| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
In this systematic review we focused on postoperative recovery and complications using four different anesthetic techniques. The database MEDLINE was searched via PubMed (1966 to June 2002) using the search words "anesthesia" and with ambulatory surgical procedures limited to randomized controlled trials in adults (>19 yr), in the English language, and in humans. A second search strategy was used combining two of the words "propofol," "isoflurane," "sevoflurane," or "desflurane". Screening and data extraction produced 58 articles that were included in the final meta-analysis. No differences were found between propofol and isoflurane in early recovery. However, early recovery was faster with desflurane compared with propofol and isoflurane and with sevoflurane compared with isoflurane. A minor difference was found in home readiness between sevoflurane and isoflurane (5 min) but not among the other anesthetics. Nausea, vomiting, headache, and postdischarge nausea and vomiting incidence were in favor of propofol compared with isoflurane (P < 0.05). A larger number of patients in the inhaled anesthesia groups required antiemetics compared with the propofol group. We conclude that the differences in early recovery times among the different anesthetics were small and in favor of the inhaled anesthetics. The incidence of side effects, specifically postoperative nausea and vomiting, was less frequent with propofol. IMPLICATIONS: A systematic analysis of the literature comparing postoperative recovery after propofol, isoflurane, desflurane, and sevoflurane-based anesthesia in adults demonstrated that early recovery was faster in the desflurane and sevoflurane groups. The incidence of nausea and vomiting were less frequent with propofol.
Ambulatory surgery has increased rapidly in the last 10 yr, and the availability of new minimally invasive surgical techniques has resulted in an increased emphasis on the expansion of day surgery. This has also been made possible by the safety of ambulatory surgery in sick patients and in those undergoing major surgery. Together with the advancement in surgical techniques has been the availability of newer and better drugs with short onset and duration of effect, resulting in quick recovery and the possibility of earlier discharge from the day surgical unit. Specifically, the advantages of IV anesthesia using propofol over inhaled anesthesia have been intensively discussed as the subject of numerous studies with opposing results. The introduction of less-soluble inhaled anesthetics, desflurane and sevoflurane, has added a new dimension to recovery and fast-tracking (13) by allowing more rapid recovery and earlier discharge home. However, these anesthetics are associated with increased cost compared with older inhaled anesthetics and may be associated with an increased incidence of side effects and complications (4,5). Additionally, Dexter and Tinker (6) concluded that the theoretical advantages of one of these anesthetics, desflurane, did not translate into rapid recovery from anesthesia partly because patients received other drugs that may have tended to equalize differences between these anesthetics. The manner in which the anesthetic drugs are delivered, including the use of concomitant medications, may therefore influence the optimal choice of anesthetic used to achieve early discharge, and an understanding of the risks and benefits of the anesthetics could assist the practitioner to determine which anesthetic to use in clinical practice. Meta-analysis has increasingly been used over the last 10 yr as one of several ways to assess the efficacy of newer drugs. The aim of this systematic review was to assess whether maintenance of anesthesia using propofol infusion, isoflurane, sevoflurane, or desflurane is associated with faster recovery and fewer side effects during ambulatory surgery in adults.
We searched MEDLINE (from 1966 to June 2002) via PubMed using the MeSH terms "ambulatory surgical procedure" or "outpatients." This was combined with the MeSH term "anesthesia" using the AND function. The results were limited to randomized controlled trials in humans, in the English language, and in the adult (>19 yr of age) patient population. We soon realized that we had missed a number of articles. Therefore, we did another search using the words "propofol," "sevoflurane," "desflurane," and "isoflurane" as text words, which produced more results. Finally, we searched the reference lists of the already published systematic reviews for any missing articles. A hand search was also done through the reference lists of included studies to further identify any articles that had been missed using earlier strategies. The following criteria were used in identifying appropriate studies that could be included in the analysis. The studies had to be randomized with the specific primary end-point of assessment of "early" recovery, "intermediate" recovery, and side effects or complications. The results had to be presented as means (and a measure of variance but not range) or numbers (yes or no) for the study to be acceptable for meta-analysis. Studies where the authors had compared inhaled anesthetics during "monitored anesthesia care" were excluded, as were studies in which halothane or enflurane was used for maintenance of anesthesia. Whenever an inhaled anesthetic was used for induction of anesthesia, the study (or the group that was induced by an inhaled anesthetic) was excluded. Finally, studies where the results were presented as graphs or histograms without digital data were also excluded. Because of the vast number of tests used for the assessment of psychomotor recovery in different studies, as well as their limited clinical significance, these tests were not included in our systematic review. Data extraction was done by two independent (primary) reviewers who were unaware of each others results (AG and TS or NS and RZ). Data included aspects of patient characteristics, anesthesia, type of surgery, "early" recovery ("time to open eyes," "time to obey commands"), "intermediate" recovery ("time to transfer from Phase 1 to Phase II," "home readiness," and "home discharge"), minor complications in the PACU or within 24 h including "pain," "nausea or vomiting," "antiemetics" used, "dizziness/giddiness," "drowsiness/somnolence," "headache," "shivering," or "coughing". Because authors had used different terminology for recovery indices, we came to an agreement that the following terms would be treated as measuring the same end-point: time to open eyes and time to awakening; time to obey commands and time to orientation or squeezing fingers. Time to "emergence" was considered too abstract to be included, and this variable was therefore excluded from the analysis. A differentiation was made between "home readiness" (when the patient was ready to be sent home) and "home discharge" (when the patient was actually sent home). Data on post-discharge nausea and vomiting (PDNV) were excluded when authors did not differentiate between nausea and vomiting. Extracted data were compared for agreement between the primary reviewers by a secondary reviewer (SP). Any discrepancies were noted and discussed further to come to a consensus among all authors.
Data extracted from the relevant studies were entered into the program RevMan 4.1 (Review Manager, Cochrane Collaboration, UK) and analyzed. For dichotomous data, the relative risk (RR) with the corresponding 95% confidence intervals (CI) were calculated for each study, and the results were pooled together using the Mantel-Haenszel Method for combining trials. For continuous data, the weighted mean difference (WMD) and its corresponding standard error (SE) were calculated. The individual effect sizes were weighted according to the reciprocal of their variance. Subsequently, the inverse variance method was used to pool the WMD. The overall estimate of pooled effect was calculated. Heterogeneity was determined under the assumption (null hypothesis) that there are no differences in treatment effect between trials. For combined dichotomous data, a fixed effect model was used when there was no significant heterogeneity (P > 0.05), otherwise a random effect model was used. For continuous data, the weighted means and their corresponding 95% CI were used. Results are presented as WMD and 95% CI for continuous data, and RR and 95% CI for dichotomous data. Whenever data were presented as mean and SE of mean, the latter was converted to standard deviation (SD) using the formula: SD =
Literature Review The MeSH term "anesthesia" revealed 127,165 articles, and "ambulatory surgical procedure" revealed 6017 articles. Combining these two terms and limiting the results to adults (>19 yr), in humans, and in English revealed 932 articles. The second search strategy using the text word "propofol" and combining it with "isoflurane" revealed 449 articles, with "sevoflurane" revealed 170 articles and with "desflurane" revealed 107 articles when limited to adults (>19 yr), in humans, and in English. Hand searching all these articles, and reviewing reference lists from previous publications including previously published systematic reviews revealed 42 articles that had extractable data. In a final search strategy, combining the MeSH terms "ambulatory surgical procedures" or "outpatient procedures" revealed 13,104 articles. The MeSH term "anesthetics, inhalation" revealed 30,538 articles. Combining these strategies using the AND function revealed 302 articles. Limiting these to adults (>19 yr), humans, and the English language revealed 180 articles. A hand search of these revealed 16 articles that could be finally included in the meta-analysis.
Propofol Versus Isoflurane
Propofol Versus Desflurane There were 13 studies (2,5,8,2534) with extractable data that were included in the meta-analysis (Table 2). Time to eye opening was significantly quicker with desflurane compared with propofol (1.3 min), as was the time to obey commands (1.3 min), with significant heterogeneity between the anesthetics (Table 5). No differences were found in home readiness or discharge between the anesthetics. The RR for postoperative complications including PON (NNT, 7) and POV (NNT, 10) was significantly greater with desflurane compared with propofol and the need for antiemetics was also increased with desflurane (RR, 3.9) (Table 5). No other differences were seen between the anesthetics.
Propofol Versus Sevoflurane There were 11 studies including 821 patients with extractable data (2,33,3543) (Table 3). No difference was found in the time to eye opening between sevoflurane and propofol, but time to obeying commands was quicker in the sevoflurane group (1.6 min), with significant heterogeneity between anesthetics (Table 5). No significant difference was found in the time to home-readiness between the anesthetics, but with significant heterogeneity. The time to home discharge was earlier with propofol compared with sevoflurane (10.3 min). The RR for postoperative complications including PON (NNT, 11) and POV (NNT, 16) was significantly greater with sevoflurane compared with propofol but with significant heterogeneity (Table 5). The need for antiemetics in the postoperative period was significantly increased with sevoflurane (RR, 4.9). There were no other significant differences between the anesthetics.
Isoflurane Versus Desflurane Four studies compared isoflurane and desflurane in the ambulatory setting (8,4446) and included 277 patients (Table 4). A statistically significant difference was found in "time to eye opening" (P < 0.004) and "time to obey commands" (P < 0.01) but in no other variable of recovery (Table 6). The WMD in early recovery between desflurane and isoflurane were modest (45 min) and in favor of desflurane. Sore throat was reported in only one study, and the incidence was more frequent in the desflurane compared with the isoflurane group (P < 0.05). No other differences were found between the anesthetics.
Isoflurane Versus Sevoflurane Six studies could be included with relevant data examining a total of 634 patients undergoing a variety of ambulatory surgical procedures (4752) (Table 4). Statistically significant differences were found in the "time to opening eyes," "time to obeying commands," "time to transfer from Phase 1 to Phase II recovery," "home readiness" (P < 0.00001), and "home discharge" (P = 0.05). The results of the latter are, however, based on only two studies. The WMD in all the recovery indices (early and intermediate recovery) between sevoflurane and isoflurane were small and in favor of sevoflurane (Table 6). Drowsiness was significantly more frequent with isoflurane compared with sevoflurane in the postoperative period (P = 0.03) but no other significant differences were found in the incidence of postoperative complications.
Sevoflurane Versus Desflurane
IV Versus Inhaled Anesthetics
In this systematic analysis of the literature, we found that early recovery, characterized by opening eyes and obeying commands, was statistically significantly different, but only marginally quicker, in desflurane and sevoflurane compared with isoflurane or propofol groups. Intermediate recovery characterized by home readiness was slightly earlier in the sevoflurane group compared with isoflurane group alone. Postoperative complications, specifically PON and POV, were significantly less frequent in the propofol group compared with isoflurane, sevoflurane, and desflurane groups. No significant difference in early recovery was found when isoflurane or propofol was used for the maintenance of anesthesia. Although we found significant differences in early recovery between sevoflurane/desflurane compared with propofol/isoflurane in favor of the former, the magnitude of these differences was small (<5 min) and therefore of doubtful clinical relevance even in a busy ambulatory unit. The small differences between these anesthetics were seen following strict protocols and not allowing stepwise reduction in anesthetic concentration towards the end of surgery, which is normal clinical practice. Despite the low blood:gas partition coefficient of desflurane and sevoflurane with the theoretical advantage of rapid recovery from anesthesia, these anesthetics have provided only marginally better early recovery characteristics compared with propofol or isoflurane. One explanation for this could be that in most studies a combination of drugs is used. Therefore, residual effects of drugs used for premedication, opiates, and muscle relaxants may interact with the anesthetics to delay the onset of early recovery. Another reason for these minor differences could be that the depth of anesthesia at the end of the operation was unknown. Most studies have used clinical assessment of anesthetic depth based on hemodynamic responses to pain during anesthesia and therefore it is easy to err on maintaining patients deeply anesthetized, which in turn affects recovery. Indeed, recent studies using bispectral index (BIS) as a guide to anesthetic depth have shown that a large number of patients can be "fast-tracked" when anesthetic depth is monitored (4). Consequently, appropriate depth of anesthesia is an important factor in quick recovery after ambulatory surgery, a factor that has not been used to advantage in most published studies. Our analysis demonstrated that sevoflurane was associated with an earlier home readiness (5 min) compared with isoflurane, with no differences among the other comparison of inhaled anesthetics or against propofol. However, propofol was associated with earlier (1015 min) home discharge compared with sevoflurane/isoflurane. Similarly, sevoflurane was associated with a difference of 25 min in home discharge when compared with isoflurane. However, the latter findings were based on two studies with wide CI, and borderline statistical significance. In interpreting differences in home readiness and discharge, it is critical to understand the factors influencing the interpretation of each outcome. "Home readiness" is the time when patients are ready to be discharged home as assessed by standardized methods such as the modified postanesthesia discharge scoring system, whereas "home discharge" is the actual time when the patient could leave the hospital. The latter is affected by many non-medical factors such as absence of an adult to accompany the patient home or waiting to meet the doctor before going home. Unfortunately, many authors have not made a distinction between these variables and it sometimes remains unclear which variable is actually being measured. Other factors such as duration and type of surgery and local hospital practices and routines may affect home discharge and may sometimes be more important than the choice of the anesthetic used. For instance, some hospitals use local anesthetics routinely into surgical wounds, which may reduce postoperative narcotic requirements, which in turn may affect discharge. Other hospitals require patients to empty their bladders before being considered home ready, which may also affect discharge times. Rapid early recovery may also be associated with a greater appreciation of pain in the early postoperative period, which increases postoperative analgesic requirement and thus delays recovery, a finding seen by Robinson et al. (56) in their meta-analysis of sevoflurane versus propofol. Recovery profiles can also be influenced by concomitant analgesics used preoperatively and the depth of anesthesia attained towards the end of the operation. For example, differences in recovery and home discharge have been found when using alfentanil or fentanyl as analgesics (57). Some authors have suggested that using BIS for depth of anesthesia monitoring may enhance recovery (4). However, using a standardized anesthetic regimen and strict discharge criteria, Ahmad et al. (58) showed that BIS monitoring does not have a significant effect on the ability to fast track outpatients. In addition, although more patients may be fast track eligible using desflurane or sevoflurane compared with propofol for maintenance of anesthesia (2), the number of patients who actually bypass the PACU is much smaller because of anesthetic-related factors such as residual sedation (34). All these confounding factors need to be considered when making a case for the use of a specific anesthetic for the maintenance of anesthesia in ambulatory surgical patients. Postoperative nausea and vomiting (PONV) were significantly more common with inhaled anesthetics compared with propofol, and the use of antiemetics was also more common with inhaled anesthetics. The incidence of PDNV was also less frequent with propofol compared with isoflurane but not desflurane or sevoflurane. Tramer et al. (59) in a meta-analysis found that maintenance of anesthesia with propofol is an advantage compared with other anesthetics when the incidence of complications is in the range of 20%60%, with NNT of <5. The incidence of PON was 25%38% in our present analysis, which would support the use of propofol infusion. However, the efficacy of propofol for reducing PONV was small compared with the inhaled anesthetics, which may question the use of propofol alone without antiemetic prophylaxis in patients at risk. Other than headache, which occurred with more frequency with isoflurane compared with propofol, and drowsiness, which occurred with a significantly more frequency with isoflurane compared with sevoflurane, no other significant differences were found in the incidence of postoperative complications. There are a number of limitations in our meta-analysis. Although we conducted a thorough review of work published in the English language, there may be additional references not identified in our search strategies. Many authors did not state whether the patients studied were inpatients or outpatients. Whenever in doubt, we came to a consensus as to the study group. When authors had not specifically stated whether the data pertained to PON or POV, we decided to exclude these data from the analysis. This limits the number of patients studied, but we chose to report PDNV as distinct end-points. We included patients who were administered nitrous oxide was administered for maintenance in the IV anesthesia group because we believe that this is a common practice used in many studies. Excluding these studies would reduce the scope of this systematic review as well as limiting our findings to a very small and distinct population. Patients induced with inhaled anesthetic (e.g., desflurane and sevoflurane) were also excluded from our review, as this could affect the primary end-points. Finally, we only analyzed studies published in the English language, which can also be considered as a bias. Many authors publish "negative" findings in local (non-English language) journals (60), which would suggest that the differences found by us may be larger than the true differences among these anesthetics, which is further support that the overall differences between the inhaled anesthetics are probably very small. In conclusion, the choice of anesthetic for maintenance of anesthesia should be guided by the training and experience of the individual physician, as well as the routines and equipment available in the hospital, because the specific anesthetic appears to play a minor role in outcome after ambulatory surgery.
Supported, in part, by grants from the Olof Norlander Memorial Fund, Karolinska Hospital, Stockholm, Sweden. We would like to thank the staff at the Welch Medical Library, Johns Hopkins Medical Institutions for the support in retrieving articles for the study. We would also like to thank the Olof Norlander Memorial Fund for financial support of this project.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|