Search

Increasing attention has been focused on health care expenditures, which include anesthetic-related drug costs. Using data from 2 large academic medical centers, we sought to identify significant contributors to anesthetic drug cost variation.Using anesthesia information management systems, we calculated volatile and intravenous drug costs for 8 types of inpatient surgical procedures performed from July 1, 2009, to December 31, 2011. For each case, we determined patient age, American Society of Anesthesiologists (ASA) physical status, gender, institution, case duration, in-room provider, and attending anesthesiologist. These variables were then entered into 2 fixed-effects linear regression models, both with logarithmically transformed case cost as the outcome variable. The first model included duration, attending anesthesiologist, patient age, ASA physical status, and patient gender as independent variables. The second model included case type, institution, patient age, ASA physical status, and patient gender as independent variables. When all variables were entered into 1 model, redundancy analyses showed that case type was highly correlated (R = 0.92) with the other variables in the model. More specifically, a model that included case type was no better at predicting cost than a model without the variable, as long as that model contained the combination of attending anesthesiologist and case duration. Therefore, because we were interested in determining the effect both variables had on cost, 2 models were created instead of 1. The average change in cost resulting from each variable compared to the average cost of the reference category was calculated by first exponentiating the β coefficient and subtracting 1 to get the percent difference in cost. We then multiplied that value by the mean cost of the associated reference group.A total of 5504 records were identified, of which 4856 were analyzed. The median anesthetic drug cost was $38.45 (25th percentile = $23.23, 75th percentile = $63.82). The majority of the variation was not described by our models-35.2% was explained in the model containing case duration, and 32.3% was explained in the model containing case type. However, the largest sources of variation our models identified were attending anesthesiologist, case type, and procedure duration. With all else held constant, the average change in cost between attending anesthesiologists ranged from a cost decrease of $41.25 to a cost increase of $95.67 (10th percentile = -$19.96, 90th percentile = +$20.20) when compared to the provider with the median value for mean cost per case. The average change in cost between institutions was significant but minor ($5.73).The majority of the variation was not described by the models, possibly indicating high per-case random variation. The largest sources of variation identified by our models included attending anesthesiologist, procedure type, and case duration. The difference in cost between institutions was statistically significant but was minor. While many prior studies have found significant savings resulting from cost-reducing interventions, our findings suggest that because the overall cost of anesthetic drugs was small, the savings resulting from interventions focused on the clinical practice of attending anesthesiologists may be negligible, especially in institutions where access to more expensive drugs is already limited. Thus, cost-saving efforts may be better focused elsewhere.

Background: The allocation of intensive care unit (ICU) beds for postoperative patients is a challenging daily task that could be assisted by the real-time detection of ICU needs. The goal of this study was to develop and validate an intraoperative predictive model for unplanned postoperative ICU use. Methods: With the use of anesthesia information management system, postanesthesia care unit, and scheduling data, a data set was derived from adult in-patient noncardiac surgeries. Unplanned ICU admissions were identified (4,847 of 71,996; 6.7%), and a logistic regression model was developed for predicting unplanned ICU admission. The model performance was tested using bootstrap validation and compared with the Surgical Apgar Score using area under the curve for the receiver operating characteristic. Results: The logistic regression model included 16 variables: age, American Society of Anesthesiologists physical status, emergency case, surgical service, and 12 intraoperative variables. The area under the curve was 0.905 (95% CI, 0.900–0.909). The bootstrap validation model area under the curves were 0.513 at booking, 0.688 at 3 h before case end, 0.738 at 2 h, 0.791 at 1 h, and 0.809 at case end. The Surgical Apgar Score area under the curve was 0.692. Unplanned ICU admissions had more ICU-free days than planned ICU admissions (5 vs. 4; P < 0.001) and similar mortality (5.6 vs. 6.0%; P = 0.248). Conclusions: The authors have developed and internally validated an intraoperative predictive model for unplanned postoperative ICU use. Incorporation of this model into a real-time data sniffer may improve the process of allocating ICU beds for postoperative patients.