Proactive Polypharmacy Management Using Large Language Models: Opportunities to Enhance Geriatric Care.

Polypharmacy remains an important challenge for patients with extensive medical complexity. Given the primary care shortage and the increasing aging population, effective polypharmacy management is crucial to manage the increasing burden of care. The capacity of large language model (LLM)-based artificial intelligence to aid in polypharmacy management has yet to be evaluated. Here, we evaluate ChatGPT's performance in polypharmacy management via its deprescribing decisions in standardized clinical vignettes. We inputted several clinical vignettes originally from a study of general practicioners' deprescribing decisions into ChatGPT 3.5, a publicly available LLM, and evaluated its capacity for yes/no binary deprescribing decisions as well as list-based prompts in which the model was prompted to choose which of several medications to deprescribe. We recorded ChatGPT responses to yes/no binary deprescribing prompts and the number and types of medications deprescribed. In yes/no binary deprescribing decisions, ChatGPT universally recommended deprescribing medications regardless of ADL status in patients with no overlying CVD history; in patients with CVD history, ChatGPT's answers varied by technical replicate. Total number of medications deprescribed ranged from 2.67 to 3.67 (out of 7) and did not vary with CVD status, but increased linearly with severity of ADL impairment. Among medication types, ChatGPT preferentially deprescribed pain medications. ChatGPT's deprescribing decisions vary along the axes of ADL status, CVD history, and medication type, indicating some concordance of internal logic between general practitioners and the model. These results indicate that specifically trained LLMs may provide useful clinical support in polypharmacy management for primary care physicians.

Do peri-operative parathyroid hormone (PTH) analogues improve bone density and decrease mechanical complications in spinal deformity correction?-a minimum 2-year radiological study measuring Hounsfield units.

OBJECTIVE: To delineate whether use of a PTH analogue in the 1-year peri-operative period improves lumbar bone density. METHODS: A prospectively collected data registry of 254 patients who underwent CMIS correction of ASD (Cobb angle > 20 or SVA > 50 mm or (PI-LL) > 10) from Jan 2011 to Jan 2020 was analysed. Patients who were placed on PTH analogues for one year in conjunction with surgery were included in the study. Ultimately, 41 patients who had pre- and two-year post-operative CT scans for review were included in this study. Hounsfield units were measured off of the L1-L3 levels for all patients before and after surgery on pre-op and post-op CT scans. RESULT: The mean age of patients in this study was 70 (52-84, SD 7). Mean follow-up was 66 (24-132, SD 33) months. Twenty-three patients met criteria for severe deformity (Cobb angle > 50 degrees or SVA > 95 mm or PI/LL mismatch > 20 or PT > 30). Based off 2-year post-op CT scan, there were significant improvements in L1 Hounsfield units when comparing pre-op values (96; SD 55) to post-op values (185 SD 102); p. < 0.05. There was no screw loosening or screw pull out. There were 2 patients with PJF (4.8%). Both these patients had not completed their PTH treatment: one only took PTH for 3 months (PJF at 2-year post-op) and the other one took it only for 1 month (PJF at 1-year post-op). No increase in bone density was noted (based off of Hounsfield units) in five patients (12%) despite completion of their PTH therapy. Only one patient experienced nausea from PTH therapy. There were no other PTH related adverse events. CONCLUSION: The incidence of PTH analogues failing to increase bone density in our series was low at 12%. This study shows that PTH analogues may be a powerful adjunct for increasing bone density and may help to mitigate the risk of mechanical complications in patients undergoing deformity correction with minimally invasive techniques. Future comparative studies are warranted to confirm these latter findings and to potentially protocolize the ideal peri-operative bone health optimization strategy.

Does the Global Alignment and Proportion score predict mechanical complications in circumferential minimally invasive surgery for adult spinal deformity?

OBJECTIVE: The Global Alignment and Proportion (GAP) score was developed to serve as a tool to predict mechanical complication probability in patients undergoing surgery for adult spinal deformity (ASD), serving as an aid for setting surgical goals to decrease the prevalence of mechanical complications in ASD surgery. However, it was developed using ASD patients for whom open surgical techniques were used for correction. Therefore, the purpose of this study was to assess the applicability of the score for patients undergoing circumferential minimally invasive surgery (cMIS) for correction of ASD. METHODS: Study participants were patients undergoing cMIS ASD surgery without the use of osteotomies with a minimum of four levels fused and 2 years of follow-up. Postoperative GAP scores were calculated for all patients, and the association with mechanical failure was analyzed. RESULTS: The authors identified 182 patients who underwent cMIS correction of ASD. Mechanical complications were found in 11.1% of patients with proportioned spinopelvic states, 20.5% of patients with moderately disproportioned spinopelvic states, and 18.8% of patients with severely disproportioned spinopelvic states. Analysis with a chi-square test showed a significant difference between the cMIS and original GAP study cohorts in the moderately disproportioned and severely disproportioned spinopelvic states, but not in the proportioned spinopelvic states. CONCLUSIONS: For patients stratified into proportioned, moderately disproportioned, and severely disproportioned spinopelvic states, the GAP score predicted 6%, 47%, and 95% mechanical complication rates, respectively. The mechanical complication rate in patients undergoing cMIS ASD correction did not correlate with the calculated GAP spinopelvic state.

Assessing the Utility of ChatGPT Throughout the Entire Clinical Workflow: Development and Usability Study.

BACKGROUND: Large language model (LLM)-based artificial intelligence chatbots direct the power of large training data sets toward successive, related tasks as opposed to single-ask tasks, for which artificial intelligence already achieves impressive performance. The capacity of LLMs to assist in the full scope of iterative clinical reasoning via successive prompting, in effect acting as artificial physicians, has not yet been evaluated. OBJECTIVE: This study aimed to evaluate ChatGPT's capacity for ongoing clinical decision support via its performance on standardized clinical vignettes. METHODS: We inputted all 36 published clinical vignettes from the Merck Sharpe & Dohme (MSD) Clinical Manual into ChatGPT and compared its accuracy on differential diagnoses, diagnostic testing, final diagnosis, and management based on patient age, gender, and case acuity. Accuracy was measured by the proportion of correct responses to the questions posed within the clinical vignettes tested, as calculated by human scorers. We further conducted linear regression to assess the contributing factors toward ChatGPT's performance on clinical tasks. RESULTS: ChatGPT achieved an overall accuracy of 71.7% (95% CI 69.3%-74.1%) across all 36 clinical vignettes. The LLM demonstrated the highest performance in making a final diagnosis with an accuracy of 76.9% (95% CI 67.8%-86.1%) and the lowest performance in generating an initial differential diagnosis with an accuracy of 60.3% (95% CI 54.2%-66.6%). Compared to answering questions about general medical knowledge, ChatGPT demonstrated inferior performance on differential diagnosis (ß=-15.8%; P<.001) and clinical management (ß=-7.4%; P=.02) question types. CONCLUSIONS: ChatGPT achieves impressive accuracy in clinical decision-making, with increasing strength as it gains more clinical information at its disposal. In particular, ChatGPT demonstrates the greatest accuracy in tasks of final diagnosis as compared to initial diagnosis. Limitations include possible model hallucinations and the unclear composition of ChatGPT's training data set.

Evaluating ChatGPT as an Adjunct for Radiologic Decision-Making.

BACKGROUND: ChatGPT, a popular new large language model (LLM) built by OpenAI, has shown impressive performance in a number of specialized applications. Despite the rising popularity and performance of AI, studies evaluating the use of LLMs for clinical decision support are lacking. PURPOSE: To evaluate ChatGPT's capacity for clinical decision support in radiology via the identification of appropriate imaging services for two important clinical presentations: breast cancer screening and breast pain. MATERIALS AND METHODS: We compared ChatGPT's responses to the American College of Radiology (ACR) Appropriateness Criteria for breast pain and breast cancer screening. Our prompt formats included an open-ended (OE) format, where ChatGPT was asked to provide the single most appropriate imaging procedure, and a select all that apply (SATA) format, where ChatGPT was given a list of imaging modalities to assess. Scoring criteria evaluated whether proposed imaging modalities were in accordance with ACR guidelines. RESULTS: ChatGPT achieved an average OE score of 1.83 (out of 2) and a SATA average percentage correct of 88.9% for breast cancer screening prompts, and an average OE score of 1.125 (out of 2) and a SATA average percentage correct of 58.3% for breast pain prompts. CONCLUSION: Our results demonstrate the feasibility of using ChatGPT for radiologic decision making, with the potential to improve clinical workflow and responsible use of radiology services.

Evaluating GPT as an Adjunct for Radiologic Decision Making: GPT-4 Versus GPT-3.5 in a Breast Imaging Pilot.

OBJECTIVE: Despite rising popularity and performance, studies evaluating the use of large language models for clinical decision support are lacking. Here, we evaluate ChatGPT (Generative Pre-trained Transformer)-3.5 and GPT-4's (OpenAI, San Francisco, California) capacity for clinical decision support in radiology via the identification of appropriate imaging services for two important clinical presentations: breast cancer screening and breast pain. METHODS: We compared ChatGPT's responses to the ACR Appropriateness Criteria for breast pain and breast cancer screening. Our prompt formats included an open-ended (OE) and a select all that apply (SATA) format. Scoring criteria evaluated whether proposed imaging modalities were in accordance with ACR guidelines. Three replicate entries were conducted for each prompt, and the average of these was used to determine final scores. RESULTS: Both ChatGPT-3.5 and ChatGPT-4 achieved an average OE score of 1.830 (out of 2) for breast cancer screening prompts. ChatGPT-3.5 achieved a SATA average percentage correct of 88.9%, compared with ChatGPT-4's average percentage correct of 98.4% for breast cancer screening prompts. For breast pain, ChatGPT-3.5 achieved an average OE score of 1.125 (out of 2) and a SATA average percentage correct of 58.3%, as compared with an average OE score of 1.666 (out of 2) and a SATA average percentage correct of 77.7%. DISCUSSION: Our results demonstrate the eventual feasibility of using large language models like ChatGPT for radiologic decision making, with the potential to improve clinical workflow and responsible use of radiology services. More use cases and greater accuracy are necessary to evaluate and implement such tools.

Empathy and Equity: Key Considerations for Large Language Model Adoption in Health Care.

The growing presence of large language models (LLMs) in health care applications holds significant promise for innovative advancements in patient care. However, concerns about ethical implications and potential biases have been raised by various stakeholders. Here, we evaluate the ethics of LLMs in medicine along 2 key axes: empathy and equity. We outline the importance of these factors in novel models of care and develop frameworks for addressing these alongside LLM deployment.

Assessing the Utility of ChatGPT Throughout the Entire Clinical Workflow.

IMPORTANCE: Large language model (LLM) artificial intelligence (AI) chatbots direct the power of large training datasets towards successive, related tasks, as opposed to single-ask tasks, for which AI already achieves impressive performance. The capacity of LLMs to assist in the full scope of iterative clinical reasoning via successive prompting, in effect acting as virtual physicians, has not yet been evaluated. OBJECTIVE: To evaluate ChatGPT's capacity for ongoing clinical decision support via its performance on standardized clinical vignettes. DESIGN: We inputted all 36 published clinical vignettes from the Merck Sharpe & Dohme (MSD) Clinical Manual into ChatGPT and compared accuracy on differential diagnoses, diagnostic testing, final diagnosis, and management based on patient age, gender, and case acuity. SETTING: ChatGPT, a publicly available LLM. PARTICIPANTS: Clinical vignettes featured hypothetical patients with a variety of age and gender identities, and a range of Emergency Severity Indices (ESIs) based on initial clinical presentation. EXPOSURES: MSD Clinical Manual vignettes. MAIN OUTCOMES AND MEASURES: We measured the proportion of correct responses to the questions posed within the clinical vignettes tested. RESULTS: ChatGPT achieved 71.7% (95% CI, 69.3% to 74.1%) accuracy overall across all 36 clinical vignettes. The LLM demonstrated the highest performance in making a final diagnosis with an accuracy of 76.9% (95% CI, 67.8% to 86.1%), and the lowest performance in generating an initial differential diagnosis with an accuracy of 60.3% (95% CI, 54.2% to 66.6%). Compared to answering questions about general medical knowledge, ChatGPT demonstrated inferior performance on differential diagnosis (ß=-15.8%, p<0.001) and clinical management (ß=-7.4%, p=0.02) type questions. CONCLUSIONS AND RELEVANCE: ChatGPT achieves impressive accuracy in clinical decision making, with particular strengths emerging as it has more clinical information at its disposal.

Natural history of Ebola virus disease in rhesus monkeys shows viral variant emergence dynamics and tissue-specific host responses.

Ebola virus (EBOV) causes Ebola virus disease (EVD), marked by severe hemorrhagic fever; however, the mechanisms underlying the disease remain unclear. To assess the molecular basis of EVD across time, we performed RNA sequencing on 17 tissues from a natural history study of 21 rhesus monkeys, developing new methods to characterize host-pathogen dynamics. We identified alterations in host gene expression with previously unknown tissue-specific changes, including downregulation of genes related to tissue connectivity. EBOV was widely disseminated throughout the body; using a new, broadly applicable deconvolution method, we found that viral load correlated with increased monocyte presence. Patterns of viral variation between tissues differentiated primary infections from compartmentalized infections, and several variants impacted viral fitness in a EBOV/Kikwit minigenome system, suggesting that functionally significant variants can emerge during early infection. This comprehensive portrait of host-pathogen dynamics in EVD illuminates new features of pathogenesis and establishes resources to study other emerging pathogens.

Adaptive substitutions underlying cardiac glycoside insensitivity in insects exhibit epistasis in vivo.

Predicting how species will respond to selection pressures requires understanding the factors that constrain their evolution. We use genome engineering of Drosophila to investigate constraints on the repeated evolution of unrelated herbivorous insects to toxic cardiac glycosides, which primarily occurs via a small subset of possible functionally-relevant substitutions to Na+,K+-ATPase. Surprisingly, we find that frequently observed adaptive substitutions at two sites, 111 and 122, are lethal when homozygous and adult heterozygotes exhibit dominant neural dysfunction. We identify a phylogenetically correlated substitution, A119S, that partially ameliorates the deleterious effects of substitutions at 111 and 122. Despite contributing little to cardiac glycoside-insensitivity in vitro, A119S, like substitutions at 111 and 122, substantially increases adult survivorship upon cardiac glycoside exposure. Our results demonstrate the importance of epistasis in constraining adaptive paths. Moreover, by revealing distinct effects of substitutions in vitro and in vivo, our results underscore the importance of evaluating the fitness of adaptive substitutions and their interactions in whole organisms.