E-drug: ISDB assess EMEA transparency performance
---------------------------------------------
WARNING: VERY LONG MESSAGE!
Dear Colleagues
'All you wanted to know about the communication policy of the European
medicines agency' could be the title of the position paper released today
by the International Society of Drug Bulletins on the occasion of the
transparency meeting at the EMEA (November 23, 2000).
Christophe Kopp
President, ISDB
Staff editor, la revue Prescrire
"Christophe Kopp" <Christophe.Kopp@wanadoo.fr
The failings of the European Medicines Evaluation Agency
Despite repeated promises to do so, the European Medicines Evaluation
Agency still fails to provide routine access to clear, reliable and
consistent information on the drugs it authorises for sale to European
citizens.
Since the creation of the European Medicines Evaluation Agency (EMEA) in
1994, the International Society of Drug Bulletins (ISDB) has been
watching closely the quality and accessibility of information
produced by EMEA. The reason is simple: if ISDB member bulletins are
themselves to provide readers with reliable information for better
use of drugs, they must have access to unpublished assessment data
contained in marketing authorisation dossiers.
EMEA has a duty to help member states provide information for health
professionals and citizens on the drugs it assesses (article 51, paragraph
i of directive 2309/93 EEC) (a). Since its inception, EMEA has
professed a willingness to encourage transparency. It has set up
procedures and
published documents translating this willingness, and has organised
meetings on the subject of transparency. Here, we examine the
transparency performance of EMEA in this new millennium.
Independent drug bulletins have repeatedly claimed more transparency
At the first annual meetings between the ISDB European group and EMEA
executive director and other persons in charge (from 1994 to 1996),
verbal promises were made that access would be provided to the
following documents: EMEA assessment reports on all drugs authorised
through the centralised procedure; pharmacovigilance data (with
creation of a European database); and substantial information on
dossiers processed through the mutual recognition procedure.
In 1996, ISDB appraised the first European Public Assessment Report
(EPAR) published by EMEA. ISDB suggested a number of improvements
that would make these documents more reliable and more useful to
patients and health professionals alike; EMEA appeared to accept
them, in principle. On 30 October 1997, at a first public meeting
organised by EMEA on transparency and access to documentation, ISDB,
noting that little real progress had been made in these areas,
restated the legitimate demands of European health professionals and
citizens, together with the limitations of confidentiality (including
manufacturing and commercial secrets; protection of private
individuals, etc.) (1), along the lines of the
Uppsala Declaration (2). Similar demands were made by the Swedish
Medical Products Agency and Health Action International (3,4).
Given the lack of improvement in EPAR quality, as well as EMEA's almost
total lack of transparency on pharmacovigilance data and mutual
recognition procedures, the ISBD European group conducted an analysis
of 9 EPARs released in 1996 and 1997. The results of this study were
presented at a new public meeting jointly organised by ISDB and EMEA
on 26 June 1998 (5,6). The existence of assessment reports such as
EPARs was praised (even though they are the least one would expect),
but it was pointed out that their quality was clearly inadequate:
they lacked clarity and consistence regarding the sources of clinical
data; their presentation was highly disparate, making them difficult
to read and hindering the search for precise information; the degree
of detail in the sections on efficacy and adverse effects varied
widely; certain points were either glossed over or dealt with in
excessive detail; finally, they were hard to understand, even for
health professionals (5).
ISDB requested that a uniform style be adopted for EPARs, together with
editorial guidelines for summaries of product characteristics (SPCs) and
patient information leaflets; the quality of all these documents was also
highly variable, probably because of the important role played by
manufacturers in their writing.
EMEA breaks its promises
In response to this analysis, EMEA promised in December 1998 to make
the following improvements to EPARs:
-greater quality, legibility and consistence;
-provision of sufficient information to understand EMEA decisions;
-presentation of information in a logical order, in increasing complexity;
-creation of a guidance document helping reporters analyse and present
dossiers to the European Committee for Proprietary Medicinal Products
(CPMP), EPARs being based on their assessment reports;
-clearer presentation of the risk-benefit assessment and the decision
process leading to marketing authorisation being granted (7).
Unfortunately, two years after these promises were made, EMEA's
transparency has not markedly increased, and the quality of its work, far
from improving, has further deteriorated.
It is true that EPARs are now available on the Internet. It is also true
that EMEA now publishes, before the EPAR, the CPMP's positive or
negative opinion.
The drugs industry may well praise EMEA for its transparency policy
(a poor signal by the way), but seen from our experience the
situation is
revolting.
Increasingly hazy and irrelevant information
The ISDB European group carried out a new analysis in November 2000. It
shows the following failings.
- Access to EPARs: a maze of information
Each EPAR is now composed of several parts comprising the SPC on the
one hand, and the various sections of the assessment report proper on
the other hand.
Thus, to access all the documentation relevant to a given substance, one
must consult several different computer files. This involves a risk of
losing track of data, each part having its own pagination and making no
reference to the other parts.
It is becoming increasingly difficult to identify the version of an EPAR,
especially as several versions can succeed each other at short intervals:
-the words "revision, followed by a number, no longer appear
routinely at the top of a report, or on the various sections of the
document, including the SPC;
-the date (of writing or revision) often disappeared from many
documents, and we have to rely on more or less obscure reference
numbers;
-there are discrepancies between files received automatically by
e-mail and the corresponding files on the agency's website;
-changes made to an EPAR are still not highlighted, despite repeated
requests, since 1994, for the use of simple editorial tools such as bold
type or italics. One recent exception to this rule demonstrates its
feasibility: additions to the adverse effects section on the patient
information leaflet for etanercept dated 6 November 2000 are underlined.
Finally, it is unacceptable that drugs should still usually be identified
by their brand names on EMEA documents, the international non
proprietary name (INN) being more informative.
- EPAR contents: a noteworthy loss of information
The increasingly obscure and coded presentation of EPARs makes their
content less and less reliable. Here are some recent examples:
Calcitonin-Forcaltonine EPAR
The lack of a date makes it impossible to know whether or not general
measures involving a given therapeutic class are taken into account in an
EPAR (note that these measures are not rendered public either).
For example, it is impossible to know if the undated Forcaltonine=B0 EPAR
posted on the EMEA site on 20 October 2000 takes into account the
revisions envisaged for all calcitonins, which were announced
(without no details
(b)) in an EMEA press release (CPMP/941/00) dated 17 April 2000. The
EPAR, which is divided into 8 sections (on different computer files)
that are either undated or dated 3 June 1999, does not clearly show
any
modifications.
Verteporfin-Visudyne. The wording of the Scientific Discussion (i.e.
the assessment part of the EPAR) makes it impossible to retrieve
reliable
information on the adverse effects of some drugs.
Thus, in the case of verteporfin, the EPAR dated 27 July 2000 states that
46% of patients treated with verteporfin and 36% of patients on placebo
experienced adverse effects attributed to the treatment by the
investigators. This information disappears from the version dated 12
October 2000. For a detailed list of adverse effects occurring either
systemically or at the injection site, we have to refer to the SPC, which
is more precise than the Safety section of the Scientific Discussion.
Likewise, the SPC gives the best review of ocular adverse effects due to
verteporfin: the corresponding section in the Scientific Discussion is
either vague and brief (version dated 27 July 2000) or a hotch-potch list
(version dated 12 October 200).
Sodium phenylbutyrate-Ammonaps. We are entitled to expect exhaustive
reports on orphan drugs, given the relative paucity of clinical data.
Yet this is not the case.
True, evidence on sodium phenylbutyrate is limited because it addresses a
rare condition (urea cycle disorders). Nevertheless, the EPAR, dated 8
December 1999, is hazy. The section on clinical efficacy presents the
results of the IND (investigational new drug) programme in the United
States, which involved 208 patients. But instead of presenting the overall
results for the 208 patients, the EPAR first gives details on the first 148
patients, followed by a more short account of the relevant results for all
208 patients, which were available as early as 1996. It is easy to
confuse the results for the 148 patients with those for all 208
patients.
Furthermore, a study of 32 girls included in the American cohort, all of
whom had an ornithine transcarbanamylase (OTC) deficit, has been
published (N Engl J Med 1996; 335: 855-859). The study of this
subgroup is interesting because OTC deficiencies are a distinct
entity among urea cycle disorders, yet the EPAR does not even mention
it.
Raloxifene-Evista. Some EPARs make believe that certain therapeutic
indications have existed for several years, although they only recently
appeared in the SPC.
Thus, according to the revision of the raloxifene EPAR dated 14 June
2000, this drug was granted EU marketing authorisation on 22 April
1998 for preventive and curative treatment of osteoporosis in
postmenopausal women.
However, in the EPAR version dated 5 August 1998, the SPC only
mentions prevention of non traumatic vertebral fracture in
postmenopausal women at increased risk of osteoporosis. The curative
indication only occurs in the SPC dated 14 June 2000, which also
states that a significant reduction in the incidence of fractures has
been demonstrated for vertebral fractures, but not for hip fractures.
In addition, in the presentation of clinical trial results the wording is
uninformative. The report of the MORE trial in the raloxifene EPAR
dated 14 June 2000 gives the relative risk (of new fractures) but not
the absolute values, even thought the latter are available in the
paper published in JAMA (1999; 282 (7): 637-645 + (22): 2124). Thus,
the risk reduction appears far more impressive than it actually is
(47% in the EPAR, only 2% in absolute values). This makes the EPAR
read like a promotional brochure.
Cetrorelix-Cetrotide=B0. Some EPARs still omit data made public by
other drugs agencies. Thus, the cetrorelix EPAR dated 22 September
2000 fails to mention a follow-up study of 316 babies born to women
treated with this drug. Yet the data are presented in the "final
draft package insert" available on the FDA website. Admittedly the
FDA report is dated 8 November 2000, but it is very unlikely that
EMEA was not aware of the existence of this follow-up study a month
and half earlier.
It is becoming difficult, and sometimes impossible, to know what has
changed in a set of indications, an adverse effect profile, etc., or why.
The editor of an ISDB bulletin, with lengthy experience in analysing
EPARs, recently said: "All I can say about the new, undated EPAR on
toremifene is that it contains less information than a previous
version, and that the section addressing the patient is shorter and
apparently more favourable to the drug".
It is a flagrant waste of European taxpayers' money to produce and
translate such confusing and sometimes useless documents. In these
conditions the notion of transparency is mere rhetoric.
- Other information distributed by EMEA
CPMP opinions, which precede the publication of the corresponding EPAR,
often suffer from the same defects. It is difficult to determine the date
and order of versions, and to determine whether they concern first
opinions on new drugs, opinions given in the setting of five-yearly
revisions, or opinions on new indications.
Secrecy still prevails regarding the mutual recognition procedure.
The EMEA website only gives the rapporteur country, although this
allows accessing information from rare countries actively applying
freedom of information.
No information is available on EMEA processing of the dossiers, or on
possible objections raised by member states.
Information on adverse effects is still inadequate. In the last five years,
the only information that has appeared on the section of the EMEA website
devoted to pharmacovigilance is the concise CPMP opinion about 5 drugs
(naftidrofuryl, sparfloxacin, chlormezanone, terfenadine,
vigabatrin); one opinion on a drug family (appetite suppressants);
and position statements on oral contraceptives and the HibTiter
vaccine. Information about particular side effects sometimes appears
in other sections of the EMEA website, but not in the
pharmacovigilance section, hence a risk of overlooking the problem
(e.g. the position paper dated 31 October 2000 on adverse reactions
associated with herbal products containing Aristolochia species).
It is difficult for the public to understand the respective parts played by
EMEA and national agencies in the field of pharmacovigilance. Some drug
bulletins actually feel that less information is available now than before.
Many member states are loath to answer questions, referring them to
EMEA, which itself speaks through pursed lips!
The king has no clothes
Despite claimed intentions, communique, and meetings on transparency
and access to data, the European medicines evaluation agency remains
half opaque or entirely opaque depending on the field. As a result,
health professionals and citizens of the European Union cannot trust
EMEA to provide reliable drug information. Is this because of poor
management of communication policy, or is it a deliberate strategy?
No one can say. But the fact that the pharmaceutical industry
provides the European Medicines Evaluation Agency with 60% of its
running costs may have something to do with it (9).
ISDB European group
a- Note that some national agencies do not have this mission, or do not
take it into account. For instance, the founding text of the French agency
for health product safety introduced the concept of the need to inform
health professionals and the public, notably through the publication of a
"summary of the authorisation files on all new drugs" (article L.793-1 of
the Public Health Code), which was not the case of the former drugs
agency (Agence du Medicament). In practice, however, very little
information is
provided.
b- Being a revision procedure (according to article 12 of directive
75/319/EEC) relating to the efficacy of calcitonins in osteoporosis and
other indications, it could be matters. La revue Prescrire, an ISDB
member bulletin, asked EMEA for further information in a letter dated
3 May 2000.
The reply, received on 7 June 2000, stated as a matter of course that a
procedure was underway but gave no further details.
1- Workshop on transparency and access to documents of the EMEA
Contribution of the International Society of Drug Bulletins (ISDB)
London, 30 October 1997 : 5 pages.
2- Statement of the International working group on transparency and
accountability in drug regulation Uppsala 14 September 1996: 18 pages.
3- E4kemedelsverket Preparation of Workshop on transparency and
access to documents of the EMEA (EMEA/MB/052/97) Uppsala 16 October
1997: 3 pages.
4- Health Action International Openness the rule, secrecy the exception
Contribution to the Workshop on transparency and access to documents
of the EMEA, London, 30 October 1997: 5 pages.
5- ISDB Assessment of nine European public assessment reports
published by the European medicines evaluation agency (EMEA) 26 June
1998: 12 pages.
6- Abbasi K and Herxheimer A The European Medicines Evaluation
Agency: open to criticism Br Med J 1998; 317: 898-900.
7- EMEA Response to the International Society of Drug Bulletins (ISDB)
London, 17 December 1998: 12 pages.
8- EMEA Outcome of public consultation on new EMEA transparency
initiatives London, 9 June 2000: 1 page + 23 pages (Annexes).
10- Prescrire E daction Le financement des agences du medicament Rev
Prescr 1999; 19 (201): 867-868 ; Prescrire Editorial Staff Funding of
medicines agencies Prescr Int 2000 ; 9 (46) : 34.
--
Send mail for the `E-Drug' conference to `e-drug@usa.healthnet.org'.
Mail administrative requests to `majordomo@usa.healthnet.org'.
For additional assistance, send mail to: `owner-e-drug@usa.healthnet.org'.