Competency-Based Assessments: Leveraging Artificial Intelligence to Predict Subcompetency Content.

PURPOSE: Faculty feedback on trainees is critical to guiding trainee progress in a competency-based medical education framework. The authors aimed to develop and evaluate a Natural Language Processing (NLP) algorithm that automatically categorizes narrative feedback into corresponding Accreditation Council for Graduate Medical Education Milestone 2.0 subcompetencies. METHOD: Ten academic anesthesiologists analyzed 5,935 narrative evaluations on anesthesiology trainees at 4 graduate medical education (GME) programs between July 1, 2019, and June 30, 2021. Each sentence (n = 25,714) was labeled with the Milestone 2.0 subcompetency that best captured its content or was labeled as demographic or not useful. Inter-rater agreement was assessed by Fleiss' Kappa. The authors trained an NLP model to predict feedback subcompetencies using data from 3 sites and evaluated its performance at a fourth site. Performance metrics included area under the receiver operating characteristic curve (AUC), positive predictive value, sensitivity, F1, and calibration curves. The model was implemented at 1 site in a self-assessment exercise. RESULTS: Fleiss' Kappa for subcompetency agreement was moderate (0.44). Model performance was good for professionalism, interpersonal and communication skills, and practice-based learning and improvement (AUC 0.79, 0.79, and 0.75, respectively). Subcompetencies within medical knowledge and patient care ranged from fair to excellent (AUC 0.66-0.84 and 0.63-0.88, respectively). Performance for systems-based practice was poor (AUC 0.59). Performances for demographic and not useful categories were excellent (AUC 0.87 for both). In approximately 1 minute, the model interpreted several hundred evaluations and produced individual trainee reports with organized feedback to guide a self-assessment exercise. The model was built into a web-based application. CONCLUSIONS: The authors developed an NLP model that recognized the feedback language of anesthesiologists across multiple GME programs. The model was operationalized in a self-assessment exercise. It is a powerful tool which rapidly organizes large amounts of narrative feedback.

An evidence-based oxytocin protocol for the third stage of labor to improve hemorrhage outcomes.

OBJECTIVE: To assess the use of an evidence-based oxytocin protocol for management of the third stage of labor to decrease non-beneficial clinical variation and improve clinical outcomes. DESIGN: This is a cohort study of pregnant patients delivering before implementation of an evidence-based oxytocin protocol compared to patients delivering after implementation of an evidence-based oxytocin protocol. SETTING: A level III maternal care referral hospital with an average delivery volume of approximately 3000 deliveries. PARTICIPANTS: Pregnant patients delivering over a 60-month period from January 2013 to December 2017. INTERVENTION: An evidence-based oxytocin protocol: 3 units of oxytocin administered over 3 minutes, with a second 3-unit bolus if inadequate tone, then oxytocin infusion at 18 units/hour × 1 hour and then 3.6 units/hour for 3 hours. MAIN OUTCOME MEASURES: Postpartum hemorrhage (PPH) rate (EBL ≥500 ml for vaginal and ≥1000 ml for cesarean). RESULTS: Data from 14 603 deliveries were analyzed, 8408 pre-protocol and 6195 post-protocol. We demonstrated a significant decrease in PPH from 5.2% to 2.9% (P < 0.001) and a small but non-significant increase in the transfusion rate from 1.8% to 2.3% (P = 0.11). CONCLUSION: A standardized oxytocin infusion protocol in the third stage of labor resulted in a significant decrease in PPH for both vaginal and cesarean deliveries.

A Case Report of Combined Ultrasound and Fluoroscopic-Guided Percutaneous Radiofrequency Lesioning of the Obturator and Femoral Articular Branches in the Treatment of Persistent Hip Pain in a Pediatric Patient.

Hip denervation comprising radiofrequency lesioning of the obturator and femoral articular branches is used in adults with refractory hip pain who are not surgical candidates. Persistent hip pain occurs infrequently in pediatric patients, and there are limited data on the safety and efficacy of this procedure in a pediatric population. We provide a case report of a successful ultrasound and fluoroscopic-guided hip denervation procedure in an 11-year-old girl with persistent right hip pain after septic arthritis refractory to conservative and surgical management. At an 18-week follow-up, hip denervation provided improvement in pain, mobility, and reduced opioid consumption by 20%.

The Intraoperative Effect of Methadone on Somatosensory Evoked Potentials.

BACKGROUND: Evoked potentials (EP), both somatosensory evoked potentials (SSEP) and transcranial motor evoked potentials (TcMEP), are often used during complex spine surgery to monitor the integrity of spinal pathways during operations in or around the spine. Changes in these monitored EP signals (increased latency and decreased amplitude) may result from ischemia, direct surgical injury, changes in blood pressure, hypoxia, changes in CO2 tension, and anesthetic agents. Typically, a clinically significant change for SSEPs is defined as an increase in latency >10% or a decrease of amplitude >50%. A clinically significant change for TcMEPs is much more complex but is also described in terms of large signal loss or decrease. Opioids have been shown to both increase latency and decrease the amplitude of SSEPs, although this change is usually not clinically significant. There has been a renewed interest in methadone for use in spine and other complex surgeries. However, the effect of methadone on intraoperative monitoring of SSEPs and TcMEPs is unknown. We present the first study to directly look at the effects of methadone on SSEP and TcMEP monitoring during complex spine surgery. METHODS: The goal of this study was to observe the effect of methadone on an unrandomized set of patients. The primary endpoint was methadone's effect on SSEPs, and the secondary endpoint was methadone's effect on TcMEPs. Adult patients undergoing spine surgery requiring intraoperative neuromonitoring were induced with general anesthesia and had a baseline set of SSEPs and TcMEPs recorded. Next, methadone dosed 0.2 mg/kg/lean body weight was given. Repeat SSEPs and TcMEPs were recorded at 5, 10, and 15 minutes, with the timing based on distribution half-life of methadone between 6 and 8 minutes. Postoperatively, adverse events from methadone administration were collected. RESULTS: There was a statistically significant difference found in SSEPs for N20 latency (95% confidence interval [CI], 0.17-0.53; P=0.028), P37 latency (95% CI, 0.65-1.25; P=0.001), and N20 amplitude (95% CI, 0.09-0.32; P=<0.001), but not for P37 amplitude (95% CI, -0.19 to 0.00; P=0.634). There was no significant effect found for TcMEPs, the secondary endpoint of the study, and there were minimal adverse events recorded postoperatively. CONCLUSIONS: The data demonstrate that a single intravenous dose of methadone has a statistically significant difference on the amplitude and latency of SSEPs. However, this statistical difference does not translate into a clinical significance.

A Case Report of Post-Operative Jöd-Basedow Phenomennon Following Oral and IV Iodine Contrast Administration.

This is a case of thyrotoxicosis, due to the Jöd-Basedow phenomenon following administration of oral and IV iodinated contrast in a patient with history of gastrointestinal stromal tumor (GIST) and small bowel obstruction. The patient developed atrial fibrillation and had an extended stay in the intensive care unit. Given the aging population with possible subclinical hyperthyroidism, multinodular goiter, and the rise in contrast administration for routine diagnostic studies, this case serves to raise awareness of the risks of "routine" tests administered to our aging patient population.