Ambulatory surgery is growing in popularity worldwide. For example, 50-70% of surgical procedures in North America are performed on an ambulatory basis. Use of Bispectral Index (BIS) monitoring for titration of general anesthesia may allow use of less anesthetics, reduction in side effects, and faster patient recovery.
MEDLINE and other databases were searched for randomized controlled trials examining the use of BIS monitoring versus standard practice in ambulatory surgery patients. Outcomes were extracted from these articles, and a meta-analysis was performed.
One thousand three hundred eighty subjects from 11 trials were included in the meta-analysis. The use of BIS monitoring significantly reduced anesthetic consumption by 19%, reduced the incidence of nausea/vomiting (32% vs. 38%; odds ratio, 0.77), and reduced time in the recovery room by 4 min. However, these benefits did not result in significant reduction in time until patient discharge from the ambulatory surgery unit. Cost analysis using pooled costs to reflect North America, Europe, and Asia indicated that use of BIS monitoring increased the cost per patient by 5.55 US dollars because of the cost of BIS electrodes.
The use of BIS monitoring modestly reduced anesthetic consumption, risk of nausea and vomiting, and recovery room time. These benefits did not reduce time spent in the ambulatory surgery unit, and cost of the BIS electrode exceeded any cost savings.
AMBULATORY surgery is rapidly growing worldwide. For example, more than 21 million ambulatory surgical procedures are annually performed in the United States alone,1and 50–70% of all surgical procedures in North America are performed on an ambulatory basis.2,3A primary goal for ambulatory anesthesia is rapid recovery from anesthesia leading to rapid patient discharge with minimal side effects. Reduction of time spent in the postanesthesia care unit (PACU) and ambulatory surgery unit (ASU) may be economically advantageous, especially if the PACU can be bypassed altogether.2Despite their greater cost, short-acting anesthetic agents (e.g. , propofol, desflurane) have become popular for ambulatory anesthesia to facilitate rapid patient recovery. In addition to pharmacology of anesthetic agents, surveys have demonstrated that postoperative pain and nausea are common reasons for delayed patient discharge, unplanned hospital admission, and patient concern.3,4Therefore, enhanced ability to titrate general anesthetics and minimize side effects may be useful for facilitating patient recovery from ambulatory surgical procedures.
The Bispectral Index (BIS) is a derived electroencephalographic parameter that has been extensively validated as a monitor for depth of anesthesia.5BIS values in the range of 40–60 are recommended as indicators of adequate depth of general anesthesia and may be used to guide titration of general anesthesia instead of routine clinical signs. Ability to consistently administer the adequate amount of general anesthetics may be especially useful in the ambulatory anesthesia setting to minimize side effects (nausea and vomiting), PACU time, and ASU time. Several individual randomized controlled trials (RCTs) have been performed but with conflicting results as to whether BIS monitoring improves ambulatory patient care.6,7Therefore, we performed this meta- and cost analysis to examine the effects of BIS monitoring on anesthetic use, incidence of nausea and vomiting, duration of PACU stay, and time to patient discharge in ambulatory anesthesia.
Materials and Methods
The National Library of Medicine’s MEDLINE database, the American College of Physicians Journal Club, the Cochrane Central Register of Controlled Trials, the Cochrane Database of Systematic Reviews, and the Database of Abstracts of Reviews of Effects were searched for the time period 1966 to January 2004. No language restrictions were used. MESH terms Ambulatory surgical procedures (n = 6619), Anesthesia recovery period (n = 2424), and Length of stay (n = 28,244) were combined with the term OR . Text word Bispectral (n = 513), MESH term electroencephalography (n = 73,346), and text word BIS (n = 26,732) were combined with the term OR . The two searches were combined with the term AND and were limited by Human and Clinical trials (n = 71).
All of the above abstracts were reviewed for potential inclusion in the meta-analysis. All RCTs that stated the use of hospitalized patients were excluded. All RCTs that compared BIS monitoring to standard clinical practice without BIS monitoring in ambulatory surgery with short-acting agents were included. RCTs that did not explicitly state the enrollment of only ambulatory surgery patients were individually considered before any statistical analyses. RCTs from this group were included if the surgical procedures were those commonly performed in ASUs or if short-acting anesthetics were used. No minimum sample sizes were invoked for inclusion of studies in the analysis, and only studies on adults (aged ≥ 18 yr) were included. After selecting the initial articles, the reference list of each of the analyzed articles was checked for any additional studies, as were the author’s personal files for additional references that met all inclusion criteria.
Data Extraction and Analysis
Each study’s methodology and results were recorded. Data were extrapolated from figures as needed. Wherever possible, data were converted to incidence for dichotomous outcomes and to mean and SD (normal distribution was assumed) for continuous outcomes. Definition of outcomes and criteria for PACU and ASU discharge were recorded as originally defined by the study. Specific outcomes that were extracted included global anesthetic consumption as a ratio of the BIS group: control group (e.g. , 1 minimum alveolar concentration [MAC] hour of desflurane required for BIS group vs. 2 MAC hours for control = anesthetic use ratio of 0.5), incidences of nausea or vomiting, incidence of ability to bypass the PACU, time in the PACU, and time until ASU discharge.
Cost data for general anesthetics were pooled from recent publications examining costs of short-acting anesthetic agents (desflurane, propofol, remifentanil, and sevoflurane) in Europe, North America, and Asia for a variety of ambulatory surgical procedures.8–10The average cost of anesthetic agents per case was 26 US dollars (USD). Cost data for prevention of nausea and vomiting were pooled from recent publications examining a variety of antiemetic agents and strategies (dexamethasone, dolasetron, droperidol, and ondansetron) in Europe and North America.11–14The average incremental cost for one additional nausea- and vomiting-free patient was 40.49 USD. PACU costs were pooled from recent publications from North America.11,15,16The average cost for each PACU minute (labor and supplies) was 0.65 USD/min. Costs for BIS monitoring only included the cost of the electrode. The capital cost of BIS platforms was not included because multiple manufacturers offer multiple types of BIS monitors with multiple financial arrangements. Our current (January 2004) institutional cost for a BIS electrode (17.50 USD for a BIS® Quattro 186-0106; Aspect Medical Systems, Newton, MA) was pooled with reported costs from recent publications originating from North America and Europe for an average cost of 15.48 USD.17,18
The meta-analysis was performed with a random effects model. The level of significance for all tests was set at a P value of 0.05, and variances were not assumed to be equal. For dichotomous outcomes, study results were pooled and odds ratios were calculated with the Mantel-Haenszel method. Therefore, odds ratios with 95% confidence intervals are displayed for effect statistic. For continuous outcomes, study results were pooled and means and SDs were calculated with the inverse variance method. Therefore, weighted mean differences and 95% confidence intervals are displayed for effect statistic. All statistical analyses were performed with Review Manager 4.2 (The Cochrane Collaboration’s Information Management System; Nordic Cochrane Centre Rigshospitalet, Copenhagen, Denmark).
Eleven RCTs enrolling 1,380 subjects met inclusion criteria (table 1).6,7,19–27Five studies were included that did not explicitly state that all patients underwent ambulatory surgery but did use short-acting anesthetics or surgical procedures commonly performed in ASUs.20–23,25Results of the meta-analysis are summarized in table 2. BIS monitoring reduced anesthetic use by 19%, reduced the risk of postoperative nausea and vomiting (32% vs. 38%; odds ratio, 0.77 [95% confidence interval, 0.56–0.99]), and reduced the duration of time in the PACU (45.2 vs. 49.1 min). BIS monitoring did not affect ability to bypass the PACU (32% vs. 29%) and did not reduce the duration of time in the ASU (159 vs. 165 min). Cost analysis indicated a net additional cost of 5.55 USD for each patient with BIS monitoring (table 3).
The use of BIS monitoring consistently reduced anesthetic use by approximately 19% compared with standard clinical practice. This finding is in accord with well-described dose–response effects for general anesthetics for different endpoints of hypnosis, movement, and hemodynamic response to surgery. Increasingly large doses of general anesthesia are required to provide hypnosis (e.g. , 0.4 MAC), immobility (e.g. , 1 MAC), and hemodynamic stability during surgery (e.g. , 1.5 MAC).28Because standard clinical practice typically includes titration to immobility or hemodynamic stability, achievement of these endpoints may easily result in a relative overdose for hypnosis. The BIS has been validated as a measure of hypnosis and anesthetic depth,5and the ability to directly titrate general anesthetics to a single endpoint of only hypnosis well explains the decreased use of anesthetics and slight reduction in cost.
The use of BIS monitoring has been proposed to provide even more valuable benefits, such as reduced adverse effects from general anesthetics, increased ability to bypass the PACU, and reduced time spent in the PACU and the ASU. BIS monitoring did reduce the risk of postoperative nausea and vomiting, probably via reduction in use of general anesthetics. This may be a worthwhile outcome in itself because surveys indicate that nausea and vomiting is the primary patient concern for the postoperative period29–31and that patients are willing to pay 100 USD out of pocket to avoid this complication.29However, the use of BIS monitoring only slightly reduced the incidence of nausea and vomiting (−6%) with a number needed to treat of 17. This magnitude of effect compares poorly with most pharmacologic antiemetic agents (ondansetron, dexamethasone, and droperidol) that have numbers needed to treat ranging from 4 to 7.29Perhaps more importantly, nausea and vomiting are major factors for prolonged PACU and ASU stay,3,4and reduced risk of nausea and vomiting coupled with decreased anesthetic use might be expected to reduce PACU and ASU stay.
Bispectral Index monitoring marginally reduced time in the PACU by 4 min and, in a purely linear system, might result in a cost savings of 2.6 USD/patient. However, even this small cost savings will be highly variable, depending on the practice and staffing patterns of each individual ASU.32More likely, day-to-day variation in clinical work and staffing will result in minimal cost savings from this small decrease in PACU time. The inability of BIS monitoring to dramatically reduce time in the PACU may reflect diminishing returns from current ambulatory anesthesia practice. A previous study examining factors affecting ability to bypass the PACU and reduced duration of stay in the PACU suggested that use of short-acting anesthetic agents and explicit processes of care to facilitate PACU bypass and reduce PACU time were key factors.2All of the RCTs from which PACU and ASU durations were extracted for this meta-analysis already incorporated at least some of these features, and the small reductions in anesthetic use and risk of nausea and vomiting may not have had sufficient ability to further improve clinical care. The inability of BIS monitoring to dramatically hasten PACU recovery probably explains the inability to reduce the duration of ASU stay with BIS monitoring in this meta-analysis. Post hoc power analysis (power = 0.8, P = 0.05) indicates sufficient subjects in this meta-analysis to detect a 10-min difference in ASU time; therefore, inclusion of additional subjects would be unlikely to change this result. One potential source of bias for analysis of ASU time is that 465 of 682 subjects were enrolled from one large study.7
Our cost analysis used pooled cost averages for anesthetic agents, incremental cost to reduce postoperative nausea and vomiting, and PACU costs from multiple countries to provide a global perspective. Overall, the use of BIS monitoring for ambulatory anesthesia is economically inefficient. The minor reductions in anesthetic consumption, prevention of postoperative nausea and vomiting, and PACU time were exceeded by the cost for BIS monitoring consumables, without even including capital costs of monitoring systems. Adding in any additional capital cost for the BIS monitoring platform would further increase the cost per patient.
Limitations of this meta-analysis reflect the rapidly changing environment of ambulatory anesthesia. Improving algorithms for prophylactic and rescue treatment of nausea and vomiting may further reduce the benefit of BIS monitoring, because prophylactic antiemetics were specified in only three RCTs.27,29,33Approximately only half of the subjects in this meta-analysis were enrolled in protocols that allowed PACU bypass, and increasing comfort with this process may change results. Because phase 1 PACU costs and care are typically the most intensive, the ability to bypass this portion of patient recovery may be the most efficient means to reduce ASU time and costs.2Discharge criteria from PACUs and ASUs are continually evolving.3Many of the included RCTs required oral intake and voiding before discharge, and adoption of more lenient discharge criteria may also affect applicability of this meta-analysis. Cost data were pooled from recent publications to reflect typical costs for all countries. Specific cost effects will vary from region to region and will depend on anesthetic cost, BIS costs, surgical scheduling, PACU staffing, and ASU staffing.
In conclusion, the use of BIS monitoring modestly to marginally reduced anesthetic consumption, risk of nausea and vomiting, and PACU time. These benefits did not reduce ASU time, and BIS monitoring resulted in a net cost of approximately 5.55 USD/patient.
The author thanks Paul I. Liu, M.D., Ph.D. (Professor and Chief, Department of Pathology, Olive View-UCLA Medical Center, Sylmar, California), for Japanese language translation and consultation.