Operational outcomes of propofol sedation versus fentanyl, midazolam and diphenhydramine sedation for endoscopies and colonoscopies at an academic medical center.

BACKGROUND: On July 1st, 2021, the University of Colorado Hospital (UCH) implemented new sedation protocols in the luminal gastrointestinal (GI) suite. GI proceduralist supervised, Nurse Administered Sedation with fentanyl, midazolam, and diphenhydramine (NAS) sedation was transitioned to Monitored Anesthesia Care with propofol under physician anesthesiologist supervision (MAC). OBJECTIVE: To determine if there are statistically significant reductions in Sedation-Start to Scope-In time (SSSI) when using Monitored Anesthesia Care with propofol (MAC) versus Nurse Administered Sedation with fentanyl, midazolam, and diphenhydramine (NAS). Secondary objectives were to determine if statistically significant improvements to other operational times, quality measures, and satisfaction metrics were present. METHOD: This study was a retrospective analysis of a natural experiment resultant of a change from NAS to MAC sedation protocols. Outcomes for NAS protocols from 1/1/21-6/30/21 were compared to outcomes of MAC protocols from the dates 8/1/21-10/31/21. Results were analyzed using Quasi-Poisson regression analysis and stratified based on upper GI, lower GI, and combined procedures. Patient demographic data including age, biological sex, comorbidities, and BMI, were adjusted for in the analysis. ASA matching was not performed as nursing sedation does not use ASA classifications. Pre-anesthesia co-morbidities were assessed via evaluation of a strict set of comorbidities abstracted from the electronic medical record. Perioperative operational outcomes include Sedation Start to Scope-In (SSSI), In-Room to Scope-In Time (IRSI), Scope Out to Out of Room (SOOR), Total Case Length (TCL), and Post Anesthesia Care Unit Length of Stay (PACU LOS). Quality outcomes include PACU Administered Medications (PAM), and Clinician Satisfaction Scores (CSS). RESULTS: A total of 5,582 gastrointestinal (GI) endoscopic cases (upper, lower, and combined endoscopies) were observed. Statistically significant decreases in SSSI of 2.5, 2.1, and 2.2 minutes for upper, lower, and dual GI procedures were observed when using MAC protocols. A statistically significant increase in satisfaction scores of 47.0 and 19.6 points were observed for nurses and proceduralists, respectively, when using MAC. CONCLUSION: MAC protocols for endoscopic GI procedures at UCH led to statistically significant decreases in the time required to complete procedures thus increasing operational efficiency.

Calcium scoring in low-dose ungated chest CT scans using convolutional long-short term memory networks.

We aimed to develop a novel deep-learning based method for automatic coronary artery calcium (CAC) quantification in low-dose ungated computed tomography attenuation correction maps (CTAC). In this study, we used convolutional long-short -term memory deep neural network (conv-LSTM) to automatically derive coronary artery calcium score (CAC) from both standard CAC scans and low-dose ungated scans (CT-attenuation correction maps). We trained convLSTM to segment CAC using 9543 scans. A U-Net model was trained as a reference method. Both models were validated in the OrCaCs dataset (n=32) and in the held-out cohort (n=507) without prior coronary interventions who had CTAC standard CAC scan acquired contemporarily. Cohen's kappa coefficients and concordance matrices were used to assess agreement in four CAC score categories (very low: <10, low:10-100; moderate:101-400 and high >400). The median time to derive results on a central processing unit (CPU) was significantly shorter for the conv-LSTM model- 6.18s (inter quartile range [IQR]: 5.99, 6.3) than for UNet (10.1s, IQR: 9.82, 15.9s, p<0.0001). The memory consumption during training was much lower for our model (13.11Gb) in comparison with UNet (22.31 Gb). Conv-LSTM performed comparably to UNet in terms of agreement with expert annotations, but with significantly shorter inference times and lower memory consumption.