EDITOR:

Implications for anaesthetic and perioperative care of severely obese patients are considerable and escalate in the presence of comorbidities [Reference Ogunnaike, Jones, Provost and Whitten1]. However, a review of anaesthetic perioperative care and pain management over a 10-yr period (1994–2004) concluded that outcome data are scarce and that literature evaluating the impact of specific care plans on short- and long-term perioperative outcome in bariatric surgical patients is lacking [Reference Schumann, Jones and Ortiz2].

The objective of the present study was to review in a retrospective manner the anaesthetic and perioperative data from 195 consecutive patients admitted to our Medical Centre for weight loss surgery between May 2000 and March 2006 (72 months). Our institutional Ethics Committee considered the study as an audit of service development, and therefore not requiring formal permission.

All the obese patients were evaluated before surgery (between 10 and 30 days) in the anaesthetic assessment clinic. Anticipated difficulties in airway management were based on the following clinical sings: Mallampati score >3, neck circumference >40 cm, thyromental distance <6.5 cm, large tongue or limitations in cervical or mandibular movements. In addition to routine intraoperative monitoring, neuromuscular transmission was assessed using acceleromyography of the adductor pollicis muscle (TOF-Guard; Organon Teknica, BV, The Netherlands). Since 2004 a bispectral index monitor was routinely introduced (BIS Monitor Model A 2000, Aspect Medical System Inc., Newton, MA, USA). For super-obese patients (body mass index (BMI) > 50) non-invasive haemodynamic monitoring was performed by HemoSonic transoesophageal echo-Doppler (HemoSonic™ 100, Arrow International Everett, MA, USA).

At the end of the surgery, trocar sites were infiltrated with ropivacaine. Extubation was performed only on fully awake patients with a train-of-four ratio of 0.9 and in 30° reverse Trendelenburg position. The anaesthetic management, postoperative care, intraoperative and postoperative complications were evaluated and recorded. The same surgeon performed all surgical procedures while anaesthetic pre-, intra- and postoperative management was by members of our anaesthesiology staff.

In order to evaluate the evolution of the surgical and perioperative management between these dates, the patients were divided into three consecutive and equal groups (Group A, B and C each with 65 patients) and compared. One-way analysis of variance was used to compare continuous variables among three groups. We used χ ² test or, when necessary, Fisher exact test, for testing univariate association between socio-demographic or clinical variables and χ ² test for trend to assess the linear trend when comparing proportions. The null hypothesis was rejected with α < 0.05. Data are presented as mean ± standard deviation (SD).

Of the 195 obese patients, 28.2% (n = 55) were male and 71.8% (n = 140) were female. Mean age was 44 ± 11 yr, mean BMI 44.8 ± 6.8 and mean weight excess at the first interview was of 55 ± 21 kg. Concomitant disease was present in 56% of patients (n = 109). Arterial hypertension was the main associated disease (21%, n = 41), followed by chronic gastritis (16%, n = 24), while diabetes was diagnosed in only 4% of patients (n = 7). There were no laboratory test abnormalities in 80.6% of patients (n = 157). Chest X-ray examination was positive for mild cardiomegaly only in two patients; the remaining films were unremarkable. Results of pulmonary function tests were moderately abnormal in 6.7% (n = 13): 2 obstructive, 10 restrictive and 1 combined. Liver steatosis was present in 16% of patients (n = 31), while chronic gastritis and gastroesophageal reflux were diagnosed by direct gastroscopy in 15% (n = 29). Using our criteria for preoperative airway assessment, we classified 5% of patients (n = 10) as at risk for problematic tracheal intubation. Preoperative evaluation did not lead in any case to delay or cancellation of scheduled surgery.

Laparoscopic adjustable gastric banding was performed in 64.6% of patients (n = 126) and laparoscopic Roux-en-Y gastric bypass in 35.4% (n = 69). The average surgical time for gastric banding was of 90 ± 12 min and that for gastric bypass was 241 ± 23 min. In four cases conversion to open surgery was required due to surgical difficulties. Other intraoperative complications included intra-abdominal adhesions (3.6%), hypoxia (2%), tooth damage (1%), severe bradycardia (1%), soft palate lesions (1%) and bronchospasm (0.5%).

In patients with anticipated high risk of a problematic tracheal intubation, awake fibreoptic intubation was performed. No difficult intubation by direct laryngoscopy occurred in the remainder 95% following rapid sequence induction. In all patients, tracheal intubation was performed in the ‘ramped’ position as suggested by Collins and colleagues [Reference Collins, Lemmens, Brodsky, Brock-Utne and Levitan3]. Narcosis was maintained with propofol infusion (44%), sevoflurane inhalation (44%), desflurane (9%) or nitrous oxide in oxygen and sevoflurane (3%); neuromuscular blockade with cisatracurium (56%), rocuronium (23%) or vecuronium (21%). Analgesia was provided by continuous infusion of remifentanil (88%) or boluses of fentanyl (12%). Extubation was performed in 96% of cases (n = 187) in the operating theatre. Neuromuscular blockade was reversed in 59% (n = 115) of patients, the others recovered spontaneously.

Direct ICU admission from the operating theatre occurred in 3.6% of patients (n = 7). The remaining 96.4% were monitored in the post-anaesthesia care unit (PACU). The mean ICU stay was 1.5 ± 0.5 days, whereas in PACU it was 213 ± 62 min. Mechanical ventilation was protracted in PACU for an average of 30 ± 7 min in 4% of patients due to extreme obesity, delayed recovery and severe hypoxia. The mean visual analogical pain score on admission to PACU was 3 ± 2, while before discharge to the surgical ward it was 1 ± 1. Postoperative analgesia was managed with meperidine (56%), ketorolac (22%), continuous infusion of remifentanil (11%), or the association of ketorolac and meperidine (9%) or morphine (2%). The most frequent complications observed in PACU were gastrointestinal (6.6%), respiratory (4.6%), cardiovascular (4.1%), bleeding (3.6%) and agitation (3.1%). Subsequently, on the surgical ward, the predominant problems encountered were pain (15%) and nausea (12%), followed by fever (8%) and persistent reflux (5%).

In patients who underwent gastric banding, mean hospitalization was 3.2 ± 1.6 days, while in patients who underwent laparoscopic Roux-en-Y gastric bypass it was 9.1 ± 3.1 days. A decreasing trend was observed for the duration of surgery of the laparoscopic gastric banding procedures (Group A 103 ± 35 vs. Group B 88 ± 31 vs. Group C 71 ± 18; P < 0.01), and also their PACU stay (Group A 313 ± 100 vs. Group B 217 ± 63 vs. Group C 200 ± 57; P < 0.01). There was a progressive decrease in the gastric banding group (P < 0.01) with an increase in Roux-en-Y gastric bypass (P < 0.01) group. There were no other significant differences among the three groups.

With the growing experience over the years of the surgical and anaesthesiological staff, we observed a reduction in the duration of laparoscopic gastric banding procedures and PACU length of stay. Probably the most important contributing factor to the observed outcome is the concept of ‘perioperative medicine’ as stressed by Dahl and Kehlet [Reference Dahl and Kehlet4], according to which the end result is determined by the management and strict interdependence among the pre-, intra- and post-operative periods. Several studies looking at the relationship between obesity and perioperative complications in patients undergoing laparoscopic surgery have reported no increased risk [Reference Birgisson, Park, Mastrangelo, Witzke and Chu5]. The principal risk factor is the presence of co-morbidities, particularly cardiorespiratory disease, and not obesity per se.

In different studies, routine laboratory preoperative tests have shown a low incidence of abnormal results (0.3–6.5%) with an even lower proportion of these results leading to modifications in patient management (0–2.6%) [Reference Ramaswamy, Gonzalez and Smith6]. Indeed, laboratory tests were abnormal in 20% of our population of obese and did not influence our perioperative management in any patient. We therefore agree with Ramaswamy and colleagues [Reference Ramaswamy, Gonzalez and Smith6] that chest X-ray examination, coagulation studies, cardiac stress tests and pulmonary function tests should be selectively performed on the basis of morbidly obese patient history and clinics. Moreover, Brodsky and colleagues [Reference Brodsky, Lemmens, Brock-Utne, Vierra and Sidman7] demonstrated that only obesity with clinical signs such as large neck circumference and a high Mallampati score are predictors of potentially difficult intubation, whereas BMI or weight per se were not. Our experience is in line with such evidence since awake fibreoptic endotracheal intubation was performed in only 5% of patients and no difficult intubation occurred in the remaining patients.

Although specific anaesthetic maintenance agents have been recommended, there is insufficient evidence to support one agent or technique over another [Reference Schumann, Jones and Ortiz2]. In fact, several different anaesthetic approaches were used over the study period, and none were associated with any increased morbidity or worse outcome.

In conclusion, anaesthesia for these patients can be safely performed when understanding that morbidly obese patients are not at risk per se due to their BMI but based on the existence and severity of co-morbidities. The concept of perioperative medicine with the strict interdependence of pre-, intra- and post-operative management is the key to a rational approach to morbidly obese patients.