Anesthetic management of patients undergoing bariatric surgery: two year experience in a single institution in Switzerland

Obesity represents a significant and growing problem around the globe [1]. Aside from the impairment of an individual patient, the negative consequences impose a significant economic burden for many health care systems [2].

For over a decade, bariatric surgical procedures have established themselves as a way to achieve a permanent weight reduction for a large number of patients [3].

This patient population represents a particular challenge for the anesthetist, with multiple publications offering a wide range of recommendations on the matter [4–8]. Additionally, there are also large differences in terms of the data that outline the perioperative process times [9–12]. All of this results in difficulty to establish such a program and predict its trajectory at the early stages.

We report the perioperative experience two years after the start of such a bariatric program at our single Swiss institution.

In early 2011 a multidisciplinary obesity program was launched at the Kantonsspital Frauenfeld (KSF) (General Hospital Level 2, 270 beds, about 8400 anesthetics per year). The program encompassed bariatric surgical care, psychiatric/psychosomatic patient guidance, nutrition counselling, gastro-enterological and cardiac work-up, as well as the follow-up and physical therapy.

Anesthesia and intensive care medicine physicians were involved early in the planning process. In collaboration with the surgical team, a perioperative approach for patient care was developed for each individual patient.

From an anesthesia perspective, the key points of this concept were the characterization of preoperative testing including laboratory tests. A detailed cardiac evaluation was obtained for all patients that were either over 55 years, had a BMI >50 kg/m², an exercise tolerance of <4 MET or any significant cardiac history. An evaluation by a pulmonologist including lung function testing and screening for obstructive sleep apnoea (OSA) were performed for all patients, if not previously diagnosed. The anesthetist in the preoperative clinic saw the patients about two weeks preoperatively or on the day before surgery. The main focus was on the detection of any form of OSA and its pre-existing treatment. All patients were informed and consented about the possibilies of an awake fiber-optic intubation, an arterial line (AL), a central venous catheter (CVC) and a postoperative stay in the intensive care unit (ICU).

The nil per os (NPO) time for all patients was six hours fasting for solid food, and two hours for clear liquids. Midazolam 7.5 mg po was given preoperatively, unless the patient had a BMI >40 kg/m², or was diagnosed with OSA, in order to minimize the risk of perioperative hypoventilation. Patients were instructed to take their baseline medication with the exception of ACE inhibitors or AT-II antagonists. In patients with a history of gastro-esophageal reflux, a therapy with proton pump inhibitors (PPI) was initiated if not already in place.

For the anesthetic induction, peripheral venous access was established. When non-invasive blood pressure measurements were reliably feasible, no AL was placed. The choice of the individual airway management technique (conventional direct laryngoscopy vs fiber-optic intubation either awake, as a rapid sequence induction or asleep) was at the discretion of the assigned anesthesia attending. All conventional intubations were performed in Anti-Trendelenburg position in order to reduce the risk of regurgitation and aspiration of gastric contents. Preoxygenation was performed with the patient breathing spontaneously via face mask (FiO₂ = 1.0) until the end-expiratory FiO₂ was at least 0.8. Propofol was used for induction, followed by BIS monitoring guided Propofol infusion and/ or Sevoflurane for anesthesia maintenance, supplemented with Fentanyl and Remifentanil. Propofol was dosed as target controlled infusion (TCI) with a range of 2.2 to 4 mcg/ml and was supplemented by Sevoflurane, as needed. Sevoflurane was used when the Propofol infusion resulted in hypotension and had to be reduced to below 2.2 mcg/ml or when despite a Propofol TCI of 4 mcg/ml the anesthetic depth was insufficient, as indicated by the BIS monitoring.

Cefazolin, 1.5 g IV served as the antibiotic prophylaxis. Paralysis for intubation was achieved with either Succinylcholine or Atracurium,. Atracurium was the paralytic of choice for subsequent paralysis. After intubation a second large bore peripheral IV catheter (≥17G) was placed. In cases where peripheral vascular access was difficult, a CVC was inserted. For the surgical procedure, the patient was positioned in a modified, semi-standing beach chair position. In order to prevent deep venous thrombosis, sequential compression devices were applied intraoperatively, followed by low molecular weight heparin sc on the first postoperative evening as prescribed by the surgeon, in combination with early mobilization of the patients. An inspiratory oxygen fraction (FiO₂) of 0.6 and a positive end expiratory pressure (PEEP) of 5 cm H₂O were the standard initial ventilator settings, while using either pressure controlled or volume controlled ventilation (target for P_max ≤30 cm H₂O, and tidal volume 6 ml/kg). Paralysis at the end of the procedure was checked with quantitative TOF-determination (TOF-Watch; Organon, Dublin, Ireland). If the TOF-ratio was less than 0.9, Glycopyrrolate combined with Neostigmine was used to reverse the neuromuscular blockade. Postoperative nausea and vomiting (PONV) was pre-emptively treated according to an escalating protocol based on the Apfel risk score [13]. In order to be extubated after the procedure, a patient’s body temperature had to be 35.0°C or higher. The emergence and extubation were performed in a separate dedicated room. The multimodal postoperative pain regime was initiated 20 min before end of surgery with a combination of 5 mg of morphine iv, paracetamol IV 4 × 1 g/d, metamizole IV 5 g/24 h, controlled-release oxycodone and naloxone (Targin®) po 2 × 10/5 mg, and morphine IV 2 – 5 mg as rescue medication. Non-steroidal anti-inflammatory drugs were omitted, and patients were kept NPO for the first postoperative night, except for the mentioned Targin®.

Elective patients with an uneventful surgical and anesthetic course were recovered in the post anesthesia care unit (PACU). Patients with OSA and no continuous positive airway pressure (CPAP) therapy were admitted to the ICU postoperatively to provide continuous monitoring for the first post-operative night. Pre-existing CPAP therapy was continued after extubation, using the patient’s usual setup and adjusting the settings as clinically needed in the PACU.

The institutional ethics committee of the Canton of Thurgau, Switzerland, approved the retrospective analysis of the anesthesia documents and medical records of all patients undergoing bariatric surgery in the first two years after the establishment of the new bariatric program at the Kantonsspital Frauenfeld (KEKTG/PEB2013/09). The study was registered with the German registry for clinical studies (http://www.drks.de; DRKS00005437). Essential anesthesia and perioperative data and complications were collected and investigated.

The surgical procedures of the patients in the bariatric program were quite heterogeneous. Therefore we mainly investigated the data of patients undergoing a gastric bypass procedure (GBP). We did not include any follow-up interventions of cosmetic nature.

The data are presented descriptively as mean [standard deviation].

During the two year observation period (2011–2013), after initiating the bariatric program, a total of 182 bariatric operations were performed, 178 of them laparascopically.

Patients had a mean age of 44 [10] years, were predominantly female (79.7%), with a BMI of 43 [8] kg/m². The surgical procedures were quite heterogeneous and had operation times of 30 min for a laparoscopic removal of a gastric band up to 425 min for a laparotomy for a complex revision case (mean 130 [51] minutes).

Because of that heterogeneity, we decided to only include patients that underwent gastric bypass procedures (GBP) in the further analysis. The total number of GBP was 147, all but one of them were performed laparascopically. 19 cases (12.9%) needed an additional procedure at the same time (18 removal of a gastric band, 1 cholecystectomy).

All patients underwent thorough preoperative evaluations. 107 patients (72.8%) were classified with an ASA physical status of 2, 39 (26.5%) with ASA 3 and 1 patient (0.7%) with ASA 4. 67 patients (45.6%) presented with arterial hypertension, 42 patients (28.6%) with OSA (29 (69.0%) of them with CPAP therapy), 33 patients (22.4%) had (mostly non-insulin dependent) diabetes mellitus, 81 (55.1%) had gastro-esophageal reflux disease, 68 patients (46.3%) were smokers, with a mean of 24 [20] pack years. 92 patients (62.6%) reported no regular medication.

In 101 patients a chest X-ray was performed as part of their preoperative work up. This resulted in 13 patients (12.8% of studies) demonstrating pathologic findings, mostly in the form of either atelectasis or left ventricular enlargement. In 143 patients a preoperative ECG was obtained, it showed abnormal findings in 26 patients (18.2%), 4 of them had a prolonged QTc interval. Forty-nine patients received a preoperative echocardiography; this resulted in 21 patients (42.9%) with a pathological result. Two patients presented with a moderate valvulopathy, one patient had a severely impaired LVEF of 25%. A stress test was performed in 96 patients, and resulted in 24 patients (25%) with abnormalities, 2 patients had clear signs of ischemia during exercise. Preoperative gastroscopy (137 patients) was pathological in 91 patients (66.4%) (e.g. axial hiatal hernia, GERD). The other preoperative investigations revealed abnormal findings in 74 patients (50.3%), the majority of them were found during the abdominal ultrasound. The majority of patients (142, 96.6% of gastric bypass patients) were admitted to the hospital at least one day prior the surgery.

In 115 patients (78.2%) the non-invasive blood pressure measurement was considered sufficient, in 32 patients (21.8%) an AL was placed for blood pressure measurement. A second large bore peripheral IV was placed in 137 patients (93.2%), 10 patients (6.8%) received a central venous catheter (CVC).

Twenty-nine patients (19.7%) were intubated fiber-optically, 26 of them awake, 3 patients after induction of general anesthesia. One of them was a difficult intubation (required 3 attempts). 117 patients (79.6%) were intubated with direct laryngoscopy, 103 of them with a rapid sequence induction. In 87.4% of the conventional intubation cases, patients presented with a Cormack-Lehane grade 1, 8.1% had a grade 2, 4.5% a grade 3 and in one case (0.9%) the airway had to be secured unplanned with the fiber-optic. One patient (0.7%) experienced an episode of aspiration of gastric contents that remained without consequences.

In 146 patients (99.3%) anesthesia was induced with Propofol and 145 patients (98.6%) received a Propofol infusion (2 [0.7] g) as main anesthetic for the case. Pain management was achieved with fentanyl (0.6 [0.2] mg) and remifentanil (2.0 [0.4] mg). 26 of the patients (17.9%) also received a supplementation of their anesthetic with Sevoflurane. In 2 patients (1.4%) the anesthesia was maintained with Sevoflurane only. Fluid homeostasis was achieved with 1700 [600] ml of Ringerfundin (B. Braun Medical, Sempach, Switzerland), and 400 [300] ml of HES 130/04 (Voluven; Fresenius Kabi, Oberdorf, Switzerland). In 59 cases (40.1%) the patient’s blood pressure had to be supported pharmacologically, mainly due to the beach chair positioning of the patients. 52 times boluses of Ephedrine were used and in 7 cases Norepinephrine was given in boluses. One patient (0.7%) received Atropine to counteract a relevant bradycardia. In 24 cases (16.3%) it was necessary to medically address hypertension. This was done using Clonidine in 21 instances, and for 3 patients Urapidil was the medication of choice.

In 63 patients (42.9%) an intra-operative PEEP in excess of 5 cm H₂O was required (maximum 11 cm H₂O).

No blood products had to be administered for any case. Four patients (2.7%) required a blood transfusion in the subsequent course of their hospital stay.

At the end of procedure 116 patients (78.9%) needed reversal of their paralysis. The core body temperature was 36 [0.5] °C. Two patients (1.4%) could not be maintained above 35°C during the procedure, one of these patients was transferred to the intensive care where he was extubated one hour later. All extubations were unproblematic, no patient had to be re-intubated.

Postoperatively 115 patients (78.2%) were transferred to the post anesthesia care unit to recover from the procedure, 32 patients (21.8%) were transferred to the ICU. 15 of those admissions (10.2% of all gastric bypass patients) were planned preoperatively due to untreated OSA. None of these patients needed anything other than supplemental application of nasal oxygen. The reasons for unplanned ICU admission was hemodynamic instability in 6 patients, insufficient oxygenation in one patient, hypothermia in another patient (<35°C). The average length of stay in the post anesthesia care unit was 4.1 [0.7] hours, patients that went to the ICU remained there for 17 [12] hours.

During the observation period 15 (10.2%) of the gastric bypass patients required additional surgery (4 major operations), two (1.4%) of them were laparotomies. Eight patients (5.4%) needed surgical revision of some form of an incisional or internal hernia. Three patients (2.0%) presented with some form of postoperative hemorrhage that required the transfusion of blood products, 2 patients (1.4%) developed pneumonia, and one of them required non-invasive ventilation. There was zero in-hospital mortality. The average length of hospital stay was 6 [2] days.

This publication reflects on the perioperative and anesthetic experience two years after starting a bariatric program. Retrospectively, the patient’s medical charts and anesthesia records were analyzed with emphasis on the bariatric surgery and any follow up procedures. All the essential process times and metrics were collected to create insight into the effects of implementing a surgical bariatric program from an anesthesia standpoint.

Our main finding was that the bariatric patient population presents with its unique challenges. When managed appropriately in an interdisciplinary fashion, anesthesia can be delivered safely to this patient population. The preoperative evaluation of these patients should be conducted based on an individual’s need rather than relying on standardized test batteries.

Anesthesia for morbidly obese patients is generally regarded as being associated with increased risks [8, 14]. Accordingly, there are various contributions on the anesthetic and perioperative management in this patient group [4–7, 15]. They include the preoperative assessment, especially with regards to the airway management and monitoring strategies, the intra-operative phase, the emergence from anesthesia and the postoperative phase. Since bariatric surgery is being used in a more widespread manner, caring for morbidly obese patients is becoming more common for many institutions [3]. The surgical procedures are well illustrated by Ogunnaike et al. [6], e.g.

Bariatric anesthesia comes with its unique challenges, but when approached in an thoughtful and interdisciplinary fashion it becomes safely manageable. In addition, the need for some urgent follow-up operations and appropriate postoperative monitoring capabilities have to be accounted for at all times.