what remains uncertain.

OBJECTIVE: To identify factors influencing implementation of machine learning algorithms (MLAs) that predict clinical deterioration in hospitalized adult patients and relate these to a validated implementation framework. MATERIALS AND METHODS: A systematic review of studies of implemented or trialed real-time clinical deterioration prediction MLAs was undertaken, which identified: how MLA implementation was measured; impact of MLAs on clinical processes and patient outcomes; and barriers, enablers and uncertainties within the implementation process. Review findings were then mapped to the SALIENT end-to-end implementation framework to identify the implementation stages at which these factors applied. RESULTS: Thirty-seven articles relating to 14 groups of MLAs were identified, each trialing or implementing a bespoke algorithm. One hundred and seven distinct implementation evaluation metrics were identified. Four groups reported decreased hospital mortality, 1 significantly. We identified 24 barriers, 40 enablers, and 14 uncertainties and mapped these to the 5 stages of the SALIENT implementation framework. DISCUSSION: Algorithm performance across implementation stages decreased between in silico and trial stages. Silent plus pilot trial inclusion was associated with decreased mortality, as was the use of logistic regression algorithms that used less than 39 variables. Mitigation of alert fatigue via alert suppression and threshold configuration was commonly employed across groups. CONCLUSIONS: : There is evidence that real-world implementation of clinical deterioration prediction MLAs may improve clinical outcomes. Various factors identified as influencing success or failure of implementation can be mapped to different stages of implementation, thereby providing useful and practical guidance for implementers.

The relationship between pain and depression and anxiety in patients with inflammatory arthritis: a systematic review protocol.

Pain is a major challenge for patients with inflammatory arthritis (IA). Depression and anxiety are common comorbidities in IA, associating with worse outcomes. How they relate to pain is uncertain, with existing systematic reviews (a) mainly considering cross-sectional studies, (b) focusing on the relationship between pain and mental health in the context of disease activity/quality of life, and (c) not specifically considering the impact of treating depression/anxiety on pain. This PROSPERO-registered (CRD42023411823) systematic review will address this knowledge-gap by synthesizing evidence to summarise the associations (and potential mediators) between pain and depression/anxiety and evaluate the impact of treating co-morbid depression/anxiety on pain in IA. Relevant databases will be searched, articles screened and their quality appraised (using Joanna Briggs Institute critical appraisal tools) by two reviewers. Eligible studies will include adults with rheumatoid arthritis or spondyloarthritis, be a clinical trial or observational study, and either (a) report the relationship between pain and depression/anxiety (observational studies/baseline trials), or (b) randomise participants to a pharmacological or psychological treatment to manage depression/anxiety with a pain outcome as an endpoint (trials). To synthesise data on the association between pain and depression/anxiety, where available adjusted coefficients from regression models will be pooled in a random-effects meta-analysis. A synthesis without meta-analysis will summarise mediators. To evaluate the impact of treating depression/anxiety on pain, endpoint mean differences between treatment arms will be combined in a random-effects meta-analysis. Through understanding how depression/anxiety contribute to pain in IA, our review has the potential to help optimise approaches to IA pain.

Risk factors predicting hospital-acquired pressure injury in adult patients: An overview of reviews.

BACKGROUND: Hospital-acquired pressure injuries remain a significant patient safety threat. Current well-known pressure injury risk assessment tools have many limitations and therefore do not accurately predict the risk of pressure injury development over diverse populations. A contemporary understanding of the risk factors predicting pressure injury in adult hospitalised patients will inform pressure injury prevention and future researchers considering risk assessment tool development may benefit from our summary and synthesis of risk factors. OBJECTIVE: To summarise and synthesise systematic reviews that identify risk factors for hospital-acquired pressure injury development in adult patients. DESIGN: An overview of systematic reviews. METHODS: Cochrane and the Joanna Briggs Institute methodologies guided this overview. The Cochrane library, CINAHL, MEDLINE, and Embase databases were searched for relevant articles published in English from January 2008 to September 2022. Two researchers independently screened articles against the predefined inclusion and exclusion criteria, extracted data and assessed the quality of the included reviews using "a measurement tool to assess systematic reviews" (AMSTAR version 2). Data were categorised using an inductive approach and synthesised according to the recent pressure injury conceptual frameworks. RESULTS: From 11 eligible reviews, 37 risk factors were categorised inductively into 14 groups of risk factors. From these, six groups were classified into two domains: four to mechanical boundary conditions and two to susceptibility and tolerance of the individual. The remaining eight groups were evident across both domains. Four main risk factors, including diabetes, length of surgery or intensive care unit stay, vasopressor use, and low haemoglobin level were synthesised. The overall quality of the included reviews was low in five studies (45 %) and critically low in six studies (55 %). CONCLUSIONS: Our findings highlighted the limitations in the methodological quality of the included reviews that may have influenced our results regarding risk factors. Current risk assessment tools and conceptual frameworks do not fully explain the complex and changing interactions amongst risk factors. This may warrant the need for more high-quality research, such as cohort studies, focussing on predicting hospital-acquired pressure injury in adult patients, to reconsider these risk factors we synthesised. REGISTRATION: This overview was registered with the PROSPERO (CRD42022362218) on 27 September 2022.

A Randomized Phase I Study of the Anti-Interleukin-33 Antibody Tozorakimab in Healthy Adults and Patients With Chronic Obstructive Pulmonary Disease.

Tozorakimab is a human monoclonal antibody that neutralizes interleukin (IL)-33. IL-33 is a broad-acting epithelial "alarmin" cytokine upregulated in lung tissue of patients with chronic obstructive pulmonary disease (COPD). This first-in-human, phase I, randomized, double-blind, placebo-controlled study (NCT03096795) evaluated the safety, tolerability, pharmacokinetics (PKs), immunogenicity, target engagement, and pharmacodynamics (PDs) of tozorakimab. This was a 3-part study. In part 1, 56 healthy participants with a history of mild atopy received single escalating doses of either intravenous or subcutaneous tozorakimab or placebo. In part 2, 24 patients with mild COPD received multiple ascending doses of subcutaneous tozorakimab or placebo. In part 3, 8 healthy Japanese participants received a single intravenous dose of tozorakimab or placebo. The safety data collected included treatment-emergent adverse events (TEAEs), vital signs, and clinical laboratory parameters. Biological samples for PKs, immunogenicity, target engagement, and PD biomarker analyses were collected. No meaningful differences in the frequencies of TEAEs were observed between the tozorakimab and placebo arms. Three tozorakimab-treated participants with COPD experienced treatment-emergent serious adverse events. Subcutaneous or intravenous tozorakimab demonstrated linear, time-independent PKs with a mean half-life of 11.7-17.3 days. Treatment-emergent anti-drug antibody frequency was low. Engagement of tozorakimab with endogenous IL-33 in serum and nasal airways was demonstrated. Tozorakimab significantly reduced serum IL-5 and IL-13 levels in patients with COPD compared with placebo. Overall, tozorakimab was well tolerated, with a linear, time-independent serum PK profile. Additionally, biomarker studies demonstrated proof of mechanism. Overall, these data support the further clinical development of tozorakimab in COPD and other inflammatory diseases.

Early mucosal events promote distinct mucosal and systemic antibody responses to live attenuated influenza vaccine.

Compared to intramuscular vaccines, nasally administered vaccines have the advantage of inducing local mucosal immune responses that may block infection and interrupt transmission of respiratory pathogens. Live attenuated influenza vaccine (LAIV) is effective in preventing influenza in children, but a correlate of protection for LAIV remains unclear. Studying young adult volunteers, we observe that LAIV induces distinct, compartmentalized, antibody responses in the mucosa and blood. Seeking immunologic correlates of these distinct antibody responses we find associations with mucosal IL-33 release in the first 8 hours post-inoculation and divergent CD8+ and circulating T follicular helper (cTfh) T cell responses 7 days post-inoculation. Mucosal antibodies are induced separately from blood antibodies, are associated with distinct immune responses early post-inoculation, and may provide a correlate of protection for mucosal vaccination. This study was registered as NCT04110366 and reports primary (mucosal antibody) and secondary (blood antibody, and nasal viral load and cytokine) endpoint data.

Analgesic prescribing in patients with inflammatory arthritis in England: observational studies in the Clinical Practice Research Datalink.

OBJECTIVES: Despite little evidence that analgesics are effective in inflammatory arthritis (IA), studies report substantial opioid prescribing. The extent this applies to other analgesics is uncertain. We undertook a comprehensive evaluation of analgesic prescribing in patients with IA in the Clinical Practice Research Datalink Aurum to evaluate this. METHODS: From 2004 to 2020, cross-sectional analyses evaluated analgesic prescription annual prevalence in RA, PsA and axial spondyloarthritis (axSpA), stratified by age, sex, ethnicity, deprivation and geography. Joinpoint regression evaluated temporal prescribing trends. Cohort studies determined prognostic factors at diagnosis for chronic analgesic prescriptions using Cox proportional hazards models. RESULTS: Analgesic prescribing declined over time but remained common: 2004 and 2020 IA prescription prevalence was 84.2/100 person-years (PY) (95% CI 83.9, 84.5) and 64.5/100 PY (64.2, 64.8), respectively. In 2004, NSAIDs were most prescribed (56.1/100 PY; 55.8, 56.5), falling over time. Opioids were most prescribed in 2020 (39.0/100 PY; 38.7, 39.2). Gabapentinoid prescribing increased: 2004 prevalence 1.1/100 PY (1.0, 1.2); 2020 prevalence 9.9/100 PY (9.7, 10.0). Most opioid prescriptions were chronic (2020 prevalence 23.4/100 PY [23.2, 23.6]). Non-NSAID analgesic prescribing was commoner in RA, older people, females and deprived areas/northern England. Conversely, NSAID prescribing was commoner in axSpA/males, varying little by deprivation/geography. Peri-diagnosis was high-risk for starting chronic opioid/NSAID prescriptions. Prognostic factors for chronic opioid/gabapentinoid and NSAID prescriptions differed, with NSAIDs having no consistently significant association with deprivation (unlike opioids/gabapentinoids). CONCLUSION: IA analgesic prescribing of all classes is widespread. This is neither evidence-based nor in line with guidelines. Peri-diagnosis is an opportune moment to reduce chronic analgesic prescribing.

Factors influencing practice choices of early-career family physicians in Canada: a qualitative interview study.

BACKGROUND: Comprehensiveness of primary care has been declining, and much of the blame has been placed on early-career family physicians and their practice choices. To better understand early-career family physicians' practice choices in Canada, we sought to identify the factors that most influence their decisions about how to practice. METHODS: We conducted a qualitative study using framework analysis. Family physicians in their first 10 years of practice were recruited from three Canadian provinces: British Columbia, Ontario, and Nova Scotia. Interview data were coded inductively and then charted onto a matrix in which each participant's data were summarized by code. RESULTS: Of the 63 participants that were interviewed, 24 worked solely in community-based practice, 7 worked solely in focused practice, and 32 worked in both settings. We identified four practice characteristics that were influenced (scope of practice, practice type and model, location of practice, and practice schedule and work volume) and three categories of influential factors (training, professional, and personal). CONCLUSIONS: This study demonstrates the complex set of factors that influence practice choices by early-career physicians, some of which may be modifiable by policymakers (e.g., policies and regulations) while others are less so (e.g., family responsibilities). Participants described individual influences from family considerations to payment models to meeting community needs. These findings have implications for both educators and policymakers who seek to support and expand comprehensive care.