Healthy Skepticism AdWatch

Dear “Senior Nexium experts”,

Thank you for your comments about AdWatch on Nexium.  If you do not tell us otherwise, we will assume that you accept all our points below.

Desire

The Nexium advertisement appeals to our desire for power

It is good that we agree that your Nexium advertisement appeals to doctors�(tm) normal desire for power.  However there are some problems:

A1.  Our main point was that appeals to desires are more powerful for influencing decision making than many people realise.

A2.  Your “power” claims have many possible meanings and connotations for the reasons explained immediately below.  One of the common ways to mislead is to use statements that have many meanings.  (See: Copi IM, Cohen C. Introduction to Logic. 9th Edn. Macmillan New York 1994)

A3.  The word “power” has many meanings and connotations.  For example the Merriam Webster Colligate Dictionary (on the Encyclopaedia Britannica 2004 CD) lists 21 meanings as a noun, 4 as a verb and 2 as an adjective as well as 34 two word expressions that include the word power.  These many meanings and connotations make the word “power” emotive and ambiguous.  By comparison word “effective” is more specific with only 5 meanings as an adjective, one as a noun and one two word expression: cost-effective.  It has been well understood, since at least the time when the ancient Greek oracle stories were first told, that it is common for people to interpret ambiguous statements to mean what they hope for.  One of the many meanings that Australian GPs (the target of your advertising) may take from your “power” claims is that Nexium has a clinically worthwhile effectiveness advantage.  Another possible meaning is that Nexium may enhance doctors�(tm) personal power especially over patients.  Such appeals to desires can influence behaviour without us being aware of it.

(See:
Mansfield PR. Healthy Skepticism�(tm)s new AdWatch: understanding drug promotion. Med J Aust 2003; 179 (11/12): 644-645 )

A4.  To choose a drug (or a car) on the basis of “power” alone without consideration of safety and costs would not be good decision making.  Health professionals, patients and the public need to know in what situations is Nexium cost-effective, if any.  Your advertisement does not provide the balanced information required for good decision making.  It is relevant that Australian state governments are considering a proposal to save lives by banning advertising for cars that focuses on speed and power without consideration of cost and safety.  (See: Move to ban irresponsible car ads. The Australian 2003: 8 November.)  Banning such advertising for drugs would probably also save lives mostly via reduced opportunity costs.

A5.  Your advertising copy does not specify the indication for Nexium 40mg but gastro-oesophageal reflux disease is mentioned in the compulsory Pharmaceutical Benefits Scheme information.  Because there is no qualification, your advertisement is open to the interpretation that the claims apply to all cases of gastro-oesophageal reflux disease.  If there is not evidence to support your claims for all such cases then the advertisement is misleading.

A6.  Your advertisement claims that “Nexium 40mg offers greater healing power that either omeprazole or lansoprazole”.  GPs may expect this claim to be based on comparison with appropriate doses of the other drugs.  The fact that the comparison was with omeprazole 20mg and lansoprazole 30mg is only disclosed as a footnote in small print in the fine print section of the advertisement.  People rarely read the fine print sections of advertisements.  (See: Ogilvy D. Ogilvy on advertising. London: Pan 1983)  The use of footnote disclaimers can breach section 52a of the Australian Trade Practices Act.  “These qualifications usually appear close to the lead selling point… the copywriter may be trying to trade on positive reactions to the selling point, while trying to keep within the law by putting the conditions in the fine print. This may not protect that business from breaching the Act.”

(See:
Advertising and Selling. Australian Competition and Consumer Corporation 2001 http://www.accc.gov.au/content/item.phtml?itemId=303213&nodeId=file3f61525563df9&fn=Advertising and Selling.pdf )

A7. An advertising claim that “Nexium 40mg may be more effective for symptom control in severe gastro-oesophageal disease than lower doses of omeprazole or lansoprazole” might be less effective for increasing sales (overuse) of Nexium but would be less misleading.

A8. In conclusion your Nexium advertisement targeting general practitioners is open to the interpretation that Nexium 40mg is the best choice for all cases of gastro-oesophageal disease because of a clinically worthwhile effectiveness advantage over appropriate comparison doses of other PPIs.  Because that is not the case for the less severe disease commonly seen in general practice (as will be discussed below) regardless of your intentions your “power” claims are misleading.

Unfair comparisons

Nexium, 40 mg has been compared with omeprazole, 20 mg, and a larger dose of omeprazole would be just as powerful.

B1.  Your advertisement does not disclose that Nexium is not a new drug but is only an isomer of omeprazole.  Regardless of your intentions that omission is misleading.

B2.  Some isomers (enantiomers) have been shown to be clinically superior to their mirrors at the same dose but some important examples have not including escitalopram and esomeprazole (Nexium).

(See: 
Svensson S, Mansfield PR. Escitalopram: superior to citalopram or a chiral chimera? Psychother Psychosom. 2004 Jan-Feb;73(1):10-6.
Mansfield PR, Henry D, Tonkin A. [editorial] Single-Enantiomer Drugs: Elegant Science, Disappointing Effects. Clin Pharmacokinet 2004; 43 (5): 1 )
B3.  We would appreciate your comments on the following:
“All proton pump inhibitors exist as two inactive enantiomers (prodrugs) that are converted to active moieties which equally inactivate the H+ /K+ -ATPase pump. Both enantiomers of omeprazole are equipotent, however, their metabolism is different. (R)-omeprazole is mainly metabolised by the polymorphic CYP2C19 enzyme. There is a 7.5-fold difference in the systemic exposure to (R)-omeprazole in patients who are poor metabolisers compared to extensive metabolisers. With (S)-omeprazole this difference is reduced to about three-fold so it was argued that use of esomeprazole would be associated with less interindividual variability in efficacy. However, there are few data to support this theoretical advantage, especially when only 3% of the Caucasian population are poor metabolisers. There may be a benefit in the Asian population where the incidence of poor metabolisers is about 20%. A rationale for chiral switching to esomeprazole might therefore be based on ethnic differences in metabolism.”

(See:
Somogyi A, Bochner F, Foster D. Inside the isomers: the tale of chiral switches. Aust Prescr 2004;27:47�“9
www.australianprescriber.com/index.php?content=/magazines/vol27no2/47_49_isomers.htm )

B4.  Do you have any evidence to show that omeprazole is less effective or clinically significantly less well tolerated at 40mg than at 20mg?

B5.  To show that a Ford car is more powerful than a Toyota car, for pulling loads, the appropriate comparison is one Toyota car (the standard dose) vs one Ford car.  Comparing one Toyota car vs two Ford cars (pulling the same load together) is not appropriate unless they are equal on adverse effect or costs.  Similarly to show that an isomer is more effective than a racemic mixture the appropriate comparison is the two drugs in the same total dose unless some other dose is equal on adverse effects or costs.  To confirm or deny the alleged kinetic advantage of S-omeprazole (Nexium) over R-omeprazole would require a comparison of Nexium with the same dose of R-omeprazole and/or omeprazole.  Similar arguments apply to the comparison of Nexium with lower doses of other drugs.  Consequently we do not agree that our comments were misleading.  By contrast, your advertisement is misleading for the reasons given here and in A6 above.

B6.  Mechanisms such as the level of acid control are only important when they lead to clinically important differences in clinically important endpoints.  Further, statistical significance is only useful if the difference is also clinically significant.

B7.  Given the findings by Andersson et al (See: Pharmacokinetics and pharmacodynamics of esomeprazole, the S-isomer of omeprazole, Aliment.Pharmacol.Ther 2001; 15:1563-69) that there is a good correlation between AUC and effect for esomeprazole, as has previously been reported for omeprazole, it follows that increasing the AUC of omeprazole (by increasing the dose) will lead to a higher AUC and greater acid suppression and thus no clinically worthwhile difference.

B8.  None of your arguments about Nexium vs omeprazole are convincing because they are all beside the point.  Our point is that “a larger dose of omeprazole would be just as powerful” [as Nexium 40mg].  To learn more about informal fallacies of irrelevance please read any introductory textbooks on logic such as:
Copi IM, Cohen C. Introduction to Logic. 9th Edn. Macmillan New York 1994

B9.  Omeprazole 80mg would be at least as effective as Nexium 40mg because it would contain Nexium 40mg (S-omeprazole).  It would also contain R-omeprazole 40mg.  R-omeprazole may not add much to effectiveness at such a high dose where the dose response curve becomes very flat.  However you have not given us any reason to believe that the presence of R-omeprazole would reduce the effectiveness of the S-omeprazole (Nexium).  Because R-omeprazole is also an active pro-drug and because of the relatively flat dose response curve it is likely that a dose of omeprazole smaller than 80mg would be clinically equivalent to Nexium 40mg.  You have not provided any evidence of any clinically significant superiority of S-omeprazole over R-omeprazole, despite having sufficient funds and significant financial incentives for doing so.  Consequently we are justified in believing that it is likely that the dose of omeprazole that reaches clinical equivalence with Nexium 40mg is close to omeprazole 40mg.

Consequently, our point that “a larger dose of omeprazole would be just as powerful” [as Nexium 40mg] is justified.  Regardless of your intentions, had we been convinced by your arguments above we would have been misled.

The difference in the percentage of oesophagitis patients healed with Nexium�’, 40 mg, compared with lansoprazole, 30 mg, is too small to be clinically significant.

C1.  It appears that you agree with us that for people with less severe disease (who are the majority of patients seen in general practice) “the difference between the percentage of oesophagitis patients healed with Nexium 40 mg, compared with lansoprazole, 30 mg, is too small to be clinically significant.”  We concede that there may be a clinically important difference in more severe disease.  We are uncertain about that because you have not provided a systematic review of all the relevant evidence.  You have not provided any evidence about whether or not that is also true for Nexium 40mg vs lansoprazole 40mg.  Your advertisement did not specify that your claim of superiority only applied to more severe disease so it would be reasonable for GPs to interpret your claims as applying to all cases especially the less severe cases that we commonly see.  We conclude that regardless of your intentions your advertisement is misleading.

C2. You have used several methods for conveying treatment effect sizes without providing 95% confidence intervals.  To help understand why doing so is misleading please read about representativeness bias.  This is the bias that arises from assuming that the results in a sample closely represent the situation in the parent population as a result of underestimating the magnitude of the contribution of chance to those results.

(See:
Elstein AS. Heuristics and biases: selected errors in clinical reasoning. Acad Med 1999;74:791-4.
Kahneman D, Slovic P, Tversky A. Judgment under uncertainty: heuristics and biases. New York: Cambridge University Press 1982. )

Missing comparisons

The advertisement does not compare Nexium with other alternatives

D1. It appears that you agree with our point that “the advertisement does not compare Nexium with other alternatives”.

D2. It is possible to produce effective advertising that is more helpful than your Nexium advertisement. We would be happy to assist you.

D3. It would be interesting to compare Nexium 40mg alone vs Nexium 20mg plus lifestyle changes because the dose response curve is flatter for the second 20mg than for the first 20mg.

D4. We concede that Nexium 40mg may have clinically worthwhile advantages over pantoprazole for more severe cases. We are uncertain about that because we are aware of contradictory evidence that you did not disclose.  (See: Scholten T, Gatz G, Hole U. Once-daily pantoprazole 40 mg and esomeprazole 40 mg have equivalent overall efficacy in relieving GERD-related symptoms. Aliment Pharmacol Ther. 2003 Sep 15;18(6):587-94. )  Either you were not up to date with the published literature on your product or you were deliberately trying to mislead us by not disclosing relevant information or there is some other explanation that is not obvious to us.

D5.  Our recommendations have taken into account the possibility that some patients’ symptoms may be better controlled on Nexium 40mg than the other PPIs available.

D6.  Many patients with mild symptoms are treated in general practice and never seen by the specialists who tend to dominate guideline writing groups.  Some patients with mild symptoms will achieve adequate symptom control with lifestyle changes such as elevating the head of the bed and where applicable quitting smoking and reducing alcohol intake. Others will do well with such lifestyle changes plus a H2 receptor antagonist.  We do not know what percentages of patients would achieve adequate symptom control with such therapy because we have not been able to find a relevant study.  One of the many adverse effects of allowing drug companies to dominate research funding is that the studies that clinicians need are often not funded because they would involve interventions that are not profitable for drug companies.  The nearest relevant study that we are aware of is a 1987 study of raising the head of the bed and/or rantidine that found that the combination was more effective than either alone for severe oesophagitis.

(See:
Harvey RF, Gordon PC, Hadley N, Long DE, Gill TR, Macpherson RI, Beats BC, Tottle AJ. Effects of sleeping with the bed-head raised and of ranitidine in patients with severe peptic oesophagitis. Lancet. 1987 Nov 21;2(8569):1200-3.)

D7.  Patients who have moderate or severe symptoms will achieve symptom control more quickly with PPIs than the interventions mentioned in D6 above.  After initial PPI therapy has controlled the symptoms then the life style changes mentioned above may provide adequate maintenance of symptom control for some.  Some will also require a H2 antagonist, others a low dose PPI and others a high dose PPI.  For those who require ongoing drug therapy many will achieve symptom control with less “powerful” drugs or doses if they implement the lifestyle changes as well.  Again we do not know the percentages.  The money saved by reducing to the cheapest effective therapy can be used elsewhere in the health system for much greater benefits.

(See:
Proton pump inhibitors: too much of a good thing? NPS News 2004; 33 http://www.nps.org.au/site.php?content=/html/news.php&news=/resources/NPS_News/news33 )

Many people with peptic ulcers are better treated with H. pylori bacteria eradication therapy. You clearly agree with us on this point.  The advertisement would be less misleading if that point was added.

Conclusions

All our concerns are justified.

We did make one error.
Instead of writing:
“The difference in the percentage of oesophagitis patients healed with Nexium 40 mg compared with lansoprazole 30 mg is too small to be clinically significant.”
we should have written:
“In many patients seen in general practice who have mild or moderate disease the difference between Nexium 40 mg and lansoprazole 30 mg is too small to be clinically significant.”

We would have made that change if you had pointed out the need to us during the months between the publication of the prototype in June 2003 and publication of AdWatch on Nexium in October 2003.

Our careful study of your advertising for Nexium leads us to conclude that regardless of your intentions your advertising is misleading.  If it is your intention to mislead then there is no need for further dialogue.  If we do not hear back from you with in 2 weeks we will report your advertising to the appropriate authorities.  However we hope that you will agree that if doctors have developed incorrect impressions about a drug as a result of advertising then measures should be taken to correct those impressions.  Therefore we would like to work with you to develop a new set of advertisements that would run until surveys of doctors show that any mistaken beliefs have been corrected.

Yours sincerely,

Dr Peter R Mansfield
on behalf of the Healthy Skepticism AdWatch Team

1. Rogers WA, Mansfield PR, Braunack-Mayer AJ, Jureidini JN.  The ethics of pharmaceutical industry relationships with medical students. Med J Aust. 2004 Apr 19;180(8):411-4. 

2. Mansfield P, Henry D, Tonkin A. Single-enantiomer drugs : elegant science, disappointing effects. Clin Pharmacokinet. 2004;43(5):287-90. 

3. Jureidini JN, Doecke CJ, Mansfield PR, Haby MM, Menkes DB, Tonkin AL. Efficacy and safety of antidepressants for children and adolescents. Brit Med J 2004;328:879-83

4. Svensson S, Mansfield PR. Escitalopram: superior to citalopram or a chiral chimera? Psychother Psychosom. 2004 Jan-Feb;73(1):10-6. 

5. Jureidini J, Mansfield P, Menkes D. The statin wars. Lancet 2003 Nov 29; 362(9395)1854  

6. Melissa Sweet. Website launched to expose “tricks” of drug ads. BMJ 2003;327:936

7. Mansfield PR. Healthy Skepticism�(tm)s new AdWatch: understanding drug promotion. Med J Aust 2003; 179 (11/12): 644-645

8. Scholten T, Gatz G, Hole U. Once-daily pantoprazole 40 mg and esomeprazole 40 mg have equivalent overall efficacy in relieving GERD-related symptoms. Aliment Pharmacol Ther. 2003 Sep 15;18(6):587-94. 

9. Labenz J on behalf of the EXPO study group, Keeling N, Eklund S. A comparison of esomeprazole 40 mg once-dail and pantoprazole 40 mg once-daily for the healing of reflux esophagitis. Gut 2003;52 Suppl VI:A241.