Prevention of postoperative nausea and vomiting by metoclopramide combined with dexamethasone: randomised double blind multicentre trial
BMJ 2006; 333 doi: https://doi.org/10.1136/bmj.38903.419549.80 (Published 10 August 2006) Cite this as: BMJ 2006;333:324All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
We note this response with surprise. The relevance of a black car is
nonsensical, if osteonecrosis is associated with corticosteroids, which it
is. 15 cases of osteonecrosis is a large number and should not be ignored.
We again, take issue with the use of the word 'safe' in the original
article. There were three facets to our letter: thromboembolism and poor
wound healing were mentioned. Are these to be discounted, or are they red
and blue cars ?
Competing interests:
None declared
Competing interests: No competing interests
Ladies and Gentlemen,
Thank you for your interest and your comment. We definitely agree
that physicians of different specialisation concerned with the treatment
of a patient should cooperate, in particular when prescribing drugs. As
well, a treatment scheme should include anamnestic issues to clarify risk
factors and possible contraindications.
However, we do not agree that the references you quoted provide
evidence that osteonecrosis was related to low-dose corticosteroids. To
understand this, imagine that I consecutively register patients of a heart
centre delivered with acute myocardial infarction. My inclusion criteria
is that they have driven a black car within the last three years. After a
while, I have collected an impressive number of cases. Now I can submit my
paper entitled "Myocardial infarction after driving a black car".
Your first reference is of this type. The personal thinking of the
authors that the number of 15 cases they presented is high is the only
argument for the increased risk of osteonecrosis associated with LOW-DOSE
corticosteroids. To assess the increased risk for an event associated with
an exposure, one needs a 2-by-2 table with exposure (yes/no) in the rows
and event (yes/no) in the columns. The authors presented only one out of
four cells of this table.
Now suppose, a patient presents with acute myocardial infarction. He
did not have any risk factors or signs predicting myocardial infarction
before. His physician read my article (see above), gets to know that his
patient has a black car, and writes a case report. The additional
"evidence" coming from this report is of the same type like your second
reference.
To avoid misunderstanding: I do not claim that osteonecrosis is not
associated with low-dose corticosteroids. I only say that evidence should
be obtained by appropriate methods. A case-control study could certainly
be carried out. Case reports are useful to trigger a systematic
observation, but they are not a substitute. Cumulative case reports are
not necessarily cumulative evidence; they may be repeated biased
observation as well.
Competing interests:
None declared
Competing interests: No competing interests
Dear Sir/Madam,
We enjoyed reading the article on the prevention of postoperative
nausea and vomiting by the intraoperative administration of metoclopramide
combined with dexamethasone. However, we would like to express
reservations about the “safe”, use of dexamethasone given the known risk
of osteonecrosis of the femoral head with short-term steroid
administration (1,2). Osteonecrosis (avascular necrosis) of the femoral
head can be crippling, especially in the younger patient and prevention is
obviously preferable. Other concerns with using this regime are that
further side-effects of corticosteroid administration include poor healing
and thromboemobolism (3), maybe indicating that this combination of drugs
wouldn’t be suitable for patients undergoing joint replacement operations
who, due to the nature of the surgery are already at increased risk of a
deep vein thrombosis and where wound breakdown would be potentially
disastrous. Patients on long term therapy with corticosteroids obviously
need steroid cover perioperatively to prevent adrenal insufficiency, but
in patients where the administration of steroids is of dubious use, the
risks and the benefits must be weighed up and alternatives used where
appropriate. It may be good practice for physicians and anaesthetists to
discuss the administration of steroids with the operating surgeon.
References:
1. McKee MD, Waddell JP, Kudo PA, Schemitsch EH, Richards RR.
Osteonecrosis of the femoral head in men following short-course
corticosteroid therapy: a report of 15 cases. CMAJ January 23, 2001; 164
(2) 205-6.
2. Gunal I, Karatosun V. Avascular necrosis of the femoral heads after
single corticosteroid injection CMAJ July 4, 2006; 175 (1).
3. British National Formulary. British Medical Association. Royal
Pharmaceutical Society of Great Britain.
Competing interests:
None declared
Competing interests: No competing interests
Dear Dr Gelbrich
Thank you for your response. It is reassuring to know that the
tachycardia and hypotension last for less than 5 minutes after high dose
metoclopromide. With regards to the incidence of these adverse events,
reading the first 2 lines of the adverse events section of your article it
appears that the incidence of tachycardia or hypertension is more than 9%
with 50mg metoclopromide.
Yours sincerely
Competing interests:
None declared
Competing interests: No competing interests
Sir,
Thank you for your comment. We think that part of the variables you
mentioned are predictors of PONV, we are still analysing this. The matter
is complicated because of the multitude of correlations between the
variables involved. To mention only a simple example, body height is
associated with PONV, because sex is associated with both height and PONV.
In multiple regression analysis, the "influence" of body height on PONV
vanishes. However, there are more tricky relationships that can not be
clarified by a single regression model. Repeated analysis and discussion
is required, and we try to avoid immature statements.
However, by confounding we understand the following. The frequencies
of a certain (side) effect in the randomised treatment groups may be
different. This may induce different concomitant treatment or diagnostic
procedures or surveillance of the subjects. If the concomitant treatment
affects the primary outcome of interest, the estimate for the effect of
the randomised treatment may be biased. As well, additional diagnostics or
more intense observation may lead to observation bias. We think that the
variables you mentioned are, in part, covariables of PONV, but not
confounders in this sense.
In particular, anaesthetic procedures may vary due to different house
policies in the participating clinics. Since we stratified randomisation
by type of surgery and centre, we indirectly stratified also by certain
procedural variables of anaesthesia. Thus such variables, possibly
affecting PONV, were not statistically associated with the dose of
metoclopramide, and hence they were not confounders in the sense described
above.
Regarding BMI, we presented (see long version on bmj.com) that
obesity (as a binary variable) was not significantly related to PONV. We
also tried to find any association of BMI and PONV, including nonlinear
and even nonmonotoneous relationships, but we did not find any.
Competing interests:
None declared
Competing interests: No competing interests
Sir,
Indeed, we also suspected that there might be superadditive effects
of the combination of dexamethasone with metoclopramide. Unfortunately, to
discover such effects (called interaction of factors by the
statisticians), one needs much larger sample sizes (for example, the
fourfold sample size for a 2-by-2 factorial design, compared to a 2-armed
design, when the superadditive gain is of the same size like the effect of
a single drug). To demonstrate superaddition of 8 mg dexamethasone and 50
mg metoclopramide, we estimated n=7000 to be needed which would have been
infeasible for us. Therefore we dropped interaction and focussed on the
issue of highest immediate clinical relevance.
The only chance to reduce sample size would be to include only
patients at high baseline risk. However, to demonstrate superadditivity
requires a factorial design including a placebo group. In view of cheap
and effective options of prophylaxis of PONV, we would consider this
unethical in a patient population with particular high risk of PONV. As a
consequence, we think we need (possibly international) megatrials.
Competing interests:
None declared
Competing interests: No competing interests
Madam,
Thank you for your comment reflecting the patient's view, supporting
the opinion that postoperative nausea and vomiting (PONV) is not at all a
minor point.
There are several options for the prophylaxis of PONV (see the
response by John B Carlisle). None of them is magic, and so isn't
metoclopramide; each of the drugs has its contraindications and side
effects. In particular, metoclopramide is contraindicated in case of
mechanic obstruction of the bowels. Furthermore, these drugs reduce the
rate of PONV, and combinations reduce it drastically, but we can not
promise anybody in advance that he or she will definitely be free of PONV
when receiving certain drugs.
However, both physicians and patients should be aware that PONV is a
major side effect of surgery but we do not need to just accept its
occurrence. By an appropriate policy of prophylaxis, beginning with
clarifying the individual risk profiles of the patients, the majority of
PONV can be avoided at reasonable cost and low rate of adverse effects.
Competing interests:
None declared
Competing interests: No competing interests
EDITOR ¡V It is interesting how anti-emesis philosophy comes full
circle with the observations by Wallenborn et al that higher doses of
metoclopramide are effective prophylaxis against postoperative nausea and
vomiting (PONV) 1. The authors comment that metoclopramide has a complex
mode of action, binding to both dopamine and 5-HT receptors. In fact it
has been known for some time that the effectiveness of higher doses of
metoclopramide against emesis induced by the chemotherapeutic agent
cisplatin is secondary to its properties as a 5-HT3 receptor antagonist.
Indeed, the development of selective 5-HT3 receptor antagonists as a new
class of drug became possible as a result. In recent years we have seen a
vast amount of research into newer more expensive anti-emetics for use in
PONV yet most people now accept that until efficacious ¡¥final pathway¡¦
anti-emetics become available, the future of PONV management lies in
combining anti-emetic drugs with different receptor actions. Doses of
metoclopramide between 1 and 3 mg/kg which have been used in cancer
chemotherapy patients 2,3 result in 5-HT3 receptor antagonism in animal
studies 4. Clearly doses in the range of 25-50 mg fall somewhat short of
this but still the 5-HT3 receptor blocking profile of metoclopramide
should not be ignored. It may even be that a cheap and effective drug for
PONV with the combined actions of a dopamine and 5-HT3 receptor antagonist
has been available for years but never been used at the right dose. On a
cautionary note however, Wallenborn¡¦s subjects all received
metoclopramide in combination with dexamethasone. Since the additive (even
synergistic) effects of combining 5-HT3 antagonists with dexamethasone are
becoming more widely appreciated, Is it possible that the differences in
effectiveness between doses of metoclopramide in the study may not have
been as marked had they been administered in isolation ?
1 Wallenborn J, Gelbrich G, Bulst D et al. Prevention of
postoperative nausea and vomiting by metoclopramide combined with
dexamethasone: randomised double blind multicentre trial. BMJ 2006;
333:324-7
2 De Mulder PH, Seynaeve C, Vermorken JB et al. Ondansetron compared
with high dose metoclopramide in prophylaxis of acute and delayed
cisplatin-induced nausea and vomiting. A multicenter, randomized, double-
blind crossover study. Ann Int Med 1990; 113:834-40
3 Harrington RA, Hamilton CW, Brogden RN. Metoclopramide. An updated
review of its pharmacological properties and clinical use. Drugs
1983;25:451-94
4 Gullikson GW, Loeffler RF, Virina MA. Relationship of serotonin-3
receptor antagonist activity to gastric emptying and motor-stimulating
actions of prokinetic drugs in dogs. J Pharmacol Exp Ther 1991;258:103-110
Competing interests:
None declared
Competing interests: No competing interests
The combination of dexamethasone and ondansetron is one which is
commonly used for the prevention of postoperative nausea and vomiting
(PONV) in UK hospitals.
Previous systematic reviews have shown metoclopramide to be
ineffective as an antiemetic at a standard dose of 10mg, with ondansetron
on the otherhand being efficacious(1)(2). Trials comparing the two agents
directly have gone largely in favour of the 5-HT antagonist, but in many
of these metoclopramide was tested at the 10mg dose only.
The efficacy of metoclopromide over ondansetron when employed in
combination with dexamethasone at doses used in the study by Wallenborn
and colleagues (3) would provide further support for a return to its use
in preventing PONV. Of note, the use of metoclopromide at the maximum dose
of 50mg as used in this study would still be approximately 3 times less
expensive than the equivalent dose of steroid and 5HT-3 antagonist (4). In
view of the increasing financial strain within the NHS and the frequency
of surgical procedures, this change in prescribing would be a significant
one.
1. Henzi I, Walder B, Tramer MR.Metoclopramide in the prevention of
postoperative nausea and vomiting: a quantitative systematic review of
randomized, placebo-controlled studies. Br J Anaesth. 1999 Nov;83(5):761-
71.
2.Tramer MR, Reynolds DJ, Moore RA, McQuay HJ.Efficacy, dose-
response, and safety of ondansetron in prevention of postoperative nausea
and vomiting: a quantitative systematic review of randomized placebo-
controlled trials. Anesthesiology. 1997 Dec;87(6):1277-89
3. Wallenborn et al, Gelbrich G, Bulst D, Behrends K, Wallenborn H,
Rohrbacn UK, Kuhnast T, Wiegel M, Olthoff D. Prevention of postoperative
nausea and vomiting by metoclopramide combined with dexamethasone:
randomised double blind multicentre trial. BMJ 2006; 333: 324 –7
4. British National Formulary 2006
Competing interests:
None declared
Competing interests: No competing interests
Author's reply II: The Concerns of The Orthopaedic Surgeon.
Ladies and Gentlemen,
We warmly recommend the review paper by Salerno and Hermann [1] which
contains a section dealing with the safety of low-dose corticosteroids,
including the problem of wound healing, as well as other popular concerns
against single low-dose corticosteroids.
Retching associated with nausea and vomiting may also, by mechanical
stress, jeopardise wound healing. Thus avoiding the benefit of a drug may
also be harmful. This should be kept in mind when speaking about the
safety of the patients.
Apfel et al. [2] randomised over 5000 patients, half of them received
dexamethasone (4 mg i.v.). They did not report delayed wound healing or
increased risk of thrombosis with dexamethasone. Of course, they could not
observe whether there is a relationship to osteonecrosis because this
would occur beyond the observation time in the study. However, if there is
a reason to believe that low-dose corticosteroids may cause osteonecrosis,
a follow-up investigation might be initiated right now. This would provide
evidence. Other suspected adverse effects could be examined at the same
time.
Evidence would arise from a statement like this: "In a consecutive
review of all cases delivered to our unit, we found 15 patients with
osteonecrosis who received and A cases who did not receive low-dose
corticosteroids. Among the patients without osteonecrosis, there were B
cases who received and C cases who did not receive low-dose
corticosteroids. The ratio 15/A is (significantly) larger than the ratio
B/C." This would not ensure causality, but at least it can demonstrate a
numerical relationship. The collection of case reports by McKee et al. [3]
is not evidence as it does not provide information about the numbers A, B
and C (where B and C obtained from a well-defined subset of control
patients would also be acceptable).
What does the matter have in common with black cars? The study in
mind "proving" inceased risk of myocardial infarction after driving a
black car [4] is, of course, nonsense. I have chosen this example because
the nonsense is obvious. In other situations, the nonsense may be less
obvious, but an analogy may help to understand the point. When replacing
"myocardial infarction" with "osteonecrosis" and "driving a black car"
with "low-dose corticosteroid therapy", we obtain reference [3]. The
methodology is exactly the same, and hence, nonsensical.
References:
1. Salerno A, Hermann R. Efficacy and safety of steroid use for
postoperative pain relief. Update and review of the medical literature. J
Bone Joint Surg Am. 2006;88: 1361-72.
2. Apfel CC, Korttila K, Abdalla M, Kerger H, Turan A, Vedder I, et
al. A factorial trial of six interventions for the prevention of
postoperative nausea and vomiting. N Engl J Med 2004;350: 2441-51.
3. McKee MD, Waddell JP, Kudo PA, Schemitsch EH, Richards RR.
Osteonecrosis of the femoral head in men following short-course
corticosteroid therapy: a report of 15 cases. CMAJ 2001;164: 205-6.
4. Gelbrich G. Author's reply: The Concerns of The Orthopaedic
Surgeon. Rapid response, bmj.com, 2 Sep 2006.
Competing interests:
None declared
Competing interests: No competing interests