| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
* Department of Anaesthesia and Intensive Care Medicine, and Department of Cardiothoracic Surgery, Mater Misericordiae University Hospital, Dublin, Ireland, and Outcomes Research InstituteTM, University of Louisville, Louisville, Kentucky
Address correspondence and reprint requests to D. J. Buggy, Department of Anaesthesia & Intensive Care Medicine, Mater Misericordiae Hospital, Eccles Street, Dublin 7, Ireland. Address email to donal.buggy{at}nbsp.ie or crina{at}ireland.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A nonpharmacological strategy for prevention of PONV is gastric decompression and evacuation of gastric content during the surgical intervention. Gastric distension increases intragastric pressure and predisposes to vomiting especially if the intragastric gas mixture contains elements of volatile anesthetics inadvertently introduced into the stomach during manual ventilation (9). Studies assessing the role of gastric decompression in the prevention of PONV after various types of surgery have reported conflicting results (10–12).
The potential role of gastric decompression as prophylaxis for PONV in cardiac surgery has never been assessed. Therefore, we tested the hypothesis that gastric decompression during elective coronary revascularization surgery with cardiopulmonary bypass (CPB) and continued afterwards until tracheal extubation would reduce the incidence of vomiting or retching and nausea in a prospective, randomized, cohort clinical study.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patients were randomly assigned (using a sequentially numbered sealed envelope technique) to receive or not receive a gastric tube at induction of anesthesia. All patients were premedicated 1–2 h before surgery with lorazepam 1–3 mg or diazepam 10–15 mg. After establishment of peripheral venous and arterial access, anesthesia was induced with fentanyl 5–20 µg/kg IV, midazolam 1–10 mg IV, and a mixture of O2/air/sevoflurane (0.75–1.0 MAC) as required. Pancuronium 0.1–0.15 mg/kg IV was administered to facilitate tracheal intubation.
After endotracheal intubation, a nasogastric tube was inserted in patients randomized to receive one and the gastric content was aspirated and recorded. The nasogastric tube was then connected to a free drainage bag for the duration of the observation period. The tube was removed at the same time as the tracheal tube postoperatively in the intensive care unit (ICU) to avoid any pharyngeal stimulation-induced vomiting. Anesthesia maintenance was based on a volatile anesthetic-opioid technique at the discretion of the anesthesiologist responsible for the case. During CPB, the patient received a further dose of midazolam 4–10 mg and pancuronium 0.05–0.1 mg/kg. Arterial blood pressure was maintained within 20% of baseline throughout induction and maintenance of anesthesia using vasopressors or vasodilators as indicated. Intraoperatively, fluid administration aimed to maintain a central venous pressure of 5–15 mm Hg. A mixture of O2 and air allowed for O2 concentrations of 30% to 50%. Ventilation was adjusted to maintain normocapnia (end-tidal CO2, 4.0–5.0 kPa). Neuromuscular block was not antagonized at the end of the surgical procedure.
Postoperatively, patients received IV infusion of fentanyl (0.5–2 µg · kg–1 · h–1) for analgesia. Boluses of morphine (1–2 mg IV) were allowed as supplementary analgesia. Midazolam boluses (1–2 mg IV) were used for sedation as required. The administration of postoperative fluids and hemodynamic therapy was aimed to maintain adequate organ perfusion and was at the discretion of the attending anesthesiologist. Weaning from ventilation and tracheal extubation were conducted as rapidly as possible within our unit’s established protocols. We used the Ramsay Sedation Scale to assess sedation level (15). A score of 6 indicates unconsciousness whereas a score of 1 indicates that the patient is awake but anxious, agitated, or restless. A score of 2 is ideal and implies that the patient is awake, cooperative, orientated, and tranquil. In our study, patients were asked to rate their nausea sensation when the sedation score was 2 or less.
Data collection included patient age, gender, weight, height, smoking status, preoperative history of nausea and vomiting/motion sickness, indication of surgery, duration of anesthesia, duration of CPB, aortic cross-clamping time, total cumulative dose of opioids, duration of mechanical ventilation, and time to discharge from ICU. The EuroSCORE was also calculated preoperatively. EuroSCORE is a score based on patient, cardiac, and operation-related risk factors (16) and is used to assess the risk of mortality in patients undergoing heart surgery.
The primary end point of this study was the incidence of vomiting or retching, which are objective events. Secondary end points were the incidence and severity of nausea on the visual analog scale (VAS). Vomiting was defined as expulsive efforts followed by elimination of gastric content. Retching was defined as expulsive efforts not followed by any expulsion of gastric content. Nausea was defined as a subjective sensation: the desire to vomit without any expulsive muscular movements.
Time zero was the time when the patient was awake enough to respond (Ramsay score of 2 or less) and indicate the level of nausea on the VAS, ranging from no nausea to the worst nausea imaginable (17). The time interval for the assessment of nausea or vomiting or both was every 1 h for the first 4 h after time zero, then every 4 h until the patient was discharged from ICU or until a 48-h period had elapsed since the first assessment. An investigator who was not involved in anesthetic care collected data. Data collection sheets were prepared for each time interval of observation. The investigator visited the patients at these times, noting the outcomes by reference to the patient and the ICU nurse.
Vomiting or retching and a score of 40 mm or more recorded on the nausea VAS triggered rescue antiemetic therapy in the form of ondansetron 4 mg IV. The number of doses of antiemetic given was also recorded.
Sample size was determined on the basis that the incidence of vomiting or retching in the control group would be approximately 40%, based on previous reports (2–4,9–12). Reduction of the incidence of vomiting to 15% was taken as being of clinical significance; therefore 48 patients would be required in each group to demonstrate this difference if a Type I error of 5% and a Type II error of 20% is accepted. We obtained Ethics Committee approval to enroll 55 patients in each group to allow for withdrawals.
Normally distributed data were analyzed using Student’s unpaired t-test and nonparametric data were compared using the Mann-Whitney U-test. Categorical variables, including the incidence of vomiting or retching were assessed by 
2 analysis of contingency tables. P < 0.05 was considered as statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Patients’ characteristics, including morphometric details are described in (Table 1. Patients in the two groups were similar in age, gender, body mass index, smoking status and Apfel, EuroSCORE, and New York Heart Association scores.
|
There were no statistically significant differences between the two cohorts of patients regarding duration of anesthesia, surgery, CPB time, and aortic cross-clamping time as per (Table 2. There were no bleeding complications associated with gastric tube insertion and no failure to insert the tube was documented. Time to commencing assessments (Ramsay Sedation Score of 2 or less) was similar in both groups. The time until tracheal extubation, the total time spent in the ICU and the total dose of postoperative opioids were also similar (Table 2).
|
The incidence of vomiting/retching and nausea is shown in (Table 3. The overall incidence of any postoperative nausea or vomiting/retching was 28.8%. There was no significant difference in the incidence of vomiting/retching between the study group and the control group (13.4% versus 11.5%, P = 0.7). Similarly, no difference was observed in the incidence of nausea during the observation period (32.7% versus 25.0%, P = 0.6). Severity of nausea on a VAS measured at 4, 8, 12, and 24 h was not significantly different between the groups at any time. The median VAS nausea score at 12 h was 25 mm (range, 0–55 mm) versus 30 mm (range, 0–60 mm), respectively, P = 0.4. The incidence of vomiting/retching and nausea was 19% and 31% among patients tracheally extubated more than 12 h after surgery compared with 9% and 23% among patients extubated within 12 h, respectively (not statistically significant).
|
All patients who vomited/retched and all those who scored more than 40 mm on VAS received antiemetic therapy (38.5% in the gastric decompression group versus 28.8% in the control group, P = 0.3). Antiemetic therapy was effective in all cases. No complications related to nausea and vomiting/retching or to the administration of the antiemetic medication were observed.
The total quantity of gastric aspirate from the time of insertion to the time of removal of gastric tube was 35 ± 10 mL in the gastric tube cohort.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Coronary revascularization surgery represents a major procedure requiring a variable period of mechanical ventilation in the stressful environment of an ICU. It is associated with many risk factors for nausea and vomiting, including prolonged duration of the surgical procedure (13), catecholamine administration (18), and large doses of opioids (19). It may be characterized by perioperative hemodynamic instability and gut mucosal hypoperfusion, which may represent other risk factors for increased incidence of PONV after cardiac surgery (20,21). It is obviously desirable to avoid the possible complications associated with PONV in cardiac surgery. Furthermore, the use of antiemetic drugs to treat PONV in cardiac surgery patients can be followed by potentially harmful arrhythmogenic effects (22). No difference in the incidence of dysrhythmia was observed in this study.
Our observed incidence of PONV was less than expected from previous publications (2–4,9–12). In a study of 398 patients after heart surgery with CPB Grebenik and Allman (2) reported an incidence of postoperative nausea of 37% and an incidence of vomiting of 47% in the first 24 hours. Similar data were reported by Woodward et al. (3) in 216 patients, the incidence of nausea being 42% and vomiting 34%. In a more recent study, Halvorsen et al. (4), looking at the effect of dexamethasone on side effects after coronary revascularization procedures, reported that 42% of patients in the control group needed antiemetic rescue therapy on the first postoperative day, similar to the incidence of PONV quoted in earlier articles.
In our study the overall incidence of vomiting/retching and nausea among our patients was only 12.5% and 28.8%, respectively, despite avoiding prophylactic antiemetics. This could have occurred as a result of our care to avoid the use of drugs with a proemetic profile including N2O, neostigmine, or etomidate. Special care was provided to control the hemodynamic characteristics with IV fluids and vasopressors or vasodilators aiming to maintain a normal organ perfusion pressure. Although it was not possible to avoid the use of opioids for postoperative analgesia, the total amount of intraoperative and postoperative opioid use was similar in the two groups.
It is possible that reducing the time to tracheal extubation reduces the incidence of PONV. Indeed, in our own study, the incidence of vomiting/retching and nausea was 19% and 31% among patients extubated more than 12 hours after surgery, compared with 9% and 23% among patients extubated within 12 hours, respectively (not statistically significant). The presence of gastric decompression had no influence on these figures. In a recent study Kogan et al. (23) reported an overall incidence of PONV during ICU stay after fast-track cardiac anesthesia of only 22%. They defined fast-track anesthesia as perioperative anesthetic management aimed at facilitating tracheal extubation within 8–10 hours after surgery and discharge from the ICU on postoperative day 1. Interestingly, they also used a gastric tube in all patients in their study population. In our study, there was no significant difference between the two cohorts in the duration of tracheal intubation or ICU length of stay; however, the duration and the length of stay in ICU for both groups was longer than that observed in similar studies (2,3,23).
Previous studies have shown that the use of opioids and the stress of surgery can be associated with decreased gastric emptying in cardiac surgery assessed on Day 1 postoperatively (24). However, Wattwil et al. (25) found no association between delayed postoperative gastric emptying and PONV in patients undergoing laparoscopic cholecystectomy, suggesting that gastric emptying may be not a predictor of PONV. In our study the total quantity of gastric aspirate from the time of insertion to the time of removal of gastric tube was only 35 ± 10 mL in the gastric tube cohort. Indeed, this small quantity of gastric aspirate suggests that there is no delayed gastric emptying in the immediate postoperative period in cardiac surgery patients. In correlation with the lack of statistically significant difference in the incidence of vomiting/retching and nausea between the two groups of patients, the above finding therefore suggests that gastric decompression may be of no use for this category of patients.
In retrospect, because our incidence of PONV after cardiac surgery was less than that reported in the literature, our study was effectively underpowered to detect a difference in PONV despite our prospective power calculation. If anything, our data suggest a trend towards a more frequent incidence of PONV with gastric decompression. There are previous suggestions in the literature that the presence of the gastric tube may stimulate mechanoreceptors in the pharyngeal area followed by an increase of the afferent input to the "vomiting" center (26). The introduction of the nasogastric tube was atraumatic in all our patients. Also, the gastric tube was removed in ICU at the same time as tracheal extubation to avoid any pharyngeal-induced stimulation of the vomiting or retching.
By choosing an observational design for this study (i.e., gastric decompression tubes were left in place on free drainage until tracheal extubation), potential for patient and observer bias exists. It is difficult to truly blind a study testing the value of gastric decompression tubes. We could have evaluated solely intraoperative use of gastric decompression by removing the tube before transfer to ICU, but because patients typically remain ventilated for 6–24 hr postoperatively, we believed that such early removal might have obviated any benefit in terms of reducing PONV.
It is interesting that our observed incidence of PONV was 28.8%, significantly lower than the 40% calculated according to Apfel’s score (13,14). This suggests that this score, validated in noncardiac surgical procedures, may overestimate the incidence of PONV in cardiac surgery patients.
In conclusion, in this prospective, randomized, cohort study, gastric decompression with a nasogastric tube left in situ until tracheal extubation did not reduce the incidence of vomiting or retching or the incidence and severity of nausea in patients undergoing CABG surgery with CPB.
|
Footnotes |
|---|
Accepted for publication August 13, 2004.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
