Search

Objective To determine if ChatGPT can generate useful suggestions for improving clinical decision support (CDS) logic and to assess noninferiority compared to human-generated suggestions. Methods We supplied summaries of CDS logic to ChatGPT, an artificial intelligence (AI) tool for question answering that uses a large language model, and asked it to generate suggestions. We asked human clinician reviewers to review the AI-generated suggestions as well as human-generated suggestions for improving the same CDS alerts, and rate the suggestions for their usefulness, acceptance, relevance, understanding, workflow, bias, inversion, and redundancy. Results Five clinicians analyzed 36 AI-generated suggestions and 29 human-generated suggestions for 7 alerts. Of the 20 suggestions that scored highest in the survey, 9 were generated by ChatGPT. The suggestions generated by AI were found to offer unique perspectives and were evaluated as highly understandable and relevant, with moderate usefulness, low acceptance, bias, inversion, redundancy. Conclusion AI-generated suggestions could be an important complementary part of optimizing CDS alerts, can identify potential improvements to alert logic and support their implementation, and may even be able to assist experts in formulating their own suggestions for CDS improvement. ChatGPT shows great potential for using large language models and reinforcement learning from human feedback to improve CDS alert logic and potentially other medical areas involving complex, clinical logic, a key step in the development of an advanced learning health system.

This briefing considers the environmental impact of pet parasiticides, which are commonly used to kill parasites such as fleas and ticks. It reviews possible routes that chemicals from veterinary parasiticides enter the environment, what impacts they may have on natural ecosystems and how to balance the needs of domestic pets, people, and the environment.

Objective We describe the Clickbusters initiative implemented at Vanderbilt University Medical Center (VUMC), which was designed to improve safety and quality and reduce burnout through the optimization of clinical decision support (CDS) alerts. Materials and Methods We developed a 10-step Clickbusting process and implemented a program that included a curriculum, CDS alert inventory, oversight process, and gamification. We carried out two 3-month rounds of the Clickbusters program at VUMC. We completed descriptive analyses of the changes made to alerts during the process, and of alert firing rates before and after the program. Results Prior to Clickbusters, VUMC had 419 CDS alerts in production, with 488 425 firings (42 982 interruptive) each week. After 2 rounds, the Clickbusters program resulted in detailed, comprehensive reviews of 84 CDS alerts and reduced the number of weekly alert firings by more than 70 000 (15.43%). In addition to the direct improvements in CDS, the initiative also increased user engagement and involvement in CDS. Conclusions At VUMC, the Clickbusters program was successful in optimizing CDS alerts by reducing alert firings and resulting clicks. The program also involved more users in the process of evaluating and improving CDS and helped build a culture of continuous evaluation and improvement of clinical content in the electronic health record.

ObjectiveTo determine if ChatGPT can generate useful suggestions for improving clinical decision support (CDS) logic and to assess noninferiority compared to human-generated suggestions.MethodsWe supplied summaries of CDS logic to ChatGPT, an artificial intelligence (AI) tool for question answering that uses a large language model, and asked it to generate suggestions. We asked human clinician reviewers to review the AI-generated suggestions as well as human-generated suggestions for improving the same CDS alerts, and rate the suggestions for their usefulness, acceptance, relevance, understanding, workflow, bias, inversion, and redundancy.ResultsFive clinicians analyzed 36 AI-generated suggestions and 29 human-generated suggestions for 7 alerts. Of the 20 suggestions that scored highest in the survey, 9 were generated by ChatGPT. The suggestions generated by AI were found to offer unique perspectives and were evaluated as highly understandable and relevant, with moderate usefulness, low acceptance, bias, inversion, redundancy.ConclusionAI-generated suggestions could be an important complementary part of optimizing CDS alerts, can identify potential improvements to alert logic and support their implementation, and may even be able to assist experts in formulating their own suggestions for CDS improvement. ChatGPT shows great potential for using large language models and reinforcement learning from human feedback to improve CDS alert logic and potentially other medical areas involving complex, clinical logic, a key step in the development of an advanced learning health system.

Non-small-cell lung cancer (NSCLC) is a heterogeneous disease, and therapeutic management has advanced to identify various critical oncogenic mutations that promote lung cancer tumorigenesis. Subsequent studies have developed targeted therapies against these oncogenes in the hope of personalized treatment based on the tumor's molecular genomics. This review presents a comprehensive review of the biology, new therapeutic interventions, and resistance patterns of two well-defined subgroups, tumors with KRAS and MET alterations. We also discuss the status of molecular testing practices for these two key oncogenic drivers, considering the progressive introduction of next-generation sequencing (NGS) and RNA sequencing in regular clinical practice.

Acute otitis media (AOM) is the most common reason for antibiotic use in children. The American Academy of Pediatrics (AAP) published its latest AOM guidelines in 2013. A safety-net antibiotic prescription (SNAP) is recommended for some patients based on age, severity, and duration of symptoms. At baseline, 78% of patients diagnosed with AOM in our general pediatrics practice met AAP guidelines, and 20% of eligible patients received a SNAP according to guidelines. We aimed to increase adherence to AAP AOM guidelines in an academic general pediatrics clinic from 78% to 90% by January 2020.A quality improvement team determined key drivers and developed interventions. Patients included were 6 months to 12 years old with AOM. Encounters were reviewed for adherence to AAP AOM guidelines. During the project, interventions included an ear pain note template, which generated guideline-based recommendations, note template education in clinic orientation sessions, a didactic session on AOM management, and reminders on workstations. Data were analyzed using P-charts.Percent of AOM encounters (n = 1266) adhering to AAP AOM guidelines increased from 78% to 92%. We also reviewed two process measures. First, the use of the ear pain note template increased from 0% to 44%. Second, the percent of AOM encounters where an eligible patient received a SNAP increased from 21% to 78% (encounters n = 421).We demonstrate increased adherence to AAP AOM guidelines, including improved use of SNAPs after introducing a note template with clinical decision support and provider educational sessions.

Objective Based on previous studies in the pediatric population, it remains unclear whether there is a difference in postoperative pain between two widely used tonsillectomy techniques: coblation and bovie electrocautery. This large prospective study investigates whether postoperative pain scores differ between these two surgical techniques for tonsillectomy. Methods Prospective, non‐randomized study of children aged 2–8 enrolled in a randomized controlled trial of single‐dose intravenous acetaminophen for pain associated with adenotonsillectomy. Included procedures occurred between October 2012 and June 2015 at a tertiary referral center. Only patients whose operations exclusively used coblation or electrocautery and who required postoperative admission for extended observation were included. Follow‐up period was the length of inpatient stay. Patients and nurses who recorded the pain scores were blinded to the tonsillectomy technique. Results A total of 183 patients were included: 117 coblation cases and 66 electrocautery cases. Pain scores in the surgical recovery unit and pain scores after admission to the floor unit were not significantly different between coblation and electrocautery, either before or after adjustment for patient age, body mass index, intravenous acetaminophen use, and surgeon. There was also no difference in length of stay, readmission rate, or post‐tonsillectomy hemorrhage. Conclusions Coblation and electrocautery tonsillectomy are associated with similar post‐operative pain scores in the recovery and inpatient units in the pediatric population. As coblation is costlier, the results of this study may affect which tool is used by otolaryngologists from a cost–benefit perspective. Level of evidence III

Long waits for diagnostic assessment prevent early identification of children suspected of having autism spectrum disorder. We evaluated the benefit of embedded diagnostic consultation within primary care clinics. Using a streamlined diagnostic model, 119 children with concerns for autism spectrum disorder were seen over 14 months. Diagnostic clarity was determined through streamlined assessment for 59% of the children, while others required follow-up. Latency from first concern to diagnosis was 55 days and median age at diagnosis was 32 months: considerably lower than national averages or comparable tertiary clinics. Findings support that embedded processes for effective triage and diagnosis within the medical home is a viable mechanism for efficient access to diagnostic services and assists in bypassing a common barrier to specialized services.