Oropharyngeal Cancer Staging Health Record Extraction Using Artificial Intelligence.

Importance: Accurate, timely, and cost-effective methods for staging oropharyngeal cancers are crucial for patient prognosis and treatment decisions, but staging documentation is often inaccurate or incomplete. With the emergence of artificial intelligence in medicine, data abstraction may be associated with reduced costs but increased efficiency and accuracy of cancer staging. Objective: To evaluate an algorithm using an artificial intelligence engine capable of extracting essential information from medical records of patients with oropharyngeal cancer and assigning tumor, nodal, and metastatic stages according to American Joint Committee on Cancer eighth edition guidelines. Design, Setting, and Participants: This retrospective diagnostic study was conducted among a convenience sample of 806 patients with oropharyngeal squamous cell carcinoma. Medical records of patients with staged oropharyngeal squamous cell carcinomas who presented to a single tertiary care center between January 1, 2010, and August 1, 2020, were reviewed. A ground truth cancer stage dataset and comprehensive staging rule book consisting of 135 rules encompassing p16 status, tumor, and nodal and metastatic stage were developed. Subsequently, 4 distinct models were trained: model T (entity relationship extraction) for anatomical location and invasion state, model S (numerical extraction) for lesion size, model M (sequential classification) for metastasis detection, and a p16 model for p16 status. For validation, results were compared against ground truth established by expert reviewers, and accuracy was reported. Data were analyzed from March to November 2023. Main Outcomes and Measures: The accuracy of algorithm cancer stages was compared with ground truth. Results: Among 806 patients with oropharyngeal cancer (mean [SD] age, 63.6 [10.6] years; 651 males [80.8%]), 421 patients (52.2%) were positive for human papillomavirus. The artificial intelligence engine achieved accuracies of 55.9% (95% CI, 52.5%-59.3%) for tumor, 56.0% (95% CI, 52.5%-59.4%) for nodal, and 87.6% (95% CI, 85.1%-89.7%) for metastatic stages and 92.1% (95% CI, 88.5%-94.6%) for p16 status. Differentiation between localized (stages 1-2) and advanced (stages 3-4) cancers achieved 80.7% (95% CI, 77.8%-83.2%) accuracy. Conclusion and Relevance: This study found that tumor and nodal staging accuracies were fair to good and excellent for metastatic stage and p16 status, with clinical relevance in assigning optimal treatment and reducing toxic effect exposures. Further model refinement and external validation with electronic health records at different institutions are necessary to improve algorithm accuracy and clinical applicability.

Design and Development of an Antigen Test for SARS-CoV-2 Nucleocapsid Protein to Validate the Viral Quality Assurance Panels.

The continuing mutability of the SARS-CoV-2 virus can result in failures of diagnostic assays. To address this, we describe a generalizable bioinformatics-to-biology pipeline developed for the calibration and quality assurance of inactivated SARS-CoV-2 variant panels provided to Radical Acceleration of Diagnostics programs (RADx)-radical program awardees. A heuristic genetic analysis based on variant-defining mutations demonstrated the lowest genetic variance in the Nucleocapsid protein (Np)-C-terminal domain (CTD) across all SARS-CoV-2 variants. We then employed the Shannon entropy method on (Np) sequences collected from the major variants, verifying the CTD with lower entropy (less prone to mutations) than other Np regions. Polyclonal and monoclonal antibodies were raised against this target CTD antigen and used to develop an Enzyme-linked immunoassay (ELISA) test for SARS-CoV-2. Blinded Viral Quality Assurance (VQA) panels comprised of UV-inactivated SARS-CoV-2 variants (XBB.1.5, BF.7, BA.1, B.1.617.2, and WA1) and distractor respiratory viruses (CoV 229E, CoV OC43, RSV A2, RSV B, IAV H1N1, and IBV) were assembled by the RADx-rad Diagnostics core and tested using the ELISA described here. The assay tested positive for all variants with high sensitivity (limit of detection: 1.72-8.78 ng/mL) and negative for the distractor virus panel. Epitope mapping for the monoclonal antibodies identified a 20 amino acid antigenic peptide on the Np-CTD that an in-silico program also predicted for the highest antigenicity. This work provides a template for a bioinformatics pipeline to select genetic regions with a low propensity for mutation (low Shannon entropy) to develop robust 'pan-variant' antigen-based assays for viruses prone to high mutational rates.

The Role of the Host Cytoskeleton in the Formation and Dynamics of Rotavirus Viroplasms.

Rotavirus (RV) replicates within viroplasms, membraneless electron-dense globular cytosolic inclusions with liquid-liquid phase properties. In these structures occur the virus transcription, replication, and packaging of the virus genome in newly assembled double-layered particles. The viroplasms are composed of virus proteins (NSP2, NSP5, NSP4, VP1, VP2, VP3, and VP6), single- and double-stranded virus RNAs, and host components such as microtubules, perilipin-1, and chaperonins. The formation, coalescence, maintenance, and perinuclear localization of viroplasms rely on their association with the cytoskeleton. A stabilized microtubule network involving microtubules and kinesin Eg5 and dynein molecular motors is associated with NSP5, NSP2, and VP2, facilitating dynamic processes such as viroplasm coalescence and perinuclear localization. Key post-translation modifications, particularly phosphorylation events of RV proteins NSP5 and NSP2, play pivotal roles in orchestrating these interactions. Actin filaments also contribute, triggering the formation of the viroplasms through the association of soluble cytosolic VP4 with actin and the molecular motor myosin. This review explores the evolving understanding of RV replication, emphasizing the host requirements essential for viroplasm formation and highlighting their dynamic interplay within the host cell.

Hybrid Closed Loop in Adults With Type 1 Diabetes and Severely Impaired Hypoglycemia Awareness.

BACKGROUND: Benefits of hybrid closed-loop (HCL) systems in a high-risk group with type 1 diabetes and impaired awareness of hypoglycemia (IAH) have not been well-explored. METHODS: Adults with Edmonton HYPO scores ≥1047 were randomized to 26-weeks HCL (MiniMed™ 670G) vs standard therapy (multiple daily injections or insulin pump) without continuous glucose monitoring (CGM) (control). Primary outcome was percentage CGM time-in-range (TIR; 70-180 mg/dL) at 23 to 26 weeks post-randomization. Major secondary endpoints included magnitude of change in counter-regulatory hormones and autonomic symptom responses to hypoglycemia at 26-weeks post-randomization. A post hoc analysis evaluated glycemia risk index (GRI) comparing HCL with control groups at 26 weeks post-randomization. RESULTS: Nine participants (median [interquartile range (IQR)] age 51 [41, 59] years; 44% male; enrolment HYPO score 1183 [1058, 1308]; Clarke score 6 [6, 6]; n = 5 [HCL]; n = 4 [control]) completed the study. Time-in-range was higher using HCL vs control (70% [68, 74%] vs 48% [44, 50%], P = .014). Time <70 mg/dL did not differ (HCL 3.8% [2.7, 3.9] vs control 6.5% [4.3, 8.6], P = .14) although hypoglycemia episode duration was shorter (30 vs 50 minutes, P < .001) with HCL. Glycemia risk index was lower with HCL vs control (38.1 [30.0, 39.2] vs 70.8 [58.5, 72.4], P = .014). Following 6 months of HCL use, greater dopamine (24.0 [12.3, 27.6] vs -18.5 [-36.5, -4.8], P = .014), and growth hormone (6.3 [4.6, 16.8] vs 0.5 [-0.8, 3.0], P = .050) responses to hypoglycemia were observed. CONCLUSIONS: Six months of HCL use in high-risk adults with severe IAH increased glucose TIR and improved GRI without increased hypoglycemia, and partially restored counter-regulatory responses. CLINICAL TRIAL REGISTRATION: ACTRN12617000520336.

Driving-Related Glucose Patterns Among Older Adults with Type 1 Diabetes.

Older adults with type 1 diabetes may face challenges driving safely. Glucose "above-5-to-drive" is often recommended for insulin-treated diabetes to minimize hypoglycemia while driving. However, the effectiveness of this recommendation among older adults has not been evaluated. Older drivers with type 1 diabetes were assessed while using sensor-augmented insulin pumps during a 2-week clinical trial run-in. Twenty-three drivers (median age 69 years [interquartile range; IQR 65-72]; diabetes duration 37 years [20-45]) undertook 618 trips (duration 10 min [5-21]). Most trips (n = 535; 87%) were <30 min duration; 9 trips (1.5%) exceeded 90 min and 3 trips (0.5%) exceeded 120 min. Pre-trip continuous glucose monitoring (CGM) was >5.0 mmol/L for 577 trips (93%) and none of these had CGM <3.9 mmol/L during driving (including 8 trips >90 min and 3 trips >120 min). During 41 trips with pre-trip CGM ≤5.0 mmol/L, 11 trips had CGM <3.9 mmol/L. Seventy-one CGM alerts occurred during 60 trips (10%), of which 54 of 71 alerts (76%) were unrelated to hypoglycemia. Our findings support a glucose "above-5-to-drive" recommendation to avoid CGM-detected hypoglycemia among older drivers, including for prolonged drives, and highlight the importance of active CGM low-glucose alerts to prevent hypoglycemia during driving. Driving-related CGM usability and alert functionality warrant investigation. Clinical trial ACTRN1261900515190.

Outcomes associated with a variable rate insulin infusion diabetic ketoacidosis protocol.

AIMS: To relate adverse events with glucose correction rates in diabetic ketoacidosis (DKA) using variable rate intravenous insulin-infusions (VRIII). METHODS: Retrospective, observational study in adults with DKA who received insulin infusions between 2012 and 2017 at St Vincent's Hospital, Melbourne. Early correction of hyperglycaemia (<10 mmol/L) was evaluated for association with hypoglycaemia (<4.0 mmol/L), hypokalaemia (potassium <3.3 mmol/L) and clinical outcomes via regression analysis. RESULTS: The study involved 97 patients, with 93 % having type 1 diabetes. The mean age was 38 years, 47 % were women and 35 % were admitted to intensive care. Hypoglycaemia rates during 12 and 24 h of treatment were 6.2 % and 8.2 %, respectively with 58 % of patients recording their first BGL <10 mmol/L within 12 h and 88 % within 24 h. Ketone clearance time averaged at 15.6 h. Hyperglycaemia correction rates to <10 mmol/L were not different in those with/without hypoglycaemia at 12/24 h, in multivariate analysis including admission BGL. Hypokalaemia occurred in 40.2 % of patients and was associated with lower pH but not BGL correction rates. CONCLUSION: The VRIII protocol achieved early hyperglycaemia correction and ketoacidosis reversal with low hypoglycaemia risk. However, high hypokalaemia rates suggest the need for aggressive potassium replacement, especially in markedly acidotic patients.

Hybrid Closed-Loop Versus Manual Insulin Delivery in Adults With Type 1 Diabetes: A Post Hoc Analysis Using the Glycemia Risk Index.

BACKGROUND: Glycemia risk index (GRI) is a novel composite metric assessing overall glycemic risk, accounting for both hypoglycemia and hyperglycemia and weighted toward extremes. Data assessing GRI as an outcome measure in closed-loop studies and its relation with conventional key continuous glucose monitoring (CGM) metrics are limited. METHODS: A post hoc analysis was performed to evaluate the sensitivity of GRI in assessing glycemic quality in adults with type 1 diabetes randomized to 26 weeks hybrid closed-loop (HCL) or manual insulin delivery (control). The primary outcome was GRI comparing HCL with control. Comparisons were made with changes in other CGM metrics including time in range (TIR), time above range (TAR), time below range (TBR), and glycemic variability (standard deviation [SD] and coefficient of variation [CV]). RESULTS: GRI with HCL (N = 61) compared with control (N = 59) was significantly lower (mean [SD] 33.5 [11.7] vs 56.1 [14.4], respectively; mean difference -22.8 [-27.2, -18.3], P = .001). The mean increase in TIR was +14.8 (11.0, 18.5)%. GRI negatively correlated with TIR for combined arms (r = -.954; P = .001), and positively with TAR >250 mg/dL (r = .901; P = .001), TBR < 54 mg/dL (r = .416; P = .001), and glycemic variability (SD [r = .916] and CV [r = .732]; P = .001 for both). CONCLUSIONS: Twenty-six weeks of HCL improved GRI, in addition to other CGM metrics, compared with standard insulin therapy. The improvement in GRI was proportionally greater than the change in TIR, and GRI correlated with all CGM metrics. We suggest that GRI may be an appropriate primary outcome for closed-loop trials.

Improved Satisfaction While Maintaining Safety and High Time in Range (TIR) With a Medtronic Investigational Enhanced Advanced Hybrid Closed-Loop (e-AHCL) System.

OBJECTIVE: To determine feasibility and compare acceptance of an investigational Medtronic enhanced advanced hybrid closed-loop (e-AHCL) system in adults with type 1 diabetes with earlier iterations. RESEARCH DESIGN AND METHODS: This nonrandomized three-stage (12 weeks each) exploratory study compared e-AHCL (Bluetooth-enabled MiniMed 780G insulin pump with automatic data upload [780G] incorporating an updated algorithm; calibration-free all-in-one disposable sensor; 7-day infusion set) preceded by a run-in (non-Bluetooth 780G [670G V4.0 insulin pump] requiring manual data upload; Guardian Sensor 3 [GS3] requiring calibration; 3-day infusion set), stage 1 (780G; GS3; 3-day infusion set), and stage 2 (780G; calibration-free Guardian Sensor 4; 3-day infusion set). Treatment satisfaction was assessed by Diabetes Technology Questionnaire (DTQ)-current (primary outcome) and other validated treatment satisfaction tools with glucose outcomes by continuous glucose monitoring metrics. RESULTS: Twenty-one of 22 (11 women) participants (baseline HbA1c 6.7%/50 mmol/mol) completed the study. DTQ-current scores favored e-AHCL (123.1 [17.8]) versus run-in (101.6 [24.2]) and versus stage 1 (110.6 [20.8]) (both P < 0.001) but did not differ from stage 2 (119.4 [16.0]; P = 0.271). Diabetes Medication System Rating Questionnaire short-form scores for "Convenience and Efficacy" favored e-AHCL over run-in and all stages. Percent time in range 70-180 mg/dL was greater with e-AHCL versus run-in and stage 2 (+2.9% and +3.6%, respectively; both P < 0.001). Percent times of <70 mg/dL for e-AHCL were significantly lower than run-in, stage 1, and stage 2 (-0.9%, -0.6%, and -0.5%, respectively; all P < 0.01). CONCLUSIONS: e-AHCL was feasible. User satisfaction increased compared with earlier Medtronic HCL iterations without compromising glucose control.

Sensor-Integrated Chairs for Lower Body Strength and Endurance Assessment.

This paper describes an automated method and device to conduct the Chair Stand Tests of the Fullerton Functional Test Battery. The Fullerton Functional Test is a suite of physical tests designed to assess the physical fitness of older adults. The Chair Stand Tests, which include the Five Times Sit-to-Stand Test (5xSST) and the 30 Second Sit-to-Stand Test (30CST), are the standard for measuring lower-body strength in older adults. However, these tests are performed manually, which can be labor-intensive and prone to error. We developed a sensor-integrated chair that automatically captures the dynamic weight and distribution on the chair. The collected time series weight-sensor data is automatically uploaded for immediate determination of the sit-to-stand timing and counts, as well as providing a record for future comparison of lower-body strength progression. The automatic test administration can provide significant labor savings for medical personnel and deliver much more accurate data. Data from 10 patients showed good agreement between the manually collected and sensor-collected 30CST data (M = 0.5, SD = 1.58, 95% CI = 1.13). Additional data processing will be able to yield measurements of fatigue and balance and evaluate the mechanisms of failed standing attempts.

Mortality in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE).

OBJECTIVE: We report mortality outcomes in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) among people with type 2 diabetes diagnosed within 10 years and no recent history of cardiovascular events or cancer. RESEARCH DESIGN AND METHODS: Overall mortality rates and major causes of death were assessed over an average of 5 years of follow-up. Cause of death was adjudicated centrally by a committee masked to treatment assignment. We examined baseline covariates and the 10-year Framingham Risk Score for associations. RESULTS: Mortality rate was low (0.59 per 100 participant-years). Participants who died during follow-up were likely to be older, be male, have a history of hypertension, have a history of smoking, and have moderate albuminuria. The two most common underlying causes of death were "cardiovascular-cause" (a composite of underlying causes) (38.6%) and cancer (26.8%). There were no differences by treatment group. CONCLUSIONS: Among people with diabetes of relatively short duration, cause of death was varied. Attention to health risks beyond cardiovascular diseases is warranted.

The diabetes management experiences questionnaire: Psychometric validation among adults with type 1 diabetes.

AIMS: To examine the psychometric properties of the Diabetes Management Experiences Questionnaire (DME-Q). Adapted from the validated Glucose Monitoring Experiences Questionnaire, the DME-Q captures satisfaction with diabetes management irrespective of treatment modalities. METHODS: The DME-Q was completed by adults with type 1 diabetes as part of a randomized controlled trial comparing hybrid closed loop (HCL) to standard therapy. Most psychometric properties were examined with pre-randomization data (n = 149); responsiveness was examined using baseline and 26-week follow-up data (n = 120). RESULTS: Pre-randomization, participants' mean age was 44 ± 12 years, 52% were women. HbA1c was 61 ± 11 mmol/mol (7.8 ± 1.0%), diabetes duration was 24 ± 12 years and 47% used an insulin pump prior to the trial. A forced three-factor analysis revealed three expected domains, that is, 'Convenience', 'Effectiveness' and 'Intrusiveness', and a forced one-factor solution was also satisfactory. Internal consistency reliability was strong for the three subscales ( α range = 0.74-0.84) and 'Total satisfaction' ( α = 0.85). Convergent validity was demonstrated with moderate correlations between DME-Q 'Total satisfaction' and diabetes distress (PAID: rs = -0.57) and treatment satisfaction (DTSQ; rs = 0.58). Divergent validity was demonstrated with a weak correlation with prospective/retrospective memory (PRMQ: rs = -0.16 and - 0.13 respectively). Responsiveness was demonstrated, as participants randomized to HCL had higher 'Effectiveness' and 'Total satisfaction' scores than those randomized to standard therapy. CONCLUSIONS: The 22-item DME-Q is a brief, acceptable, reliable measure with satisfactory structural and construct validity, which is responsive to intervention. The DME-Q is likely to be useful for evaluation of new pharmaceutical agents and technologies in research and clinical settings.

Home sleep apnea tests: Conflicts of interest and funding.

OBJECTIVE: Pediatric otolaryngologists rely on HSAT literature to guide their diagnostic methods related to obstructive sleep apnea (OSA). Our objectives were to review the rates of presence of funding and/or potential conflict of interest (COI), as well as its relationship to the overall quality of HSAT publications in the literature over the last two decades. DATA SOURCES: Medline, Web of Science and Embase databases. REVIEW METHODS: A review was performed reviewing publications from January 2000 to December 2021. Oxford Level of Evidence (OLE) was used as a quality metric. COI and funding were recorded verbatim as self-declared in the text of the manuscript. RESULTS: Literature search yielded 4257 articles with 400 articles included in final analysis. The odds of higher quality studies (LOE 1 or 2) were higher in the last five years from 2016 to 2021 (OR, 3.6; 95% CI 1.4 to 6.9). Nearly half of all articles (43.0%) lacked a statement regarding funding or COI. There was a positive correlation between level of evidence and industry funding. The largest source of funding was from industry, comprising 39.6% of all studies that had a funding statement. Of these industry-funded studies, 37.5% reported no COI or lacked a COI statement. CONCLUSION: Despite a growing interest in HSATs for OSA evaluation, there is heterogeneity in reporting of COI and high prevalence of industry funding and COI. Re-evaluation and consensus amongst journals on guidelines for reporting disclosures are needed.

What difference does sleep make? Continuous glucose monitoring metrics during fixed-overnight time versus sleep periods among older adults with type 1 diabetes.

Hypoglycaemia during sleep is a common and clinically important issue for people living with insulin-treated diabetes. Continuous glucose monitoring devices can help to identify nocturnal hypoglycaemia and inform treatment strategies. However, sleep is generally inferred, with diabetes researchers and physicians using a fixed-overnight period as a proxy for sleep-wake status when analysing and interpretating continuous glucose monitoring data. No study to date has validated such an approach with established sleep measures. Continuous glucose monitoring and research-grade actigraphy devices were worn and sleep diaries completed for 2 weeks by 28 older adults (mean age 67 years [SD 5]; 17 (59%) women) with type 1 diabetes. Using continuous glucose monitoring data from a total of 356 nights, fixed-overnight (using the recommended period of 00:00 hours-06:00 hours) and objectively-measured sleep periods were compared. The fixed-overnight period approach missed a median 57 min per night (interquartile range: 49-64) of sleep for each participant, including five continuous glucose monitoring-detected hypoglycaemia episodes during objectively-measured sleep. Twenty-seven participants (96%) had at least 1 night with continuous glucose monitoring time-in-range and time-above-range discrepancies both ≥ 10 percentage points, a clinically significant discrepancy. The utility of fixed-overnight time continuous glucose monitoring as a proxy for sleep-awake continuous glucose monitoring is inadequate as it consistently excludes actual sleep time, obscures glycaemic patterns, and misses sensor hypoglycaemia episodes during sleep. The use of validated measures of sleep to aid interpretation of continuous glucose monitoring data is encouraged.

3D Airway Epithelial-Fibroblast Biomimetic Microfluidic Platform to Unravel Engineered Nanoparticle-Induced Acute Stress Responses as Exposome Determinants.

Insights into how biological systems respond to high- and low-dose acute environmental stressors are a fundamental aspect of exposome research. However, studying the impact of low-level environmental exposure in conventional in vitro settings is challenging. This study employed a three-dimensional (3D) biomimetic microfluidic lung-on-chip (µLOC) platform and RNA-sequencing to examine the effects of two model anthropogenic engineered nanoparticles (NPs): zinc oxide nanoparticles (Nano-ZnO) and copier center nanoparticles (Nano-CCP). The airway epithelium exposed to these NPs exhibited dose-dependent increases in cytotoxicity and barrier dysregulation (dominance of the external exposome). Interestingly, even nontoxic and low-level exposure (10 µg/mL) of the epithelium compartment to Nano-ZnO triggered chemotaxis of lung fibroblasts toward the epithelium. An increase in α smooth muscle actin (α-SMA) expression and contractile activity was also observed in these cells, indicating a bystander-like adaptive response (dominance of internal exposome). Further bioinformatics and network analysis showed that a low-dose Nano-ZnO significantly induced a robust transcriptomic response and upregulated several hub genes associated with the development of lung fibrosis. We propose that Nano-ZnO, even at a no observable effect level (NOEL) dose according to conventional standards, can function as a potent nanostressor to disrupt airway epithelium homeostasis. This leads to a cascade of profibrotic events in a cross-tissue compartment fashion. Our findings offer new insights into the early acute events of respiratory harm associated with environmental NPs exposure, paving the way for better exposomic understanding of this emerging class of anthropogenic nanopollutants.

Impact of Telehealth on Failure to Attend Rates and Patient Re-Engagement in Adult Congenital Heart Disease Clinic.

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic paved the way for telehealth consultations. We aimed to determine the impact of telehealth on rates of failure to attend (FTA) in adult congenital heart disease (ACHD) clinics and whether telehealth could re-engage patients with previous FTA face-to-face appointments. METHODS: This was a retrospective audit of a tertiary ACHD clinic over a 12-month pre-telehealth (26 March 2019-17 March 2020) and 12-month post-telehealth implementation period (24 March 2020-16 March 2021). Patients with one or more FTAs during the 24-month study period were included. Our ACHD clinic is run three times per month. Patients with ACHD are offered lifelong follow-up and reviewed annually on average. Re-engagement was defined as two or more consecutive face-to-face FTAs immediately before the telehealth period with subsequent attendance of their telehealth appointment. RESULTS: A total of 359 patients with a total of 623 FTAs were included. Complexity of congenital heart disease was moderate in 56% (202/359) and severe in 19% (69/359) of patients. Overall FTA rate was 18% (623/3,452). FTA rate was significantly lower in the post-telehealth period (15%, 257/1,664) compared with the pre-telehealth period (20%, 366/1,788) (p<0.00001). At study conclusion, 1% of patients had died (5/359). Of the 354 remaining patients, 42% (150/354) were considered lost to follow-up (two or more FTAs including telehealth), 37% (132/354) missed only one clinic appointment, and 20% (72/354) previously considered lost to follow-up had re-engaged in the telehealth period. CONCLUSIONS: Rates of FTA in a tertiary ACHD clinic significantly reduced after the introduction of telehealth consultation. A fifth of patients considered lost to follow-up were re-engaged with telehealth. Additional strategies to further reduce FTA should be explored.

Usability of ENTRUST as an Assessment Tool for Entrustable Professional Activities (EPAs): A Mixed Methods Analysis.

OBJECTIVE: As the American Board of Surgery transitions to a competency-based model of surgical education centered upon entrustable professional activities (EPAs), there is a growing need for objective tools to determine readiness for entrustment. This study evaluates the usability of ENTRUST, an innovative virtual patient simulation platform to assess surgical trainees' decision-making skills in preoperative, intra-operative, and post-operative settings. DESIGN: This is a mixed-methods analysis of the usability of the ENTRUST platform. Quantitative data was collected using the system usability scale (SUS) and Likert responses. Analysis was performed with descriptive statistics, bivariate analysis, and multivariable linear regression. Qualitative analysis of open-ended responses was performed using the Nielsen-Shneiderman Heuristics framework. SETTING: This study was conducted at an academic institution in a proctored exam setting. PARTICIPANTS: The analysis includes n = 47 (PGY 1-5) surgical residents who completed an online usability survey following the ENTRUST Inguinal Hernia EPA Assessment. RESULTS: The ENTRUST platform had a median SUS score of 82.5. On bivariate and multivariate analyses, there were no significant differences between usability based on demographic characteristics (all p > 0.05), and SUS score was independent of ENTRUST performance (r = 0.198, p = 0.18). Most participants agreed that the clinical workup of the patient was engaging (91.5%) and felt realistic (85.1%). The most frequent heuristics represented in the qualitative analysis included feedback, visibility, match, and control. Additional themes of educational value, enjoyment, and ease-of-use highlighted participants' perspectives on the usability of ENTRUST. CONCLUSIONS: ENTRUST demonstrates high usability in this population. Usability was independent of ENTRUST score performance and there were no differences in usability identified in this analysis based on demographic subgroups. Qualitative analysis highlighted the acceptability of ENTRUST and will inform ongoing development of the platform. The ENTRUST platform holds potential as a tool for the assessment of EPAs in surgical residency programs.

Exploring the Use of Natural Language Processing for Objective Assessment of Disorganized Speech in Schizophrenia.

Objective: Measurement-based care tools in psychiatry are useful for symptom monitoring and detecting response to treatment, but methods for quick and objective measurement are lacking especially for acute psychosis. The aim of this study was to explore potential language markers, detected by natural language processing (NLP) methods, as a means to objectively measure the severity of psychotic symptoms of schizophrenia in an acute clinical setting. Methods: Twenty-two speech samples were collected from seven participants who were hospitalized for schizophrenia, and their symptoms were evaluated over time with SAPS/SANS and TLC scales. Linguistic features were extracted from the speech data using machine learning techniques. Spearman's correlation was performed to examine the relationship between linguistic features and symptoms. Various machine learning models were evaluated by cross-validation methods for their ability to predict symptom severity using the linguistic markers. Results: Reduced lexical richness and syntactic complexity were characteristic of negative symptoms, while lower content density and more repetitions in speech were predictors of positive symptoms. Machine learning models predicted severity of alogia, illogicality, poverty of speech, social inattentiveness, and TLC scores with up to 82% accuracy. Additionally, speech incoherence was quantifiable through language markers derived from NLP methods. Conclusions: These preliminary findings suggest that NLP may be useful in identifying clinically relevant language markers of schizophrenia, which can enhance objectivity in symptom monitoring during hospitalization. Further work is needed to replicate these findings in a larger data set and explore methods for feasible implementation in practice.

Impact of an Inpatient Allergy Guideline on ß-Lactam and Alternative Antibiotic Use.

BACKGROUND: A guideline identifying when inpatients with penicillin or cephalosporin antibiotic allergy labels (PCAAL) can receive ß-lactam antibiotics increased ß-lactam receipt at a large northeastern US health care system. OBJECTIVE: To report outcomes of implementing a similar guideline and electronic order set (OS) at an independent academic health care system. METHODS: Penicillin/cephalosporin receipt (percentage of inpatients receiving full doses) and alternative antibiotic use (days of therapy per 1000 patient-days [DOT/1000PD]) were compared over 3 periods before (February 1, 2017, to January 31, 2018) and after guideline implementation (February 1, 2018, to January 31, 2019), and after OS implementation (February 1, 2019, to January 31, 2020) among inpatients with PCAAL admitted on medical services with access to guideline/OS and education (Medical-PCAAL, n = 8721), surgical services with access to guideline/OS without education (Surgical-PCAAL, n = 5069), and obstetrics/gynecology services without interventions (Ob/Gyn-PCAAL, n = 798) and inpatients without PCAAL admitted on the same services (Medical-No-PCAAL, n = 50,840; Surgical-No-PCAAL, n = 29,845; Ob/Gyn-No-PCAAL, n = 6109). χ2 tests were used to compare categorical variables, and analysis of variance was used to compare continuous and interrupted time series analyses (ITSA) to investigate the guideline/OS implementation effect on penicillin/cephalosporin receipt. RESULTS: In the Medical-PCAAL group, penicillin/cephalosporin receipt increased (58%-68%, P < .001), specifically for cefazolin (8%-11%, P = .02) and third- to fifth-generation cephalosporins (43%-48%, P = .04), and aztreonam use decreased (12 DOT/1000PD, P = .03). In the Medical-No-PCAAL group, penicillin/cephalosporin receipt increased (88%-90%, P = .004), specifically for penicillin (40%-44%, P < .001), without changes in aztreonam use. Significant changes were not observed in these outcomes on surgical or obstetrics/gynecology services. Per ITSA, guideline/OS implementation was associated with increased penicillin/cephalosporin receipt in the Medical-PCAAL group only. CONCLUSION: Guideline and OS implementation was associated with improved antibiotic stewardship on inpatient services that also received allergy education.