A Machine Learning Application to Classify Patients at Differing Levels of Risk of Opioid Use Disorder: Clinician-Based Validation Study.

Background: Despite restrictive opioid management guidelines, opioid use disorder (OUD) remains a major public health concern. Machine learning (ML) offers a promising avenue for identifying and alerting clinicians about OUD, thus supporting better clinical decision-making regarding treatment. Objective: This study aimed to assess the clinical validity of an ML application designed to identify and alert clinicians of different levels of OUD risk by comparing it to a structured review of medical records by clinicians. Methods: The ML application generated OUD risk alerts on outpatient data for 649,504 patients from 2 medical centers between 2010 and 2013. A random sample of 60 patients was selected from 3 OUD risk level categories (n=180). An OUD risk classification scheme and standardized data extraction tool were developed to evaluate the validity of the alerts. Clinicians independently conducted a systematic and structured review of medical records and reached a consensus on a patient's OUD risk level, which was then compared to the ML application's risk assignments. Results: A total of 78,587 patients without cancer with at least 1 opioid prescription were identified as follows: not high risk (n=50,405, 64.1%), high risk (n=16,636, 21.2%), and suspected OUD or OUD (n=11,546, 14.7%). The sample of 180 patients was representative of the total population in terms of age, sex, and race. The interrater reliability between the ML application and clinicians had a weighted kappa coefficient of 0.62 (95% CI 0.53-0.71), indicating good agreement. Combining the high risk and suspected OUD or OUD categories and using the review of medical records as a gold standard, the ML application had a corrected sensitivity of 56.6% (95% CI 48.7%-64.5%) and a corrected specificity of 94.2% (95% CI 90.3%-98.1%). The positive and negative predictive values were 93.3% (95% CI 88.2%-96.3%) and 60.0% (95% CI 50.4%-68.9%), respectively. Key themes for disagreements between the ML application and clinician reviews were identified. Conclusions: A systematic comparison was conducted between an ML application and clinicians for identifying OUD risk. The ML application generated clinically valid and useful alerts about patients' different OUD risk levels. ML applications hold promise for identifying patients at differing levels of OUD risk and will likely complement traditional rule-based approaches to generating alerts about opioid safety issues.

Post-Hospitalization Home Monitoring Programs During the COVID-19 Pandemic: Survey Results from the Hospital Medicine Re-engineering Network (HOMERuN).

BACKGROUND: During the coronavirus disease 2019 (COVID-19) pandemic, hospitals and healthcare systems launched innovative responses to emerging needs. The creation and use of programs to remotely follow patient clinical status and recovery after COVID-19 hospitalization has not been thoroughly described. OBJECTIVE: To characterize deployment of remote post-hospital discharge monitoring programs during the COVID-19 pandemic METHODS: Electronic surveys were administered to leaders of 83 US academic hospitals in the Hospital Medicine Re-engineering Network (HOMERuN). An initial survey was completed in March 2021 with follow-up survey completed in July 2022. RESULTS: There were 35 responses to the initial survey (42%) and 15 responses to the follow-up survey (43%). Twenty-two (63%) sites reported a post-discharge monitoring program, 16 of which were newly developed for COVID-19. Physiologic monitoring devices such as pulse oximeters were often provided. Communication with medical teams was often via telephone, with moderate use of apps or electronic medical record integration. Programs launched most commonly between January and June 2020. Only three programs were still active at the time of follow-up survey. CONCLUSIONS: Our findings demonstrate rapid, ad hoc development of post-hospital discharge monitoring programs during the COVID-19 pandemic but with little standardization or evaluation. Additional study could identify the benefits of these programs, instruct their potential application to other disease processes, and inform further development as part of emergency preparedness for upcoming crises.

A Patient-Centered Environmental Scan of Inpatient Visitor Policies During the COVID-19 Pandemic.

Researchers and patients conducted an environmental scan of policy documents and public-facing websites and abstracted data to describe COVID-19 adult inpatient visitor restrictions at 70 academic medical centers. We identified variations in how centers described and operationalized visitor policies. Then, we used the nominal group technique process to identify patient-centered information gaps in visitor policies and provide key recommendations for improvement. Recommendations were categorized into the following domains: 1) provision of comprehensive, consistent, and clear information; 2) accessible information for patients with limited English proficiency and health literacy; 3) COVID-19 related considerations; and 4) care team member methods of communication.

Duplex Ultrasound Surveillance After Uncomplicated Endovascular Abdominal Aortic Aneurysm Repair.

OBJECTIVES: Health-care costs and risks of radiation and intravenous contrast exposure challenge computed tomography angiography (CTA) as the standard surveillance method after endovascular abdominal aortic aneurysm repair (EVAR). We reviewed our experience using Duplex ultrasound scan (DUS) as an initial and subsequent surveillance technique after uncomplicated EVAR. METHODS: The medical records of patients who underwent EVAR from 2004 to 2014 with at least 1 postoperative imaging study were retrospectively reviewed. Duplex ultrasound scan was the primary modality, with CTA reserved for patients with suspicious findings. RESULTS: Mean follow-up was 3.2 years for 266 patients. Fifty-seven endoleaks (7 type I, 50 type II) were detected in 51 patients (19%). Nineteen (33%) endoleaks were identified and monitored by DUS alone. Nine (16%) endoleaks were identified on CTA without prior DUS. Twenty-two (39%) endoleaks were identified on DUS and confirmed by CTA; 6 of these patients had a secondary intervention. When compared to subsequent CTA, there were 7 discordant results: 4 false-negative and 3 false-positive endoleaks on DUS. Two of these patients with discordant results required intervention. Follow-up CTA was not obtained for the other 2 patients due to severe comorbidities including renal disease. One of these patients eventually developed abdominal aortic aneurysm rupture and death. Among 88 patients with both DUS and CTA, positive predictive value and negative predictive value for DUS were 0.88 and 0.94, respectively. Sac size on DUS compared to CTA resulted in an interclass correlation coefficient of r = .84. CONCLUSIONS: In our experience, DUS was safe and effective for initial and follow-up surveillance after uncomplicated EVAR.

Computed tomography for the diagnosis and management of abdominal aortic aneurysms.

Computed tomography (CT) is useful in the detection and diagnosis of abdominal aortic aneurysms (AAA). Rupture risk can be assessed by accurately measuring diameter, tortuosity, thrombus extent, and wall stress. CT can aid in accurately determining anatomic variants as well as AAA etiology. Evaluation for surgical intervention is made by close examination of AAA morphology and specific anatomic features.

Carotid-subclavian arterial reconstruction: concomitant ipsilateral carotid endarterectomy increases risk of perioperative stroke.

BACKGROUND: Carotid-subclavian bypass (CSB) and carotid-subclavian transposition (CST) have excellent long-term patency with low perioperative mortality and morbidity. Carotid endarterectomy (CEA) is necessary for severe ipsilateral internal carotid artery stenosis in a small subset of these patients. CEA can be performed as a combined or separate procedure. This study was undertaken to delineate the results of CSB and CST at our institution and to determine if concomitant CEA with CSB or CST is safe. METHODS: We evaluated the outcome of 36 patients with symptomatic subclavian artery stenosis treated surgically at a single institution during a 22-year period. Outcomes of patients undergoing CSB or CST with concomitant CEA were compared with those of patients undergoing CSB or CST alone. Available literature was reviewed to compare the rate of perioperative stroke following CSB or CST with concomitant CEA versus CSB or CST alone. RESULTS: Twenty-one patients underwent CST and 15 patients underwent CSB. There were 2 (5.6%) deaths and 2 (5.6%) strokes within 30 days of surgery. Concomitant CEA was performed in 6 CST patients and 2 CSB patients. Both perioperative strokes occurred in patients who had concomitant CEA. There were no strokes in the CST or CSB alone group (P = .044). In a collected review of 12 evaluable studies plus our experience, the rate of perioperative stroke was 0.32% in 617 patients who underwent CSB or CST alone versus 4.73% in 148 patients who had concomitant CEA with CSB or CST (P < .001). CONCLUSIONS: Both CSB and CST are safe and effective for symptomatic subclavian artery stenosis, with excellent long-term results. In patients also requiring CEA, the rate of perioperative stroke is significantly higher with a combined procedure. Consideration should be given to performing CEA separately from CSB or CST.