1 Department of Gynaecology, University College Hospitals London, The United Elizabeth Garrett Anderson Hospital & Hospital for Women Soho, 144 Euston Road, London NW1 7AJ, UK, and 2 The Uppsala Monitoring Centre, WHO Collaborating Centre for International Drug Monitoring, Stora Torget 3, S-753 20 Uppsala, Sweden
The constant negative drip of unbalanced information from the lay press makes it impossible for the professionals and consumers to interpret the true risks and benefits of drugs. This can have a serious effect on the quality of life, not only in the target population. A recent, almost weekly, newspaper campaign surrounding the combined oral contraceptive pill (COC) claimed that in the past 4 years, 50 deaths of women had been associated with the use of the oral contraceptive pill in the UK (Davies, 1998). The message once again is that the pill kills. This figure, even if true, would be the equivalent of approximately four deaths per million women pill users per year in the UK. This would equate with the death rate from terminations of pregnancy (HMSO, 1996) and be considerably lower than the death rate from other life events such as pregnancywhich is 98 per million maternities in the UK (HMSO, 1996
), smoking, road traffic accidents, and household accidents. Women seem unaware of the reported health benefits of the COC, which include the reduction in risk of developing an ectopic pregnancy, anaemia, ovarian cancer or endometrial cancer (Vessey, 1989
). Thus, the constant drip of adverse reporting about a highly effective reversible method of contraception continues to be presented to women and doctors in a way that prevents them making a truly informed contraceptive choice.
A number of recent publications (Edwards et al., 1996; Edwards, 1997
; Edwards and Hugman, 1997
) and initiatives taken under the World Health Organization (WHO) and the Council for the International Organisations of Medical Science (CIOMS) have attempted to promote better communication of the benefits and risks of pharmaceutical preparations to the health professions and the public. This culminated in a multi-disciplinary, international statement of intent known as the Erice Declaration (McNamee, 1997
; the full declaration can be found on: http://www.who.pharmasoft.se/communications.htm). These publications will be used as a basis for reviewing and suggesting alternative strategies for communications affecting COC use.
Benefit and risk balance
A benefit to risk balance is context dependent. A simple overall public health perspective which merely sets the benefit of reliable contraception against a very low risk of death does not take account of the risk of death in pregnancy, or from abortion, the reduction in stress from fear of an unwanted pregnancy, or a possible improvement in menstrual cycle and bleeding, amongst other issues. It certainly does not take into account the health requirements of an individual patient perhaps having a disease, which may make pregnancy particularly hazardous. Most women are not aware of the importance, or given the information to allow them to evaluate the risk benefit ratio of their choice of contraception. The importance of this process has been emphasized by several authors (Edwards et al., 1997).
One is not surprised that acceptance of individual risk is variable, being influenced by many factors and perceptions. Adverse drug reactions are seldom quantified into numbers that can be compared with other readily recognizable risks that are taken in everyday life. It is clear that to provide a public health benefit risk analysis requires that a total benefitrisk profile should be created for the drug, set in context with the risks of the treated condition (or the condition to be prevented, in this case pregnancy). Any special subgroups need to be identified. Alternative therapies, including abortion in this case, should be analysed in the same way and compared.
It is easy to take the simplistic view and concentrate on a `risk driver', i.e. some new finding of risk that concentrates public and professional attention. This is what has happened with the COC issue, currently and for many years in the past, concentrating mainly on vascular disease and breast cancer. However, the more useful approach is to re-evaluate the total benefitrisk relationship (which includes the strength of any new evidence). This was apparently not done, or not communicated, during the latest `pill scare' about venous thromboembolism (VTE) (Committee on Safety of Medicines, 1995). In considering the total benefitrisk profile of a drug usually a single (or very few) indication for a drug has to be compared with a number of adverse reactions of differing frequency, duration and seriousness. Frequently key information on adverse drug reactions is missing. Studies of rare but serious adverse drug reaction phenomena are difficult to perform without bias and confounding arising because of the live clinical setting for the studies. It is clear that interpretations of such data are likely to be controversial.
It is then essential that all the data on which the benefitrisk judgements are made should be available for peer scrutiny. At the start of the `pill scare', suggesting that third generation COCs had twice the odds ratio that second generation COCs had for VTE, public communications (Committee on Safety of Medicines, 1995) were made on unpublished data (Bloemenkamp et al., 1995
; Jick et al., 1995
; WHO, 1995; Spitzer et al., 1996
): neither the health professionals nor the public could examine the foundation for a published but incomplete benefitrisk analysis, with recommendations to turn to a medical profession which had no access to further information. It has been pointed out that this action, whilst technically not withdrawing third generation pills, came dangerously close by putting the responsibility on the physician (Benagiano, 1998
).
Furthermore, a recent review was conducted on epidemiological information on COCs with a view to answering the clinically relevant question: `What is the risk of cardiovascular disease amongst the users of the currently available, low dose, combined oral contraceptives, who are aged <35 years, do not smoke and do not have a medical condition known to increase the risk'. The author's conclusion was: `much of the epidemiological data about the risk of cardiovascular disease in users of COCs is not useful to clinicians. Some of the discarded data could be made more useful to clinicians by re-analysis'. Furthermore authors of the studies on VTE and COCs continue to debate whether this increase in risk is a true finding or due to bias and confounders within the studies (Spitzer, 1997). This emphasizes the need for great circumspection in public announcements even on published data (Hannaford and Owen-Smith, 1998
).
The players
The Erice Declaration states in its preamble: `Monitoring, evaluating, and communicating drug safety is a public-health activity with profound implications that depend on the integrity and collective responsibility of all partiesconsumers, health professionals, researchers, academia, media, pharmaceutical industry, drug regulators, governments and international organisationsworking together. High scientific, ethical and professional standards and a moral code should govern this activity. The inherent uncertainty of the risks and benefits of drugs needs to be acknowledged and explained. Decisions and actions that are based on this uncertainty should be informed by scientific and clinical considerations and should take into account social realities and circumstances'.
The active players in the latest `pill scare' were firstly the regulatory body and then the media. Most new drug safety information is published by drug regulatory bodies and manufacturers. Some comes from academic research. Health professionals at large and patients/consumers do not appear amongst those consulted at an early stage of any benefitrisk analysis, in spite of the obvious fact that they are the groups to be most affected by such analyses.
At least one major medical journal, the British Medical Journal, has a policy for early release to journalists some days before the Journal is released for the medical profession (personal communication). Some academics release the substance of their work to the general media before it is published (Lancet, 1997; McNamee, 1997
). These actions exclude one or other of the key players whose role may be very beneficial in interpreting new safety information and deciding upon the best ways to communicate it to the public in a balanced way.
There is a limit to the numbers of groups which can be consulted over new drug safety issues but, in the UK, there are the various Colleges which represent specialist interests and general practice as well as many established patient and consumer groups which could be consulted. The general media have played a large part in the creation and continuation of concern about COCs. Our experience in talking to the media is that they desire to inform the public about important issues and strongly resist any hint of keeping any information from the public. The issue of providing a balanced view of a subject is dealt with by providing two viewpoints on the same story, not always in the same issue or programme. There is also an acknowledged requirement to `sell newspapers'. It is clear that stories which can be turned into controversy sell newspapers and programme time, but the publication of opposing views does not, to us, constitute a balanced argument, it merely serves to reinforce the view that experts do not know what they are talking about. If there was space (or time) to present opposing views with full justification, then members of the public who were sufficiently interested and knowledgeable could certainly `make up their own minds'. Unfortunately, this is not the situation and the public is forced to seek advice from the increasingly beleaguered medical profession.
Communication
Good communication is both an art and a science. Among the many core questions which require consideration in the planning of any communication activity are: (i) the content of the message; (ii) the purpose of the message; (iii) the state of mind of the intended recipient(s); (iv) the general context or climate in which the message will be perceived; (iv) the medium or media of communication to be used; (v) feedback mechanisms to assess the extent of receipt of the message and its subsequent effects; and (vi) monitoring and evaluation.
Clarity about these questions is essential if any communication is to have a chance of achieving its desired impact. In addition to the above, one might also add `a consideration of the benefits and risks of any given communication', which is essentially a summary of the above points which will allow for a clear conclusion to be drawn on the wisdom of sending a particular message in a particular way. The various headings are used below to analyse the overall situation with respect to the COCs.
Content of the message
It is essential to consider the basis of the information to be transmitted. It is particularly important to decide whether the information is factual or whether there is a level of uncertainty associated with the message. This uncertainty has been emphasized in the case of third generation COCs and VTE, by the court ruling on 19th December 1997 that Germany's regulatory health authority (BfArM) should remove the prescribing restriction which prevented first time COC users receiving third generation COCs.
Purpose of message and state of mind of the intended recipient(s)
When new information came to light on a probable higher risk of third generation, in comparison with second generation COCs in respect of VTE, there was a need to communicate it to the health professions and thence the public, in a way which would allow reasoned decision making. A concentration on this particular risk driver lead to the magnitude of the putative risk being quoted as a relative risk: third generation COCs having twice the risk of the second generation COCs.
The term `relative risk' is poorly understood by many of the players, does not put the risk in a day-to-day context, nor give any possible advantages of COCs. In this case the estimated risk of non-fatal VTE was estimated to be 30/100 000 compared with 15/100 000 women years of use for third versus second generation COCs (Mills et al., 1996).
Women who are healthy expect their chosen method of contraception to be safe. Relative risks quoted as double for one product will be easily interpreted as unacceptable. This situation has been reinforced by high media exposure of cases where VTE and death have actually occurred in women taking the COC.
The Erice Declaration states: `Education in the appropriate use of drugs, including interpretation of safety information, is essential for the public at large, as well as for patients and health care providers. Such education requires special commitment and resources. Drug information directed to the public in whatever form should be balanced with respect to risks and benefits'.
The medical profession and all other health workers have a responsibility to make themselves aware of the principles of benefitrisk evaluation and perception and try to educate the public as much as possible, certainly by trying to put drug risks in to an everyday perspective. This should certainly have been done in respect of the COCs and thrombo-embolism by all the active players.
General context or climate in which the message will be perceived
There is a growing mistrust by the public of government agencies, the pharmaceutical industry and the medical profession. Secrecy by regulators and the industry has not helped, and is certainly the very opposite of the freer consultation envisaged by the Erice Declaration. The medical profession has lost public confidence because of the inability of the profession to provide time and explanations to the patient.
Lack of time is an increasing problem in an under-resourced situation and is compounded by a huge general information overload, which can hide key information needed for the prescriber and patient alike (Hannaford and Owen-Smith, 1998).
The COC scare has clearly exacerbated the situation, leaving all parties lacking credibility in the public's eyes. It was, however, clear from the outset that once the public had controversial information over a serious adverse drug reactions they would turn to their doctors for an explanation.
Communication media to be used
There are many media for communication, ranging from letters to e-mail. `Dear Doctor' letters have been the main way to transmit important and urgent information from regulators and industry to the medical profession. Less urgent information on drugs is transmitted by occasional publications and changes in data sheets.
`Dear Doctor' letters are synonymous with urgency, and therefore raise concern per se. Although the finding of possible increased mortality associated with the third generation COCs affects large numbers of women, the very low absolute level of increased risk which makes an urgent transmission of such information debatable.
An argument arises as to whether information should go to the public at the same time as the health professionals. It would seem that the use of, say, television would be a suitable way to reach the largest segment of the population with an important, clearly defined message. The increased risk with third generation COCs is, however, not by any means clear-cut and the need for professional interpretation for an individual patient very strong (Mills, 1997). In this instance it was unfortunate that the public had the information before many members of the profession, but it is quite clear that a secure method of communication to avoid such a situation is difficult. Drug data sheets are not a suitable medium to communicate complex issues either. They are more of an aide memoire.
It is a shame that the only medium open for mass communication to the medical profession (the medical journals in paper copy or soft copy via the internet) are offering special early access to the general media before the professions they serve.
Feedback mechanisms
An essential part of any communication is to check that the message is received and understood in the way it was intended. Reviewing events after a communication allows for more messages to be given to `fine tune' the response and deal with unforeseen circumstances.
Important communications should have a feedback mechanism devised for each occasion. Firstly, so that the sender can be sure that the message has been received and secondly, so that he/she can obtain early information about problems and response to that communication.
Monitoring and evaluation
A consequence analysis does not alter adverse effects of a current communication, in the way early feedback may do. The more long-term monitoring and evaluation of a total event provide lessons to be learned for the future. It is important that this is done in a spirit of learning rather than recrimination.
Discussion and conclusions
Given that the risk of VTE was lower than some previous reports (Vessey, 1989) and only a small increase in the anticipated number of women with thromboembolism occurred with the third generation COCs when compared with second generation COCs, it seems that it was unwise to use an urgent communication to warn doctors. It was difficult in a short communication to give a benefitrisk picture, and concentration on a risk driver is bound to increase public concern. The publication of the study which initiated the controversy was subsequent to the `Dear Doctor' letter. All this has undermined confidence in the medical profession, and unnecessary elevation of concern over COCs and a fall in their use. The obvious consequences have been an increase in birth rate and abortions in the UK as well as many other European countries (Skejeldestad, 1997
). The latest figures show the abortion rate for England and Wales to be the highest for 30 years (Office for National Statistics, 1997
).
The pharmaceutical industry and regulators do have a duty to warn of serious issues when they occur, and they feel themselves under pressure when they have such information. Part of their problem is their own secrecy and a paternalistic approach to their work. There should be much more general and specific consultation with the health professions and society at large. This could lead to `Good Regulatory Practice' guidelines that would, perhaps, state that regulators need not warn prior to publication unless there is a public health emergency and to define what constitutes the latter. There should also be consultation with physicians and relevant patient groups over public health crises. Such guidelines would go a long way in rebutting any accusation of `sitting on information'.
The medical profession needs independent sources which help to reduce the information overload into a easily accessible drug formulary which quantifies, as far as possible, the adverse drug reaction risks and benefits, in absolute terms, of all pharmaceutical products. In our view, much could be done to improve the content and design of formularies.
The medical media could also adopt an improved practice which would involve giving the relevant expert professional group (ideally not individuals) the chance to see papers which are likely to cause public concern, at least as early as the general media so that they may prepare comments. Again one can envisage a simple code of practice to cover this.
Professional organizations, e.g. specialist colleges, should be prepared to be much more pro-active with the media over general media debates on health topics. This must include forging good relationships with the media in advance of crises.
It must be said that the general media have been responsible for a continuing negative message being delivered to the public about COCs without any thought to a balance of any sort, for example in the 6 months following the Committee on Safety of Medicines (1995) announcement, the Swedish press published 554 articles about the `dangerous pills' (Anonymous, 1997). We strongly recommend that the general media look very carefully into their ethical stance in relation to balanced reporting. The setting of a confrontation between dissenting experts, shocking headlines and limited time and space for reasoned argument are the facts of life in the media. We wish this could be changed, and do not believe that what sells newspapers is synonymous with `what the public want' and less still with what the public need. We accept that this is our opinion only, but it is born out in this case by the harm caused by many unintended pregnancies and abortions.
Our suggestion would be that the media seek the opinion of patient groups and health specialist groups to add consensus opinion on critical health issues. These groups should be used to aid in editorial material that will add benefitrisk balance to other published stories.
Our overall conclusion is that much could be achieved by respect for the principles embodied in the Erice Declaration, the development of some good practice guidelines, for regulators and the media in particular, and that use of good communications practice should be the norm in releasing information to the public, including a carefully considered feedback and consequence analysis.
Notes
This debate was previously published on Wetrack 31, August 28, 1998
To whom correspondence should be addressed
References
Anonymous (1997) The media magnified the false alarm about the pill; abortions increased in Sweden. Journalisten, 30th January 1997.
Benagiano, G. (1998) Learning from the past: venous thromboembolism and the pill: an endless saga. Hum. Reprod., 13, 11151116.
Bloemenkamp, K.W.M., Rosendaal, F.R., Helmerhorst, F.M. et al. (1995) Enhancement of factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a third generation progestogen. Lancet, 346, 15931596.[ISI][Medline]
Committee on Safety of Medicines (1995) Combined oral Contraceptives and Thromboembolism. Letter 17 October 1995. Committee on Safety of Medicines, London, UK.
Edwards, I.R. (1997) Who cares about pharmacovigilance? Eur. J. Clin. Pharm., 53, 8388.[ISI][Medline]
Edwards, I.R and Hugman, B. (1997) The challenge of effective benefitrisk information communication. Drug Safety, 17, 216227.[ISI][Medline]
Edwards, I.R., Wilholm, B. and Martinez, C. (1996) Concepts in benefitrisk assessment a simple merit analysis of a medicine? Drug Safety, 15, 17.[ISI][Medline]
Edwards, R.G., Beard, H.K. and Bradshaw, J.P. (1997) Balancing the risks and benefits of oral contraception. [Editorial.] Hum. Reprod., 12, 23392340.
Erice Declaration (1997) HYPERLINK http://www.who.pharmasoft.se/communications.htm http://www.who.pharmasoft.se/communications.htm
Hannaford, P.C. and Owen-Smith V. (1998) Using epidemiological data to guide clinical practice: review of studies on cardiovascular disease and use of combined oral contraceptives. Br. Med. J., 316, 984987.
HMSO (1996) Confidential Enquiries into Maternal Deaths in the United Kingdom, 19911993. HMSO, London, UK.
Jick, H., Jick, S., Gurewich, V. et al. (1995) Risk of idiopathic cardiovascular death and nonfatal venous thromboembolism in women using oral contraceptives with differing progestogen components. Lancet, 346, 15891593.[ISI][Medline]
Lancet (1997) Time to act to avoid drug scares. [Editorial.] Lancet, 350, 1041.[ISI][Medline]
McNamee, D. (1997) Communicating drug safety information. Lancet, 350, 1646.
Mills, A. (1997) Combined oral contraception and the risk of venous thromboembolism. Hum. Reprod., 12, 25952598.
Mills, A.M., Wilkinson, C.L., Bromham, D.R. et al. (1996) Guidelines for prescribing the combined oral contraceptives. Br. Med. J., 312, 121122.
Office for National Statistics (1997) ONS Monitor AB 97. London, UK.
Skejeldestad, F.E. (1997) Increased number of induced abortions in Norway after media coverage of adverse vascular events from the use of third-generation oral contraceptives.Contraception, 55, 1114.[ISI][Medline]
Spitzer, W.O. (1997) The 1995 pill scare revisited: anatomy of a non-epidemic. Hum. Reprod., 12, 23472357.[Abstract]
Spitzer, W.O., Lewis, M., Heinemann, L.A.J. et al. (1996) on behalf of the Transnational Research Group on oral contraceptives and the health of young women. Third generation oral contraceptives and risk. Br. Med. J., 312, 8388.
Vessey, M.P. (1989) The Jeffcott lecture: An overview of the benefits and risks of combined oral contraceptives. In Mann, R.D. (ed.), Oral Contraceptives and Breast Cancer: The Implications of the Present Findings for Informed Consent and Informed Choice. Parthenon Publishing Group, New Jersey, USA, pp. 121135.
World Health Organization (1995) Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effect of different progestogens in low dose oral contraceptives on venous thromboembolic disease. Lancet, 346, 15821588.[ISI][Medline]