Introduction
Obesity is a worldwide health problem with clearly established health implications1. During the past three decades, the obesity prevalence in Mexico has had an unprecedented increase, and its rate of growth is one of the highest in the world. However, this issue has been recognized only recently as a health problem, requiring immediate attention to improve prevention, early diagnosis, and control in the general population2.
Patients with obesity have an increased risk for developing comorbidities. One of the therapeutic options for patients with morbid obesity is bariatric surgery, which encompasses a variety of surgical weight loss procedures3.
Bariatric surgery has been associated with beneficial effects on patient’s comorbidities such as improvement health and control of diabetes mellitus, cardiovascular diseases, sleep apnea, joint pain, and life quality4.
Anesthetic management of the patient with morbid obesity undergoing to bariatric surgery represents a challenge to the anesthesiologist because they can be particularly sensitive to the respiratory depressant effect of opioid analgesic drugs and subsequent hypoxic episodes5.
Some anesthesiologists have recommended the use of opioid-free anesthetic techniques in patients with morbid obesity6–8; this requires the use of adjuvant drugs such as dexmedetomidine or ketamine instead of intravenous opioids9. In addition, some studies suggest that use of post-operative opioids for analgesia should be avoided in patients with morbid obesity due to the risk of respiratory depression6,8. Nowadays, there is no scientific evidence about which anesthetic technique would be better for a safe management of patients with morbid obesity undergoing bariatric surgery.
The aim of this study is to evaluate the safety and the adverse effects of anesthesia based on a continuous fentanyl infusion for patients with morbidly obesity supplemented with two different anesthetic techniques minimum alveolar concentrations (MACs): MAClow (0.8-1.0 MAC) versus MACstandard (more than 1.0 MAC) undergoing bariatric surgery. Clinical endpoints were complication rates (respiratory depression, airway obstruction, desaturation, and apnea), post-operative nausea vomiting (PONV) rates and post-operative pain control (measured as analgesic drugs requirement) in post-anesthesia care unit (PACU), and during the first 24 h of the post-operative period.
Materials and methods
With the Approval of The Committee Investigation on February, 2017 in The American British Cowdray Medical Center (ABCMC), we retrospectively reviewed the medical charts of 812 consecutive of patients with morbid obesity undergoing bariatric surgery from June 2010 to January 2014 at (ABCMC) in Mexico City. Only 374 of these anesthesia records and medical charts were complete, so we reviewed and included them in the study.
Anesthesia technique
Once patients arrived at the operating room, they were monitored with non-invasive arterial pressure every 5 min, continuous electrocardiography, and pulse oximetry; patients were denitrogenated with a oxygen inspiratory fraction (FiO2) of 100% and a continuous positive airway pressure ventilation during 10 min in a “ramped” upper-body, reverse Trendelenburg position. The induction was made with midazolam (30-45 μg/kg total body weight), propofol (2 mg/kg total body weight), and a non-depolarizing muscle relaxant (rocuronium), adjusted according to ideal body weight. After induction the trachea was intubated maintaining the “ramped” position and patients were assigned to one of two alternatives, according to the anesthesiologist’s in charge preference: desflurane 0.8-1.0 MAC (MAClow group) or desflurane more than > 1.0 MAC (MACstandard group), combined with a continuous intravenous fentanyl infusion in both groups. Intravenous fentanyl infusion was initiated after a fentanyl bolus of 3 μg/kg of pharmacokinetic mass10, at 10 μg/kg of pharmacokinetic mass during the first 15 min. Fentanyl was reduced 20% every 15 min according to blood pressure and/or heart rate (with the aim of maintaining ± 20% of baseline values) and a mean arterial pressure ≥ 60 mmHg, and suspended 40 min before the surgery was over. Patients also received an IV clonidine infusion (300 μg total dose) during the 1st h of surgery.
Other medications used were prophylactic antibiotics, omeprazole 40 mg, ketorolac 60 mg, and ondansetron 8 mg IV given 30 min before the end of the surgery. In addition, in all our patients we used, acetaminophen 1 g IV tid ketorolac 30 mg IV tid stat and tramadol 50 mg IV bid to tid stat for postoperative pain control.
At the beginning of the surgery, we initiated with medium compression socks and intermittent mechanical compression in both legs to prevent deep vein thrombosis and pulmonary embolism; an appropriate and comfortable positioning was used to avoid skin or nerve injuries. When the extubation criteria were met, we extubated and moved the patient to PACU, for post-operative surveillance.
Statistical analysis
The statistical analyses were performed with StatPlus: MAC. Pro.V-5.9.6 (AnalystSoft Inc.). Continuous data for each group were compared utilizing an unpaired, two-sided Student’s t-test whereas categorical data were compared utilizing a χ2 test, with a p < 0.05 considered as significant; U Mann–Whitney and Fisher’s exact tests for data with a non-normal distribution were also used. All data are reported as means ± SDs, percentages, and CI95% as needed.
Results
We presented some preliminary results in a Poster Scientific Presentation at the World Congress International Federation for Surgery of Obesity and Metabolic Disorders (IFSO Obesity Week) 2013 in Atlanta, so we expanded the review of the total medical charts and have the following data, and a second review as a Thesis Scientific Presentation as a Poster for Bariatric Anesthesia Fellowship in 2014, in Universidad Nacional Autónoma de México (UNAM), Mexico City, Mexico.
As it is shown in table 1, the demographic characteristics of patients in the MAClow group were heavier and taller than patients in the MACstandard group but had a similar body mass index (BMI). In contrast, patients in the MACstandard group had a greater frequency of snoring and a greater frequency of polysomnographic studies, but not a different frequency of obstructive sleep apnea than the MAClow group. The rest of patients’ characteristics were not significantly different between groups.
Table 1. Demographic characteristics
| Variable | MAClow group | MACstandard group | p value |
|---|---|---|---|
| Age (year)* | 40 ± 11.5 | 41 ± 11.0 | 0.605 |
| Gender % (F/M)* | 56.6/43.4 | 64.5/35.5 | 0.904 |
| Weight (kg) | 118.1 ± 24.28 | 110.1 ± 20.92 | 0.042 |
| Height (cm) | 168 ± 0.1 | 165 ± 0.1 | 0.047 |
| BMI (kg/m2) | 41.6 ± 6.65 | 40.5 ± 5.25 | 0.163 |
| ASA % (II/III) | 86.1/13.9 | 85.5/14.5 | 0.51 |
| Snore % (Y/N) | 41.1/58.9 | 64.7/35.3 | 0.001 |
| Polisomnography % (Y/N) | 4.6/95.4 | 15.2/84.8 | 0.01 |
| OSA % (Y/N) and | 29.4/70.6 | 41.3/58.7 | 0.074 |
| Duration of surgery (min) | 187.6 ± 54.11 | 199.5 ± 89.86 | 0.331 |
|
* Continuous variables were compared utilizing an unpaired Student’s t-test while categorical variables were compared utilizing χ2; F: female; M: male; BMI: body mass index; ASA: American Society of Anesthesiologists; Y: Yes; N: No; OSA: obstructive sleep apnea. |
|||
When we divided patients into two groups according to the MAC received (MAClow group vs. MACstandard group) during anesthesia, we found statistically significant differences in: total fentanyl dose (1.563 ± 532.8 [CI95% 1.505.5-1.621.0] μg vs. 1.343 ± 256.6 [CI95% 1.214.3-1,472. 8 μg p = 0.003), total propofol dose (131 ± 35.6 mg vs. 144 ± 41.6 mg p = 0.008), and total muscle relaxant dose (59 ± 22.7 mg vs. 40 ± 30.2 mg p = 0.00002). There were no statistically significant differences in total ondansetron, benzodiazepine, clonidine, or steroid doses.
When we adjusted the mean total fentanyl consumption by actual body weight, we found that the mean total fentanyl actual body weight adjusted dose was 13 ± 4.1 μg/kgactual body weight in the MAClow group versus 11 ± 3.8 μg/kgactual body weight in the MACstandard group (p = 0.082). When we used the pharmacokinetic mass as an adjusting factor, the mean total pharmacokinetic mass adjusted fentanyl consumption results is shown in table 2.
Table 2. Fentanyl consumption by actual body weigh
| Variable | MAClow group | MACstandard group | p value |
|---|---|---|---|
| Mean total fentanyl actual body weight-adjusted dose | 13 ± 4.1 µg/kg | 11 ± 3.8 µg/kg | 0.082 |
| Mean total pharmacokinetic mass-adjusted fentanyl consumption | 17 ± 5.3 µg/kg (CI95% 16.8–17.9) | 15 ± 5.0 µg/kg (CI95% 13.9-16.5) | 0.005 |
| Mean total fentanyl dose per hour* | 4.5 ± 1.50 µg/kg actual body weight/h | 4.0 ± 1.43 µg/kg actual body weight/h | 0.03 |
| Pharmacokinetic mass and time as adjusting factors the mean total fentanyl dose per hour | 5.8 ± 1.84 µg/kg pharmacokinetic mass/h (CI95% 5.6-5.9) | 5.0 ± 1.82 µg/kg pharmacokinetic mass/h (CI95% 3.7-6.3) | 0.007 |
|
* Adjusted the mean total fentanyl consumption by actual body weight and time. |
|||
The mean tramadol total dose used in the MAClow group patients was 103 ± 4.8 (CI95% 92.5-113.0) mg, while the mean total dose used in MACstandard group patients was 96 ± 2.3 (CI95% 77.4-115.2) mg being unable to reach statistical significance (p = 0.439). When analyzing postoperative ketorolac doses, we found that in the MAClow group, the total mean dose was 114 ± 37.7 (CI95% 109.7-118.0) mg, while in the MACstandard group, the total mean dose was 120 ± 36.0 (CI95% 111.2-128.8) mg (p = 0.363). Since there was no documented pain intensity measure during the post-operative period, we used the analgesic consumption as a surrogate during the first post-operative 24 h. There were no statistically significant differences in post-operative analgesic rescue during the first post-operative 24 h, but when we compared the number of analgesic drugs consumed in the post-operative period, we found that the use of none or one versus two or more analgesic drugs was statistically significantly different between groups: 82.9/17.1% in the MAClow group versus 71.2/28.8% in the MACstandard group (p = 0.034) (Fig. 1).
Figure 1. Use of analgesics in MAClow and MACstandard groups during the first 24 h in patients undergoing bariatric surgery.
*MAC: minimum alveolar concentration.
When the frequency of PONV was examined, there were no statistically significant differences between groups in the incidence of PONV during the first 24 h (Table 3 and Fig. 2). The frequency of PONV in the group as a whole was 23.5% (CI95% 19.3-27%) and when comparing between groups, it was 24.2% (CI95% 19.5-28.8%) in the MAClow group versus 20.3% (CI95% 10.1–30.6%) in the MACstandard group (p = 0.524). We also searched for associations between PONV and genre (p = 0.959), BMI (p = 0.156), desflurane concentration received (p = 0.954), fentanyl dose (p = 0.812), use of metoclopramide (p = 0.908), steroids (p = 0.480), benzodiazepines (p = 0.608), and 5-HT3 antagonists (p = 0.424) and we did not find any. There were no cases of respiratory depression, desaturation, reintubation, or post-operative mechanical ventilation.
Table 3. Presence of PONV in two groups. MAClow and MACstandard. First 24 h after bariatric surgery (divided in 0-4 h, 4-12 h, 12-24 h)
| Variable | PONV 0-4 h | PONV 4-12 h | PONV 12-24 h |
|---|---|---|---|
| MAClow | |||
| Without PONV | 97.86% | 83.18% | 90.52% |
| With PONV | 2.14% | 16.82% | 9.48% |
| MACstandard | |||
| Without PONV | 93.55% | 88.71% | 91.94% |
| With PONV | 6.45% | 11.29% | 8.06% |
Figure 2. Presence of PONV in MAClow and MACstandard groups during the first 24 h in patients undergoing bariatric surgery.
*MAC: minimum alveolar concentration;
PONV: post-operative nausea vomiting.
There were no documented cases of post-operative respiratory complications (i.e., bronchoaspiration, atelectasis, pneumonia, or pulmonary embolism).
Discussion
Patients undergoing bariatric surgery are often at high risk for complications due to the obesity itself and related comorbidities. Therefore, careful patient selection, together with well-designed strategies for preventing and managing complications, including anesthesia technique, are the key to success3.
Although the Sjöström et al.4, the study demonstrated that bariatric surgery (gastric bypass and vertical banded gastroplasty) was associated with a marked reduction in overall mortality as compared with control subjects (diet plus changes in lifestyle), surgery is still associated with risks inherent to every major surgery, that is, the metabolic response to surgical stress11 and the complications associated with anesthetic management.
It is well documented that surgical trauma is followed by a period of metabolic acute phase response. The levels of biochemical markers of metabolic response to surgical trauma have been shown to correlate with the degree of surgical trauma. It also has been shown that the intensity of the metabolic surgical stress response is correlated with post-operative morbimortality.
Systemic stress response after bariatric laparoscopic surgery is similar after a bariatric open surgery, except that concentration of norepinephrine, Adreno Cortico Trophic Hormone (ACTH), C-reactive protein, and cytokine IL-6 were lower after laparoscopic than after open surgery. This suggests a lower degree of surgical trauma after laparoscopic bariatric surgery11, but still a significant one.
The metabolic stress response blunting is an objective in major surgery due to the association with post-operative complications. Several studies have proven that high-dose fentanyl anesthesia (50 μg/kg) can control stress more effectively than lower doses can by diminishing the metabolic and endocrine response to surgery12, block hypothalamic-pituitary function13, and lower blood lactate concentrations14, decrease plasma glucose and non-esterified fatty acids, and low heart rate and mean arterial pressure; this changes may be the result of decreased cortisol, catecholamines, insulin, growth hormone, and glucagon secretion. Cooper et al.15 describe that the use of high-dose opioid anesthesia (the so-called ‘stress-free-anesthesia’) confers cardiovascular stability, improvement in post-operative nitrogen balance, and abolishes hormonal and metabolic changes during major surgery12. Fujita et al.16 study demonstrates that high-dose fentanyl anesthesia (a 30 μg/kg group vs. 75 μg/kg group) for cardiac surgery causes only small decreases in heart rate and arterial blood pressure, but that in the 75 μg/kg group, the epinephrine level elevation was totally suppressed; the use of 30 μg/kg dose was still an attractive anesthesia technique for patients with valvular disease. Liu et al.17 used fentanyl (30-100 μg/kg) and demonstrated that these doses can completely suppress the stress response induced by intubation and intense surgical stimulus before cardiopulmonary bypass-induced stress response began and that the effect was not dose-dependent. They concluded that a 60 μg/kg fentanyl total dose seemed to be an ideal one. Duncan et al.18 described that a balanced anesthetic containing fentanyl 25-50 μg/kg is sufficient to obtund hemodynamic and stress responses to the pre-bypass part of cardiac surgery; higher doses of fentanyl (100-150 μg/kg) offered little advantage over 50 μg/kg and were associated with more pharmacologic interventions to prevent hypotension. Philbin et al.19 found that use of high-dose opioid infusions (sufentanil 30 μg/kg and fentanyl 100 μg/kg) to maintain high plasma concentrations did not provide better control of hemodynamic responses to noxious stimuli. Furthermore, these very high opioid doses may produce acute tolerance to the analgesic and hypnotic effects, and although they are well tolerated, they may not actually produce more stable anesthetics than lower doses.
In our patients, we used a mean (± SD) total fentanyl dose of 13 (± 4.1) μg/kg in the MAClow versus 11 (± 3.8) μg/kg in the MACstandard group (p = 0.017), about half to one tenth of the doses used by the authors mentioned before.
Although based on its pharmacokinetic properties, the best semisynthetic narcotic to be used for general anesthesia would be sufentanyl20; its congener fentanyl is widely used around the world including our country for balancing general inhaled anesthesia with the aim of controlling the sympathetic responses to tracheal intubation and surgical stress. Its use has been widely studied in different surgical populations including patients with morbid obesity. Shibutani et al.10 developed a novel parameter for adjusting doses of fentanyl in patients with morbid obesity called the pharmacokinetic mass; this parameter was used for calculating fentanyl doses in our patients.
When compared with the doses administered in other studies of patients with morbid obesity undergoing bariatric surgery, we found that our patients received a mean total fentanyl dose (± SD) greater than that reported in other studies 17.3 (± 5.32) μg/kgpharmacokinetic mass in the MAClow group versus 15.2 (± 5.02) μg/kgpharmacokinetic mass in the MACstandard group. In spite of this, none of our patients had adverse respiratory events, such as a diminished respiratory rate or desaturation, neither reintubation nor post-operative mechanical ventilation support. When very high fentanyl doses (50 μg/kg actual weight) are used as Cooper et al.15 did there is severe respiratory depression in all patients and require reversal with naloxone (0.04-0.16 mg) and an intravenous infusion (0.4 mg in 1 L 0.9% sodium chloride solution) maintenance over the following 6 h.
Comparing the fentanyl doses used in our patients with those used in other patients reported in the literature, we found that in a randomized controlled clinical trial comparing fentanyl, enflurane, and halothane in patients with morbid obesity scheduled for elective gastric stapling, Cork et al.21 used an average total dose of fentanyl of 328.6 μg (3.46 μg/kgpharmacokinetic mass) and did not find any difference neither in the extubation time nor in the expended time at the PACU between groups. In another randomized controlled clinical trial, Naja et al.22 studied patients with morbid obesity undergoing laparoscopic gastric sleeve and compared the use of a clonidine (0.8–1.2 μg/kg actual weight) infusion versus a dexmedetomidine infusion; patients in the clonidine group consumed a mean fentanyl dose of 3.9 μg/kg pharmacokinetic mass while patients in the dexmedetomidine group consumed a mean fentanyl dose of 3.5 μg/kg pharmacokinetic mass; PONV frequency was 37.5% in the clonidine group and 52% in the dexmedetomidine group. Strum et al.23 studied 50 patients with morbid obesity requiring gastrointestinal bypass through open laparotomy to determine if awakening and recovery times differed between desflurane and sevoflurane anesthesia. Their patients received a total fentanyl dose of 3.02 μg/kg versus 2.7 μg/kg (they did not mention if doses were calculated using actual weight, adjusted weight, or ideal weight) in the desflurane versus sevoflurane group, respectively.
PONV is the most frequent side effect after anesthesia24 and also a frequent problem after laparoscopic bariatric surgery; in a study conducted by Ziemann-Gimmel et al.25 the frequency of patients requiring antiemetic rescue medication using triple prophylaxis (dexamethasone, ondansetron, and a scopolamine patch) was up to 42.7%. The present study showed that the incidence of PONV during the first 24 h was 24.2% in the MAClow group and 20.3% in the MACstandard group (23.5% in the group as a whole) administering to each patient an 8 mg ondansetron dose 30 min before the end of surgery and in some patients metoclopramide and/or dexamethasone. In a prospective, randomized, placebo-controlled, and double-blinded study, Moussa et al.17 studied the incidence of PONV in 120 patients who received either granisetron 1 mg, granisetron 1mg plus droperidol 1.25 mg, granisetron 1 mg plus dexamethasone 8 mg or intravenous saline solution immediately before induction of anesthesia, and they found an incidence of 30% in the granisetron group, 30% in the granisetron plus droperidol group, 20% in the granisetron plus dexamethasone group, and 76% in the placebo group during the first 24 h after surgery. In a prospective, randomized study, Ziemann-Gimmel et al.25 compared the effect of two anesthetic techniques (general balanced anesthesia vs opioid-free total intravenous anesthesia) on PONV frequency, finding an incidence of 37.3% in the general balanced anesthesia group vs an incidence of 20.0% in the opioid-free total intravenous anesthesia26,27. As we can observe in our results, a high to moderate opioid dose anesthesia technique did not increase significantly the incidence of PONV in the post-operative period (Table 3 and Fig. 2).
Post-operative pain control in patients undergoing laparoscopic bariatric surgery is generally achieved using a multimodal approach trying to avoid postoperative opioid use; we used this approach in our patients and the moderate to high fentanyl doses helped us in achieving this goal. In a study made by Bakhamees et al.27, 80 morbidly obese patients were randomized to one of two groups, one that received dexmedetomidine (0.8 μg/kg bolus, 0.4 μg/kg/h infusion) and the other one that received normal saline (placebo); total post-operative morphine consumption in the first group was 35.4 ± 6.4 mg (equivalent to 177 mg of tramadol) while in the second one was 47.8 ± 8 mg (equivalent to 239 mg of tramadol); in average, these patients required more narcotic analgesic than our patients did. Sollazzi et al.6 found in a study of 50 patients with morbid obesity undergoing open biliopancreatic diversion randomly allocated into a study group receiving an infusion of ketamine-clonidine before the anesthesia induction and a control group who received standard anesthesia, a lower tramadol consumption in the study group (138 ± 57mg) than in the control group (252 ± 78 mg); these doses were again higher than the doses that our patients required.
In recent years, some investigators have suggested that opioid-free anesthesia would be the best anesthetic technique for patients with morbid obesity undergoing bariatric surgery; to date as far as we know there is only one study comparing an opioid-free anesthesia technique versus a standard technique28 but it was a small study (28 patients) and it did not measure important outcomes such as pain intensity and frequency of complications such as PONV and respiratory complications.
Finally, we considered in this study desflurane parameters ideal for maintenance of anesthesia in morbidly obese patients undergoing bariatric surgery between end-tidal anesthesic gas 0.8-1.3 MAC as a secure target to avoid patient transoperatory awareness without using bispectral index (BIS) as secure anesthesia BIS range between 40 and 6029,30. This allows the patient’s extubation phase performed earlier and attain verbal contact faster, without compromising safety. The benefits of better recovery extend into the immediate post-operative phase31.
Conclusions
The main drawback of our study is its retrospective design that impedes to obtain detailed information about important outcomes because the anesthetic and medical charts were not specifically designed to record this information.
This anesthetic technique using moderate fentanyl doses is safe and feasible in patients with morbid obesity undergoing bariatric surgery.
It would be desirable to conduct a randomized clinical trial comparing moderate fentanyl dose anesthetic technique with an opioid-free anesthetic technique to settle down the discussion about the best anesthetic technique in this population.
Implication statement
This study compare two bariatric anesthesia technique for patients with morbidly obesity, based on a novel parameter for adjusting doses of fentanyl in these patients called pharmacokinetic mass10; which is used as a parameter calculating a Fentanyl infusion doses (In Mexicans hospitals is the most common and available opioid) and two balanced general anesthetic techniques using two desflurane MAC dose concentrations: desflurane standard (MACstandard) or desflurane MAClow dose. Our technique is safe and well tolerated in patients for bariatric surgery.
Acknowledgments
The authors thank to Department of Nutrition, Obesity & Metabolic Alterations Center of The American British Cowdray Medical Center: with the support of Maureen M. Mosti RN, Bariatric Certificate Nurse; The Clinical Medical Files Department of The American British Cowdray Medical Center; and the support of Degree in Information Science Micaela Ayala-Picazo Head Library of The American British Cowdray Medical Center.
Funding
The authors declare that they have not received funding.
Conflicts of interest
The authors declare no conflicts of interest.
Ethical disclosures
Protection of humans and animals. The authors declare that no experiments on humans or animals have been performed for this research.
Confidentiality of data. The authors declare that no patient data appear in this article.
Right to privacy and informed consent. Right to privacy and informed consent. The authors have obtained the approval of the Ethics Committee for the analysis and publication of routinely obtained clinical data. The informed consent of the patients was not required because this was a retrospective observational study.
Use of artificial intelligence to generate texts. The authors declare that they have not used any type of generative artificial intelligence in the writing of this manuscript or for the creation of figures, graphs, tables, or their corresponding captions or legends.
