Pharmacovigilance and communication failure

In FOCUS March 2013, Antonio Addis discusses the inadequacies of information and communication in pharmacovigilance and also challenges us, ‘Do try and find someone asserting that information in pharmacovigilance is useless’. In FOCUS July 2013, Richard Smith talks about peoples’ perception of drug safety and ends, ‘My conclusion is that Britain—like most countries—is awash with drugs, and that although Britons are well aware that some drugs can be dangerous they are generally hazy on the ratio of benefits to risks. This means that many people may be exposing themselves to the dangers of drugs when they can expect little or no benefit’.
These two eminent and thoughtful people are politely telling us that, if we have not failed, we are certainly not doing well in our communication! And I agree with them, but the question is, ‘Why?’ and, ‘What can we do better?’
We are overloaded by information, but paradoxically we often do not have enough. Every time we read something, particularly perhaps on the web, new questions are often raised because much is simple summary that all can understand. It disappoints us when we can’t explore further.
Pharmacovigilance poses basic challenges in communication management. Adverse effects from medicines are relatively rare from most individual medicines. The majority of users of frequently used medicines take them and get better without problems (often the underlying disease is self-limiting anyway!). This reinforces the concept by the public that medicines are safe – that is until they read the formidable product information with the listings of all possible adverse effects. The result of this situation can only be a dichotomous and confused general view about medicines. That is If they read and understand the product information at all!

The situation is worsened by the nature of post-marketing information on harm. the first tentative signals right through to the availability of pharmacoepidemiological analyses of the more extensive date, various levels of causal doubt exist about even serious adverse effects. This is a problem of the rarity of adverse effects due to medicines versus the background rates of the same conditions caused by other disease processes. Doubt, and the corresponding concept probability, is very difficult to communicate. To complicate the matter further probability is used by epidemiologists interchangeably with risk, but for many, risk also includes the degree of seriousness of an outcome. There are striking differences in the meaning of probability descriptions, e.g. ‘rare’ or ‘frequent’ in mathematical terms: in one group study ‘rare’ ranged from 1:100 to 1:10 000! Moreover, all of us respond to risk differently, our perceptions based on our general nature and background experience: this includes both health professionals and their patients.
As Antonio Addis points out, our pharmacovigilance information does improve all the time. Nevertheless, there is still a huge mismatch in the money and effort in finding out about the effectiveness and risk of medicines used in day-to-day clinical practice and making sure that the people who need the information get it, understand it and use it. True communication (as opposed to information) requires a dialogue between the person/ organisation giving any message and the recipient(s). In a conversation between two people this is often, but by no means always, a natural process. It is our common experience that we can quickly tell if someone has heard what we say and quite often we find out if they have understood and even acted on what we say in an appropriate fashion.
To test the success of a dialogue between an organisation and populations, the sender of the message must go out actively and find out by survey if the message has been heard. The organisation must find by survey if the message has been understood (if not, why not) and find out whether it has changed behaviour either by a population study or by monitoring key success factors. Richard Smith tells us we have not done well overall, but the real issue is that there are relatively few medicines where there is a total evaluation of the success or otherwise of communication based on outcomes. This is done with some medicines: it should be done with all.
I mention ‘end users’ in the last paragraph. Who are the end users? Most information about medicines has, in the past, been directed to health professionals as ‘learned intermediaries’. It was assumed they had knowledge and hopefully wisdom to interpret complex science for the individual patient’s situation. This situation has changed radically and rapidly over a couple of decades. More and more people get information from the internet, and what a confusing mixture of good and bad that is. We cannot truly check the credentials of the provider or have them take responsibility for misleading or missing information. My own trawls through the web have shown great differences in the level of information provided and much that is contradictory.
We have two ‘end users’. Health professionals need technical information for use in their professional practice; the public is a much, much more diverse group in their needs. The internet provides very complex information as well as summarised more basic medication guidelines. Health professionals should understand nearly all the information they can find; the public – who knows? What we are fairly sure of is that the personal dialogue between health professionals and their patients has become richer perhaps and certainly more complex. I must also say that more complex dialogue is usually more time consuming and therefore, in a professional situation, more expensive.
I need to mention three voices that in my opinion often have a negative effect on good communication in pharmacovigilance, they are lawyers, regulators and the pharmaceutical industry. First, however I would like to defend them. All three of these groups argue that they have the public good in mind in their work, and I am sure that they have clear consciences as they make that statement. The problem is that their view of maintaining the public good is that some things can be misunderstood by the public and therefore are better hidden. With this paternalism, transparency is frequently expressed as, ‘We must avoid creating a drug scare’ or ‘The public will not understand’ or ‘We must not damage the vaccination programme’. These actual examples of stated views are against true transparency: where there is doubt - don’t, is the thinking behind it. The result is that harm by medicines must be, ‘Proven beyond reasonable doubt’, to use a legal expression, before talking to the public. For pharmacovigilance this has been a real challenge because we cannot easily prove things finally when adverse effects are rare and the data evolving. It may take large numbers of patients exposed to medicines before we have epidemiological ‘proof’. Before that ‘proof’ is clear many argue that silence is better. I have said that communication of doubt is challenging, but if we remain silent, or in some other way try to bury suspicions of risk in small print and jargon we will have more Vioxx-like situations where there was suspicion of excess harm for years before the extent was established and made public¹. Lawyers often find the need to offer different advice, which is to tell everything in principle but not any more detail than necessary to satisfy the law. This can have the effect of describing many clinical events that would not be really suspicions of risk. These clinical events find themselves added to medicines information, albeit under a section of ‘other reports’, but without further information.
Suggestions for ways forward.
We need a paradigm shift to where communication and dialogue is a prime focus of our work: telling people repeatedly about the risk/effectiveness balance of medicines is as important as studying it. We must fight for a sufficient budget for communication.
We are at the beginnings of much better communication strategies. We must evaluate them and their results in improving therapeutics and particularly preventing harm. No study of the benefits and risks of medicines should be complete without a review of the communication issues involved. Pharmacocommunication should be as important a topic as pharmacoepidemiology.
Audit and litigation should be pursued, more to assess the quality and content of communications about harm caused and the outcomes resulting from those communications. Audit of input of data and litigation around instances of failure to warn seems to be of less importance than measuring the overall effectiveness of pharmacovigilance.
There are multiple audiences for pharmacocommunication, and we must use media, language and methodology that are appropriate for each situation. It would not be practical to consider all audiences, but one is critical. Health professionals need to have all information that is useful for the diagnosis and prevention of harm from medicines. For diagnosis those needs include suspicions, as well as quantitative information on known harms. Qualitative clinical information surrounding the diagnoses of suspect and confirmed cases is also essential, as also is information on the seriousness of harms and their management. Such information needs to be accessible quickly and easily as it is often needed during a patient consultation. Information presented must be up-to-date, noting any changes when they occur, and available on the internet, at various levels of complexity using e.g. drop-down boxes. To ensure the validity of information it should come from an approved regulator and industry source. There are an increasing number of internet resources e.g. phone apps. developing this way, but a fully optimised information source has yet to be developed. I suspect that cost and the political will are barriers here.
The public need some very important messages about the safe, optimum use of medicines and warnings of the most likely and serious risks particular to each medicine, and what they should do about them. Patients can best obtain information about their medicines from health professionals which is tailored to each individual’s situation.
My hope is that patients should have a full, satisfactory dialogue with their health professionals about the effectiveness, expected benefits and risks of their treatment. This should involve periodic review of treatment outcomes, answering patient queries and generally monitoring patients’ progress. Of course this is done now but I doubt if the pressures and limited consultation times allows this do be done well. As health care costs more, economic pressures mitigate against good communication time with patients. Help can be obtained from, imaginative communication approaches like patient group sessions with health professionals for different treatment categories, and patient self –help groups. General education about medicines’ potential benefits and risks could also be offered by local health professionals as a complement to the increasing teaching that goes on in schools.

Finally, we must realise that our work in pharmacovigilance will fail if our communication is not excellent, repeated and outcomes monitored.