Aakre, Christopher A.

Primary care is a language-intensive environment where clinicians frequently synthesize, document, and communicate large amounts of information. Given the amount of information in primary care, large language models (LLMs) have emerged as tools to support primary care clinician workflows, aiming to reduce administrative burden and augment patient care. LLMs are artificial intelligence tools that can generate text and assist with chart summarization, message drafting, and clinical decision support. However, because LLMs are probabilistic, identical inputs may yield inconsistent responses. The LLMs output can be significantly improved through prompt engineering. Prompt engineering is the intentional design of text entered into the LLM to help produce better LLM results. Prompting is best approached as an iterative process where a clinician defines the task, the LLM generates output, and the clinician critically reviews and refines the results. Low-quality prompts are often vague and produce generic or misleading responses. High-quality prompts are specific, context-rich, and well-structured. Using structured prompting frameworks can improve the LLMs response. Including core elements, such as the LLMs role, the clinical context, the requested task, and the desired output formatting, can help improve the LLMs response by providing situational background and framing response formatting. Additional frameworks, such as Ask-Context-Expectation and PICO+O, can be utilized for specific patient circumstances and cases. However, despite their promise, LLMs have important limitations. They may generate inaccurate or hallucinated information and produce inconsistent results. Given these limitations, LLMs should not be viewed as knowledge authorities but rather as cognitive assistants. Thoughtful prompt design can help reduce these limitations and improve response accuracy, always keeping in mind that LLMs should not replace human judgment or clinical reasoning. As LLMs continue to be integrated into primary care, prompt engineering will become an increasingly important clinical skill. Careful prompt design and critical appraisal of LLM responses will be key to maintaining excellent patient care in the AI era.

Background: Artificial intelligence (AI) platforms can potentially enhance clinical decision-making (CDM) in primary care settings. OpenEvidence (OE), an AI tool, draws from trusted sources to generate evidence-based medicine (EBM) recommendations to address clinical questions. However, its effectiveness in real-world primary care cases remains unknown. Objective: To evaluate the performance of OE in providing EBM recommendations for five common chronic conditions in primary care: hypertension, hyperlipidemia, diabetes mellitus type 2, depression, and obesity. Methods: Five patient cases were retrospectively analyzed. Physicians posed specific clinical questions, and OE responses were evaluated on clarity, relevance, evidence support, impact on CDM, and overall satisfaction. Four independent physicians provided ratings using a 0 to 4 scale. Results: OE provided accurate, evidence-based recommendations in all cases, aligning with physician plans. OE was scored on a scale of zero to four, where zero was very unclear, and four was very clear. Mean scores across cases were clarity (3.55 ± 0.60), relevance (3.75 ± 0.44), support (3.35 ± 0.49), and satisfaction (3.60 ± 0.60). However, the impact on CDM was limited (1.95 ± 1.05), as OE primarily reinforced rather than modified plans. Conclusion: OE was rated high in clarity, relevance, and evidence-based support, reinforcing physician decisions in common chronic conditions. While the impact on CDM was minimal due to the study’s retrospective nature, OE shows promise in augmenting the primary care physician. Prospective trials are needed to evaluate its utility in complex cases and multidisciplinary settings.

Objective: This study evaluated the feasibility and satisfaction of using a wearable brain activity sensing device for stress reduction among patients experiencing Long COVID (LC). Patients and Methods: Patients with LC (N = 45) were invited to participate in an open-label pilot study. Participants were asked to use a brain-sensing electroencephalogram (S-EEG) wearable device (Muse-S™) daily for 90 days and followed for an additional 90 days (180 days total participation). Study enrollment began June 28, 2022, and ended July 28, 2023. Results: Patients were predominantly female (82%), white (96%), and with an average age of 47.6 (±12.5) years. At the end of treatment (90 days), cognition, sleep, self-efficacy, and quality-of-life quality of life improved ( P < .001 to P = .04). In addition, stress and anxiety were significantly reduced ( P < .001). These significant changes persisted by end-of-study (180 days). For stress ( P = .023) and anxiety ( P = .033), a change from the baseline during follow-up was significantly associated with the amount of S-EEG usage. The end-of-study satisfaction survey indicated that patients were very satisfied with the S-EEG device, and 97% planned to continue using it. Conclusion: Results suggest that the S-EEG device improved cognition in LC, which is one of the main severe symptoms we see in our LC clinic. In addition, it helped reduce stress and anxiety. These findings have important clinical implications for LC patients using S-EEG as an adjunct to concurrent LC clinical care which includes pharmacotherapies.

Background: In addition to the morbidity and mortality associated with acute infection, COVID-19 has been associated with persistent symptoms (>30 days), often referred to as Long COVID (LC). LC symptoms often cluster into phenotypes, resembling conditions such as fibromyalgia, postural orthostatic tachycardiac syndrome (POTS), and myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). LC clinics have been established to best address the needs of LC patients and continuity of care. We developed a cross-sectional survey to assess treatment response through our LC Clinic (LCC). Methods: A 25-question survey (1-10 Likert scale) was expert- and content-validated by LCC clinicians, patients, and patient advocates. The survey assessed LC symptoms and the helpfulness of different interventions, including medications and supplements. A total of 852 LCC patients were asked to complete the survey, with 536 (62.9%) responding. Results: The mean time from associated COVID-19 infection to survey completion was 23.2 ± 6.4 months. The mean age of responders was 52.3 ± 14.1 (63% females). Self-reported symptoms were all significantly improved ( P < .001) from the initial visit to the LCC (baseline) to the time of the follow-up survey. However, only 4.5% (24/536) of patients rated all symptoms low (1-2) at the time of the survey, indicating low levels of full recovery in our cohort. The patients rated numerous interventions as being helpful, including low-dose naltrexone (45/77; 58%), vagal nerve stimulation (18/34; 53%), and fisetin (28/44; 64%). Conclusions: Patients report general improvements in symptoms following the initial LCC visit, but complete recovery rates remain low at 23.2 ± 6.4 months.

Background: Over 30% of patients with COVID-19 have persistent symptoms that last beyond 30 days and referred to as Long COVID. Long COVID has been associated with a persistent elevation in peripheral cytokines including interleukin-6, interleukin-1β, and tumor necrosis factor-α. This study reports cytokine profiles of patients in our clinic across SARS-COV-2 variant epochs. Methods: The clinical cytokine panel was analyzed in patients with Long COVID during periods that were stratified according to variant epoch. The 4 variant epochs were defined as: (1) wild-type through alpha, (2) alpha/beta/gamma, (3) delta, and (4) omicron variants. Results: A total of 390 patients had the clinical cytokine panel performed; the median age was 48 years (IQR 38-59) and 62% were female. Distribution by variant was wild-type and alpha, 50% (n = 196); alpha/beta/gamma, 7.9% (n = 31); delta, 18% (n = 72); and omicron, 23% (n = 91). Time to cytokine panel testing was significantly longer for the earlier epochs. Tumor necrosis factor-α ( P < .001) and interleukin 1β ( P < .001) were significantly more elevated in the earlier epochs (median of 558 days in wild-type through Alpha epoch vs 263 days in omicron epoch, P < .001)). Nucleocapsid antibodies were consistently detected across epochs. Discussion: When stratified by variant epoch, patients with early epoch Long COVID had persistently elevated peripheral pro-inflammatory cytokine levels when compared to later epoch Long COVID. Patients with Long COVID have similar clusters of symptoms across epochs, suggesting that the underlying pathology is independent of the peripheral cytokine signature.

Abstract Introduction/Objective: To assess the clinical utility of the FM/a® chemokine-cytokine assay in diagnosing fibromyalgia at an academic medical center. Methods: We performed a cross-sectional study on 50 patients diagnosed with fibromyalgia at a specialty fibromyalgia clinic between January 1, 2021 through July 31, 2021. Patients completed questionnaires and provided a venous blood sample sent to EpicGenetics, to complete the FM/a® test. Demographic, symptom, and historical data was obtained from chart review. Statistical analysis was performed. Results: Of 50 patients with a clinical diagnosis of fibromyalgia, the FM/a® test was positive in 45 (90%). Performance of the FM/a® test compared to the 2016 ACR criteria yielded an odds ratio of 3.5 with sensitivity of 0.91, specificity of 0.25. Univariate regression demonstrated an area under the curve (AUC) of 0.7337, which improved to 0.89 when adjusted for age, gender, and race. When compared to the 1990 ACR criteria, a positive FM/a® test had an odds ratio of 2.33 with sensitivity 0.92 and specificity 0.17. Univariate regression analysis demonstrated an AUC of 0.571, when adjusted for age, gender and race, AUC was similar at 0.585. Conclusion: The FM/a® test performed well overall, though inferiorly compared to the 1990 and 2016 ACR diagnostic criteria. When adjusted for age, gender, and race the test performed almost equivalently to the 2016 ACR criteria. The FM/a® test may be useful in general clinical practices to differentiate patients who are more likely to have FM.

Respiratory disorders, being one of the leading causes of disability worldwide, account for constant evolution in management technologies, resulting in the incorporation of artificial intelligence (AI) in the recording and analysis of lung sounds to aid diagnosis in clinical pulmonology practice. Although lung sound auscultation is a common clinical practice, its use in diagnosis is limited due to its high variability and subjectivity. We review the origin of lung sounds, various auscultation and processing methods over the years and their clinical applications to understand the potential for a lung sound auscultation and analysis device. Respiratory sounds result from the intra-pulmonary collision of molecules contained in the air, leading to turbulent flow and subsequent sound production. These sounds have been recorded via an electronic stethoscope and analyzed using back-propagation neural networks, wavelet transform models, Gaussian mixture models and recently with machine learning and deep learning models with possible use in asthma, COVID-19, asbestosis and interstitial lung disease. The purpose of this review was to summarize lung sound physiology, recording technologies and diagnostics methods using AI for digital pulmonology practice. Future research and development in recording and analyzing respiratory sounds in real time could revolutionize clinical practice for both the patients and the healthcare personnel.

Abstract Bowel sounds have been previously used to study intestinal motility and overall digestive health in various clinical settings. However, the blurred definition of bowel sounds and their subtypes, limited resources for interpretation, poor sensitivity, and low positive predictive value led to their restricted utility. Recent advances in artificial intelligence and machine learning have steered interest in developing unique tools using the phonoenterogram to analyze diverse bowel sounds. In our study, bowel sounds were recorded from eight healthy volunteers using the Eko Duo stethoscope. A novel deep-learning algorithm was designed to classify the recordings into baseline or prominent bowel sounds. A total of 11,210 data points (5,605 balanced sounds) were used to train and test the model to yield an accuracy of 0.895, a precision of 0.890, and a recall of 0.854 reflecting successful segregation of these sounds into respective groups. More extensive studies enrolling healthy and diseased subjects with a device specifically tailored to record bowel sounds are needed to generalize these results and determine their application in the patient population.

Abstract Blood pressure measurement in current medical practice relies on manual methods with the most widely used modality being sphygmomanometers. Utilizing the principle of Photoplethysmography, it is possible to provide an accurate reading of one’s blood pressure through light signals and photodetector devices. This research paper introduces a new Artificial Intelligence driven approach to predict Blood pressure levels and classify them according to the updated ACC (American College of Cardiology) criteria as Normal, Elevated, Stage I, and II Hypertension from the given PPG signal values using Machine Learning Models. This research paper aims to accurately read the Systolic and Diastolic Blood Pressure using Artificial Intelligence, place them into the correct value bins and further prove that the blood pressure values differ based on different skin tones in different light wavelengths such as red, infrared, and green. Machine Learning models such as the Support Vector Machine have shown an accuracy of 70.58% for Systolic Blood Pressure and Decision Tree with an accuracy of 74.4% for Diastolic Blood Pressure classification. The models used in this research are Support Vector Machine, Decision Tree and K-Nearest Neighbor. This research study has future applications and extensions to predict blood pressure levels for patients with different skin tones under different light radiations and PPG signal readings. Neural Network models will be developed to compare the blood predictions from this work.