Motor vehicle collision characteristics and hospitalization outcomes associated with mild traumatic brain injury and concomitant whiplash injury.

OBJECTIVE: Few large studies have investigated the factors and outcomes related to concomitant injuries occurring alongside mild traumatic brain injury (mTBI) after motor vehicle collisions (MVCs). Thus, the objective of this study was to assess whether MVC characteristics predict which patients with mTBI will have concomitant whiplash injury, and whether concomitant whiplash injury affects care utilization for these patients. METHODS: This retrospective cohort study included 22,213 patients with mTBI after MVC identified from the American College of Surgeons Trauma Quality Programs dataset. A hierarchical logistic regression model was constructed to investigate patient and MVC factors associated with concomitant whiplash injury. Propensity score matching on whiplash status, in conjunction with a multivariable logistic regression model, assessed if concomitant whiplash affected odds of hospitalization. In the subgroup of patients who were hospitalized, associations with hospital length of stay (LOS) and discharge disposition were investigated. RESULTS: The median (IQR) age was 34 (24-51) years, with a median Glasgow Coma Scale score at presentation of 15 (15-15). Patients with concomitant whiplash were older (median 36 years vs 34 years, p = 0.03) and had higher rates of hospitalization (75% vs 64%, p < 0.001). In the hierarchical model for associations with concomitant whiplash injury, patients with blood alcohol content (BAC) greater than the federal driving limit had lower odds of concomitant whiplash (OR 0.63, 95% CI 0.49-0.81) along with those who had airbag deployment (OR 0.80, 95% CI 0.68-0.95), but seatbelt use was associated with greater odds (OR 1.41, 95% CI 1.16-1.71). After matching, concomitant whiplash was independently associated with increased odds of hospitalization (OR 1.67, 95% CI 1.40-1.99) while seatbelt use was associated with decreased odds (OR 0.88, 95% CI 0.81-0.95). Among hospitalized patients, concomitant whiplash was not associated with hospital LOS or discharge disposition. CONCLUSIONS: MVC characteristics such as alcohol consumption and airbag deployment were protective toward development of concomitant whiplash for mTBI patients, while seatbelt use was associated with higher risk. Concomitant whiplash increases the odds of hospitalization for mTBI patients but does not affect hospital LOS or discharge disposition, while seatbelt use is associated with lower rates of hospitalization and a more favorable hospital course. These findings provide context to injury patterns and care provision after a common mechanism of injury.

Large Language Model Capabilities in Perioperative Risk Prediction and Prognostication.

Importance: General-domain large language models may be able to perform risk stratification and predict postoperative outcome measures using a description of the procedure and a patient's electronic health record notes. Objective: To examine predictive performance on 8 different tasks: prediction of American Society of Anesthesiologists Physical Status (ASA-PS), hospital admission, intensive care unit (ICU) admission, unplanned admission, hospital mortality, postanesthesia care unit (PACU) phase 1 duration, hospital duration, and ICU duration. Design, Setting, and Participants: This prognostic study included task-specific datasets constructed from 2 years of retrospective electronic health records data collected during routine clinical care. Case and note data were formatted into prompts and given to the large language model GPT-4 Turbo (OpenAI) to generate a prediction and explanation. The setting included a quaternary care center comprising 3 academic hospitals and affiliated clinics in a single metropolitan area. Patients who had a surgery or procedure with anesthesia and at least 1 clinician-written note filed in the electronic health record before surgery were included in the study. Data were analyzed from November to December 2023. Exposures: Compared original notes, note summaries, few-shot prompting, and chain-of-thought prompting strategies. Main Outcomes and Measures: F1 score for binary and categorical outcomes. Mean absolute error for numerical duration outcomes. Results: Study results were measured on task-specific datasets, each with 1000 cases with the exception of unplanned admission, which had 949 cases, and hospital mortality, which had 576 cases. The best results for each task included an F1 score of 0.50 (95% CI, 0.47-0.53) for ASA-PS, 0.64 (95% CI, 0.61-0.67) for hospital admission, 0.81 (95% CI, 0.78-0.83) for ICU admission, 0.61 (95% CI, 0.58-0.64) for unplanned admission, and 0.86 (95% CI, 0.83-0.89) for hospital mortality prediction. Performance on duration prediction tasks was universally poor across all prompt strategies for which the large language model achieved a mean absolute error of 49 minutes (95% CI, 46-51 minutes) for PACU phase 1 duration, 4.5 days (95% CI, 4.2-5.0 days) for hospital duration, and 1.1 days (95% CI, 0.9-1.3 days) for ICU duration prediction. Conclusions and Relevance: Current general-domain large language models may assist clinicians in perioperative risk stratification on classification tasks but are inadequate for numerical duration predictions. Their ability to produce high-quality natural language explanations for the predictions may make them useful tools in clinical workflows and may be complementary to traditional risk prediction models.

Predicting Out-of-Hospital Cardiac Arrest in the General Population Using Electronic Health Records.

BACKGROUND: The majority of out-of-hospital cardiac arrests (OHCAs) occur among individuals in the general population, for whom there is no established strategy to identify risk. In this study, we assess the use of electronic health record (EHR) data to identify OHCA in the general population and define salient factors contributing to OHCA risk. METHODS: The analytical cohort included 2366 individuals with OHCA and 23 660 age- and sex-matched controls receiving health care at the University of Washington. Comorbidities, electrocardiographic measures, vital signs, and medication prescription were abstracted from the EHR. The primary outcome was OHCA. Secondary outcomes included shockable and nonshockable OHCA. Model performance including area under the receiver operating characteristic curve and positive predictive value were assessed and adjusted for observed rate of OHCA across the health system. RESULTS: There were significant differences in demographic characteristics, vital signs, electrocardiographic measures, comorbidities, and medication distribution between individuals with OHCA and controls. In external validation, discrimination in machine learning models (area under the receiver operating characteristic curve 0.80-0.85) was superior to a baseline model with conventional cardiovascular risk factors (area under the receiver operating characteristic curve 0.66). At a specificity threshold of 99%, correcting for baseline OHCA incidence across the health system, positive predictive value was 2.5% to 3.1% in machine learning models compared with 0.8% for the baseline model. Longer corrected QT interval, substance abuse disorder, fluid and electrolyte disorder, alcohol abuse, and higher heart rate were identified as salient predictors of OHCA risk across all machine learning models. Established cardiovascular risk factors retained predictive importance for shockable OHCA, but demographic characteristics (minority race, single marital status) and noncardiovascular comorbidities (substance abuse disorder) also contributed to risk prediction. For nonshockable OHCA, a range of salient predictors, including comorbidities, habits, vital signs, demographic characteristics, and electrocardiographic measures, were identified. CONCLUSIONS: In a population-based case-control study, machine learning models incorporating readily available EHR data showed reasonable discrimination and risk enrichment for OHCA in the general population. Salient factors associated with OCHA risk were myriad across the cardiovascular and noncardiovascular spectrum. Public health and tailored strategies for OHCA prediction and prevention will require incorporation of this complexity.

Evaluation of the Glasgow Coma Scale-Pupils score for predicting inpatient mortality among patients with traumatic subdural hematoma at United States trauma centers.

OBJECTIVE: The Glasgow Coma Scale-Pupils (GCS-P) score has been suggested to better predict patient outcomes compared with GCS alone, while avoiding the need for more complex clinical models. This study aimed to compare the prognostic ability of GCS-P versus GCS in a national cohort of traumatic subdural hematoma (SDH) patients. METHODS: Patient data were obtained from the National Trauma Data Bank (2017-2019). Inclusion criteria were traumatic SDH diagnosis with available data on presenting GCS score, pupillary reactivity, and discharge disposition. Patients with severe polytrauma or nonsurvivable head injury at presentation were excluded. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of GCS-P versus GCS scores for inpatient mortality prediction were evaluated across the entire cohort, as well as in subgroups based on age and traumatic brain injury (TBI) type (blunt vs penetrating). Calibration curves were plotted based on predicted probabilities and actual outcomes. RESULTS: A total of 196,747 traumatic SDH patients met the study inclusion criteria. Sensitivity (0.707 vs 0.702), specificity (0.821 vs 0.823), and AUC (0.825 vs 0.814, p < 0.001) of GCS-P versus GCS scores for prediction of inpatient mortality were similar. Calibration curve analysis revealed that GCS scores slightly underestimated inpatient mortality risk, whereas GCS-P scores did not. In patients > 65 years of age with blunt TBI (51.9%, n = 102,148), both GCS-P and GCS scores underestimated inpatient mortality risk. In patients with penetrating TBI (2.4%, n = 4,710), the AUC of the GCS-P score was significantly higher (0.902 vs 0.851, p < 0.001). In this subgroup, both GCS-P and GCS scores underestimated inpatient mortality risk among patients with lower rates of observed mortality and overestimated risk among patients with higher rates of observed mortality. This effect was more pronounced in the GCS-P calibration curve. CONCLUSIONS: The GCS-P score provides better short-term prognostication compared with the GCS score alone among traumatic SDH patients with penetrating TBI. The GCS-P score overestimates inpatient mortality risk among penetrating TBI patients with higher rates of observed mortality. For penetrating TBI patients, which comprised 2.4% of our SDH cohort, a low GCS-P score should not justify clinical nihilism or forgoing aggressive treatment.

Machine Learning Identifies Variation in Timing of Palliative Care Consultations Among Traumatic Brain Injury Patients.

Background and Objective Timely palliative care involvement offers demonstrable benefits for traumatic brain injury (TBI) patients; however, palliative care consultations (PCCs) are used inconsistently during TBI management. This study aimed to employ advanced machine learning techniques to elucidate the primary drivers of PCC timing variability for TBI patients. Methods Data on admission, hospital course, and outcomes were collected for a cohort of 232 TBI patients who received both PCCs and neurosurgical consultations during the same hospitalization. Principal Component Analysis (PCA) and K-means clustering were used to identify patient phenotypes, which were then compared using Kaplan-Meier analysis. An extreme gradient boosting model (XGBoost) was employed to determine drivers of PCC timing, with model interpretation performed using SHapley Additive exPlanations (SHAP). Results Cluster A (n = 86) consisted mainly of older (median [IQR] = 87 [78, 94] years), White females with mild TBIs and demonstrated the shortest time-to-PCC (2.5 [1.0, 7.0] days). Cluster B (n = 108) also sustained mild TBIs but comprised moderately younger (81 [75, 86] years) married White males with later PCC (5.0 [3.0, 10.8] days). Cluster C (n = 38) represented much younger (46.5 [29.5, 59.8] years), more severely injured, non-White patients with the latest PCC initiation (9.0 [4.2, 17.0] days). The clusters did not differ by discharge disposition (p = 0.4) or frequency inpatient mortality (p > 0.9); however, Kaplan-Meier analysis revealed a significant difference in the time from admission to PCC (p < 0.001), despite no differences in time from admission to mortality (p = 0.18). SHAP analysis of the XGBoost model identified age, sex, and race as the most influential drivers of PCC timing. Conclusions This study highlights crucial disparities in PCC timing for TBI patients and underscores the need for targeted strategies to ensure timely and equitable palliative care integration for this vulnerable population.

Impact of COVID-19 Medication (Favipiravir and Hydroxychloroquine) on High Frequency Hearing.

During the initial phase of the COVID-19 pandemic, there was ongoing investigation into potentially effective treatments. Antiviral medications such as Favipiravir and Hydroxychloroquine were employed to treat COVID-19 infections. However, limited studies have examined the adverse effects of these medications on hearing, particularly at extended high frequencies. This study included 10 subjects who had received medications like Azithromycin, a combination of Favipiravir and Hydroxychloroquine, and Hydroxychloroquine alone as part of their COVID-19 treatment. These subjects had previously undergone extended high-frequency audiometry testing (from 8 to 20 kHz) as part of another project conducted by the same department before contracting COVID-19. Post-COVID-19 extended high-frequency audiometry was performed 1 month after the patients received a negative RT-PCR report. The results were then compared using a Paired t-test. A significant shift in the thresholds of high frequencies above 8-20 kHz is found in subjects who had received Favipiravir and Hydroxychloroquine medications. We observed a significant impact of COVID-19 medications on high-frequency hearing, which tends to go unnoticed in regular pure-tone audiometry evaluations. Therefore, our study emphasizes the need for regular follow-ups, including detailed audiological assessments that incorporate extended high-frequency testing, at least once every 3 months for patients who have taken medications for COVID-19 treatment.

Resection of an incidentally discovered spinal arachnoid web: illustrative case.

BACKGROUND: Spinal arachnoid webs (SAWs) are rare pathologies of the spinal meninges often associated with syringomyelia and the radiographic "scalpel sign." Patients can experience pain, numbness, gait disturbances, or no symptoms at all. They are typically diagnosed via magnetic resonance imaging and treated with laminectomy and excision. OBSERVATIONS: A 61-year-old male presented after a mechanical fall and had an incidentally discovered SAW on imaging. He was initially asymptomatic and was therefore conservatively managed. Several years later, however, the patient experienced new-onset back pain, paresthesia, and balance problems, with interval imaging demonstrating worsening of the edema surrounding his SAW. The patient subsequently underwent resection of the SAW, which led to significant resolution of his symptoms. LESSONS: An SAW can be asymptomatic or can manifest with a wide variety of symptoms. When this condition is incidentally discovered in asymptomatic patients, neurosurgeons should guide these patients to follow-up urgently if they develop any neurological symptoms. At that time, further imaging can be performed to determine if surgical treatment is indicated. Although SAW is rare, clinicians should be aware of the signs and symptoms, because prompt surgical intervention can significantly improve neurological symptoms.

Concomitant Traumatic Brain Injury Delays Surgery in Patients With Traumatic Spinal Cord Injury.

BACKGROUND AND OBJECTIVES: Growing evidence supports prompt surgical decompression for patients with traumatic spinal cord injury (tSCI). Rates of concomitant tSCI and traumatic brain injury (TBI) range from 10% to 30%. Concomitant TBI may delay tSCI diagnosis and surgical intervention. Little is known about real-world management of this common injury constellation that carries significant clinical consequences. This study aimed to quantify the impact of concomitant TBI on surgical timing in a national cohort of patients with tSCI. METHODS: Patient data were obtained from the National Trauma Data Bank (2007-2016). Patients admitted for tSCI and who received surgical intervention were included. Delayed surgical intervention was defined as surgery after 24 hours of admission. Multivariable hierarchical regression models were constructed to measure the risk-adjusted association between concomitant TBI and delayed surgical intervention. Secondary outcome included favorable discharge status. RESULTS: We identified 14 964 patients with surgically managed tSCI across 377 North American trauma centers, of whom 2444 (16.3%) had concomitant TBI and 4610 (30.8%) had central cord syndrome (CCS). The median time to surgery was 20.0 hours for patients without concomitant TBI and 24.8 hours for patients with concomitant TBI. Hierarchical regression modeling revealed that concomitant TBI was independently associated with delayed surgery in patients with tSCI (odds ratio [OR], 1.3; 95% CI, 1.1-1.6). Although CCS was associated with delayed surgery (OR, 1.5; 95% CI, 1.4-1.7), we did not observe a significant interaction between concomitant TBI and CCS. In the subset of patients with concomitant tSCI and TBI, patients with severe TBI were significantly more likely to experience a surgical delay than patients with mild TBI (OR, 1.4; 95% CI, 1.0-1.9). CONCLUSION: Concomitant TBI delays surgical management for patients with tSCI. This effect is largest for patients with tSCI with severe TBI. These findings should serve to increase awareness of concomitant TBI and tSCI and the likelihood that this may delay time-sensitive surgery.

Public Health Informatics and the Perioperative Physician: Looking to the Future.

The role of informatics in public health has increased over the past few decades, and the coronavirus disease 2019 (COVID-19) pandemic has underscored the critical importance of aggregated, multicenter, high-quality, near-real-time data to inform decision-making by physicians, hospital systems, and governments. Given the impact of the pandemic on perioperative and critical care services (eg, elective procedure delays; information sharing related to interventions in critically ill patients; regional bed-management under crisis conditions), anesthesiologists must recognize and advocate for improved informatic frameworks in their local environments. Most anesthesiologists receive little formal training in public health informatics (PHI) during clinical residency or through continuing medical education. The COVID-19 pandemic demonstrated that this knowledge gap represents a missed opportunity for our specialty to participate in informatics-related, public health-oriented clinical care and policy decision-making. This article briefly outlines the background of PHI, its relevance to perioperative care, and conceives intersections with PHI that could evolve over the next quarter century.

Craniotomy versus craniectomy for traumatic acute subdural hematoma-coarsened exact matched analysis of outcomes.

BACKGROUND AND OBJECTIVES: Acute subdural hematoma (aSDH) after traumatic brain injury frequently requires emergent craniotomy (CO) or decompressive craniectomy (DC). We sought to determine the variables associated with either surgical approach and to compare outcomes between matched patients. METHODS: A multi-center retrospective review was used to identify traumatic aSDH patients who underwent CO or DC. Patient variables independently associated with surgical approach were used for coarsened exact matching.Multivariate logistic regression and multivariate Cox proportional-hazards regression wereconducted on matched patients to determine independent predictors of mortality. RESULTS: Seventy-six patients underwent CO and sixty-two underwent DC for aSDH evacuation. DC patients were21.4 years younger (P < 0.001), more likely to be male (80.6 % vs 60.5 %,P = 0.011), and present with GCS ≤ 8 (64.5 % vs 36.8 %,P = 0.001). Age (P < 0.001), epidural hematoma (P = 0.01), skull fracture (P = 0.001), and cisternal effacement (P = 0.02) were independently associated with surgical approach. After coarsened exact matching, DC (P = 0.008), older age (P = 0.007), male sex (P = 0.04), and intraventricular hemorrhage (P = 0.02), were independently associated with inpatient mortality. Multivariate Cox proportional-hazards regression demonstrated that DC was independently associated with mortality at 90-days (P = 0.001) and 1-year post-operation (P = 0.003). CONCLUSION: aSDH patients who receive surgical evacuation via DC as opposed to CO are younger, more likely to be male, and have worse clinical exam. After controlling for patient differences via coarsened exact matching, DC is independently associated with mortality.

Perioperative Mortality of the COVID-19 Recovered Patient Compared to a Matched Control: A Multicenter Retrospective Cohort Study.

BACKGROUND: Surgical procedures performed on patients with recent exposure to COVID-19 infection have been associated with increased mortality risk in previous studies. Accordingly, elective surgery is often delayed after infection. The study aimed to compare 30-day hospital mortality and postoperative complications (acute kidney injury, pulmonary complications) of surgical patients with a previous COVID-19 infection to a matched cohort of patients without known previous COVID-19. The authors hypothesized that COVID-19 exposure would be associated with an increased mortality risk. METHODS: In this retrospective observational cohort study, patients presenting for elective inpatient surgery across a multicenter cohort of academic and community hospitals from April 2020 to April 2021 who had previously tested positive for COVID-19 were compared to controls who had received at least one previous COVID-19 test but without a known previous COVID-19-positive test. The cases were matched based on anthropometric data, institution, and comorbidities. Further, the outcomes were analyzed stratified by timing of a positive test result in relation to surgery. RESULTS: Thirty-day mortality occurred in 229 of 4,951 (4.6%) COVID-19-exposed patients and 122 of 4,951 (2.5%) controls. Acute kidney injury was observed in 172 of 1,814 (9.5%) exposed patients and 156 of 1,814 (8.6%) controls. Pulmonary complications were observed in 237 of 1,637 (14%) exposed patients and 164 of 1,637 (10%) controls. COVID-19 exposure was associated with an increased 30-day mortality risk (adjusted odds ratio, 1.63; 95% CI, 1.38 to 1.91) and an increased risk of pulmonary complications (1.60; 1.36 to 1.88), but was not associated with an increased risk of acute kidney injury (1.03; 0.87 to 1.22). Surgery within 2 weeks of infection was associated with a significantly increased risk of mortality and pulmonary complications, but that effect was nonsignificant after 2 weeks. CONCLUSIONS: Patients with a positive test for COVID-19 before elective surgery early in the pandemic have an elevated risk of perioperative mortality and pulmonary complications but not acute kidney injury as compared to matched controls. The span of time from positive test to time of surgery affected the mortality and pulmonary risk, which subsided after 2 weeks.

Impact of Antithrombotic Medications and Reversal Strategies on the Surgical Management and Outcomes of Traumatic Acute Subdural Hematoma.

OBJECTIVE: Careful hematologic management is required in surgical patients with traumatic acute subdural hematoma (aSDH) taking antithrombotic medications. We sought to compare outcomes between patients with aSDH taking antithrombotic medications at admission who received antithrombotic reversal with patients with aSDH not taking antithrombotics. METHODS: Retrospective review identified patients with traumatic aSDH requiring surgical evacuation. The cohort was divided based on antithrombotic use and whether pharmacologic reversal agents or platelet transfusions were administered. A 3-way comparison of outcomes was performed between patients taking anticoagulants who received pharmacologic reversal, patients taking antiplatelets who received platelet transfusion, and patients not taking antithrombotics. Multivariable regressions, adjusted for injury severity, further investigated associations with outcomes. RESULTS: Of 138 patients who met inclusion criteria, 13.0% (n = 18) reported taking anticoagulants, 16.7% (n = 23) reported taking antiplatelets, and 3.6% (n = 5) reported taking both. Patients taking antiplatelets who received platelet transfusion had longer intraoperative times (P = 0.040) and higher rates of palliative care consultations (P = 0.046) compared with patients taking anticoagulants who received pharmacologic reversal and patients not taking antithrombotics. Across groups, no significant differences were found in frequency of in-hospital intracranial hemorrhage and venous thromboembolism, length of hospital stay, rate of inpatient mortality, or follow-up health status. In multivariable analysis, intraoperative time remained longest for the antiplatelets with platelet transfusion group. Other outcomes were not associated with patient group. CONCLUSIONS: Among surgical patients with traumatic aSDH, those taking antiplatelet medications who receive platelet transfusions experience longer intraoperative procedure times and higher rates of palliative care consultation. Comparable outcomes were observed between patients receiving antithrombotic reversal and patients not taking antithrombotics.

Exoscope Use in Spine Surgery: A Systematic Review of Applications, Benefits, and Limitations.

BACKGROUND: Exoscopes were recently developed as an alternative to the operative microscope (OM) and endoscope for intraoperative visualization during neurosurgery. Prior reviews studying mixed cranial and spinal surgical cohorts reported advantages with exoscope use, including improved ergonomics and teaching. In recent years, there has been an increase in exoscope research, with no updated systematic review focused exclusively on the benefits and limitations of exoscope use in spine surgery. Thus, we sought to systematically synthesize the literature related to exoscope-assisted spine surgery. METHODS: A literature search was conducted using the PubMed, Embase, Scopus, Cochrane, and Web of Science databases to identify relevant studies reported between 2010 and September 2023. Data, such as the exoscope model used, procedure types performed, and user observations, were then collected. RESULTS: A total of 31 studies met our inclusion criteria, including 481 patients with spine pathologies who underwent a surgical procedure using 1 of 9 exoscope models. The lumbar region was the most frequently operated area (n = 234; 48.6%), and discectomies comprised the most overall procedures (n = 273; 56.8%). All patients benefited clinically. The reported advantages of exoscopes compared with OMs or endoscopes were improved focal distance, surgeon posture, trainee education, compactness, and assistant participation. Other aspects such as stereopsis, illumination, and cost had various observations. CONCLUSIONS: Exoscopes have advantages compared with OMs or endoscopes during spine surgery. The user learning curve is minimal, and no negative patient outcomes have been reported. However, some aspects of exoscope use necessitate longer term prospective research before exoscopes can be considered a standard tool in the armamentarium of intraoperative visualization strategies.

A future blood test for acute traumatic spinal cord injury.

Acute spinal cord injury (SCI) requires prompt diagnosis and intervention to minimize the risk of permanent neurologic deficit. Presently, SCI diagnosis and interventional planning rely on magnetic resonance imaging (MRI), which is not always available or feasible for severely injured patients. Detection of disease-specific biomarkers in biofluids via liquid biopsy may provide a more accessible and objective means of evaluating patients with suspected SCI. Cell-free DNA, which has been used for diagnosing and monitoring oncologic disease, may detect damage to spinal cord neurons via tissue-specific methylation patterns. Other types of biomarkers, including proteins and RNA species, have also been found to reflect neuronal injury and may be included as part of a multi-analyte assay to improve liquid biopsy performance. The feasibility of implementing liquid biopsy into current practices of SCI management is supported by the relative ease of blood sample collection as well as recent advancements in droplet digital polymerase chain reaction technology. In this review, we detail the current landscape of biofluid biomarkers for acute SCI and propose a framework for the incorporation of a putative blood test into the clinical management of SCI.

Prediction of American Society of Anesthesiologists Physical Status Classification from preoperative clinical text narratives using natural language processing.

BACKGROUND: Electronic health records (EHR) contain large volumes of unstructured free-form text notes that richly describe a patient's health and medical comorbidities. It is unclear if perioperative risk stratification can be performed directly from these notes without manual data extraction. We conduct a feasibility study using natural language processing (NLP) to predict the American Society of Anesthesiologists Physical Status Classification (ASA-PS) as a surrogate measure for perioperative risk. We explore prediction performance using four different model types and compare the use of different note sections versus the whole note. We use Shapley values to explain model predictions and analyze disagreement between model and human anesthesiologist predictions. METHODS: Single-center retrospective cohort analysis of EHR notes from patients undergoing procedures with anesthesia care spanning all procedural specialties during a 5 year period who were not assigned ASA VI and also had a preoperative evaluation note filed within 90 days prior to the procedure. NLP models were trained for each combination of 4 models and 8 text snippets from notes. Model performance was compared using area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (AUPRC). Shapley values were used to explain model predictions. Error analysis and model explanation using Shapley values was conducted for the best performing model. RESULTS: Final dataset includes 38,566 patients undergoing 61,503 procedures with anesthesia care. Prevalence of ASA-PS was 8.81% for ASA I, 31.4% for ASA II, 43.25% for ASA III, and 16.54% for ASA IV-V. The best performing models were the BioClinicalBERT model on the truncated note task (macro-average AUROC 0.845) and the fastText model on the full note task (macro-average AUROC 0.865). Shapley values reveal human-interpretable model predictions. Error analysis reveals that some original ASA-PS assignments may be incorrect and the model is making a reasonable prediction in these cases. CONCLUSIONS: Text classification models can accurately predict a patient's illness severity using only free-form text descriptions of patients without any manual data extraction. They can be an additional patient safety tool in the perioperative setting and reduce manual chart review for medical billing. Shapley feature attributions produce explanations that logically support model predictions and are understandable to clinicians.

Cardiometabolic disease and obesity patterns differentially predict acute kidney injury after total joint replacement: a retrospective analysis.

BACKGROUND: Acute kidney injury (AKI) is a frequent yet understudied postoperative total joint arthroplasty complication. This study aimed to describe cardiometabolic disease co-occurrence using latent class analysis, and associated postoperative AKI risk. METHODS: This retrospective analysis examined patients ≥18 years old undergoing primary total knee or hip arthroplasties within the US Multicenter Perioperative Outcomes Group of hospitals from 2008 to 2019. AKI was defined using modified Kidney Disease: Improving Global Outcomes (KDIGO) criteria. Latent classes were constructed from eight cardiometabolic diseases including hypertension, diabetes, and coronary artery disease, excluding obesity. A mixed-effects logistic regression model was constructed for the outcome of any AKI and the exposure of interaction between latent class and obesity status adjusting for preoperative and intraoperative covariates. RESULTS: Of 81 639 cases, 4007 (4.9%) developed AKI. Patients with AKI were more commonly older and non-Hispanic Black, with more significant comorbidity. A latent class model selected three groups of cardiometabolic patterning, labelled 'hypertension only' (n=37 223), 'metabolic syndrome (MetS)' (n=36 503), and 'MetS+cardiovascular disease (CVD)' (n=7913). After adjustment, latent class/obesity interaction groups had differential risk of AKI compared with those in 'hypertension only'/non-obese. Those 'hypertension only'/obese had 1.7-fold increased odds of AKI (95% confidence interval [CI]: 1.5-2.0). Compared with 'hypertension only'/non-obese, those 'MetS+CVD'/obese had the highest odds of AKI (odds ratio 3.1, 95% CI: 2.6-3.7), whereas 'MetS+CVD'/non-obese had 2.2 times the odds of AKI (95% CI: 1.8-2.7; model area under the curve 0.76). CONCLUSIONS: The risk of postoperative AKI varies widely between patients. The current study suggests that the co-occurrence of metabolic conditions (diabetes mellitus, hypertension), with or without obesity, is a more important risk factor for acute kidney injury than individual comorbid diseases.

Helicobacter Pylori-Negative MALT Lymphoma: A Series of Two Cases Presenting with Life-Threatening Upper Gastrointestinal Bleeding.

Mucosa-associated lymphoid tissue (MALT) lymphoma is a common cause of gastric lymphoma. Although most cases are associated with an H. pylori infection, approximately 10% are H. pylori-negative. Patients with gastric MALT lymphoma are usually asymptomatic or present with nonspecific symptoms such as abdominal pain, dyspepsia, weight loss, and occult gastrointestinal bleeding. In this report, we describe two patients with H. pylori-negative MALT lymphoma who both presented with acute upper gastrointestinal bleeding that led to hemodynamic instability. After resuscitation, emergent endoscopy was performed. Both patients had the t (11; 18) (q21; q21) translocation, which prompted direct treatment by radiotherapy.

Evaluating the role for regional analgesia in children with spina bifida: a retrospective observational study comparing the efficacy of regional versus systemic analgesia protocols following major urological surgery.

INTRODUCTION: Regional techniques are a key component of multimodal analgesia and help decrease opioid use perioperatively, but some techniques may not be suitable for all patients, such as those with spina bifida. We hypothesized peripheral regional catheters would reduce postoperative opioid use compared with no regional analgesia without increasing pain scores in pediatric patients with spina bifida undergoing major urological surgery. METHODS: A retrospective review of a multicenter database established for the study of enhanced recovery after surgery was performed of patients from 2009 to 2021 who underwent bladder augmentation or creation of catheterizable channels. Patients without spina bifida and those receiving epidural analgesia were excluded. Opioids were converted into morphine equivalents and normalized to patient weight. RESULTS: 158 patients with pediatric spina bifida from 7 centers were included, including 87 with and 71 without regional catheters. There were no differences in baseline patient factors. Anesthesia setup increased from median 40 min (IQR 34-51) for no regional to 64 min (IQR 40-97) for regional catheters (p<0.01). The regional catheter group had lower median intraoperative opioid usage (0.24 vs 0.80 mg/kg morphine equivalents, p<0.01) as well as lower in-hospital postoperative opioid usage (0.05 vs 0.23 mg/kg/day morphine equivalents, p<0.01). Pain scores were not higher in the regional catheters group. DISCUSSION: Continuous regional analgesia following major urological surgery in children with spina bifida was associated with a 70% intraoperative and 78% postoperative reduction in opioids without higher pain scores. This approach should be considered for similar surgical interventions in this population. TRIAL REGISTRATION NUMBER: NCT03245242.

Completion of an Enhanced Recovery Program in a Pediatric Ambulatory Surgery Center: A Quality Improvement Initiative.

BACKGROUND: Enhanced Recovery After Surgery (ERAS) was first established in 2001 focusing on recovery from complex surgical procedures in adults and recently expanded to ambulatory surgery. The evidence for ERAS in children is limited. In 2018, recognized experts began developing needed pediatric evidence. Center-wide efforts involving all ambulatory surgical patients and procedures have not previously been described. METHODS: A comprehensive assessment and gap analysis of ERAS elements in our ambulatory center identified 11 of 19 existing elements. The leadership committed to implementing an Enhanced Recovery Program (ERP) to improve existing elements and close as many remaining gaps as possible. A quality improvement (QI) team was launched to improve 5 existing ERP elements and to introduce 6 new elements (target 17/19 ERP elements). The project plan was broken into 1 preparation phase to collect baseline data and 3 implementation phases to enhance existing and implement new elements. Statistical process control methodology was used. Team countermeasures were based on available evidence. A consensus process was used to resolve disagreement. Monthly meetings were held to share real-time data, gather new feedback, and modify countermeasure plans as needed. The primary outcome measure selected was mean postanesthesia care unit (PACU) length of stay (LOS). Secondary outcomes measures were mean maximum pain score in PACU and patient/family satisfaction scores. RESULTS: The team had expanded the pool of active ERP elements from 11 to 16 of 19. The mean PACU LOS demonstrated significant reduction (early in phase 1 and again in phase 3). No change was seen for the mean maximum pain score in PACU or surgical complication rates. Patient/family satisfaction scores were high and sustained throughout the period of study (91.1% ± 5.7%). Patient/family and provider engagement/compliance were high. CONCLUSIONS: This QI project demonstrated the feasibility of pediatric ERP in an ambulatory surgical setting. Furthermore, a center-wide approach was shown to be possible. Additional studies are needed to determine the relevance of this project to other institutions.