Colonoscopy - Is Sedation Necessary And Is There Any Role For Intravenous Propofol?

Colonoscopy remains a difficult and at times frustratingly lengthy procedure, and it has been suggested that a trainee endoscopist may need to perform over 200 supervised examinations to become reasonably proficient [1]. The passage of a flexible endoscope along a floppy, highly distensible and mobile length of bowel frequently causes painful looping of the instrument [2]. As stated in a recent edition of a standard textbook, “some stretching of peritoneal attachments is inevitable, at least transiently, during colonoscopy. This may cause gnawing or acute unpleasant visceral pain. Coupled with the sensation of rectal fullness simulating the desire to defaecate, this makes a slow examination difficult to bear for some patients …” [3].

If we accept that a degree of pain and discomfort will accompany colonoscopy, why does the provision of sedation and analgesia vary so much throughout the world? In Great Britain and the United States of America, the vast majority of colonoscopies are carried out using some form of intravenous sedation, most commonly a benzodiazepine in combination with an opioid [2] [3] . In France about 80 % of colonoscopies are performed under general anaesthesia [4] while, in contrast, in Germany [5] and Finland [6] most examinations are conducted without any form of sedation whatsoever.

National differences in sedation practice may reflect many different factors. These include the personal preference and training of the endoscopist, the availability of anaesthetic services, the need to train colleagues in endoscopic techniques, the cost and availability of monitoring equipment, differences in the availability and use of the commonly used drugs and, importantly, the expectation of the patient. In the UK there has been a reduction in the dose of intravenous sedatives used by many endoscopists compared with even a few years ago, following, among other things, publication of guidelines [7] and a survey that looked at the morbidity and mortality associated with flexible upper gastrointestinal endoscopy [8]. In the case of colonoscopy, the combination of a benzodiazepine and an opioid remains the most common form of sedo-analgesia [9]. Conscious sedation may be defined as the use of sedative techniques to produce a sedated patient who is able to respond to verbal commands. If the patient is unresponsive, by definition, a general anaesthetic has been administered, with all the clinical and legal implications that this holds.

Do agents such as propofol have any place in endoscopic practice? Propofol (2,6-diisopropylphenol), a chemically inert phenol derivative, is a short-acting anaesthetic agent with a rapid onset of action. It produces unconsciousness within 30 seconds and is used by anaesthetists to induce and maintain general anaesthesia, and for conscious sedation [9]. Propofol is administered intravenously either in bolus doses or by continuous infusion. Use of propofol by nonanaesthetists to produce sedation remains controversial. The concentrations of propofol required for sedation range from 1.0 to 2.5 μg/ml. Concentrations needed for the maintenance of anaesthesia in normal individuals vary between 3.0 and 6.0 μg/ml [10]. When administered as a sole agent, single intravenous bolus doses of 0.75 - 1.0 mg/kg are required for sedation, whilst for the induction of anaesthesia the dose would be 2.0 - 3.0 mg/kg. The use of bolus doses will necessitate repetition every few minutes. Practically, for anything other than short procedures a continuous infusion will be required [9].

Propofol has a short recovery time that makes it suitable as an intravenous agent for day-case anaesthesia and sedation. In practical terms, no matter how long the infusion period, recovery will occur within 10 - 20 minutes once it is discontinued. This is because propofol is rapidly metabolized, principally by conjugation in the liver [10]. There may even be some metabolism within the gastrointestinal tract [11].

Anaesthetists have used propofol in combination with low doses of midazolam fentanyl and alfentanil to produce effective sedo-analgesia [9]. However, non-anaesthetists need to be extremely wary of this sort of practice since propofol acts synergistically with both midazolam [12] and alfentanil [13]. Such synergism can result in the unplanned induction of anaesthesia, with potentially disastrous consequences for those without training in anaesthesia and experience in airway maintenance and manual inflation of the lungs [9]. In addition, propofol used as a sole agent can cause both cardiovascular and respiratory depression and, if it is used in combination with drugs such as fentanyl which may stimulate the vagus, there can be severe or even life-threatening bradycardias [14].

Freeman [15], quoting the work of Carlsson and Grattidge [16] on the use of propofol for upper gastrointestinal endoscopy, concluded that “… this agent produces significant respiratory depression, pain on injection, and usually requires continuous infusion. Its narrow therapeutic range and demands associated with its administration probably render propofol ill-suited to use by gastrointestinal endoscopists …”

Others take an opposing view. Roseveare and colleagues published two studies in this journal in 1998 on the use of intermittent patient-controlled sedation and analgesia using propofol and alfentanil for colonoscopy [17] [18] . In the latter study [18], patients undergoing colonoscopy were randomized either to receive an intravenous bolus of pethidine (50 mg) and Diazemuls (10 - 20 mg), or to receive self-administered 0.5 ml boluses of a mixture of propofol (10 mg/ml) and alfentanil (25 μg/ml) as required. An anaesthetist was present throughout the procedure. The median dose of the drugs infused in the patient-controlled sedation (PCS) group was 105 mg of propofol and 0.13 mg of alfentanil. The PCS group reported significantly more pain than the pethidine-Diazemuls group, but patient satisfaction in both groups was good, and predictably the recovery period was much shorter in the propofol group than in the pethidine-Diazemuls group (median 10 minutes, compared with 40 minutes; P < 0.0001). The median cost of the drugs, tubing and syringes was ¥ 5.46 for the PCS group and ¥ 2.94 for the pethidine-Diazemuls group. However, the authors felt it prudent to have an anaesthetist present during the use of propofol [18]. If this sort of cover is thought to be necessary when propofol is used, it represents an additional expense.

In this edition of Endoscopy, Reimann and colleagues report their experience of propofol; this time in combination with a benzodiazepine rather than an opioid [19]. Their paper entitled “Synergistic sedation with low-dose midazolam and propofol for colonoscopy” reports a study of 79 patients undergoing colonoscopy who were randomly assigned to receive either midazolam with or without nalbuphine (group I), or midazolam (2 mg) and repetitive increments of propofol (group II). Unsurprisingly, patients in group I required a relatively large dose of midazolam (mean dose 9 mg), and 59 % subsequently needed additional analgesia with nalbuphine (mean dose 20 mg) to ensure completion of the colonoscopy. Equally predictably, group II patients in the “synergistic sedation” group who received only 2 mg of midazolam and a relatively modest dose of propofol recovered much sooner after the procedure (mean time 17 minutes) compared with recovery times in group I (mean time 93 minutes). Perhaps of more interest is that patients treated with the propofol-low-dose midazolam combination rated the procedure as more comfortable than did those in the midazolam-nalbuphine group (81 % vs. 50 %), despite a lower proportion of patients having complete amnesia. The endoscopists rated the quality of the sedation to be similar in the two groups [19].

Are there wider lessons to be learnt from this study? The normal recommendations for endoscopists using an opioid-benzodiazepine combination are that the opioid drug should be administered first, in a dose of about one-quarter of that which might be administered if the drug was to be used on its own, and that the benzodiazepine should then be given by careful titration [9] [20] . Had Reimann and colleagues [19] given nalbuphine to all the group I patients first and then titrated the midazolam it might have been safer and would, almost certainly, have led to the use of a dose of midazolam considerably less than the median of 9 mg used in their study. This, in turn, might have reduced the recovery time in that group. It is considered unacceptable in anaesthetic circles for oxygen saturation levels to fall below 85 % [21], as happened to 22 % of patients in the midazolam-nalbuphine group and 19 % in the midazolam-propofol group. These levels are frankly dangerous, and the routine use of supplemental oxygen would have greatly reduced this unnecessary risk to the patients [21] [22] [23] .

Also in this edition of Endoscopy is a paper by Jung and colleagues [24] on the use of propofol as an alternative to midazolam for sedation in endoscopic retrograde cholangiopancreatography (ERCP). We were pleased to see that all patients received routine supplemental oxygen and an anaesthetist administered the propofol in every case [24]. This was wise since “… one patient in the propofol group suffered a protracted apneic phase accompanied by hypotension that was managed by manual ventilation and drug therapy …” [24].

If sedation is used for any endoscopic technique, it should be as safe as possible. It is easy for an endoscopist to administer an “anaesthetic” dose unwittingly, rather than a “conscious sedation” dose, when using i.v. benzodiazepines either alone or in combination with an opioid drug. This is clearly demonstrated by the episodes of desaturation seen in the study by Reimann et al. [19]. The smaller therapeutic ratio seen with propofol means that, in our opinion, there is an inadequate margin of safety when this drug is used by non-anaesthetists. In addition, as anaesthetists are rarely, if ever, available to supervise endoscopic sessions in many parts of the world, this makes the widespread adoption of propofol in endoscopic practice unlikely for safety reasons.

There is still a need for studies that look at the 30-day morbidity and mortality associated with colonoscopy and ERCP with the same detail as those previously published about flexible upper gastrointestinal endoscopy [8]. In the meantime, there are simple clinical practices that can increase patient safety. These include oxygen supplementation, adequate periprocedural monitoring and the use of carefully selected sedation techniques which follow basic principles [7] [9] [21] . With regard to colonoscopy, perhaps the optimal clinical situation is for the procedure to be carried out without either sedation or general anaesthesia, and there have been recent studies which compared sedated with unsedated colonoscopy [5] [6] [25] . An alternative would be to offer brief analgesia when needed, with the use of patient-administered inhalation of nitrous oxide-oxygen mixtures [26]. Avoiding the use of i.v. sedation and/or general anaesthesia would reduce cost.

Many now believe that if the cost of colonoscopy could be reduced, then a one-time colonoscopy at age 60 might be the most cost-effective method for colorectal cancer screening [27]. For this policy to be adopted we need to have a better understanding of where, why and how pain is produced during colonoscopy. We need to design endoscopes which are ergodynamically more efficient for the endoscopist and more comfortable for the patient. Non-radiological magnetic endoscope imaging systems [28] [29] [30] , in combination with the use of thinner and floppier endoscopes [31] which can be intermittently stiffened [32], may be one way forward.

Sedation for any form of endoscopy carries a cost to the patients in terms of morbidity and mortality. Death rates quoted in connection with these procedures [8] remain at levels which horrify anaesthetists [21], and we have to continue our efforts to improve patient safety. Unfortunately, many endoscopists (and their patients) are likely to need a lot more persuading before giving up the idea of routine sedation and analgesia prior to colonoscopy.

References

• 1 Parry B R, Williams S M. Competency and the colonoscopist: a learning curve.  Aust NZ J Surg. 1991;  61 419-422
• 2 Williams C B. Colonoscopy and flexible sigmoidoscopy. In: Cotton PB, Williams CB (eds). Practical gastrointestinal endoscopy.  Oxford; Blackwell Science Ltd, 1996 Fourth edition: 187-274
• 3 Waye J D, Williams C B. Colonoscopy and flexible sigmoidoscopy. In: Yamada T, Alpers DH, Laine L, et al. (eds). Textbook of gastroenterology.  Philadelphia, Lippincourt; Williams and Wilkins, 1999 Third edition: 2701-2717
• 4 Canard J M, Carayon P, Dumas R, et al. La colonoscopie en France en 1998: résultat d'une enquête prospective nationale de la Société Française d'Endoscopie Digestive (S.F.E.D.).  Endoscopy. 1999;  31 6
• 5 Eckardt V S, Kanzler G, Schmitt T, et al. Complications and adverse effects of colonoscopy with selective sedation.  Gastrointest Endosc. 1999;  49 560-565
• 6 Ristikankare M, Hartikainen J, Heikkinen M, et al. Is routinely given conscious sedation of benefit during colonoscopy?.  Gastrointest Endosc. 1999;  49 566-572
• 7 Bell G D, McCloy R F, Charlton J E, et al. Recommendations for standards of sedation and patient monitoring during gastrointestinal endoscopy.  Gut. 1991;  32 823-827
• 8 Quine M A, Bell G D, McCloy R F, et al. Prospective audit of upper gastrointestinal endoscopy in two regions of England: safety, staffing and sedation methods.  Gut. 1995;  36 462-467
• 9 Whitwam J G, McCloy R F. Principles and practice of sedation.  Oxford; Blackwell Science Ltd, 1998 Second edition
• 10 Cockshott I D. Propofol (diprivan): pharmacokinetics and metabolism - an overview.  Postgrad Med J. 1985;  61 (Suppl. 3) 45-50
• 11 Gray P A, Park G, Cockshott I D, et al. Propofol metabolism in man during the anhepatic and reperfusion phases of liver transplantation.  Xenobiotica. 1992;  22 105-114
• 12 Short T G, Chiu P G. Propofol and midazolam act synergistically in combination.  Br J Anaesth. 1992;  67 539-545
• 13 Vinik H R. Propofol-alfentanil hypnotic interaction.  Anesth Analg. 1990;  71 65-69
• 14 Guise P A. Asystole following propofol and fentanyl in an anxious patient.  Anaesth Intens Care. 1991;  19 116-118
• 15 Freeman M L. Sedation and monitoring for gastrointestinal endoscopy. In: Yamada T, Alpers DH, Laine L, et al. (eds). Textbook of gastroenterology.  Philadelphia, Lippincourt; Williams and Wilkins, 1999 Third edition: 2655-2667
• 16 Carlsson U, Grattidge P. Sedation for upper gastrointestinal endoscopy: a comparative study of propofol and midazolam.  Endoscopy. 1995;  27 240-245
• 17 Roseveare C, Seavell C, Patel P, et al. Patient-controlled sedation with propofol and alfentanil during colonoscopy: a pilot study.  Endoscopy. 1998;  30 482-483
• 18 Roseveare C, Seavell C, Patel P, et al. Patient-controlled sedation and analgesia, using propofol and alfentanil, during colonoscopy: a prospective randomized controlled trial.  Endoscopy. 1998;  30 768-773
• 19 Reimann F M, Samson U, Derad I, et al. Synergistic sedation with low dose midazolam and propofol for colonoscopy.  Endoscopy. 2000;  32 240-254
• 20 Ben-Shlomo I, Abd-El-Khalim H, Ezry J, et al. Midazolam acts synergistically with fentanyl for induction of anaesthesia.  Br J Anaesth. 1990;  64 45-47
• 21 Charlton J E. Monitoring and supplemental oxygen during endoscopy (editorial).  BMJ. 1995;  310 886-887
• 22 Bell G D, Bown N S, Morden A, et al. Prevention of hypoxaemia during upper gastrointestinal endoscopy using supplemental oxygen via nasal cannulae.  Lancet. 1987;  1 1022-1023
• 23 Bell G D, Jones J G. Routine use of pulse oximetry and supplemental oxyen during endoscopic procedures under conscious sedation: British beef or common sense (editorial)?.  Endoscopy. 1996;  28 718-721
• 24 Jung M, Hofmann C, Kiesslich R, Brackertz A. Improved sedation in diagnostic and therapeutic ERCP: propofol is an alternative to midazolam.  Endoscopy. 2000;  32 233-239
• 25 Rex D J, Imperiale T F, Portish V. Patients willing to try colonoscopy without sedation: associated clinical factors and results of a randomized controlled trial.  Gastrointest Endosc. 1999;  49 554-559
• 26 Saunders B P, Fukumoto M, Halligan S, Williams C B. Patient-administered nitrous oxide/oxygen inhalation provides effective sedation and analgesia for colonoscopy.  Gastrointest Endosc. 1994;  40 418-421
• 27 Lieberman D A. Cost-effectiveness model for colon cancer screening.  Gastroenterology. 1995;  109 1781-1790
• 28 Bladen J S, Anderson A P, Bell G D, et al. Non-radiological technique for three dimensional imaging of intestinal endoscopes.  Lancet. 1993;  341 719-722
• 29 Williams C B, Saunders B P, Bell G D, et al. Real-time magnetic three-dimensional imaging of flexible endoscopy. In: Rex DK (ed). Evolving issues in colon endoscopy. Gastrointest Endosc Clin N Am 1997 7 (3): 469-476
• 30 Bell G D, Rowland R S, Rutter M, et al. Colonoscopy aided by magnetic 3-D imaging - would the routine use of a stiffening tube speed up the procedure?.  Med Biol Eng Comput. 1999;  37 605-611
• 31 Bell G D, Painter J, Rowland R S, et al. Preliminary experience comparing two longer thinner prototype Olympus endoscopes with a standard 60 cm flexible sigmoidoscope.  Gut. 1999;  44 (Suppl. 1) 22
• 32 Brooker J C, Saunders B P, Williams C B. A new variable-stiffness colonoscope makes insertion quicker and may reduce patient discomfort.  Gut. 1999;  44 (Suppl. 1) 21