Boussina et al. recently evaluated a deep learning sepsis prediction model (COMPOSER) in a prospective before-and-after quasi-experimental study within two emergency departments at UC San Diego Health, tracking outcomes before and after deployment. Over the five-month implementation period, they reported a 17% relative reduction in in-hospital sepsis mortality and a 10% relative increase in sepsis bundle compliance. This editorial discusses the importance of shifting the focus towards evaluating clinically relevant outcomes, such as mortality reduction or quality-of-life improvements, when adopting artificial intelligence (AI) tools. We also explore the ecosystem vital for AI algorithms to succeed in the clinical setting, from interoperability standards and infrastructure to dashboards and action plans. Finally, we suggest that algorithms may eventually fail due to the human nature of healthcare, advocating for the need for continuous monitoring systems to ensure the adaptability of these tools in the ever-evolving healthcare landscape.

The utilization of artificial intelligence (AI) in diabetes care has focused on early intervention and treatment management. Notably, usage has expanded to predict an individual’s risk for developing type 2 diabetes. A scoping review of 40 studies by Mohsen et al. shows that while most studies used unimodal AI models, multimodal approaches were superior because they integrate multiple types of data. However, creating multimodal models and determining model performance are challenging tasks given the multi-factored nature of diabetes. For both unimodal and multimodal models, there are also concerns of bias with the lack of external validations and representation of race, age, and gender in training data. The barriers in data quality and evaluation standardization are ripe areas for developing new technologies, especially for entrepreneurs and innovators. Collaboration amongst providers, entrepreneurs, and researchers must be prioritized to ensure that AI in diabetes care is providing quality and equitable patient care.

Large language models (LLMs) have shown promise in reducing time, costs, and errors associated with manual data extraction. A recent study demonstrated that LLMs outperformed natural language processing approaches in abstracting pathology report information. However, challenges include the risks of weakening critical thinking, propagating biases, and hallucinations, which may undermine the scientific method and disseminate inaccurate information. Incorporating suitable guidelines (e.g., CANGARU), should be encouraged to ensure responsible LLM use.

We explore the evolving landscape of diagnostic artificial intelligence (AI) in dermatology, particularly focusing on deep learning models for a wide array of skin diseases beyond skin cancer. We critically analyze the current state of AI in dermatology, its potential in enhancing diagnostic accuracy, and the challenges it faces in terms of bias, applicability, and therapeutic recommendations.

This piece critiques the exclusion of healthcare practitioners (HCPs) from the digital health innovation process. Drawing on “Sync fast and solve things—best practices for responsible digital health” by Landers et al., the editorial argues for the importance of inclusive co-creation, in which clinicians play an active role in developing digital health solutions. It emphasizes that without the meaningful involvement of HCPs, digital health tools risk being clinically irrelevant.

Background: As the global prevalence of stroke continues to rise, it becomes increasingly pressing to investigate digital health behaviour change interventions that promote physical activity and reduce sedentary behaviour for stroke patients to support active lifestyles. Purpose: The primary aim of this study is to investigate the effectiveness of digital health interventions in promoting physical activity and reducing sedentary behaviour for stroke patients. The secondary aim is to investigate the intervention components that explain intervention effectiveness to further inform intervention development and policy making. Methods: A systematic search of the literature was conducted in four databases (Scopus, MEDLINE (PubMed), Web of Science, and PsychINFO) to identify the most robust evidence in the form of randomised controlled trials of digital interventions for patients with stroke. A random-effects meta-analysis were utilized to quantify the intervention effects on behaviour change, and subgroup analyses to characterise intervention effective components. Results: In total, 16 RCTs were deemed eligible and included in the systematic review. Meta-analyses suggested significant improvements in physical activity (SMD = 0.39, 95% CI 0.17, 0.61, N = 326, p < 0.001, I₂ = 0%), and reductions in time of sedentary behaviour (SMD= −0.45, 95% CI −0.76, -0.14, N = 167, p = 0.00, I₂ = 0%) after stroke. The 10 m walk test for physical activity, and the timed up and go test for sedentary behaviour, were the objective outcome measures in the most effective behavioural change interventions. Subgroup analyses found that most effective interventions were underpinned by theories of self-regulation and utilised interactive functions to engage patients with the processes of behaviour change. Conclusions: Digital self-monitoring behavioural interventions are effective in promoting physical activity for stroke patients in adjunct to usual care clinical practice and rehabilitation programmes. Rigorous studies are required to provide evidence to disentangle the most effective intervention components for preventative practices and rehabilitation programs and to inform policymaking for stroke treatment.