Anesthesia Information Management System Data Research Articles

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

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. 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. 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. 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.

Incidence of Artifacts and Deviating Values in Research Data Obtained from an Anesthesia Information Management System in Children.

Vital parameter data collected in anesthesia information management systems are often used for clinical research. The validity of this type of research is dependent on the number of artifacts. In this prospective observational cohort study, the incidence of artifacts in anesthesia information management system data was investigated in children undergoing anesthesia for noncardiac procedures. Secondary outcomes included the incidence of artifacts among deviating and nondeviating values, among the anesthesia phases, and among different anesthetic techniques. We included 136 anesthetics representing 10,236 min of anesthesia time. The incidence of artifacts was 0.5% for heart rate (95% CI: 0.4 to 0.7%), 1.3% for oxygen saturation (1.1 to 1.5%), 7.5% for end-tidal carbon dioxide (6.9 to 8.0%), 5.0% for noninvasive blood pressure (4.0 to 6.0%), and 7.3% for invasive blood pressure (5.9 to 8.8%). The incidence of artifacts among deviating values was 3.1% for heart rate (2.1 to 4.4%), 10.8% for oxygen saturation (7.6 to 14.8%), 14.1% for end-tidal carbon dioxide (13.0 to 15.2%), 14.4% for noninvasive blood pressure (10.3 to 19.4%), and 38.4% for invasive blood pressure (30.3 to 47.1%). Not all values in anesthesia information management systems are valid. The incidence of artifacts stored in the present pediatric anesthesia practice was low for heart rate and oxygen saturation, whereas noninvasive and invasive blood pressure and end-tidal carbon dioxide had higher artifact incidences. Deviating values are more often artifacts than values in a normal range, and artifacts are associated with the phase of anesthesia and anesthetic technique. Development of (automatic) data validation systems or solutions to deal with artifacts in data is warranted.

Influencing Anesthesia Provider Behavior Using Anesthesia Information Management System Data for Near Real-Time Alerts and Post Hoc Reports

Influencing Anesthesia Provider Behavior Using Anesthesia Information Management System Data for Near Real-Time Alerts and Post Hoc Reports

Influencing Anesthesia Provider Behavior Using Anesthesia Information Management System Data for Near Real-Time Alerts and Post Hoc Reports.

In this review article, we address issues related to using data from anesthesia information management systems (AIMS) to deliver near real-time alerts via AIMS workstation popups and/or alphanumeric pagers and post hoc reports via e-mail. We focus on reports and alerts for influencing the behavior of anesthesia providers (i.e., anesthesiologists, anesthesia residents, and nurse anesthetists). Multiple studies have shown that anesthesia clinical decision support (CDS) improves adherence to protocols and increases financial performance through facilitation of billing, regulatory, and compliance documentation; however, improved clinical outcomes have not been demonstrated. We inform developers and users of feedback systems about the multitude of concerns to consider during development and implementation of CDS to increase its effectiveness and to mitigate its potentially disruptive aspects. We discuss the timing and modalities used to deliver messages, implications of outlier-only versus individualized feedback, the need to consider possible unintended consequences of such feedback, regulations, sustainability, and portability among systems. We discuss statistical issues related to the appropriate evaluation of CDS efficacy. We provide a systematic review of the published literature (indexed in PubMed) of anesthesia CDS and offer 2 case studies of CDS interventions using AIMS data from our own institution illustrating the salient points. Because of the considerable expense and complexity of maintaining near real-time CDS systems, as compared with providing individual reports via e-mail after the fact, we suggest that if the same goal can be accomplished via delayed reporting versus immediate feedback, the former approach is preferable. Nevertheless, some processes require near real-time alerts to produce the desired improvement. Post hoc e-mail reporting from enterprise-wide electronic health record systems is straightforward and can be accomplished using system-independent pathways (e.g., via built-in e-mail support provided by the relational database management system). However, for some of these enterprise-wide systems, near real-time data access, necessary for CDS that generates concurrent alerts, has been challenging to implement.

Increased Mean Time from End of Surgery to Operating Room Exit in a Historical Cohort of Cases with Prolonged Time to Extubation

Prolonged time to extubation has been defined as the occurrence of a ≥ 15-minute interval from the end of surgery to removal of the tracheal tube. We quantified the increases in the mean times from end of surgery to exit from the OR associated with prolonged extubations and tested whether the increases were economically important (≥ 5 minutes). Anesthesia information management system data from 1 tertiary hospital were collected from November 2005 through December 2012 (i.e., sample sizes were N = 22 sequential quarters). Cases were excluded in which the patient's trachea was not intubated or extubated while physically in the operating room (OR). For each combination of stratification variable (below) and quarter, the mean time from end of surgery to OR exit was calculated for the extubations that were not prolonged and for those that were prolonged. Results are reported as mean ± SEM, with "at least" denoting the lower 95% confidence interval. The mean times from end of surgery to OR exit were at least 12.6 minutes longer for prolonged extubations when calculated with stratification by duration of surgery and prone or other positioning (13.0 ± 0.1 minutes), P < 0.0001 compared to 5 minutes (i.e., times were substantively long economically). The mean times were at least 11.7 minutes longer when calculated stratified by anesthesia procedure code (12.4 ± 0.4, P < 0.0001) and at least 11.3 minutes longer when calculated stratified by surgeon (12.4 ± 0.6, P < 0.0001). We recommend that anesthesia providers document the times of extubations and monitor the incidence of prolonged extubations as an economic measure. This would be especially important for providers at facilities with many ORs that have at least 8 hours of cases and turnovers.

Estimate of the Relative Risk of Succinylcholine for Triggering Malignant Hyperthermia

Facilities with volatile anesthetic agents stock dantrolene for the treatment of malignant hyperthermia (MH). The availability of dantrolene at these facilities satisfies cost-utility norms even for sites with as few as 1 anesthetic per workday, based on the overall incidence of MH per anesthetic. We considered the stocking of dantrolene at facilities with succinylcholine alone (i.e., where volatile anesthetics are not available), by using registry data and estimates of the frequency of administration of succinylcholine during anesthesia. We determine the magnitude of the relative risk of the administration of succinylcholine for triggering MH. The relative risk of triggering MH by succinylcholine versus volatile agents was calculated using data from 2 sources. The ratio of the number of cases of MH among patients receiving succinylcholine to number among patients not receiving succinylcholine was estimated from the previously published cohort of 284 cases of MH from the North American MH Registry of the MH Association of the United States (MHAUS). The percentage of anesthetics with succinylcholine was estimated using anesthesia information management system data from a typical North American hospital comprising tertiary operating rooms, obstetrics unit, ambulatory surgical center, and endoscopy and radiological suites. The relative risk of MH with versus without succinylcholine was 19.6 (lower 95% confidence limit > 16.1). Limiting to cases with volatile anesthetics, the relative risk was 9.1 (>7.5). Both relative risks exceed 1.0 (P < 0.0001). Because more than half of the reported cases of MH included the use of succinylcholine, the relative risk exceeded 1.0 provided fewer than half of anesthetics in North America included the use of succinylcholine. The incidences of succinylcholine use at the hospital were 5.8% and 11.6% for all anesthetics and for anesthetics with volatile agents, respectively. Our results provide no insight into the triggering mechanism for MH (i.e., succinylcholine could in isolation have an extremely low incidence of inducing MH, yet markedly increase the risk when administered in combination with volatile anesthetics). Until more epidemiologic data are collected and analyzed, having dantrolene available, where succinylcholine may be used, is reasonable, and this practice should be maintained.

Systematic Criteria for Type and Screen Based on Procedure’s Probability of Erythrocyte Transfusion

Background At many hospitals, the type and screen decision is guided by the hospital's maximum surgical blood order schedule, a document that includes for each scheduled (elective) surgical procedure a recommendation of whether a preoperative type and screen be performed. There is substantial heterogeneity in the scientific literature for how that decision should be made. Methods Anesthesia information management system data were retrieved from the 160,207 scheduled noncardiac cases in adults of 1,253 procedures at a hospital. Results Neither assuming a Poisson distribution of mean erythrocyte units transfused, nor grouping rare procedures into larger groups based on their anesthesia Current Procedural Terminology code, was reliable. In contrast, procedures could be defined to have minimal estimated blood loss (less than 50 ml) based on low incidence of transfusion and low incidence of the hemoglobin being checked preoperatively. Among these procedures, when the lower 95% confidence limit for erythrocyte transfusion was less than 5%, type and screen was shown to be unnecessary. The method was useful based on including multiple differences from the hospital's maximum surgical blood order schedule and clinicians' test ordering (greater than or equal to 29% fewer type and screen). Results were the same with a Bayesian random effects model. Conclusions We validated a method to determine procedures on the maximum surgical blood order schedule for which type and screen was not indicated using the estimated blood losses and incidences of transfusion.

THIS MONTH IN Anesthesiology

The concept that the μ-opioid receptor may be involved in tumor progression is discussed.This lecture promotes pursuing a redesign of perioperative healthcare.This study suggests that exploration of the μ-opioid receptor in non-small cell lung carcinoma merits further study both as a diagnostic and therapeutic option.Anesthetic- or surgical-related etiologies of perioperative nerve injury are reviewed.Unbundling of consciousness, connectedness, and responsiveness is critical for the future development of anesthesia.Previous studies have suggested interactions between opioid genetics, tumor growth, and clinical effects of morphine. In this long term study, 2,039 patients with breast cancer were genotyped and followed for clinical outcomes. Patients genotyped with the A118G μ-opioid receptor polymorphism had less breast cancer–specific mortality at 10 yr. Breast cancer–specific mortality was significantly reduced in patients with one or more copies of the G allele (P &lt; 0.001) for invasive cases. Mortality was significantly decreased in patients with A/G and G/G phenotypes compared with A/A (P = 0.006). This study supports further investigation between opioid genetics, pain management, and cancer survival.There is conflicting literature regarding a patient's need for preoperative blood typing and screening. This study reviewed anesthesia information management system data from 160,207 scheduled noncardiac adult cases at a single hospital. Procedures could be defined to have minimal estimated blood loss (&lt; 50 ml) based on low incidence of transfusion and low incidence of hemoglobin being checked preoperatively. Type and screen was unnecessary when the confidence intervals for erythrocyte transfusion were less than 5%. These data may allow development of a new method for determining which patients require preoperative type and screen. See the accompanying Editorial View on page 749 Use of perioperative epidural anesthesia and analgesia during cancer resections may avoid immunosuppressive recurrence triggers. Using the Medicare-Surveillance, Epidemiology, and End Results database, the authors compared survival and cancer recurrence rates for resection of colorectal cancer in patients who received perioperative epidural anesthesia and analgesia to those who did not. Of 42,151 patients, 5-yr survival was similar between groups (61% and 55% in the epidural group and the nonepidural groups, respectively). Although there was a significant association between epidural use and improved survival (hazard ratio = 0.91), after adjustment for covariates there was no significant reduction of recurrence in this group (odds ratio = 1.05). This study does not support an association between perioperative epidural use and decreased cancer recurrence in patients with nonmetastatic colorectal cancer undergoing resection.

Journal-related Activities and Other Special Activities at the 2011 American Society of Anesthesiologists Annual Meeting

Journal-related Activities and Other Special Activities at the 2011 American Society of Anesthesiologists Annual Meeting

The Design and Implementation of an Automated System for Logging Clinical Experiences Using an Anesthesia Information Management System

Residents in anesthesia training programs throughout the world are required to document their clinical cases to help ensure that they receive adequate training. Current systems involve self-reporting, are subject to delayed updates and misreported data, and do not provide a practicable method of validation. Anesthesia information management systems (AIMS) are being used increasingly in training programs and are a logical source for verifiable documentation. We hypothesized that case logs generated automatically from an AIMS would be sufficiently accurate to replace the current manual process. We based our analysis on the data reporting requirements of the American College of Graduate Medical Education (ACGME). We conducted a systematic review of ACGME requirements and our AIMS record, and made modifications after identifying data element and attribution issues. We studied 2 methods (parsing of free text procedure descriptions and CPT4 procedure code mapping) to automatically determine ACGME case categories and generated AIMS-based case logs and compared these to assignments made by manual inspection of the anesthesia records. We also assessed under- and overreporting of cases entered manually by our residents into the ACGME website. The parsing and mapping methods assigned cases to a majority of the ACGME categories with accuracies of 95% and 97%, respectively, as compared with determinations made by 2 residents and 1 attending who manually reviewed all procedure descriptions. Comparison of AIMS-based case logs with reports from the ACGME Resident Case Log System website showed that >50% of residents either underreported or overreported their total case counts by at least 5%. The AIMS database is a source of contemporaneous documentation of resident experience that can be queried to generate valid, verifiable case logs. The extent of AIMS adoption by academic anesthesia departments should encourage accreditation organizations to support uploading of AIMS-based case log files to improve accuracy and to decrease the clerical burden on anesthesia residents.

Meta-Analysis of Average and Variability of Time to Extubation Comparing Isoflurane with Desflurane or Isoflurane with Sevoflurane

We recently determined how to use anesthesia information management system data to model the time from end of surgery to extubation. We applied that knowledge for meta-analyses of trials comparing extubation times after maintenance with desflurane and sevoflurane. In this study, we repeated the meta-analyses to compare isoflurane with desflurane and sevoflurane. A Medline search through December 2009 was used to identify studies with (1) humans randomly assigned to isoflurane or desflurane groups without other differences (e.g., induction drugs) between groups, and (2) mean and SD reported for extubation time and/or time to follow commands. The search was repeated for random assignment to isoflurane or sevoflurane groups. We considered extubation times >15 minutes (representing 15% of cases in the anesthesia information management system data) to be prolonged. Desflurane reduced the mean extubation time by 34% and reduced the variability in extubation time by 36% relative to isoflurane. These reductions would reduce the incidence of prolonged extubation times by 95% and 97%, respectively. Sevoflurane reduced the mean extubation time by 13% and reduced the SD by 8.7% relative to isoflurane. These reductions would reduce the incidence of prolonged extubation times by 51% and 35%, respectively. The pharmacoeconomics of volatile anesthetics are highly sensitive to measurement of relatively small time differences. Therefore, surgical facilities should use these values combined with their local data (e.g., mean baseline extubation times) when making evidence-based management decisions regarding pharmaceutical purchases and usage guidelines.

Operating room utilization: information management systems

Advances during the past year in operational decision making using information management systems data have been predominantly in better understanding of how to allocate operating room time based on operating room efficiency, not just operating room utilization. Each quarter, operating room allocations based on maximizing operating room efficiency can be calculated with 6-10 months of data. Analysis can be performed using operating room or anesthesia information management system data. Uncertainty in knowing the operating rooms in which cases were performed (i.e. in knowing turnover times) has little effect on operating room allocations or efficiency. It is important, however, that when a service has filled its allocated operating room time and has a new case to schedule, the case be scheduled into the operating room time of another service. The service whose operating room time is released can be the service with the largest difference between allocated and scheduled operating room time at the time when the new case is scheduled. If optimal operating room allocations and case scheduling based on maximizing operating room efficiency are not implemented, the resulting increase in anesthesia group labor costs can be used in negotiations as a statistically sound estimate for the increased labor cost to the group. During the last couple of years there has been continual advancement in our understanding of how to analyze operating room information system data based on operating room efficiency. Work has expanded from operating room allocations to case scheduling, releasing allocated operating room time, and addressing implementation challenges including poor data quality or lack of interest in change.