Special Meeting 28 October 2009
Convened pursuant to Rule 3 of the
Constitution of the Advertising Standards Complaints Board
Complainant: AFT Pharmaceuticals Limited
Advertisement: GlaxoSmithKline Consumer Healthcare Australia
Complaint: Two print and one website advertisement promoted the new product
Advertisement 1: The magazine advertisement appeared in “Pharmacy Today”. It
contained text, in bold letters at the head of the advertisement, which said “37%
STRONGER THAN REGULAR PARACETAMOL1”. Directly below this heading was
a large pack shot of the new product. Fine print at the foot of the advertisement said,
in part, “Reference: 1. Laska EM et al. JAMA 1984; 251:1711-18”.
Advertisement 2: The flyer advertisement contained a pack shot of the Panadol
Extra product on one side of the flyer, and the other side was headed “37% MORE
POWERFUL THAN REGULAR PARACETAMOL1”. Text beneath this said “Because
Panadol Extra is 37% more powerful than regular paracetamol it provides extra pain
relief and helps you break through the pain barrier”. Fine print at the foot of the
advertisement said, in part, “Reference: 1. Laska EM et al. JAMA 1984; 251:1711-
Advertisement 3: The website advertisement was headed “BREAK THROUGH THE
PAIN BARRIER AND YOU COULD WIN $5,000”. A smaller heading beneath this
said “ABOUT PANADOL EXTRA”. Text beneath this included “Panadol Extra has
been formulated to relieve strong pain. It combines paracetamol with caffeine for
37% extra pain relief…”
Complainant, AFT Pharmaceuticals Limited, said:
“AFT Pharmaceuticals has been in discussion with GSK for well over 3 weeks with
regards to advertising claims for Panadol Extra.
AFT has also consulted additional independent experts such as an Associate
Professor of Statistics and a very experienced consultant who was previously head of
Global Research for a major multinational Pharmaceutical Company.
These experts concur with AFT's view that the advertising is misleading
However GSK is of the opinion that its position is correct and neither AFT nor
independent experts that we have consulted are correct in relation to the key concern
of the 37% stronger claim being misleading.
Therefore AFT sees no option other than to refer the complaint to an independent
arbitrator such as the ASA to seek a resolution.
This advertising campaign is currently being run by GSK so we seek an urgent
hearing since GSK have declined to halt the advertising.
GSK have raised a couple of points which we feel have no relevance to our
The 37% stronger claim (or one essentially similar to it) ……… has been made
Our contention is that this does not invalidate or diminish a valid complaint. In fact if
the claim is incorrect then this makes the issue even more serious given the length of
GSK raises a concern that AFT has no apparent competitor
AFT does not wish to discuss any details of its commercial activities but
nevertheless we contend that there is still a serious issue to be considered. If it is
true that consumers are being subjected to misleading advertising then I am sure
that GSK would also agree that this in itself would be a serious issue. AFT contends
that its concerns cannot be validly dismissed on this basis....
The crucial issue is that the use of potency to imply efficacy is misleading in GSK's
advertising. We believe that this breaches the Advertising Standards Code.
The specific subject of our complaint is
ITEM 1: Advertising to Health. Professionals in September Pharmacy Today
ITEM 2: Advertising to Consumers sourced from Foodtown Takapuna (Attachments
We believe that the advertising breaches the below sections of the code:
R4.3 Scientific information
Scientific information within an advertisement must be presented in an accurate
manner. Scientific terminology must be appropriate, clearly communicated and able
to be readily understood by the audience to whom it is directed.
Advertisements must not directly nor by implication, omission, ambiguity,
exaggerated claim or comparison:
(a) mislead or deceive, or be likely to mislead or deceive
ADVERTISING DIRECTED TO HEALTHCARE PRACTITIONERS
Advertisements must contain truthful and balanced representations and claims
that are valid and have been substantiated
Advertisements must not directly nor by implication, omission, ambiguity or
comparison mislead or deceive, or be likely to mislead or deceive. Claims and
representations made in advertisements must be truthful and have been
Potency refers to a ratio of doses to achieve the same analgesic effect rather than
any improved (stronger) efficacy result. To claim that Panadol Extra is 30% (we
understand to be historical advertising claims) or 37% stronger (current claims) is
misleading to either healthcare professionals or consumers in that there is a clear
inference that it would provide 37% more efficacy which is normally perceived to
mean analgesia. This is not the case since the claim is only supported by a paper
reporting relative potency.
This concern is strongly supported by the consumer promotion in Attachment 2 which
refers to a supporting website (Attachments. 34). The website contains the claim
"Panadol Extra …… it combines paracetamol with caffeine for 37% extra pain relief
1" Reference 1 in turn refers to the key reference Laska EM 1984 JAMA 251: 1711-l8
which estimates relative potency.
However it seems clear that GSK either is deliberately confusing the potency and
efficacy claims or the promotion has confused even the GSK script writers into
confusing efficacy with potency which we contend is the whole point of this
Efficacy is commonly considered as TOTPAR results which are derived from head to
head studies. This is considered more relevant to results a patient would receive i.e.
if a patient took Paracetamol or Paracetamol + Caffeine then what sort of pain relief
would they receive. In this context relative potency has no relevance whatsoever as
patients take the doses manufactured by GSK i.e. Paracetamol 500mg or
Paracetamol 500mg + Caffeine 65mg or multiples thereof.
If we consider other literature as provided by GSK including GSK sponsored studies
(Ali Z et al 2007 Curr Med Res 23(4): 841-851) we get the following results:
Results listed in Table 2, page 846
TOTPAR Paracetarnol +Caffeine- = 2.89
TOTPAR Paracetarnol Alone = 2.62
Patients with a higher TOTPAR have better pain relief. This could be expressed as
2.89/2.62 = 1.10 whereby the patients treated with the combination experience a
10% improvement in TOTPAR than patients treated with paracetamol alone.
This is more representative of improved pain relief than a relative potency claim of
37%. GSK advertising should state 10% better pain relief than standard paracetamol
not 37% stronger: AFT contends that GSK should use this figure in their advertising
and not 37%.
GSK have provided various calculations which are not helpful in presenting the
relative analgesic relief that a consumer would experience. The calculation presented
by GSK in their letter of 9th September (AFT Attachment 8 and further detailed in their
Attachment 1) is misleading in this respect.
GSK state the following "The Ali et al paper further supports the GSK claim by
showing a 90%, 143% and 56% adjuvant effect for total pain relief cramping pain and
GSK support this contention by subtracting the placebo TOTPAR results from the
paracetamol or combination results and then using these transformed figures to
arrive at the above “justification". This is again misleading when used in the context
of presenting the additive effect of a combination analgesic to a consumer or health
To illustrate this further, this is similar logic as saying the average person runs 100M
in 12 seconds. If the GSK competitor can run it in 14 seconds and AFT competitor 13
seconds, then subtract 14-12 and compare it with 13-2 and claim that AFT is 50%
faster. We contend that this or any similar permutation of this is misleading
manipulation of data.
GSK supplied AFT with a number of papers to support their position as follows
(Included in AFT Attachment 8). We discuss each in turn as agreed.
Reference: Lasko EM 1984 JAMA 251: 1711-1.8
Rather than "just" rely on AFT we asked an experienced statistician to comment. We
showed Dr Frampton the advertisement and the paper. His comments are in
Attachment 5. It is interesting to note that his comments on the paper and your use
thereof are definitely not supportive of the GSK case. We also append a copy of Dr
Frampton's CV to Attachment 5 for your information.
Regardless of minor points later made by GSK over his analysis, the key point is still
the issue of potency and strength as outlined above.
Notwithstanding this there are still serious methodological flaws in the paper. To
quote one example - Table 4 lists Acetaminophen-caffeine relative potency as being
derived from studies 24-29. Table 2 then appears to list study 29 as the largest study
of the group 24-29. However the note against it relating to the footnote then states
"Deviations from parallelism"
What is also interesting is that the copy that GSK provided has somehow managed
to delete this important footnote in the copy that it supplied to AFT (see Attachment 6
& 7). This was certainly unhelpful to any attempt to address the paper regardless of
protestations of clerical errors.
The relevance of the above point is that the relative potencies from SPID cannot be
estimated in the event that there are not parallel dose response curves. As Dr
Frampton states anyway SPID is not a suitable way to derive this data.
There are a number of genuine methodological concerns and problems with the
approach of this paper. This is all consistent with our contention that this paper and
its approach is out of date. Furthermore it is unusual that a company should choose
to rely on such an old publication when newer publications are available which
provide a more accurate and accepted representation of endpoints which in this case
is analgesia rather than an irrelevant endpoint to patient pain relief such as potency.
Reference: Zhang WY & Li Wan Po A (1996). Analgesic Efficacy of paracetamol and
its combination with codeine and caffeine in surgical pain - a meta-analysis. J Clin
Pharm Ther 21: 261-282
The conclusion of this meta-analysis is as follows: "While the meta-analysis does not
exclude the possibility of small effects in favor of paracetamol-codeine combinations
and paracetamol-caffeine, it can be concluded that the additive do not add much to
the value of paracetamol.”
Although in the GSK letter of 9th September, they attempt to make the case that
somehow AFT has attempted to selectively quote from the paper we are merely
quoting the penultimate conclusion which is by no means selective. This conclusion
is inconsistent with the inferred (and in one place actual claim of analgesia) of 37%
Reference: Zhang WY (2001) A Benefit-Risk Assessment of Caffeine as an
Analgesic Adjuvant Drug Safety 2.4(15): 1127-1142
Again Iooking at the conclusions seems to put this into perspective
"Caffeine enhances pain relief in headache, as well as increases the risk of
nervousness and dizziness. In other types of pain such as postpartum pain, dental
pain, postoperative pain, rheumatic and cancer pain, the adjuvant effects of caffeine
Furthermore the risks of caffeine are also put into perspective in the conclusion
"There is some epidemiological evidence regarding increased risks of drug-induced
headache, withdrawal headache and neuropathy associated with long-term overuse
of analgesics, particularly phenacetin combinations. However little is known about
caffeine effects in the development of these conditions except that it is widely co
formulated with analgesics"
AFT notes that the comment on phenacetin has no relevance to Panadol Extra.
However the review is consistent with a small positive increase in efficacy with some
downside due to adverse events. However again this is again inconsistent with the
37% claim made by GSK.
Reference: Laska EM et al 1983. Effect of caffeine on acetaminophen analgesia.
Clin Pharmacol Ther 33(4): 498-509
The reference reports the results of 4 studies and relative potencies. Table III does
give some indication of pain relief through "TOTAL" which is the weighted sum of the
relief observations. Reported treatments means for the combination were 9.6, 9.3,
9.4 & 12.1 and for acetaminophen alone 8.1, 8.8, 9.1 and 10.9
Calculating the mean improvement in pain relief (TOTAL combo/TOTAL para) results
in a value of about 1.096 or a % improvement of 9.6% which again is considerably
different from a value of 37%
Reference: Winter L et al 1983. A double blind comparative evaluation of
acetaminophen, caffeine, and the combination of acetaminophen and caffeine in
outpatients with post-operative oral surgery pain. Curr Ther Res 33(1): 115-122.
A single study compared the combination of paracetamol + caffeine versus individual
components and placebo. This time a small improvement in pain scores as
expressed by TOTPAR was shown for paracetamol i.e. MEAN TOTPAR = 7.1
(combination) vs 7.4 (paracetamol).
This again does not support the 37% stronger claim since in this study the
combination produced inferior pain relief.
Reference: Ali Z et al (2007) Efficacy of a paracetamol and caffeine combination in
the treatment of the key symptoms of primary dysmenorrhoea. Curr Med Res
Opinions 23(4), 841-851 (NB: this was a GSK study).
This study again looks at 4 groups, paracetamol + caffeine vs paracetamol vs
caffeine vs placebo. Pain relief is expressed as TOTPAR values in Table 2. The
value for the combination is 2.89 and paracetamol alone is 2.62. The improvement
A review of the data supplied by GSK show a number of studies that show a modest
improvement in pain relief of up to 10% although it is still important to note that this is
not a consistent result since some studies show an inferior result. However
nevertheless it is not AFT's aim to debate these points but rather the misleading
nature of the 37% stronger claim.
Table: Para + Caffeine Combination vs Paracetamol Alone (Pain Relief)
10.3% Ali Z etal 2007
-4.1% Winter et aI 1983
9.6% Laska et al 1983
We raised a concern with GSK that proceeding with the campaign without taking into
account the above information would result in further serious misinformation to the
market place including consumers and health professionals. However GSK
We do believe that the evidence set out is clear and we have also provided support
from an independent expert, Dr Frampton.
However, as outlined in detail above we contend that the issue is simple and should
not be blurred by complex discussions on methodology. The claim of 37% stronger is
misleading to consumers and health professionals since "potency" is a theoretical
parameter unrelated to analgesic efficacy. As opposed to Pharmacologists,
consumers and health professionals have no use for the concept of potency since
the tablet strength is already set by GSK and this is the key relevance of potency in
the actual market place. However consumers and health professionals are interested
in pain relief.
Consistent with this misinformation, GSK also include the totally incorrect statement
in their advertising to consumers:
"Panadol Extra.... it combines paracetamol with caffeine for 37% extra pain relief 1"
However we are not just focusing on this statement and it is important to note that the
whole campaign is misleading without significant qualification. AFT would propose
that GSK should use the results obtained in clinical comparative studies of
paracetamol vs paracetamol + caffeine i.e. Panadol Extra is 10% stronger than
regular paracetamol or Panadol Extra provides up to 10% more pain relief than
The Chairman ruled that the following provisions were relevant:
Therapeutic Products Advertising Code
Part B1 Requirement 4 (a)
Advertisements must not directly nor by implication, omission, ambiguity,
exaggerated claim or comparison:
(a) mislead or deceive, or be likely to mislead or deceive;
Part B1 Requirement 4.3 Scientific information
Scientific information within an advertisement must be presented in an
accurate manner. Scientific terminology must be appropriate, clearly
communicated and able to be readily understood by the audience to whom it
Publication of research results in an advertisement must identify the
researcher and the financial sponsor of the research.
Part B3 Requirement 3
Advertisements must contain truthful and balanced representations and
claims that are valid and have been substantiated, and:
Part B3 Requirement 4
Advertisements must not directly nor by implication, omission, ambiguity or
comparison mislead or deceive, or be likely to mislead or deceive.
Procedure: The Chairman ruled to deal with the matter by adjudication with
attendance of the parties pursuant to Rule 3 of the Complaints Procedures of the
Advertising Standards Complaints Board. This system was designed to resolve
disputes between competitors. Accordingly, the Chairman appointed a Panel.
The Panel: Mr E Abernethy, Chairperson of the Advertising Standards Complaints
Appeal Board. Co-panelists Mr R Moffat, Member of the Advertising Standards
Complaints Appeal Board, and Dr G Simmons, Member of the Advertising
Standards Complaints Board.
Dr Tim Maling appeared as an expert adviser to the Panel.
The Complainant, AFT Pharmaceuticals Limited, was represented by Mr H
Atkinson, Managing Director, and Associate Professor C Frampton, Department of
Biostatistics - Otago University.
The Advertiser, GlaxoSmithKline Consumer Healthcare Australia was represented
by Mr P Thanopoulos, Associate Director, Medical and Scientific Affairs, and Mr R
Bycroft, Legal Counsel – Chapman Tripp.
The Advertiser, GlaxoSmithKline Consumer Healthcare Australia, said:
1 Thank you for the opportunity to make a written submission on this complaint.
2 The complaint concerns the following advertising claims that GSK makes
regarding Panadol Extra ("the 37% claims"):
2.1 "37% stronger than regular paracetamol";
2.2 "37% more powerful than regular paracetamol". Because Panadol
Extra is 37% more powerful than regular paracetamol it provides extra
pain relief and helps you break through the pain barrier..."; and
2.3 "...It combines paracetamol with caffeine for 37% extra pain relief…”
3 The claims in paragraphs 2.1 and 2.2 ("the Strength and Power Claims") are
made in print advertisements for Panadol Extra (such as in the Magazine and
the Flyer that are the subject of this complaint). These print advertisements
are aimed at either healthcare professionals or consumers.
4 The claim in paragraph 2.3 ("the Pain Relief Claim") is made on the website
www.panadol.co.nz/panadolextra under the heading "About Panadol Extra".
This statement is aimed at consumers.
5 GSK remains firmly of the view that the claims made in the advertisements
are truthful and not misleading, and that therefore GSK has complied with
Requirement 4 of Part B1, and Requirements 3 and 4 of Part B3, of the
Therapeutic Products Advertising Code.
6 For the reasons set out in this submission, GSK believes that it can provide a
firm scientific basis for the 37% claims.
7 GSK substantiates the Strength and Power Claims with data taken directly
from the Laska et al 1984 clinical review paper, published in JAMA (a
respected first tier scientific journal).
8 Furthermore the 37% claims are supported by data from other valid published
papers: Migliardi 1994 and Ali 2007, and Zhang 2001.
9 In bringing this complaint against GSK, AFT has provided no scientifically
supported arguments to counter GSK's position. None of AFT's arguments
invalidate the Laska 1984 data. Moreover, AFT has chosen not to respond to
GSK's detailed explanations of the data.
History of the 37% claims, and TAPS approval
10 The "37% claims" (or an essentially similar "30% claim") have been used in
the New Zealand marketplace in a wide variety of healthcare professional and
consumer media since 2002 and have been vetted by TAPS on each
occasion. Each printed advertisement which is the subject of this complaint
has TAPS approval (Magazine advertisement TAPS PP7301, Flyer
advertisement TAPS PP7356).
11 AFT first complained to GSK on this issue at the end of August 2009 during a
time when GSK was in the middle of preparing additional campaign materials.
AFT provided a running commentary of correspondence to ASA without
actually making a formal complaint. TAPS became aware of the matter and
took the precaution of ceasing any further approvals of Panadol Extra
12 GSK took the step of making its responses to AFT's allegations available to
TAPS. On the basis of the evidence provided by GSK, TAPS has once again
elected to approve Panadol Extra Advertising.
Background to this complaint
13 Over a period of 4 weeks (31st August to 26th September 2009),
GlaxoSmithKline Consumer Healthcare (GSK) received a series of four letters
of complaint from AFT about advertisements for Panadol Extra (paracetamol
14 The first AFT letter contained errors, and AFT withdrew it immediately and
replaced it with the letter of 1st September 2009. GSK responded to this
second letter on 9th September 2009.
15 GSK then received a third letter from AFT intended to provide support for
AFT's allegations. GSK provided a detailed response on 22nd September
16 These two GSK letters have been supplied to the ASA (refer to Attachments 8
and 9, respectively, included in the AFT letter to ASA dated 26th September).
17 On 26th September, GSK received advice from AFT that it would proceed with
a complaint via ASA. Thereafter, GSK was copied in on numerous AFT
emails to ASA which resulted in what is referred to as the fourth complaint
provided by ASA on 30th September 2009.
18 This written submission summarises GSK's position on the fourth complaint.
Should the ASA Panel wish, it can refer to GSK's earlier responses to the
complainant for additional detail.
Motivation behind the AFT complaint
19 Given what we now know about AFT's own advertising claims for its new
combination analgesic, Maxigesic, the nature of AFT's complaint appears
designed to disrupt GSK's ongoing campaign for Panadol Extra. AFT is trying
to remove a competitor claim from the marketplace prior to the launch of
Maxigesic, the advertising slogan for which is Maxigesic® can give you 32%
more pain relief than Paracetamol") (Attachment 1).i
The Strength and Power Claims
20 GSK bases the Strength and Power Claims on a relative potency study by
Laska et al (1984) published in the respected medical journal JAMA.
21 This study, which pooled data from 6 double-blind, randomised controlled
clinical studiesii involving 2625 individuals, found that the combination of
paracetamol + caffeine has a relative potency of 1.37 compared to
22 Using this published data as supporting evidence, GSK chose to use the
phrases "stronger than regular paracetamol" and “more powerful than regular
paracetamol" to communicate the idea of increased pharmacological potency.
23 By using the Strength and Power Claims, GSK intends to convey Laska's
1.37 relative potency result to consumers, so as to communicate that:
23.1 taking a standard dose of Panadol Extra is like taking 37% more of a
standard dose of paracetamol alone; and
Attachment 1: Print advert for Maxigesic in Life Pharmacy.
Laska 1984; Page 1712, Column 1, Last 4 lines.
Laska 1984, Page 1715, Table 4, Row 5
23.2 consumers taking the standard dose of Panadol Extra can expect to
experience a reduction in their subjective pain level that is larger than
would be experienced by taking the equivalent standard dose of
The Pain Relief Claim
24 The complainant asserts readers will understand the Strength and Power
Claims to mean "37% more pain relief than regular paracetamol".
25 However, GSK contends that even if a reader of the Strength and Power
Claims interprets them in the way that AFT suggest, the claim is still not
misleading, as there are other studies (described below) that do support this
26 GSK relies on the results of these other studies (in addition to Laska et al
1984) to support the Pain Relief Claim.
The studies by Ali et al and Migliardi et al showing increased efficacy
27 Further clinical studies on a variety of pain states show that participants who
were given a combination of 1000mg paracetamol + 130mg caffeine, showed
a reduction in subjective pain level (as rated on reliable pain scales) of much
more than 37% when compared to the reduction achieved by equivalent
doses of regular paracetamol.
28 Studies by Ali et al (2007) and Migliardi et al (1994) show that consumers
experience increased pain relief of 56% - 143% by taking a standard dose of
Panadol Extra, when compared with the equivalent dose of regular
29 Therefore, the results of these other studies support the Extra Pain Relief
Claim (and, if interpreted that way, the Strength and Power Claims).
30 Like the Laska study, the Ali et al and Migliardi et al studies are of high
quality, use sample sizes that are large enough to draw statistically significant
conclusions, and have been peer reviewed and published in respected
31 The Ali study was based on pain caused by primary dysmenorrhoea (period
pain). The Migliardi study was based on pain caused by episodic tension-type
32 In GSK's response to AFT of 9th September, a simple calculation using the
data published in the Ali et al study was provided to demonstrate the
increment in analgesic effect achieved by combining caffeine with
paracetamol compared to the effect of paracetamol alone.iv Please see
Attachment 2.v These calculations set out how levels of increased pain relief
of 90% (TOTPAR - Total Pain Relief), 143% (SACID - Sum of Abdominal
GSK response letter 16 September 2009, Attachment 1.
Attachment 2: GSKs calculations based on Ali 2007
Cramp Intensity Difference) and 56% (SBID - Sum of Backache Intensity
Difference Score) can be attributed to use of caffeine as an additive
33 Similarly, a study by Migliardi et al shows that the caffeine adjuvant effect
(increased pain relief) for TOTPAR is 89% greater than the analgesia
produced by paracetamol alone.vi
Why has GSK used the Laska study for support instead of the All and
34 GSK has adopted the 37% potency figure for use in advertisements (instead
of results from the All and Migliardi studies), as the Laska et al study that
generated this 37% figure pools the largest number of studies and
participants, and reports on more than one cause of pain (postpartum and
dental pain). The Laska study also gives a lower percentage result than the
other papers highlighted above, and it was GSK's preference to use the more
35 GSK is not required to reference all studies that support factual claims in its
advertising, rather it must simply be able to substantiate the factual claims
made. Therefore, we have selected the Laska study as being the most
The complainant's criticisms and our rebuttal
GSK's calculations are supported in the published literature
36 AFT agrees with GSK that taking a combination of paracetamol plus caffeine
provides increased pain relief as compared to paracetamol alone. However,
the complainant contends that the scientific literature supports only a 10%
increase in pain relief.vii To support this contention, AFT only provides its own
calculation based on the data in the Ali 2007 paper. However AFT has never
provided any independent scientific evidence to support this calculation. This
same unsupported calculation is provided in complaint letters 3 and 4. This
calculation does not take into consideration the placebo effect.
37 Contrary to AFT's view, calculations based on data from placebo controlled,
cross-over trials (such as the Ali study) need to take account of the placebo
effect, in order to correctly compare the pain relief relationship between
paracetamol + caffeine and paracetamol alone.
38 In responding to complaint letter 3, GSK provided AFT with published
evidence of the validity of this calculation. Applying the calculation method in
the Migliardi study achieves exactly the same result as that which GSK uses
in Attachment 2 to demonstrate that the increment in analgesic effect
Please see page 57, column 2, 3 paragraph AND page 582 bottom paragraphs of the
study by Migliardi et al.
26 September 2009 letter from AFT to GSK, pages 3 and 6. The complainant takes the
TOTPAR results from the study by Ali et al, and divides the TOTPAR value for paracetamol +
caffeine (2.89) by the TOTPAR value for paracetamol alone (2.62) = 1.10. The complainant
states that this expresses a 10% improvement in TOTPAR and that therefore GSK’s
advertising “should state 10% better pain relief than standard paracetamol, not 37% stronger”.
achieved by combining caffeine with paracetamol is 90% of the net
paracetamol effect for TOTPAR at 0-2 hours.
39 Migliardi et al state: "The magnitude or clinical significance of the caffeine
adjuvancy effect (the analgesic scores for the combination minus those for
acetaminophen [paracetamol]) can best be estimated by comparison with the
net analgesic effect of 1000 mg acetaminophen [paracetamol] (the analgesic
scores for the acetaminophen [paracetamol] minus those for placebo).”viii
40 Using this calculation, Migliardi et at state; "for the pooled APAP/CAF
[combination analgesic] study data, the increment in analgesic effect
achieved by combining caffeine with acetaminophen [paracetamol] was 83%,
76% and 89% of the net acetaminophen [paracetamol] effect for SPID,
%SPID and TOTPAR, respectively".ix
41 AFT has provided no argument to dispute this evidence. Its most recent
complaint letter ignores the fact that GSK has already provided valid, peer-
reviewed evidence to support the methodology of calculating the incremental
analgesic effect achieved by combining caffeine with paracetamol.
42 In previous correspondence, AFT provided a 'simple' analogy (about runners
in a race) to attempt to dispute GSKs calculations of the data in the All 2007
paper. This analogy serves no purpose and is flawed. AFT's analogy is
constructed using three separate runners; one representing each medication
(the combination, paracetamol alone and the placebo). But the trials in
question (Ali 2007 and Migiiardi 1994) were both of a crossover design. To
relate this back to AFT's analogy, instead of three people running in a
single race there was one person running in three separate races and
that one person took a different medication (the combination, paracetamol
alone or the placebo) before each race. It is therefore necessary to account
for the placebo effect in that one person to determine the true contribution of
caffeine over and above paracetamol alone. GSK's calculation does this
using a methodology which is supported by published peer-reviewed
evidence (Migliardi et al, 1994).
The Laska et al 1984 study is based on a published, validated
43 In AFT's complaint letter 3 it provided GSK with the opinions of an
independent statistician. The statistician had been provided with a copy of the
Laska 1984 paper.
44 Most of the statistician's comments are also raised by AFT, and are dealt with
by GSK in other parts of this submission. Two points by the statistician
remain, and GSK's responses to these are set out below.
45 The statistician comments that limited pain states were used in Laska and
that the result is therefore not generic to all pain states.
Migliardi et al, 1994. Page 582, column 1.
Migliardi et99, Page 579, Column 2, Paragraph 3.
46 In response GSK observes that post-partum uterine cramping and episiotomy
pain are recognised as established, frequently used, clinical pain models for
efficacy trials of investigational new analgesic agents. They are established
as pain models for assessing analgesic efficacy and therefore the results are
generally perceived to be transferable to other pain states. In any case the
37% claim is further supported by data derived from a variety of pain states
(headache, dysmenorrhoea, and dental pain) in the Laska, Ali and Migliardi
47 The statistician contends that as SPID is based on ratings of 0-3 on a scale, it
should not be used to estimate relative potency even where the association
between log-dose and response is linear.
48 GSK maintains that the appropriateness of the use of SPID as a measure
from which to estimate relative potency is supported in the peer-reviewed
published literature. There is no evidence in the public domain that the use of
SPID to generate potency is flawed.
49 The methodology of the Laksa study already accounts for the assumption that
"the response variable for both the standard and test preparations can be
represented by parallel lines on a log dose scale" which AFT argues was not
accounted for. The statistical methodology used to determine relative potency
in the 1984 paper is published in a previous paper by Laska et al. This paper
describes a computer program specifically developed to take standard
endpoints from head to head clinical trials of analgesic efficacy (SPID and
TOTPAR) and convert these into relative potencies. In this publication this
methodology is referred to as a bioassay method. This same bioassay
methodology is reconfirmed in more recent statistics literature (Attachment 3) x
where it is stated that "Multivariate techniques can be used to combine
multiple responses such as SPID and TOTPAR."
50 In the current complaint AFT still contends that Laska 1984 is flawed,xi but
provides no reasoning in support and nothing to counter the explanation set
out above in paragraph 49 above, which was given to AFT in GSK's letter of
22nd September. GSK questions why neither AFT nor its independent
statistician have commented on the explanation provided by GSK in our
correspondence of 22nd September 2009. Indeed, AFT provides no comment
on this correspondence at all.
Zhang 2001 supports GSKs claim
51 AFT attempts to use safety conclusions in a 2001 review paper by Zhang to
cast doubt as to the safety of paracetamol when combined with caffeine.xii It is
unclear to GSK what relevance the safety of paracetamol plus caffeine has in
the context of a complaint about its efficacy.
Attachment 3: Statistics in the Pharmaceutical Industry, 3 edn. CR Buncher and JY Tsay
(eds), 2006. New York, Chapman and Hall – CRC. Page:187. Cited online and accessed:
11 October 2009.
“Complaint over GSK advertising for Panadol Extra. Statisical considerations”, C Frampton,
23 August 2009, para 3.
AFT Complaint # 4, 26 September 2009, Page 5 top three paragraphs.
52 In prior correspondence with AFT on this issue, GSK provided a copy of
Zhang 2001 to demonstrate that it supported the GSK claims being made.xiii
AFT responded by trying to discredit the safety of the paracetamol-caffeine
combination instead. Such an argument has no place in the current
53 The facts remain that in the Zhang 2001 paper the GSK position is supported:
53.1 Six of the seven paracetamol plus caffeine studies (at doses used in
Panadol Extra or very similar doses) showed a statistically significant
difference in favour of the paracetamol/caffeine combination in terms
of caffeine analgesic adjuvant effect (increased pain relief), as did a
meta-analysis of 4 studies.xiv
53.2 The only study of paracetamol plus caffeine within Zhang 2001 that
provides sufficient data for analysis of complete pain relief showed a
200% (2-fold) increase in favour of the combination.xv
Zhang and Po 1996 does not override the Laska 1984 data
54 AFT has suggested that the Laska 1984 paper which GSK relies on is out of
date, and cites the conclusions in a 1996 paper by Zhang and Po to support
this contention.xvi AFT relies on a statement by Zhang and Po, 1996, that “it
can be concluded that the additives [i.e. caffeine] do not add much to the
value of paracetamol"xvii to infer that the 37% figure is incorrect and outdated.
55 However, Zhang and Po 1996 is a meta-analysis of the data contained within
two previously published papers (Laska 1983 and Winter 1983). All of the
original data from which Zhang and Po 1996 draw their conclusions were
published before 1984, and in any case are included in the Laska 1984
analysis upon which GSK relies. Importantly, the Laska 1984 paper includes
additional data over and above that which was analysed by Zhang and Po in
their 1996 meta-analysis. Therefore, although the publication dates may
make it appear that Laska 1984 is outdated by Zhang and Po 1996 (as AFT
claims), the details of the actual studies indicate otherwise,
56 AFT provide no further argument as to why Zhang and Po is a more relevant
reference, and have also not responded to any of GSK's concerns about the
methods used in the Zhang and Po study nor about the errors in it, that were
previously raised by GSK regarding this particular paper.xviii
Winter et al 1983 study does not override the Laska 1984 data: it is older
than Laska and uses a much smaller sample size
GSK response to AFT complaint # 2, 9 September 2009, Page 5.
Zhang 2001, table I: Wojicki (1 study) and Migliardi (2 studies), Table III: Laska (3 Studies)
and Zhang & Po meta-analysis and Table IV: Winter (1 study). Only the Winter study has a p
value of .0.05.
Zhang 2001, Table II, Row 1.
AFT Complaint # 2, 1 September 2009, Page 1 paragraphs 3 and 4.
Zhang and Po, 1996 page 279.
GSK response to AFT complaint # 2, 9 September 2009, page 3.
57 AFT asserts that a study by Winter et al overrides the data obtained in the
Laska study and presents its own calculated conclusion that the data
represents a 4.1% reduction in efficacy with the combination compared to
58 The relative efficacy figure of -4.1% does not appear anywhere in the Winter
at al 1983 paper. We only have AFT's word that this is a valid calculation of
the data contained in the paper.
59 Critically, the data from Winter 1983 is included in the pooled data that is
used in the Laska 1984 analysis, therefore this data is fully accounted for by
60 Just as importantly, the study by Winter et al used only 164 patients, as
compared to the 2625 patients included in the Laska 1984 study.xix The Laska
1984 study is more recent than Winter and is more reliable due to a larger
sample size and the fact that it included evaluations of efficacy in more than
one pain state (dental and post-partum pain).
Laska et al 1983 study does not override the Laska 1984 data: the data
is included in the Laska study that GSK relies on
61 The complainant asserts that a 1983 study by Laska et alxx (prior to Laska's
1984 study) shows a mean improvement in pain relief of around 9.6%.
62 The relative efficacy figure of around 9.6% does not appear anywhere in the
Laska at al 1983 paper, Again this is AFTs own calculation of the data
contained in the paper. Again, AFT has provided no published source to
demonstrate the validity of its calculations.
63 The Laska 1983 paper itself states relative potency estimates of between 1.5
(50% Increase) and 1.7 (70% increase).xxi The paper also states: "Thus,
based on the summary variables, almost twice as much of acetaminophen
[paracetamol] would be required to give the same response as a given dose
of the combination."xxii
64 The results from this earlier study were included in the pooled study results
that Laska used a year later in the 1984 study,xxiii so the Laska 1984 study
takes account of this 1983 result.
Laska et al, 1984, page 1715, Table 4 (“No. of Subjects”).
Laska EM et al. Effect of caffeine on acetaminophen analgesia. Clin Pharmacol Ther 1983;
Laska et al, 1983. Page 503, Table IV.
Laska et al, 1983. Page 501, column 2, first paragraph.
Please turn to page 1711 of the Laska 1984 study where the Laska 1983 study is
specifically mentioned (viz: “Many of these are relevant to the issue of the adjuvancy of
caffeine, but only the four mentioned previously herein have appeared in the literature”
[Reference 7 = Laska 1983]).
65 For the reasons set out above GSK believes that it has provided a firm
scientific basis for the 37% claims.
66 GSK can substantiate the "37% stronger" claim with data taken directly from
the Laska 1984 clinical review paper, published in JAMA (a respected first tier
scientific journal). Furthermore it is supported by data from Migliardi 1994
and Ali 2007, and other valid published papers (Zhang 2001).
67 In bringing this complaint against GSK, AFT has provided no scientifically
supported arguments to counter GSK's position. None of the arguments
presented by AFT invalidate the Laska 1984 data. Moreover, AFT has chosen
not to respond to GSK's detailed explanations and additional supporting
68 GSK goes to great lengths to ensure that its advertising materials are
appropriately substantiated. This is supported by the fact that TAPS have
been kept apprised of the complaint while continuing to sign off materials in
the current "37%" campaign.
69 It is clear that AFT, in choosing to continue with a complaint without fully
considering the responses previously provided by GSK, has chosen to use
the ASA process as a tactical means through which to try and force GSK to
remove the long-standing 37% claim from the marketplace simply because
the GSK claim is higher than that the 32% of AFT's soon to be launched
70 Thank you for the opportunity to respond to this complaint.”
A table entitled “Calculation of caffeine adjuvant effect” was tabled by the
Advertiser at the meeting and accepted by all parties.
The Agency, Ogilvy New Zealand, said:
“In this instance I believe the client is dealing with this complaint.
Please let me know if you require any help from me”.
The Media, CMP Medica NZ Limited, said:
“I wish to acknowledge receipt of the correspondence concerning the above
complaint (referenced above) from AFT Pharmaceuticals against GSK’s Panadol
advertisement appearing in Pharmacy Today magazine (published by CMPMedica
CMPMedica has fulfilled its responsibility as publisher in ensuring that this
advertisement has received TAPS approval. Consequently there is no further
contribution to the dispute that CMPMedica can provide at this point.
I will be happy to assist further if required by the Complaints Board.”
Mr H Atkinson and Associate Professor C Frampton representing the Complainant,
and Mr P Thanopoulos and Mr R Bycroft representing the Advertiser presented oral
submissions and responded to questions posed by members of the Panel, and Dr
Maling, advisor to the Panel.
The Panel confirmed that prior to its deliberation, it had taken into account all
submissions made in relation to the complaint. It then proceeded to identify the
advertisements subject to complaint, and the provisions from the Therapeutic
Products Advertising Code against which these advertisements were to be
Turning to advertisement 1, the Panel noted that this was an advertisement directed
at healthcare practitioners, and that it appeared in the magazine “Pharmacy Today”.
Accordingly, it confirmed that it was to consider this advertisement against Part B3
Requirements 3 and 4 of the Therapeutic Products Advertising Code. This required
the Panel to consider whether or not the advertisement contained truthful and
balanced representations and claims that were valid and had been substantiated,
and whether the advertisement contained anything which either directly or by
implication, was likely to deceive or mislead practitioners.
The Panel then turned to advertisements 2 and 3, and noted that these were
directed at consumers and were a retail flyer, and a website advertisement. It
confirmed that the appropriate provisions to consider these advertisements against
were Part B1 Requirement 4(a) and 4.3, also of the Therapeutic Products
Advertising Code. This required the Panel to consider whether or not the
advertisements contained anything which, directly or by implication, omission,
exaggerated claim or ambiguity, would be likely to deceive or mislead consumers.
The Panel also had to consider whether scientific information within the
advertisements was presented in an accurate manner.
Before turning to an individual consideration of the advertisements before them, the
Panel considered the product being marketed in the advertisements, Panadol Extra.
It noted that all three advertisements contained the claim that this product was either
“37%” stronger or more powerful than paracetamol, or that it “combines paracetamol
with caffeine for 37% extra pain relief…” The Panel noted, from advice from the
Advertiser, that this was a product that combined paracetamol, the ingredient used
in Panadol, with caffeine. The Panel then noted two different measures that could be
relevant in describing the pain relief product - potency, and analgesic efficacy. The
Panel, on the advice of both the Advertiser and the Complainant understood
potency to mean the ratio of doses of an analgesic formula required to achieve the
same analgesic effect. It understood that analgesic efficacy referred to the “ratio of
effectiveness per unit of analgesic drug dose”, effectiveness being the pain relief
(analgesia) received by the patient.
The Panel noted advice received from the Advertiser, that the claims in the
advertisements, were intended to convey the message that the new product was
37% more potent than regular paracetamol. It noted the Complainant’s assertion,
however, that the claims in the advertisements were likely to be interpreted by
consumers as referring to analgesic efficacy. The Complainant asserted that a claim
of 37% more analgesic efficacy than regular paracetamol could not be substantiated
by the Advertiser, and that, at most, the studies provided by the Advertiser in
substantiation of their claim showed about a 10% increased analgesic efficacy. The
Panel noted that the Advertiser said that although their intention was not to promote
a claim of analgesic efficacy, they were comfortable in the event that consumers
interpreted the claim in this way, as they were of the opinion that they could
substantiate a claim of an increased analgesic efficacy of 37%.
As a further preliminary matter, the Panel also noted that all three advertisements
had been submitted to, and approved by, the Therapeutic Advertising Pre-vetting
System (TAPS). It noted that all advertisements contained a TAPS approval
Having noted all of the above, the Panel confirmed that its role, in considering the
advertisements, was not to be an arbiter of scientific fact. It confirmed that rather its
role was to consider the advertisements, and the claims made in them, from the
perspective of their likely audience. It then had to consider the information provided
to it and decide, when taken at face value, whether in its mind, this information went
far enough to substantiate the claims made in the advertisement. It also noted that
the onus fell on the Advertiser to substantiate the claims in the advertisement.
Keeping this in mind, the Panel elected to consider each of the advertisements in
turn, and first turned to advertisement 1.
It noted the claim made in the advertisement, in bold letters at the head of the
advertisement, “37% STRONGER THAN REGULAR PARACETAMOL1”. It observed
that directly below this heading was a large pack shot of the new product. The Panel
also observed the fine print at the foot of the advertisement where it said, in part
“Reference: 1. Laska EM et al. JAMA 1984; 251:1711-18” (Laska 1984). The Panel
noted again that this advertisement was directed at healthcare practitioners, and
that it was published in the magazine “Pharmacy Today”. The Panel noted that the
target audience for this magazine was pharmacists. It appreciated that others
outside this target audience would also be likely to read this publication and the
advertisement, for example, retail assistants who were not pharmacy trained. Even
taking this into account, however, the Panel considered the environment in which
the publication and advertisement was read to be one more informed than one in
which a general and untargeted publication would be likely to be read.
The Panel was of the view that within this informed environment, there would be a
greater awareness and familiarity with analgesics, the difference between analgesic
effect and potency, and a level of comfort with references to scientific studies and
the capacity and the ability to access these studies, if further clarification was
required of the reference to them.
Having made these observations, the Panel was of the view that medical
practitioners reading the advertisement would understand the word “STRONGER” in
the advertisement to mean potency. It then noted the Laska 1984 study provided by
the Advertiser in substantiation of this claim of 37% more potency. Turning to the
Laksa 1984 study, the Panel noted that it was peer reviewed and published in a
respected medical journal. It noted that the Complainant had tacitly raised some
objections to the study, however that, on balance, they were prepared to accept its
findings. The Panel also said that, in the absence of any good reason not to, it were
entitled to rely on the findings of the study. It noted that the study found that the
relative potency of a caffeine and paracetamol formula as opposed to a caffeine
alone formula was 1.37, or 37% more potent. Having made these observations, the
Panel was of the view that the Advertiser had substantiated the claim that Panadol
Extra was 37% more potent than paracetamol alone. It was of the view that, noting
its view that the target audience would understand “STRONGER” to mean potency,
that advertisement 1 contained truthful and balanced representations, and claims
that were valid and had been substantiated. It also was of the view that the
advertisement did not contain anything which was likely to deceive or mislead.
Accordingly, it said that advertisement 1 was not in breach of either Part B3
Requirements 3 or 4 of the Therapeutic Products Advertising Code.
The Panel then turned to advertisement 2. It noted that this flyer advertisement
contained two claims. It noted the first claim headed the advertisement and said
“37% MORE POWERFUL THAN REGULAR PARACETAMOL1”, and that the
second was in text beneath this and said “Because Panadol Extra is 37% more
powerful than regular paracetamol it provides extra pain relief and helps you break
through the pain barrier”. The Panel observed the fine print at the foot of the
advertisement where it said, in part “Reference: 1. Laska EM et al. JAMA 1984;
251:1711-18” (Laska 1984). Turning first to the first claim in the advertisement, the
Panel noted that it was similar to the claim in advertisement 1, but that it said “37%
MORE POWERFUL THAN REGULAR PARACETAMOL”. The Panel noted the use
of the word powerful. It was of the view that this was an accurate description of
potency and that, reading this statement alone, consumers would understand it to
mean that Panadol Extra was 37% more potent than regular paracetamol. The
Panel referred to its discussion above where it confirmed that it was satisfied that
this claim of 37% extra potency had been substantiated by the Laska 1984 study.
Accordingly, the Panel was of the view that the advertisement did not contain
anything which, directly or by implication, omission, exaggerated claim or ambiguity,
would be likely to deceive or mislead consumers, with respect to this claim. The
Panel was also of the view that the scientific information within the advertisement,
including the claim and the reference to the Laska 1984 study, was presented in an
accurate manner. Accordingly, it said that the claim “37% MORE POWERFUL
THAN REGULAR PARACETAMOL” was not in breach of either Part B1
Requirements 4(a) or 4.3 of the Therapeutic Products Advertising Code.
The Panel then considered the second claim in advertisement 2, “Because Panadol
Extra is 37% more powerful than regular paracetamol it provides extra pain relief
and helps you break through the pain barrier” (emphasis added). The Panel noted
that this claim repeated the text in the first claim, however included the words
“provides extra pain relief”. The Panel noted that this directly referred to pain relief,
and the analgesic efficacy, rather than the potency, of the product. The Panel
considered the consumer take-out of this claim. It was of the view that the claim
would be understood by consumers to mean that Panadol Extra Panadol Extra
provided 37% more pain relief than regular paracetamol. Providing further
clarification, the Panel said that consumers would understand the advertisement to
mean that if you had two people with identical pain (and had the same therapeutic
response to analgesics), and one took Panadol Extra and one took paracetamol,
then the person who took Panadol Extra would have a 37% greater decrease in their
pain than that of the person who took paracetamol.
The Panel then looked for substantiation of this claim, and noted the reference in the
advertisement to the Laska 1984 study. The Panel noted the advice from the
Advertiser that this was a study which measured potency, and which concluded that
that caffeine and paracetamol was 37% more potent than paracetamol alone. The
Panel then again acknowledged the difference between potency and analgesic
effect. It noted that in their response the Advertiser had referred to a range of other
studies which had measured the analgesic effect of paracetamol and caffeine as
opposed to paracetamol alone, and that some of these had shown an increase in
effect of up to 90%. It also noted, though, that on the Complainant’s interpretation of
these studies, that figure was, at most, 10%. The Panel noted however that none of
these studies were referenced in the advertisements containing the claim regarding
analgesic effect, and that the figure, 37%, was derived directly from a potency study.
The Panel had reservations that a figure derived directly from a potency study was
being used in reference to what the Panel was of the view was a clear analgesic
efficacy claim, noting that the 37% figure was a very exact one, and would be likely
to suggest to consumers that it was the exact result of a test comparing the
analgesic effect of paracetamol and caffeine as opposed to paracetamol alone. The
Panel noted that this was not the case, and said that this claim, through implication
and ambiguity, would be likely to deceive or mislead consumers.
The Panel was also of the view that the scientific information, specifically the
reference to 37% within the second claim, was presented in a way that was not
accurate. Accordingly, it said that the claim “Because Panadol Extra is 37% more
powerful than regular paracetamol it provides extra pain relief and helps you break
through the pain barrier” was in breach of both Part B1 Requirement 4(a) and Part
B1 Requirement 4.3 of the Therapeutic Products Advertising Code.
Finally, the Panel turned to advertisement 3, the website advertisement. It noted that
the claim in this advertisement 3 was “Panadol Extra…combines paracetamol with
caffeine for 37% extra pain relief”. The Panel considered the consumer take-out of
this claim, and was of the view that it would be similar to the second claim in
advertisement 2. The Panel was of the view that, like the second claim in
advertisement 2, consumers would understand the claim in advertisement 3 to mean
that Panadol Extra provided 37% more pain relief to consumers than regular
paracetamol. The Panel then referred to its discussions of the substantiation of this
claim, above. It was of the view that this discussion applied to the advertisement
before it, and accordingly said that advertisement 3 was also in breach of both Part
B1 Requirement 4(a) and Part B1 Requirement 4.3 of the Therapeutic Products
Having made all the above observations, the Panel ruled to uphold the complaint in
part, in relation to the claims in advertisements 2 and 3, “Because Panadol Extra is
37% more powerful than regular paracetamol it provides extra pain relief and helps
you break through the pain barrier” and “Panadol Extra…combines paracetamol with
caffeine for 37% extra pain relief”.
Decision: Complaint Upheld (in part)