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.

Early Cardiac Evaluation, Abnormal Test Results, and Associations with Outcomes in Patients with Acute Brain Injury Admitted to a Neurocritical Care Unit.

Background: to examine factors associated with cardiac evaluation and associations between cardiac test abnormalities and clinical outcomes in patients with acute brain injury (ABI) due to acute ischemic stroke (AIS), spontaneous subarachnoid hemorrhage (SAH), spontaneous intracerebral hemorrhage (sICH), and traumatic brain injury (TBI) requiring neurocritical care. Methods: In a cohort of patients ≥18 years, we examined the utilization of electrocardiography (ECG), beta-natriuretic peptide (BNP), cardiac troponin (cTnI), and transthoracic echocardiography (TTE). We investigated the association between cTnI, BNP, sex-adjusted prolonged QTc interval, low ejection fraction (EF < 40%), all-cause mortality, death by neurologic criteria (DNC), transition to comfort measures only (CMO), and hospital discharge to home using univariable and multivariable analysis (adjusted for age, sex, race/ethnicity, insurance carrier, pre-admission cardiac disorder, ABI type, admission Glasgow Coma Scale Score, mechanical ventilation, and intracranial pressure [ICP] monitoring). Results: The final sample comprised 11,822 patients: AIS (46.7%), sICH (18.5%), SAH (14.8%), and TBI (20.0%). A total of 63% (n = 7472) received cardiac workup, which increased over nine years (p < 0.001). A cardiac investigation was associated with increased age, male sex (aOR 1.16 [1.07, 1.27]), non-white ethnicity (aOR), non-commercial insurance (aOR 1.21 [1.09, 1.33]), pre-admission cardiac disorder (aOR 1.21 [1.09, 1.34]), mechanical ventilation (aOR1.78 [1.57, 2.02]) and ICP monitoring (aOR1.68 [1.49, 1.89]). Compared to AIS, sICH (aOR 0.25 [0.22, 0.29]), SAH (aOR 0.36 [0.30, 0.43]), and TBI (aOR 0.19 [0.17, 0.24]) patients were less likely to receive cardiac investigation. Patients with troponin 25th-50th quartile (aOR 1.65 [1.10-2.47]), troponin 50th-75th quartile (aOR 1.79 [1.22-2.63]), troponin >75th quartile (aOR 2.18 [1.49-3.17]), BNP 50th-75th quartile (aOR 2.86 [1.28-6.40]), BNP >75th quartile (aOR 4.54 [2.09-9.85]), prolonged QTc (aOR 3.41 [2.28; 5.30]), and EF < 40% (aOR 2.47 [1.07; 5.14]) were more likely to be DNC. Patients with troponin 50th-75th quartile (aOR 1.77 [1.14-2.73]), troponin >75th quartile (aOR 1.81 [1.18-2.78]), and prolonged QTc (aOR 1.71 [1.39; 2.12]) were more likely to be associated with a transition to CMO. Patients with prolonged QTc (aOR 0.66 [0.58; 0.76]) were less likely to be discharged home. Conclusions: This large, single-center study demonstrates low rates of cardiac evaluations in TBI, SAH, and sICH compared to AIS. However, there are strong associations between electrocardiography, biomarkers of cardiac injury and heart failure, and echocardiography findings on clinical outcomes in patients with ABI. Findings need validation in a multicenter cohort.

Utility of Routine Surveillance Head Computed Tomography After Receiving Therapeutic Anticoagulation in Patients with Acute Traumatic Intracranial Hemorrhage.

INTRODUCTION: Patients with traumatic intracranial hemorrhage (tICH) are at increased risk of venous thromboembolism and may require anticoagulation. We evaluated the utility of surveillance computed tomography (CT) in patients with tICH who required therapeutic anticoagulation. METHODS: This single institution, retrospective study included adult patients with tICH who required anticoagulation within 4 weeks and had a surveillance head CT within 24 hours of reaching therapeutic anticoagulation levels. The primary outcome was hematoma expansion (HE) detected by the surveillance CT. Secondary outcomes included 1) changes in management in patients with HE on the surveillance head CT, 2) HE in the absence of clinical changes, and 3) mortality due to HE. We also compared mortality between patients who did and did not have a surveillance CT. RESULTS: Of 175 patients, 5 (2.9%) were found to have HE. Most (n = 4, 80%) had changes in management including anticoagulation discontinuation (n = 4), reversal (n = 1), and operative management (n = 1). Two patients developed symptoms or exam changes prior to the head CT. Of the 3 patients (1.7%) without preceding exam changes, each had only very minor HE and did not require operative management. No patient experienced mortality directly attributed to HE. There was no difference in mortality between patients who did and those who did not have a surveillance scan. CONCLUSIONS: Our findings suggest that most patients with tICH who are started on anticoagulation could be followed clinically, and providers may reserve CT imaging for patients with changes in exam/symptoms or those who have a poor clinical examination to follow.

Clinical Experience With a Dedicated Neurocritical Care Quality Improvement Program in an Academic Medical Center.

Background Managing neurocritical care patients encompasses many complex challenges, necessitating specialized care and continuous quality improvement efforts. In recent years, the focus on enhancing patient outcomes in neurocritical care may have led to the development of dedicated quality improvement programs. These programs are designed to systematically evaluate and refine care practices, aligning them with the latest clinical guidelines and research findings. Objective To describe the structure, processes, and outcomes of a dedicated Neurocritical Care Quality Improvement Program (NCC-QIP) at Harborview Medical Center, United States; a quaternary academic medical center, level I trauma, and a comprehensive stroke center. Materials and methods We describe the development of the NCC-QIP, its structure, function, challenges, and evolution. We examine our performance with several NCC-QI quality measures as proposed by the Joint Commission, the American Association of Neurology, and the Neurocritical Care Society, self-reported quality improvement (QI) concerns and QI initiatives undertaken because of the information obtained during our event/measure reporting process for patients admitted between 1/1/2014 and 06/30/2023. Results The NCC-QI reviewed data from 20,218 patients; mean age 57.9 (standard deviation 18.1) years, 56% (n=11,326) males, with acute ischemic stroke (AIS; 22.3%, n=4506), spontaneous intracerebral hemorrhage (ICH; 14.8%, n=2,996), spontaneous subarachnoid hemorrhage (SAH; 8.9%, n=1804), and traumatic brain injury (TBI; 16.6%, n=3352) among other admissions, 37.4% (n=7,559) were mechanically ventilated, and 13.6% (n=2,753) received an intracranial pressure monitor. The median intensive care unit length of stay was two days (Quartile 1-Quartile 3: 2-5 days), and the median hospital length of stay was seven days (Quartile 1-Quartile 3: 3-14 days); 53.9% (n=10,907) were discharged home while 11.4% (2,309) died. The three most commonly reported QI concerns were related to care coordination/communication/handoff (40.4%, n=283), medication-related concerns (14.9%, n=104), and equipment/devices-related concerns (11.7%, n=82). Hospital-acquired infections were in the form of ventilator-associated pneumonia (16.3%, n=419/2562), ventriculostomy catheter-associated infections (4%, n=102/2246), and deep venous thrombosis/pulmonary embolism (3.2%, n=647). The quality metrics documentation was as follows: nimodipine after SAH (99.8%, 1802/1810), Hunt and Hess score (36%, n=650/1804), and ICH score (58.4% n=1752/2996). In comparison, 72% (n=3244/4506) of patients with AIS had a documented National Institute of Health Stroke Scale. Admission Glasgow Coma Score was recorded in 99% of patients with SAH, ICH, and TBI. Educational modules were implemented in response to event reporting. Conclusion A dedicated NCC-QIP can be successfully implemented at a quaternary medical medical center. It is possible to monitor and review a large volume of neurocritical care patients, The three most reported NCC-QI concerns may be related to care coordination-communication/handoff, medication-related concerns, and equipment/devices-related complications. The documentation of illness severity scores and stroke measures depends upon the type of measure and ability to reliably and accurately abstract and can be challenging. The quality improvement process can be enhanced by educational modules that reinforce quality and safety.

External Ventricular Drain Placement, Critical Care Utilization, Complications, and Clinical Outcomes after Spontaneous Subarachnoid Hemorrhage: A Single-Center Retrospective Cohort Study.

BACKGROUND: To examine the association between external ventricular drain (EVD) placement, critical care utilization, complications, and clinical outcomes in hospitalized adults with spontaneous subarachnoid hemorrhage (SAH). METHODS: A single-center retrospective study included SAH patients 18 years and older, admitted between 1 January 2014 and 31 December 2022. The exposure variable was EVD. The primary outcomes of interest were (1) early mortality (<72 h), (2) overall mortality, (3) improvement in modified-World Federation of Neurological Surgeons (m-WFNSs) grade between admission and discharge, and (4) discharge to home at the end of the hospital stay. We adjusted for admission m-WFNS grade, age, sex, race/ethnicity, intraventricular hemorrhage, aneurysmal cause of SAH, mechanical ventilation, critical care utilization, and complications within a multivariable analysis. We reported adjusted odds ratios (aORs) and 95% confidence intervals (CI). RESULTS: The study sample included 1346 patients: 18% (n = 243) were between the ages of 18 and 44 years, 48% (n = 645) were between the age of 45-64 years, and 34% (n = 458) were 65 years and older, with other statistics of females (56%, n = 756), m-WFNS I-III (57%, n = 762), m-WFNS IV-V (43%, n = 584), 51% mechanically ventilated, 76% White (n = 680), and 86% English-speaking (n = 1158). Early mortality occurred in 11% (n = 142). Overall mortality was 21% (n = 278), 53% (n = 707) were discharged to their home, and 25% (n = 331) improved their m-WFNS between admission and discharge. Altogether, 54% (n = 731) received EVD placement. After adjusting for covariates, the results of the multivariable analysis demonstrated that EVD placement was associated with reduced early mortality (aOR 0.21 [0.14, 0.33]), an improvement in m-WFNS grade (aOR 2.06 [1.42, 2.99]) but not associated with overall mortality (aOR 0.69 [0.47, 1.00]) or being discharged home at the end of the hospital stay (aOR 1.00 [0.74, 1.36]). EVD was associated with a higher rate of ventilator-associated pneumonia (aOR 2.32 [1.03, 5.23]), delirium (aOR 1.56 [1.05, 2.32]), and a longer ICU (aOR 1.33 [1.29;1.36]) and hospital length of stay (aOR 1.09 [1.07;1.10]). Critical care utilization was also higher in patients with EVD compared to those without. CONCLUSIONS: The study suggests that EVD placement in hospitalized adults with spontaneous subarachnoid hemorrhage (SAH) is associated with reduced early mortality and improved neurological recovery, albeit with higher critical care utilization and complications. These findings emphasize the potential clinical benefits of EVD placement in managing SAH. However, further research and prospective studies may be necessary to validate these results and provide a more comprehensive understanding of the factors influencing clinical outcomes in SAH.

Delirium and Its Associations with Critical Care Utilizations and Outcomes at the Time of Hospital Discharge in Patients with Acute Brain Injury.

Background and Objectives: We analyzed delirium testing, delirium prevalence, critical care associations outcomes at the time of hospital discharge in patients with acute brain injury (ABI) due to acute ischemic stroke (AIS), non-traumatic subarachnoid hemorrhage (SAH), non-traumatic intraparenchymal hemorrhage (IPH), and traumatic brain injury (TBI) admitted to an intensive care unit. Materials and Methods: We examined the frequency of assessment for delirium using the Confusion Assessment Method for the intensive care unit. We assessed delirium testing frequency, associated factors, positive test outcomes, and their correlations with clinical care, including nonpharmacological interventions and pain, agitation, and distress management. Results: Amongst 11,322 patients with ABI, delirium was tested in 8220 (726%). Compared to patients 18-44 years of age, patients 65-79 years (aOR 0.79 [0.69, 0.90]), and those 80 years and older (aOR 0.58 [0.50, 0.68]) were less likely to undergo delirium testing. Compared to English-speaking patients, non-English-speaking patients (aOR 0.73 [0.64, 0.84]) were less likely to undergo delirium testing. Amongst 8220, 2217 (27.2%) tested positive for delirium. For every day in the ICU, the odds of testing positive for delirium increased by 1.11 [0.10, 0.12]. Delirium was highest in those 80 years and older (aOR 3.18 [2.59, 3.90]). Delirium was associated with critical care resource utilization and with significant odds of mortality (aOR 7.26 [6.07, 8.70] at the time of hospital discharge. Conclusions: In conclusion, we find that seven out of ten patients in the neurocritical care unit are tested for delirium, and approximately two out of every five patients test positive for delirium. We demonstrate disparities in delirium testing by age and preferred language, identified high-risk subgroups, and the association between delirium, critical care resource use, complications, discharge GCS, and disposition. Prioritizing equitable testing and diagnosis, especially for elderly and non-English-speaking patients, is crucial for delivering quality care to this vulnerable group.

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.

Neurological Pupillary Index and Disposition at Hospital Discharge following ICU Admission for Acute Brain Injury.

We examined the associations between the Neurological Pupillary Index (NPi) and disposition at hospital discharge in patients admitted to the neurocritical care unit with acute brain injury (ABI) due to acute ischemic stroke (AIS), spontaneous intracerebral hemorrhage (sICH), aneurysmal subarachnoid hemorrhage (SAH), and traumatic brain injury (TBI). The primary outcome was discharge disposition (home/acute rehabilitation vs. death/hospice/skilled nursing facility). Secondary outcomes were tracheostomy tube placement and transition to comfort measures. Among 2258 patients who received serial NPi assessments within the first seven days of ICU admission, 47.7% (n = 1078) demonstrated NPi ≥ 3 on initial and final assessments, 30.1% (n = 680) had initial NPI < 3 that never improved, 19% (n = 430) had initial NPi ≥ 3, which subsequently worsened to <3 and never recovered, and 3.1% (n = 70) had initial NPi < 3, which improved to ≥3. After adjusting for age, sex, admitting diagnosis, admission Glasgow Coma Scale score, craniotomy/craniectomy, and hyperosmolar therapy, NPi values that remained <3 or worsened from ≥3 to <3 were associated with poor outcomes (adjusted odds ratio, aOR 2.58, 95% CI [2.03; 3.28]), placement of a tracheostomy tube (aOR 1.58, 95% CI [1.13; 2.22]), and transition to comfort measures only (aOR 2.12, 95% CI [1.67; 2.70]). Our study suggests that serial NPi assessments during the first seven days of ICU admission may be helpful in predicting outcomes and guiding clinical decision-making in patients with ABI. Further studies are needed to evaluate the potential benefit of interventions to improve NPi trends in this population.

Caseworker Cultural Mediator Involvement in Neurocritical Care for Patients and Families With Non-English Language Preference: A Quality Improvement Project.

Objective To describe Harborview Medical Center's experience with the involvement of caseworker cultural mediators (CCM) for patients requiring neurocritical care. Methods Using univariate and multivariate analysis (model adjusted for age, Glasgow Coma Scale score (GCS), Sequential Organ Failure Assessment (SOFA) Scores, mechanical ventilation, transition to comfort measures only (CMO), and death by neurologic criteria), we examined CCM team members' involvement in the care of Amharic/Cambodian/Khmer/Somali/Spanish/Vietnamese patients admitted to our neurocritical care service between 2014-2022, factors associated with CCM utilization, and changes in CCM utilization after a QI initiative was implemented in 2020 to encourage healthcare providers to consult the CCM team. Results Compared to eligible patients (n=827) who did not receive CCM referral, patients with CCM involvement (n=121) were younger (49 [interquartile range, IQR 38,63] vs. 56 [IQR 42,68] years, p = 0.002), had greater illness severity (admission GCS 8.5 [IQR 3,14] vs. 14 [IQR 7,15], p < 0.001, SOFA scores (5 [IQR 2,8] vs. 4 [IQR2,6], p = 0.007), and more frequently required mechanical ventilation (67% vs. 40%, odds ratio, OR 3.07, 95% CI 2.06,4.64), with higher all-cause mortality (20% vs. 12%, RR 1.83, 95% CI 1.09, 2.95), and with a higher rate of transition to CMO (11.6% vs. 6.2%, OR 2.00, 95% CI 1.03;3.66). The CCM QI initiative was independently associated with increased CCM involvement (aOR 4.22, 95% CI [2.32;7.66]). Overall, 4/10 attempts made by CCMs to reach out to the family to provide support were declined by the family. CCMs reported providing cultural/emotional support (n=96, 79%), end-of-life counseling (n=16, 13%), conflict mediation (n=15, 12.4%), and facilitating goals of care meetings (n=4, 3.3%). Conclusions Among eligible patients, CCM consultations appeared to occur in patients with higher disease severity. Our QI initiative increased CCM involvement.

Anesthesiology Performance Improvement and Reporting Exchange (ASPIRE) Quality Metrics in Patients Undergoing Decompressive Craniectomy and Endoscopic Clot Evacuation after Spontaneous Supratentorial Intracerebral Hemorrhage: A Retrospective Observational Study.

BACKGROUND: We report adherence to 6 Anesthesiology Performance Improvement and Reporting Exchange (ASPIRE) quality metrics (QMs) relevant to patients undergoing decompressive craniectomy or endoscopic clot evacuation after spontaneous supratentorial intracerebral hemorrhage (sICH). METHODS: In this retrospective observational study, we describe adherence to the following ASPIRE QMs: acute kidney injury (AKI-01); mean arterial pressure < 65 mm Hg for less than 15 minutes (BP-03); myocardial injury (CARD-02); treatment of high glucose (> 200 mg/dL, GLU-03); reversal of neuromuscular blockade (NMB-02); and perioperative hypothermia (TEMP-03). RESULT: The study included 95 patients (70% male) with median (interquartile range) age 55 (47 to 66) years and ICH score 2 (1 to 3) undergoing craniectomy (n=55) or endoscopic clot evacuation (n=40) after sICH. In-hospital mortality attributable to sICH was 23% (n=22). Patients with American Society of Anesthesiologists physical status class 5 (n=16), preoperative reduced glomerular filtration rate (n=5), elevated cardiac troponin (n=21) and no intraoperative labs with high glucose (n=71), those who were not extubated at the end of the case (n=62) or did not receive a neuromuscular blocker given (n=3), and patients having emergent surgery (n=64) were excluded from the analysis for their respective ASPIRE QM based on predetermined ASPIRE exclusion criteria. For the remaining patients, the adherence to ASPIRE QMs were: AKI-01, craniectomy 34%, endoscopic clot evacuation 1%; BP-03, craniectomy 72%, clot evacuation 73%; CARD-02, 100% for both groups; GLU-03, craniectomy 67%, clot evacuation 100%; NMB-02, clot evacuation 79%, and; TEMP-03, clot evacuation 0% with hypothermia. CONCLUSION: This study found variable adherence to ASPIRE QMs in sICH patients undergoing decompressive craniectomy or endoscopic clot evacuation. The relatively high number of patients excluded from individual ASPIRE metrics is a major limitation.

The Association Between Hypophosphatemia and Lactic Acidosis After Cardiac Surgery With Cardiopulmonary Bypass: A Retrospective Cohort Study.

OBJECTIVES: The clinical significance of hypophosphatemia in cardiac surgery has not been investigated extensively. The aim of this study was to evaluate the association of postoperative hypophosphatemia and lactic acidosis in cardiac surgery patients at the time of intensive care unit (ICU) admission. DESIGN: A retrospective cohort study. SETTING: At a single academic center. PARTICIPANTS: Patients who underwent nontransplant cardiac surgery with cardiopulmonary bypass between August 2009 and December 2020. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Serum phosphate and lactate levels were measured upon ICU admission in patients undergoing nontransplant cardiac surgery with cardiopulmonary bypass. There were 681 patients in the low-phosphate (<2.5 mg/dL) group and 2,579 patients in the normal phosphate group (2.5-4.5 mg/dL). A higher proportion of patients in the low phosphate group (26%; 179 of 681; 95% CI: 23-30) had severe lactic acidosis compared to patients in the normal phosphate group (16%; 417 of 2,579; 95% CI: 15-18). In an unadjusted logistic regression model, patients in the low phosphate group had 1.9-times the odds of having severe lactic acidosis (serum lactate ≥4.0 mmol/L) when compared to patients in the normal phosphate group (95% CI: 1.5-2.3), and still 1.4-times the odds (95% CI: 1.1-1.7) after adjusting for several possible confounders. CONCLUSIONS: Hypophosphatemia is associated with lactic acidosis in the immediate postoperative period in cardiac surgery patients. Future studies will need to investigate it as a potential treatment target for lactic acidosis.

Associations Between Transcranial Doppler Vasospasm and Clinical Outcomes After Aneurysmal Subarachnoid Hemorrhage: A Retrospective Observational Study.

OBJECTIVE: The objective is to examine the relationship between transcranial Doppler cerebral vasospasm (TCD-vasospasm), and clinical outcomes in aneurysmal subarachnoid hemorrhage (aSAH). METHODS: In a retrospective cohort study, using univariate and multivariate analysis, we examined the association between TCD-vasospasm (defined as Lindegaard ratio >3) and patient's ability to ambulate without assistance, the need for tracheostomy and gastrostomy tube placement, and the likelihood of being discharged home from the hospital. RESULTS: We studied 346 patients with aSAH; median age 55 years (Interquartile range IQR 46,64), median Hunt and Hess 3 [IQR 1-5]. Overall, 68.6% (n=238) had TCD-vasospasm, and 28% (n=97) had delayed cerebral ischemia. At hospital discharge, 54.3% (n=188) were able to walk without assistance, 5.8% (n=20) had received a tracheostomy, and 12% (n=42) had received a gastrostomy tube. Fifty-three percent (n=183) were discharged directly from the hospital to their home. TCD-vasospasm was not associated with ambulation without assistance at discharge (adjusted odds ratio, aOR 0.54, 95% 0.19,1.45), tracheostomy placement (aOR 2.04, 95% 0.23,18.43), gastrostomy tube placement (aOR 0.95, 95% CI 0.28,3.26), discharge to home (aOR 0.36, 95% CI 0.11,1.23). CONCLUSION: This single-center retrospective study finds that TCD-vasospasm is not associated with clinical outcomes such as ambulation without assistance, discharge to home from the hospital, tracheostomy, and gastrostomy feeding tube placement. Routine screening for cerebral vasospasm and its impact on vasospasm diagnostic and therapeutic interventions and their associations with improved clinical outcomes warrant an evaluation in large, prospective, case-controlled, multi-center studies.

Impact of COVID-19 on neurocritical care delivery and outcomes in patients with severe acute brain injury - Assessing the initial response in the first US epicenter.

To investigate the pandemic's impact on critically ill patients with neurological emergencies, we compared care metrics and outcomes of patients with severe acute brain injury (SABI) before and during the initial COVID-19 surge at our institution. We included adult patients with SABI during two separate three-month time periods: 'pre-COVID vs COVID'. We further stratified the COVID cohort to characterize outcomes in patients requiring COVID-19 precautions (Patient Under Investigation, 'PUI'). The primary endpoint was in-hospital mortality; secondary endpoints included length of stay (LOS), diagnostic studies performed, time to emergent decompressive craniectomies (DCHC), ventilator management, and end-of-life care. We included 394 patients and found the overall number of admissions for SABI declined by 29 % during COVID (pre-COVID n = 231 vs COVID, n = 163). Our primary outcome of mortality and most secondary outcomes were similar between study periods. There were more frequent extubation attempts (72.1 % vs 76 %) and the mean time to extubation was shorter during COVID (55.5 h vs 38.2 h). The ICU LOS (6.10 days vs 4.69 days) and hospital LOS (15.32 days vs 11.74 days) was shorter during COVID. More PUIs died than non-PUIs (51.7 % vs 11.2 %), but when adjusted for markers of illness severity, this was not significant. We demonstrate the ability to maintain a consistent care delivery for patients with SABI during the pandemic at our institution. PUIs represent a population with higher illness severity at risk for delays in care. Multicenter, longitudinal studies are needed to explore the impact of the pandemic on patients with acute neurological emergencies.

Comparison of PACU length of stay and opioid requirements of patients maintained on buprenorphine or methadone for opioid use disorder.

OBJECTIVES: Literature supporting best practice of perioperative buprenorphine management for opioid use disorder is evolving with more recent studies trending toward maintenance of home dose. To guide treatment protocols at our institution, we evaluated patients taking medication for opioid use disorder (MOUD) undergoing similar surgeries. Patients were maintained on either their outpatient buprenorphine or methadone. PATIENTS AND PARTICIPANTS: Data were collected on 46 patients maintained on buprenorphine MOUD who underwent surgery. A subset of these patients (n = 24) was compared with 24 patients maintained on methadone MOUD, matched on surgical procedure, admission date, age, and sex. DESIGN: This is a retrospective matched control study. SETTING: An academic, tertiary, Level 1 trauma center. MAIN OUTCOME MEASURES: Primary outcomes were post-operative opioid use and post-anesthesia care unit (PACU) length of stay. RESULTS: No significant differences in demographic characteristics, physical status, comorbid psychiatric diagnoses, or illicit substance use history were observed between patient groups. A higher proportion of patients taking methadone was admitted due to infection (41.7 percent vs 16.7 percent, p = 0.031) and underwent nonelective surgery (75.0 percent vs 45.8 percent, p = 0.039). No significant differences were observed between patients taking buprenorphine versus methadone with respect to PACU length of stay, post-operative opioid consumption, time-to-transition to oral opioids, or discharge opioid prescriptions. Patients taking buprenorphine were more likely to receive intravenous lidocaine (25.0 percent vs 0.0 percent, p = 0.031) and ketamine (83.3 percent vs 54.2 percent, p = 0.039). CONCLUSION: Findings from this study support accumulating evidence that patients should be maintained on buprenorphine MOUD throughout the perioperative period.

The Prevalence and Factors Associated with the Prescription of Opioids for Head/Neck Pain after Elective Craniotomy for Tumor Resection/Vascular Repair: A Retrospective Cohort Study.

Background and objective: There is no report of the rate of opioid prescription at the time of hospital discharge, which may be associated with various patient and procedure-related factors. This study examined the prevalence and factors associated with prescribing opioids for head/neck pain after elective craniotomy for tumor resection/vascular repair. Methods: We performed a retrospective cohort study on adults undergoing elective craniotomy for tumor resection/vascular repair at a large quaternary-care hospital. We used univariable and multivariable analysis to examine the prevalence and factors (pre-operative, intraoperative, and postoperative) associated with prescribing opioids at the time of hospital discharge. We also examined the factors associated with discharge oral morphine equivalent use. Results: The study sample comprised 273 patients with a median age of 54 years [IQR 41,65], 173 females (63%), 174 (63.7%) tumor resections, and 99 (36.2%) vascular repairs. The majority (n = 264, 96.7%) received opioids postoperatively. The opiate prescription rates were 72% (n = 196/273) at hospital discharge, 23% (19/83) at neurosurgical clinical visits within 30 days of the procedure, and 2.4% (2/83) after 30 days from the procedure. The median oral morphine equivalent (OME) at discharge use was 300 [IQR 175,600]. Patients were discharged with a median supply of 5 days [IQR 3,7]. On multivariable analysis, opioid prescription at hospital discharge was associated with pre-existent chronic pain (adjusted odds ratio, aOR 1.87 [1.06,3.29], p = 0.03) and time from surgery to hospital discharge (compared to patients discharged within days 1−4 postoperatively, patients discharged between days 5−12 (aOR 0.3, 95% CI [0.15; 0.59], p = 0.0005), discharged at 12 days and later (aOR 0.17, 95% CI [0.07; 0.39], p < 0.001)). There was a linear relationship between the first 24 h OME (p < 0.001), daily OME (p < 0.001), hospital OME (p < 0.001), and discharge OME. Conclusions: This single-center study finds that at the time of hospital discharge, opioids are prescribed for head/neck pain in as many as seven out of ten patients after elective craniotomy. A history of chronic pain and time from surgery to discharge may be associated with opiate prescriptions. Discharge OME may be associated with first 24-h, daily OME, and hospital OME use. Findings need further evaluation in a large multicenter sample. The findings are important to consider as there is growing interest in an early discharge after elective craniotomy.

Development and Validation of ARC, a Model for Anticipating Acute Respiratory Failure in Coronavirus Disease 2019 Patients.

OBJECTIVES: To evaluate factors predictive of clinical progression among coronavirus disease 2019 patients following admission, and whether continuous, automated assessments of patient status may contribute to optimal monitoring and management. DESIGN: Retrospective cohort for algorithm training, testing, and validation. SETTING: Eight hospitals across two geographically distinct regions. PATIENTS: Two-thousand fifteen hospitalized coronavirus disease 2019-positive patients. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Anticipating Respiratory failure in Coronavirus disease (ARC), a clinically interpretable, continuously monitoring prognostic model of acute respiratory failure in hospitalized coronavirus disease 2019 patients, was developed and validated. An analysis of the most important clinical predictors aligns with key risk factors identified by other investigators but contributes new insights regarding the time at which key factors first begin to exhibit aberrency and distinguishes features predictive of acute respiratory failure in coronavirus disease 2019 versus pneumonia caused by other types of infection. Departing from prior work, ARC was designed to update continuously over time as new observations (vitals and laboratory test results) are recorded in the electronic health record. Validation against data from two geographically distinct health systems showed that the proposed model achieved 75% specificity and 77% sensitivity and predicted acute respiratory failure at a median time of 32 hours prior to onset. Over 80% of true-positive alerts occurred in non-ICU settings. CONCLUSIONS: Patients admitted to non-ICU environments with coronavirus disease 2019 are at ongoing risk of clinical progression to severe disease, yet it is challenging to anticipate which patients will develop acute respiratory failure. A continuously monitoring prognostic model has potential to facilitate anticipatory rather than reactive approaches to escalation of care (e.g., earlier initiation of treatments for severe disease or structured monitoring and therapeutic interventions for high-risk patients).

Machine learning approach to predict postoperative opioid requirements in ambulatory surgery patients.

Opioids play a critical role in acute postoperative pain management. Our objective was to develop machine learning models to predict postoperative opioid requirements in patients undergoing ambulatory surgery. To develop the models, we used a perioperative dataset of 13,700 patients (≥ 18 years) undergoing ambulatory surgery between the years 2016-2018. The data, comprising of patient, procedure and provider factors that could influence postoperative pain and opioid requirements, was randomly split into training (80%) and validation (20%) datasets. Machine learning models of different classes were developed to predict categorized levels of postoperative opioid requirements using the training dataset and then evaluated on the validation dataset. Prediction accuracy was used to differentiate model performances. The five types of models that were developed returned the following accuracies at two different stages of surgery: 1) Prior to surgery-Multinomial Logistic Regression: 71%, Naïve Bayes: 67%, Neural Network: 30%, Random Forest: 72%, Extreme Gradient Boost: 71% and 2) End of surgery-Multinomial Logistic Regression: 71%, Naïve Bayes: 63%, Neural Network: 32%, Random Forest: 72%, Extreme Gradient Boost: 70%. Analyzing the sensitivities of the best performing Random Forest model showed that the lower opioid requirements are predicted with better accuracy (89%) as compared with higher opioid requirements (43%). Feature importance (% relative importance) of model predictions showed that the type of procedure (15.4%), medical history (12.9%) and procedure duration (12.0%) were the top three features contributing to model predictions. Overall, the contribution of patient and procedure features towards model predictions were 65% and 35% respectively. Machine learning models could be used to predict postoperative opioid requirements in ambulatory surgery patients and could potentially assist in better management of their postoperative acute pain.

Forecasting a Crisis: Machine-Learning Models Predict Occurrence of Intraoperative Bradycardia Associated With Hypotension.

BACKGROUND: Predictive analytics systems may improve perioperative care by enhancing preparation for, recognition of, and response to high-risk clinical events. Bradycardia is a fairly common and unpredictable clinical event with many causes; it may be benign or become associated with hypotension requiring aggressive treatment. Our aim was to build models to predict the occurrence of clinically significant intraoperative bradycardia at 3 time points during an operative course by utilizing available preoperative electronic medical record and intraoperative anesthesia information management system data. METHODS: The analyzed data include 62,182 scheduled noncardiac procedures performed at the University of Washington Medical Center between 2012 and 2017. The clinical event was defined as severe bradycardia (heart rate <50 beats per minute) followed by hypotension (mean arterial pressure <55 mm Hg) within a 10-minute window. We developed models to predict the presence of at least 1 event following 3 time points: induction of anesthesia (TP1), start of the procedure (TP2), and 30 minutes after the start of the procedure (TP3). Predictor variables were based on data available before each time point and included preoperative patient and procedure data (TP1), followed by intraoperative minute-to-minute patient monitor, ventilator, intravenous fluid, infusion, and bolus medication data (TP2 and TP3). Machine-learning and logistic regression models were developed, and their predictive abilities were evaluated using the area under the ROC curve (AUC). The contribution of the input variables to the models were evaluated. RESULTS: The number of events was 3498 (5.6%) after TP1, 2404 (3.9%) after TP2, and 1066 (1.7%) after TP3. Heart rate was the strongest predictor for events after TP1. Occurrence of a previous event, mean heart rate, and mean pulse rates before TP2 were the strongest predictor for events after TP2. Occurrence of a previous event, mean heart rate, mean pulse rates before TP2 (and their interaction), and 15-minute slopes in heart rate and blood pressure before TP2 were the strongest predictors for events after TP3. The best performing machine-learning models including all cases produced an AUC of 0.81 (TP1), 0.87 (TP2), and 0.89 (TP3) with positive predictive values of 0.30, 0.29, and 0.15 at 95% specificity, respectively. CONCLUSIONS: We developed models to predict unstable bradycardia leveraging preoperative and real-time intraoperative data. Our study demonstrates how predictive models may be utilized to predict clinical events across multiple time intervals, with a future goal of developing real-time, intraoperative, decision support.

Perioperative Opioid Use and Chronic Post-Surgical Pain after Liver Transplantation: A Single Center Observational Study.

OBJECTIVES: To investigate the opioid requirements and prevalence of chronic postsurgical pain (CPSP) in liver transplant (LT) recipients and to evaluate the association of opioid use with postoperative survival. DESIGN: Retrospective analysis. SETTING: A large academic medical center. PATIENTS: Cadaveric liver transplants recipients from 2008 to 2016. INTERVENTIONS: Analysis of demographic, perioperative, and outcome data. MEASUREMENTS AND MAIN RESULTS: This study measured the incidence and quantity of preoperative opioid use, postoperative opioid requirements, the incidence of CPSP, and survival in patients with and without CPSP. Opioid requirements were calculated in morphine milligram equivalents. In total, 322 LT recipients satisfied the inclusion criteria. The cohort of interest included 61 patients (18.9%) who were prescribed opioids before LT, compared to the control group of 261. Postoperative opioid requirements were significantly higher in the cohort of interest in the first 24 hours (205.9 ± 318.5 v 60.4 ± 33.6 mg, p < 0.0001) and at 7 days after transplant (57.0 ± 70.6 mg v 19.2 ± 15.4 mg, p < 0.0001). Incidence of CPSP was significantly higher in the cohort of interest at 3 months (70.5% v 45.5%, p < 0.0001), at 2 years (38% v 12%), and at 5 years (29.8% v 6.9%) postoperatively. CPSP was a significant risk factor for patient mortality after transplantation (p = 0.038, HR 1.26). CONCLUSIONS: Opioid use is relatively frequent in patients waiting for LT. It significantly affects the postoperative opioid requirements and the incidence of CSPS. CPSP may significantly affect survival after LT.