Gastrointestinal Motility Following Thoracic Surgery: The Effect of Thoracic Epidural Analgesia

A Randomised Controlled Trial

Argyro Zoumprouli; Aikaterini Chatzimichali; Stamatios Papadimitriou; Alexandra Papaioannou; Evaghelos Xynos; Helen Askitopoulou

Disclosures

BMC Anesthesiol. 2017;17(139) 

In This Article

Methods

This prospective randomised controlled study, was approved by the ethics committee of the University Hospital of Heraklion, Greece (No 3197, 19 March 2002). The study was performed between March 2002 and January 2009 and was registered retrospectively (20/03/2017) at ISRCTN registry system with registration number: ISRCTN11953159. All patients provided written informed consent. Only patients undergoing major thoracic surgery (thoracotomy) were recruited in order to avoid the confounding direct effect of intra-abdominal surgery on the GI system. Furthermore, a standardised postoperative recovery regimen of feeding, ambulation and pain score targets was followed, with the purpose of controlling and optimising factors that affect the recovery of GI function. The exclusion criteria were diabetes mellitus, history of chronic pain, drug/alcohol dependence, corticosteroid use, treatment with drugs known to affect GI motility, inflammatory bowel disease, previous bowel surgery, previous history of abdominal radiation, morphine or local anaesthetic allergy, ASA physical status > III, age younger than 30 or older than 85 years, presence of contraindications to insertion of an epidural catheter and severe renal and liver disease. All patients had a history of normal bowel habits. Eligible patients were randomly assigned to three groups by opening opaque sealed envelops in the anaesthetic room prior to surgery. The groups were pre-determined by a computer-generated list of random numbers (block of 6 with 1:1:1 allocation).

Preoperative Period

All patients underwent standard preoperative assessment and received teaching on how to score pain and to report side effects. In addition, they were informed about postoperative tests, feeding and early ambulation, as well as postoperative visits from different teams. 1 h before surgery, all patients received premedication with intramuscular 0.07 mg.kg−1 midazolam. On arrival to the anaesthetic room, each patient was randomised in one of three analgesic groups: group Ep-R, ropivacaine epidurally, group Ep-RM a combination of ropivacaine and morphine epidurally and group IV-M IV morphine by PCA. In groups Ep-R and Ep-RM a thoracic epidural was performed before induction of general anaesthesia between the levels T5–9 using loss of resistance technique with an 18G Tuohy needle and a 20G epidural catheter was inserted 3–5 cm into the epidural space. After negative aspiration of blood and cerebrospinal fluid, a test dose of 3 ml of lidocaine 2% containing adrenaline 5 mcg.ml−1 was injected through the catheter.

Intra-operative Period

General anaesthesia was induced in all patients with intravenous fentanyl 1.5–2 mcg.kg−1, propofol 1.5–3 mg.kg−1 followed by rocuronium 0.6 mg.kg−1 or cis-atracurium 0.1 mg.kg−1. Intubation of the trachea was performed using a Robertshaw left-sided double lumen endotracheal tube. Following the placement of the patient in the lateral right or left decubitus position, the correct tracheal and bronchial tube position was confirmed by a fibreoptic bronchoscope.

Anaesthesia was maintained with an oxygen/air mixture and either propofol continuous infusion or sevoflurane administration. 20 min before the surgical incision group Ep-R received the first epidural bolus of 5 ml of 0.5% ropivacaine (5 mg.ml−1), group Ep-RM received 5 ml of 0.5% ropivacaine and 3 mg of morphine in 8 ml of sterile normal saline 0.9%, and group IV-M received an IV bolus of 0.05 mg.kg−1 morphine. Neuromuscular blockade was maintained with rocuronium or cisatracurium boluses, as needed and monitored with a train of four stimuli from a peripheral nerve stimulator. Intraoperative monitoring also included pulse oximetry, electrocardiography, end-tidal CO2, invasive arterial pressure and urinary output. All patients were placed on a water-warming mattress. Blood pressure and heart rate were maintained at ±20% of preoperative baseline values throughout the operation with the use of phenylephrine boluses of 40 mcg, as required. Patients received further epidural or IV boluses according to analgesic requirements and group allocation. Intraoperative blood loss (from suction and weighted surgical dressings) was recorded and replaced by crystalloids, colloids and blood products according to individual needs and departmental policy.

At the end of surgery neuromuscular blockade was reversed using neostigmine 2.5 mg in combination with glycopyrrolate 0.5 mg, and the double lumen tube was removed. Patients were then transferred to the post-anaesthesia care unit, where they remained for at least 2 h for monitoring and clinical observation. Invasive arterial pressure monitoring was removed before patients' transfer to the surgical ward. The same surgical team performed the surgery on all patients and did not take part in the collection of study data.

Postoperative Analgesia Regimes

At the end of the surgical procedure, an epidural or an IV infusion was started according to group allocation. Group Ep-R received a continuous epidural infusion of ropivacaine 0.2% (2 mg.ml−1) at a rate of 5–8 ml.h−1, with boluses of 2 ml of the same solution and a 20-min lockout interval via a PCA pump. Group Ep-RM received a continuous epidural infusion of ropivacaine 0.15% (1.5 mg.ml−1) with morphine 0.05 mg.ml−1 at a rate of 5–7 ml.h−1, with boluses of 2 ml of the same solution and a 20-min lockout interval via a PCA pump. Group IV-M received a continuous IV infusion of morphine of 1 mg.ml−1.h−1, with boluses of 0.5–1 mg and a 15-min lockout interval also via a PCA pump.

Both patients and the research team were double blinded for the epidural groups and unblinded for the group IV-M. Only the Acute Pain Team was aware of the solutions administered and was allowed to give extra boluses accordingly, if and when needed, so that analgesia was titrated to a VAS score at rest of <5 (on a 10-point scale, where 0 = no pain, 10 = worst possible pain). The level of epidural block was evaluated by loss of pain sensation to pinprick. No other analgesics were used. The analgesic regimes were continued until at least the third postoperative day.

Postoperative Evaluation of GI Regime

On the first postoperative morning, all patients were started on a standardised low-fat diet. Mobilisation was commenced from the first postoperative day. The surgical, acute pain and research teams visited the patients independently twice a day to optimise and monitor the postoperative course and care. The surgical team decided patients' discharge.

In all patients GI motility was assessed on the first and third postoperative days by two objective tests: (a) OCTT using the hydrogen (H2) breath test and (b) CTT using radiopaque markers, and also by subjective tests: (a) the passage of flatus (b) defecation and (c) the presence of bowel sounds.

The OCTT was evaluated by the lactulose - H2 breath test, a non-invasive method widely used for quantifying the OCTT. The OCTT test measures the time (in minutes) taken for lactulose to reach the caecum. Lactulose is a synthetic disaccharide that cannot be absorbed in humans and therefore passes unchanged to the colon where it is fermented by bacteria. The H2 produced by fermentation, passes into the blood stream and therefore can be measured in the exhaled breath from the lungs. The OCCT the point at which H2 is increased in the exhaled breath.

Patients were advised to fast for at least 12 h prior to the breath test and to avoid fibre rich food for 24 h. The day of the test, the subjects were fasting and the exhaled H2 concentrations in breath were measured in parts per million (ppm) by an electrochemical detector (Lactoscreen, Hoek Loos, Schiedam, The Netherlands) following an oral load of 10 g lactulose (diluted in 200 ml of water) at time 0 and every 15 min thereafter for up to 4 h, or when the subject reached an increment of 10 ppm H2. The detector was calibrated using samples of room air (undetectable H2) and a standard gas mixture containing 100 ppm H2 (automated process). Before the test the patient was rested for at least 1 min. Breath samples were collected by aspirating aliquots of end-expiratory air into 20 ml plastic syringes with a three-way stopcock. The measurements were conducted in the sitting position in a quiet, well-ventilated room, without physical activity during the test. The OCTT test was performed at three times: (a) 1 week before surgery (OCTT0), (b) on the first postoperative morning (OCTT1) and (c) on the third postoperative morning (OCTT3).

The CTT was evaluated by plain abdominal radiographs taken 4 days after the ingestion of 20 radiopaque markers consisting of 1–2 mm of hollow radiopaque polyethylene tubing. The markers were ingested 1 week before surgery (CCT0) and on the first postoperative morning (CCT1). The measurement of the total number of markers retained after 96 h assessed the progression of radiopaque markers along the large bowel, giving an indication of CTT.

Data collection. OCTT was recorded 1 week and the CTT 4 days preoperatively. Patient demographics were also recorded during the preoperative period. The total intraoperative amounts (in mg) of epidural ropivacaine and epidural or IV morphine were recorded, as well as the level of the thoracic epidural catheter placement, the duration of surgery, blood loss, and intraoperative fluid replacement.

During the postoperative period, both the acute pain and the research teams collected data independently until patient discharge. Postoperative colonic motility was evaluated by the OCTT and CTT tests as well as by the clinical signs of the first passage of flatus, faeces and the first presence of bowel sounds. The amount (in mg) of epidural ropivacaine and epidural or IV morphine administered, the VAS pain scores and the upper and lower sensory levels (pinprick sensation) of the epidural blockade were recorded, as well as blood pressure, heart rate and respiratory rate measurements. The VAS scores were measured at rest (VAS rest) and on mobilisation (VAS dynamic), on the first, second, and third postoperative days. The patients were monitored for side effects such as nausea, vomiting, pruritus, sedation, motor or sensory block, hypotension, bradycardia and respiratory depression and treated appropriately.

Statistical analysis. A small pilot study demonstrated that OCTT time was increased by 270% in patients receiving continuous IV infusion of morphine 1 ml.h−1 and PCA boluses. Based on the assumption that a 30% difference in the time of recovery of the GI function (assessed by OCTT) between the epidural groups and the IV morphine group was of clinical significance, a power analysis estimated that nine patients per group were needed to provide 80% power and 0.05 alpha error.

Data are presented as the mean ± SD, numbers (%), or median (interquartile range). Statistical analyses were performed using the SPSS software version 24 (IBM Corp., USA). Between-group differences for the OCTT test were analysed using General Linear Model (GLM) – repeated measures. One-way ANOVA (3 groups comparison) for parametric variables or the Kruskal–Wallis test for non-parametric variables were also used. The Fisher exact test was used to analyse categorical variables, where appropriate. Pearson's and Spearman's rho rank correlation coefficients were used to assess the degree of association between variables. Values for p < 0.05 were considered statistically significant.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....