Chapter 4 - COMMUNICATING RISK
From:
Always Read the Leaflet – Getting the Best Information with Every
Medicine
Report
of the Committee on Safety and Medicines
Working
Group on Patient Information
Medicines
and Healthcare products Regulatory Agency
2005
See http://medicines.mhra.gov.uk/inforesources/publications/Alwaysreadtheleaflet.pdf
The law requires each PIL to provide
information on all the side effects that have been identified for the medicine
concerned, to present the information in a logical order and to include a
description of the side effects, estimates of their frequency (or probability
of occurring) and advice on any necessary actions. In practice there is a vast
range of differing procedures and standards, indicating a need for more
detailed guidance in this area.
No EU guidelines exist specifically for
presentation of risk information in PILs, and current practice is highly variable.
It will never be possible to cater for every aspect of every individual’s
perceptions about risk in a single PIL, but it should be possible to improve
general standards of risk communication. Patient organisations and other
experts have helped the Group to identify the main problems with risk
communication in PILs, and we have developed flexible, practical guidance to
the pharmaceutical industry which, together with user testing of all new PILs,
will help to ensure that PILs meet medicine users’ needs.
Key areas for guidance have been
considered:
■ putting the most important information first;
■ including information on benefit;
■ using the right words;
■ using numbers appropriately to convey risk.
In addition to providing guidance on the
optimal presentation of risk in PILs, the Working Group has considered the
potential for developing a supplementary leaflet about the risks and benefits
of medicines, which can be read in addition to PILs accompanying medicines and
which provides general background information on the safe use of medicines.
4.1 RISK COMMUNICATION ABOUT MEDICINES – POTENTIAL
IMPACT ON PUBLIC HEALTH
Patients naturally want to be involved
in decisions relating to their health. Decisions such as whether to undertake a
course of treatment, and which treatment to choose, can only be truly shared if
the patient has a similar understanding of the possible advantages and
disadvantages of each option as the clinician. Accurate and effective risk
communication is therefore of great importance in establishing trust, reaching
shared agreements and developing concordant clinical management plans.
In recent years, the safety of medicines
has become the focus of intense media attention. High profile ‘scare’ stories
about medicines have become frequent. The failure of such stories to
communicate risk information in an accurate and balanced way can have serious
public health consequences. One example has been the increased number of
unwanted pregnancies after ‘scares’ about the contraceptive pill. Therefore,
high quality risk communication has an important role for public health, both
at an individual and population level.
4.1.1 Challenges in risk communication
for medicines
Risk is the probability or likelihood of
a negative or undesirable outcome. For medicines, a negative outcome usually
means a harmful side effect. Risk communication in this context therefore
provides understandable information on:
(a) the harmful effect itself;
(b) the probability of it occurring; and
ideally
(c) how to minimise this risk and what
actions to take in the event of a problem.
The level of trust placed in the
information source is also an important factor in determining the effectiveness of the information provided2 4. Trust in large organisations such as governments and
pharmaceutical companies is far from universal, but transparency and a
willingness to share uncertainties about information are likely to be helpful.
Patient information leaflets cannot
possibly deal with every nuance of risk perception or cater for the needs of
every patient. However, good quality risk communication is worth striving
towards and should be achievable. For patients to make informed choices about
the use of medicines, they should also understand the probability and nature of
the beneficial effects of the medicine, and consider possible benefit and harm
side by side. Achieving this within a single document is far from
straight-forward. Providing an accessible format and concise descriptions to
convey drug side effects and their seriousness without being alarmist can be
difficult. Likewise, great care is needed when providing statistical
information to ensure that it presents risks in a clear and unbiased way.
4.2
RISK COMMUNICATION IN PATIENT INFORMATION LEAFLETS – CURRENT PRACTICE
The meeting with patients’ organisations
and the reports described in Chapter 3 indicate that one of the key problems
with patient information leaflets is the way in which the inherent risks with a
medicine are communicated. Particular concerns related to the lack of balanced
information in relation to benefit and the lack of clarity or hierarchy in the
way in which the side effects are set out. A major contributory factor may be
the lack of specific guidance in this area.
4.2.1 Existing guidance on writing
patient information leaflets
Little guidance on risk communication is
available to those writing patient information leaflets and there are no formal
EU guidelines. Some information on how to express side effects is included in
the readability guideline which recommends that information about side effect incidence
should be conveyed to the patient where these data are available. In addition,
the guidance also recommends that very serious, typical undesirable effects
should be mentioned first or specially highlighted, irrespective of their
frequency. This applies in particular where there is an urgent need for the
patient to take action. Another method of presenting this section of the
leaflet, if frequency data are not available, is according to the seriousness
of the side effects or the body system affected.
The measures to be taken to remedy or
alleviate undesirable effects should be mentioned. If a patient needs to seek
help urgently, the term “immediately” is recommended. For less urgent
conditions, use of the phrase “as soon as possible” may be used. At the end of
the side effects section, the patient should be invited to communicate any
undesirable effect, especially if it is not mentioned in the leaflet, to a
doctor or pharmacist.
Nevertheless, it is fair to say that,
although the advice offered within this guidance document is helpful, it is far
from comprehensive and is certainly not applied consistently across all patient
information leaflets currently available.
4.3 IMPROVING THE COMMUNICATION OF RISK
IN PATIENT INFORMATION LEAFLETS
In considering how best to improve the
way in which risk is communicated in patient information, the Working Group
undertook a number of fact finding exercises. A literature review was conducted
to identify concepts associated with risk and risk expression and to bring
members to a common understanding of the factors involved. Separately, expert
advice was sought from members and external advisors, including patient group
representatives.
A meeting with patient groups in July
2004 (see Chapter 3) helped to bring the users’ perspective into the Group’s
thinking. At this meeting the MHRA also introduced the proposals set out below
to improve risk communication in PILs by including key points as a summary at
the start, giving information on the benefits of taking the medicines and
guidance on presenting statistical information. The views of attendees on the
proposals were generally positive and a number of constructive suggestions were
made on how to improve the comprehensibility of the information, on important
information that needed to be conveyed and proposals for how to achieve this.
These were taken into account in finalising the Working Group’s guidance.
4.4 REVIEW OF THE RELEVANT LITERATURE
4.4.1 General aspects of risk perception
Risk perception is a complex cognitive
function and subject to many influences. Not only does it (usually) require
some understanding of statistical probability, but is likely to be influenced
by beliefs and values. Different personality types may also affect the way
individuals approach the subject of risk. For example, some individuals might
generally take a ‘fatalistic’ approach to risk (i.e., assume that it is futile
to attempt to control risk), while others may prefer to manage risks actively
(e.g., by rules and regulations, public consultation or individual risk
assessment)1 2.
4.4.2 Biases in thinking about risk
Common biases in thinking about risk
have been identified in the literature. These include:
Awareness bias: whereby increased awareness of an issue
(e.g. through media reports) can lead to an exaggerated perception of risk;
Optimistic bias: whereby individuals tend to rate their
chances of avoiding mishap as ‘better than average’;
New risk bias: new risks generate more fear than risks
that have been experienced for some time;
Catastrophic bias: risks that are perceived as
catastrophic, e.g. those killing many people at once in one place, are feared
more than those that are chronic,
killing an equal or greater number of people but over time and in
scattered locations.
4.4.3 Fright factors
Risks are generally more worrying3 (and less acceptable) if perceived as:
■ involuntary rather than voluntary;
■ inequitably distributed (some benefit while others suffer
the consequences);
■ inescapable by taking personal precautions;
■ arising from an unfamiliar or novel source;
■ resulting from man-made, rather than natural sources;
■ causing hidden and irreversible damage, e.g. through
illness after many years of exposure;
■ posing some particular danger to small children or
pregnant women or, more generally, to future generations;
■ threatening a form of death (or illness/injury) arousing
particular dread;
■ damaging to identifiable rather than anonymous victims;
■ poorly understood by science;
■ subject to contradictory statements from responsible
sources (or, even worse, from the same source).
4.4.4 Sources of information and trust
Whether or not information on risk is
heeded can be heavily influenced by the level of trust that is put in the
source of the information. In general terms, information arising from large
organisations, corporations, governments and their Agencies is trusted less
than that from individual credible professionals, such as doctors or
scientists.
4.4.5 Expression of risk from a
regulatory perspective
At present, the main emphasis in
expressing risk from a regulatory perspective (SPCs and PILs) is in terms of
statistical probability and corresponding verbal descriptors – e.g. ‘very rare’
corresponds to ‘up to 0.01% (less than 1 per 10,000)’.
Verbal Descriptor EU assigned frequency*
Very common > 10% (more than 1 per
10)
Common >1% and <10% (less than 1
per 10 but more than 1 per 100)
Uncommon 0.1% to 1% (less than 1 per 100
but more than 1 per 1000)
Rare 0.01% to 0.1% (less than 1 per 1000
but more than 1 in 10,000)
Very rare up to 0.01% (less than 1 per
10,000)
* Key: > more than < less than
The available literature suggests that a
statistical approach to describing risk is often met with satisfaction by the
recipients. However, research into what individuals understand by terms such as
‘very rare’, ‘common’ etc5 6, suggests that the current EU
guidelines on verbal descriptors are not correctly matched with statistical
probabilities. In general, it appears that the public equate the verbal descriptors
(very rare, common etc) to risks that are substantially higher than those
defined in regulatory documents. Perceiving very small risks is particularly
problematic and a number of models have been proposed in the literature to help
with this. One scale7 is based on a different set of verbal
descriptors (high, moderate, low, very low and minimal), but this too may not
be in accord with people’s actual interpretations. Other scales attempt to
relate potentially hazardous events to familiar concepts such as:
■ scales relating the size of various communities as risk
comparators one/street/town etc7;
■ a scale using the risk of other events such as car
accident, murder, lightning as comparators8 9.
Meaningful interpretation, however,
necessitates a reasonable familiarity with the comparator.
4.4.6 Other issues relating to
statistical information
Other issues relating to effective
expression of the statistical magnitude of risk have been discussed in the
literature. The main points are summarised below:
■ denominators: it is better to use the same denominator
throughout when
describing more than one risk8;
■ relative and absolute risk: it is helpful to express
information on relative and absolute risks, so that the baseline risk and risk
attributable to the medicine are clearly identified10;
■ framing: perceptions of statistical risk can be
influenced by either positive or negative ‘framing’ (e.g. a 1/100 risk of
adverse outcome is equivalent to a 99/100 chance of no harmful effect, but the
perception of these two statements may be quite different)8 11;
■ use of diagrams (e.g., bar charts) and pictures can be
helpful in describing statistical probabilities8 12.
4.4.7 Format: word, number and pictures
There is no single format of risk
description that caters optimally for every situation. Choice of words and
style of language is clearly important in conveying qualitative and
quantitative aspects of risk. In some situations, pictures and numbers can
assist understanding.
4.5 MOVING TOWARDS BETTER RISK COMMUNICATION
The main areas considered by the Working
Group are discussed below. In some areas an iterative approach was taken to
developing and reviewing ideas. In this way, the Group was able to try out new
concepts (with worked examples) to see how risk communication principles
identified in the literature translated into PILs. In some cases, the concepts
considered were not seen to be useful in PILs and were consequently not taken
forward into the final guidance document.
4.5.1 General aspects of risk perception
and risk individualisation
Trust and recognition
PILs are produced by the relevant MA
holder and PILs frequently have the logo of the relevant pharmaceutical
company. Whether or not an MHRA logo or ‘seal of approval’/quality mark would
influence patients’ views on the validity or trustworthiness of the document is
unclear. Much would depend on the public reputation and standing of the MHRA.
This is one possible area for consideration in the context of the MHRA’s
proposals to increase public recognition of its role.
Transparency is an effective method of
improving trust in many situations. Improved transparency over the sources of
data and certainty/uncertainty of risk estimates presented in PILs might
therefore be helpful in establishing trust. Whilst it would not be possible to
reference every source of information in PILs, directing the reader to
additional background information could be helpful in establishing trust,
especially if the additional source would generally
be regarded as being highly credible.
Attention to ‘fright factors’
The literature suggests that a number of
‘fright factors’ (see section 4.4.3) might generally cause exaggerated
concerns. Providing clear information on the true scale and nature of such
risks in PILs is therefore important. The Working Group has considered a number
of approaches which might be particularly valuable when referring to potential
‘fright factors’. Such approaches include:
■ use of analogies and alternative risk scales to represent
rarity of risk;
■ describing baseline risk and increased risk with the
drug;
■ provision of further information sources on these risks.
Individualising risks
Inevitably, PILs can only describe risks
as they apply to the population as a whole. Patients might therefore benefit
from some guidance in the PIL about factors that could modify particular risks,
so that they can interpret the general information in a way that is relevant
for each individual. The Group therefore supported the development of generic
supplementary information on risks and benefits, which could contain a range of
information regarding safe use of medicines, including general information
about risk factors for side effects, and risk minimisation. This is further
discussed below at section 4.5.5.
4.5.2 Access to the most important
information for safe and effective use: headlines
The length and apparent complexity of
PILs is likely to be a disincentive to read the document13. Although European and national legislation demand that PILs
contain comprehensive clinical information, extra-statutory information is
permitted and this provides an opportunity to improve the accessibility of key
messages.
Headline information, presented
prominently at the beginning of a PIL and summarising a few key messages for
safe and effective use, is one option that has been considered in depth by the
Working Group. Such ‘headlines’ could not attempt to summarise all the
information in the PIL, but could include carefully selected messages that are
important.
The Group took into account the
development of “Drug Facts” established in the
The Group looked at whether or not the
use of such headlines needed to have a sound evidence base and considered the
need for user testing on their content and impact on the patient reading the
rest of the leaflet. Although there appears to be no published evidence to
support this additional information within the PIL, the requirement placed on
marketing authorization holders to undertake consultations with target patient
groups would provide evidence on individual PILs which will go some way to
addressing this concern.
The proposals for headline information
were also endorsed at the patient meeting in July 2004 and tested in practice
during discussion of the Seroxat PIL by a focus group of representatives from
SSRI patient interest and user groups (see section 3.1.3 for details).
Information that could be included in
headlines
There is no set formula or list of
issues that should be considered for headlines for every product but, for most
products, it is likely that at least one or two key messages can be identified
from the following:
■ why the patient should take the product;
■ the maximum dose or duration of treatment;
■ potential side effects/withdrawal reactions (symptoms to
look out for, especially for common or serious side effects);
■ contraindications;
■ important drug interactions;
■ circumstances in which the drug should be stopped;
■ what to do if the medicine doesn’t work; or
■ where to find further information.
As with other aspects of the PIL, it
would be important to ensure that headlines do not appear alarmist or overly
‘negative’. Careful wording and the use of a non-promotional statement of the
licensed indication (for example, “Your doctor has prescribed [PRODUCT]
because it is a treatment for migraine headaches”) at the start should help
to ensure a balance is achieved.
It would also be important that
headlines include a firm encouragement for the
patient to read the rest of the leaflet. The following concluding
headline was proposed. Inclusion of the date of latest revision may be helpful
to long-term medicines users who would not otherwise realise that the PIL has
changed since they last read it.
“Now read the rest of this leaflet. It includes other important information
on the safe and effective use of this medicine that might be especially
important for you. This leaflet was last updated on xx/xx/xx.”
How headlines might appear – an example
On the following page is an example of
six headlines which the Group decided were possible for the anticonvulsant
carbamazepine. As a general rule, it is unlikely that more than six headlines
would be helpful.
Carbamazepine 200mg Tablets
Important things that you need to know:
■ Carbamazepine tablets are prescribed for different
illnesses including epilepsy, manic-depression and neuralgia.
■ Take carbamazepine regularly to get the maximum benefit. You should not stop taking the medicine
without talking to your doctor. Sometimes stopping the medicine can cause
problems.
■ Carbamazepine can cause side effects, although most
people do not have serious problems – see page 2 for details. If you have
fever, sore throat, skin rashes or skin yellowing, mouth ulcers, bruising or
bleeding, see your doctor immediately.
■ Some side effects may occur early in treatment. These
often disappear after a few days as your body gets used to the drug (for
example dizziness, drowsiness or clumsiness).
■ Taking other medicines, including other anti-epilepsy
drugs, may sometimes cause problems. Check with your doctor or pharmacist
before taking any other medicines.
■ If you are (or might become) pregnant while taking
carbamazepine, it is important to talk to your doctor about this.
Now read the rest of this leaflet. It includes other important information
on the safe and effective use of this medicine that might be especially
important for you. This leaflet was last updated on xx/xx/xx
4.5.3 Striking a balance – conveying
information on benefits as well as risks
A common criticism of PILs is that much
of the advice and information relates to possible side effects and other
warnings, and that this can appear frightening and might dissuade some patients
from taking their medicines14.
Clearly, a balance should be struck
between providing information on all possible side effects (as is required in
EU legislation) and being non-alarmist. To some extent the problem can be
tackled by ensuring that information on side effects is presented optimally
(see section 4.5.4), but a further option is to include ‘positive’ information
in the PIL about the potential benefits of taking the medicine.
Currently, the section of the PIL
entitled “What is your medicine and how does it work” includes
information on the pharmacotherapeutic group to which the medicine belongs and
the indications for which it is authorised. The inclusion of more information
about the benefits of taking the medicine in this section would be
extra-statutory, but consistent with current legislation. In addition, there is
evidence from the literature that short factual statements on the
benefits of medicines can help patients
weigh up the risks and benefits of the medicine.15.
As with ‘headline’ information, there is
no single formula for including benefit information that is likely to be
appropriate for all medicines. Any additional information should be compatible
with the Summary of Product Characteristics, useful to the patient and should
not be promotional. The Group considered that some of the following issues
could be covered in a few sentences (about 80 words or less):
■ why it is important to treat the disease and what the
likely clinical outcome would be if the disease remained untreated;
■ whether the treatment is for short term or chronic use;
■ whether the medicine is being used to treat the
underlying disease (i.e., curative) or for control of symptoms. If the latter,
which symptoms will be controlled and how long the effects will last;
■ whether the effects will last after the medication is
stopped;
■ where the medicine is used to treat two or more discrete
indications, all should be succinctly described as above;
■ where to obtain more information on the condition.
It would seem likely that ‘benefit’
information would be most helpful for prescription medicines and, in
particular, preventative or long-term treatments.
On the following page is an example of
how the expanded information might appear for an inhaled corticosteroid for the
prevention of asthma. This concept has gained support from the Group and from a
number of patient group representatives who attended the meeting at the MHRA in
July 2004.
Without benefit information
PRODUCT contains beclometasone
propionate which is one of a group of medicines called corticosteroids. These
have an anti-inflammatory action and are used to treat asthma.
With benefit information
PRODUCT contains beclometasone
propionate which is one of a group of medicines called corticosteroids, or
“steroids”. Corticosteroids prevent attacks of asthma by reducing swelling of
the air passages and are sometimes called “preventers”. You should take this
medicine regularly every day even if your asthma is not troubling you. Using
PRODUCT can help prevent severe asthma attacks which sometimes need hospital
treatment and if left untreated could even be life-threatening.
This medicine should not be used to
treat a sudden asthma attack; it will not help. You will need to use a
different inhaler (“reliever”) to deal with these attacks.
4.5.4 Better information about side
effects
Many patients want to know about the
risks attached to taking their medicine and whether symptoms that they
experience might be a side effect. Crucially, all patients need to know what to
do if they encounter serious problems.
Ideally, PILs should provide information
on all the side effects that have been identified for a particular medicine.
The information should be provided in a logical order and include a description
of the side effects, estimates of their frequency (or probability of occurring)
and advice on any necessary actions.
However, although these principles are
straightforward, examination of current PILs reveals a vast range of differing
practice and standards and the need for more detailed guidance in this area.
The sections below cover the main areas and principles that the Working Group
decided to take forward with advice from the patient groups.
Putting the most important information
first
It is likely that the order and style of
presentation of side effects can greatly affect a patient’s perception of the
risks. The current guidelines16
require a description of
possible side effects consistent with the SPC and recommend that the most
serious side effects requiring action are presented first. However, in many
current PILs, side effects are simply put in order of body system, exactly as
in the SPC.
There are many possible ways to
categorise side effect information. The most important information for patients
relates to those situations where they may need to take action, such as
stopping the medicine or seeking medical help.
Separating this type of information from
less important issues and placing it first therefore seems logical. In some
cases, early identification and prompt action could avert major consequences.
Examples of side effects that would fall into this category are:
■ gastrointestinal bleeding with non-steroidal
anti-inflammatory medicines;
■ angioedema/facial swelling with any medicine;
■ tendon pain with fluoroquinolones;
■ unexplained muscle pain with statins;
■ painful swollen leg (possible deep venous thrombosis)
with oral
contraceptives.
Usually, the most serious side effects
are also the rarest and in order to avoid unnecessary alarm, it would be
important to include information on the frequency of such side effects wherever
possible.
Using the right words
For all side effects, the description is
critical to patients’ understanding. Ideally, the description should convey an
accurate impression of the side effect, including the symptoms that patients
are likely to experience. For example, gastrointestinal bleeding would be
recognised by the patient as either black or blood-stained vomiting or stools,
often with abdominal pain.
Words should be carefully chosen not
only to describe the side effect, but also to convey seriousness or severity.
This is particularly important for conditions that are unfamiliar to most
patients and might otherwise be misunderstood.
For example, any description of
rhabdomyolysis should not simply describe muscle breakdown, but should also
mention the severity of symptoms and the possible serious complications.
Many side effects are dose-related. PILs
should advise patients that higher doses, needed to achieve full
benefit/efficacy in some patients, may be associated with an increased risk of
side effects. A general warning statement may suffice in some circumstances,
but care is needed to ensure that the warning is not alarmist to those who have
been prescribed high doses.
Specific statements relating to
individual side effects may be appropriate if an important dose-relationship
exists (e.g. muscle side effects with statins).
Glossary of lay terms
There are many factors to consider when
describing side effects. Currently, descriptions of side effects are submitted
by companies and assessed individually for each PIL, resulting in differing and
inconsistent terminology. For patients, who may read about the same side effect
described in two or more quite different ways, this inconsistency is likely to
be unhelpful. Standardisation of side effect terminology would therefore seem desirable,
and adoption of ‘preferred lay terms’ for specific side effects would also be
helpful to industry and regulators. The Group has endorsed a proposal to
develop a glossary of side effect terms, and a current draft of the first 56
terms is at Annex 8. It is envisaged that further terms will be added to the
glossary.
Conveying risk with numbers
Conveying the concept of small risks has
been discussed in section 4.4.5 and options include reference to scales such as
those discussed there. All these scales suffer a limitation however in that
meaningful interpretation necessitates a reasonable familiarity with the
comparator, and an individual’s perception of the risk of a comparator event
may be heavily influenced by their own experiences. For example, someone who
has witnessed or been involved in a car accident might have a completely
different perception of this risk compared to someone who has not. Thus,
although conceptually attractive, the use of analogies to convey the magnitude
of risk is itself prone to bias and has not been taken forward into guidelines.
The Working Group and patient groups
have considered the suitability of a variety of approaches to expressing
statistical risk in PILs. A number of key principles have been identified from
worked examples.
i. Quantifying risk: use of
absolute numbers e.g,, 1 in 10,000 patients. If possible, baseline risk and
absolute excess risks should be presented.
ii. Verbal descriptors of risk (e.g.,
‘very rare’) should only be used if accompanied by the equivalent statistical
information. For example: “Very rarely (fewer than 1 in 10,000 patients
treated)…”.
iii. Conveying uncertainty around
risk estimates: imprecision of point estimates should be conveyed using
terms such as ‘approximately’/’about’/’around’ when referring to estimates for
major safety issues (for example “about 5 extra cancers for every 1000
patientstreated”). The Group considered an alternative approach of
including a range of values (for example “between 3-7 extra cancers for
every 1000 patients treated”), but worked examples of PILs have shown this
can lead to cumbersome statistics, and it is not clear that showing such ranges
of risk would improve patient understanding, or modify perceptions.
iv. Frequency ranges: to simplify
descriptions, it is preferable to use only the upper bound for each range. For
example, use ‘fewer than 1 in every 1,000’ rather than ‘between 1 in
10,000 and 1 in 1,000’. Worked examples have shown that this approach lessens
the burden of statistical information in PILs and may therefore improve
readability.
v. Duration of risk: it is
important to state the duration over which the excess risk applies if this is
known. For example, the risk of serious blood disorders with the antipsychotic
medicine clozapine is known to differ during the first 18 weeks versus weeks
19-52, and weeks 53 and above.
Another example is the excess of risk of
cancer associated with hormone replacement therapy (HRT) that can be stated in
relation to the number of years of treatment. Similarly, if it is known that
specific side effects may occur shortly after starting the drug and are likely
to be transient, this information is helpful to include in the PIL.
vi. Frequency estimates based on
spontaneous adverse drug reaction (ADR) data: reporting rates are likely to
be an underestimate of true incidence or risk. This should be stated in the PIL
when referring to data based only on spontaneous ADR data.
vii. Constant denominators: in
some cases, it may be helpful to express the risk of adverse reactions using a
‘constant denominator’, for instance when expressing small differences in risk.
Worked examples based on existing PILs have suggested that constant
denominators might appear confusing when comparing greatly differing risks; for
example, comparing a risk of 1
in 100 versus 1 in 10,000 would be
represented as 100 in 10,000 versus 1 in 10,000. Consultation with
patient groups suggested the use of constant denominators might occasionally be
appropriate; however, user testing will be key to ensuring that this concept is
understandable.
Constant numerator (1) Constant
denominator (10,000)
1 in 10,000 1 in 10,000
1 in 1,000 10 in 10,000
1 in 100 100
in 10,000
Risk
using constant numerators and constant denominators
A number of concepts have been
considered to be inappropriate to take forward for general guidance. Some of
these are discussed below.
■ NNT/NNH. Numbers needed to treat or harm17 18 are
calculated as the reciprocal of absolute reduction or increase in risk. Such
calculations are most accurately obtained from clinical trial data, and the
most useful comparisons are with placebo. However, robust, placebo-controlled
clinical trial data regarding important risks are not always available. A
further disadvantage of NNT/NNHs is that they are usually dependent upon
duration of treatment. This means that an NNT based on a clinical trial of two
years duration is only relevant for patients who take the medicine for this
period of time. As it is not routinely possible to calculate NNTs in PILs, and
as the calculations may be subject to misinterpretation, this concept has not
been taken forward into guidelines.
■ Positive framing and negative framing. This concept was also informally tested
in worked examples and found to be too cumbersome and lengthy for PILs. In
addition to these concerns, patient groups did not find the concept helpful and
were concerned that ‘positive framing’ might resemble a marketing ploy rather
than a genuine attempt to provide balanced information. An example of positive
and negative framing is shown below. The following side effects may affect
fewer than 1 in 10,000 people. This means that at least 9,999 out of 10,000
people are not expected to experience one of these side effects….
■ Use of diagrams: diagrams can be helpful but constraints on size means that these
could only rarely be used in PILs and no formal guidance on their presentation
has therefore been taken forward.
4.5.5 Supplementary information – a
leaflet about risks and benefits
PILs provide comprehensive information
that is specific for one medicine. Understanding this information and putting
it to best use may require at least a rudimentary prior understanding about
medicines and their side effects, and some patients may lack this
understanding. In order to address this, the following general leaflet on risks
and benefits of medicines was developed by the Working Group.
Benefits and side effects of medicines –
some questions and answers
What is this leaflet about?
This leaflet aims to answer some
questions you may have about taking medicines and the risk of side effects. You
will find more information about your particular medicine in the patient
information leaflet provided with your medicine.
If you have received a medicine but no
leaflet, please ask your pharmacist to get one for you.
Need further advice?
If you need further advice about
medicines remember you can ask your doctor or pharmacist or call NHS Direct
& NHS Wales/Galw lechyd Cymru on 0845 46 47 (text phone 0845 606 46 47) or
NHS 24 (Scotland) on 08454 24 24 (textphone 18001 08454 24 24).
1. How do medicines work?
The medicine you are taking may:
• cure your condition - for
example an antibiotic, which is used to treat an infection;
• control your condition - for
example a medicine to lower your blood pressure;
• treat the symptoms of your
condition - for example a painkiller to take for
toothache;
• prevent you from becoming unwell -
for example a vaccination against disease.
2. Will my medicine cause side effects?
• The expected benefit of your medicine
will usually be greater than the risk of suffering any harmful side effects.
• Most people take medicines without
suffering any side effects.
• However, all medicines can cause
problems. Your patient information leaflet will list all the known side effects
linked to your medicine.
Important: most people take medicines without
suffering any side effects.
3. What is meant by a "common"
or "rare" side effect?
The chance (the risk) of having a side
effect can be described using words or figures or both. This is how risk may be
described in your patient information leaflet:
• Very common means that more
than 1 in 10 people taking the medicine are likely
to have the side effect.
• Common means that between 1 in
10 and 1 in 100 people are affected.
• Uncommon means that between 1
in 100 and 1 in 1,000 people are affected.
• Rare means that between 1 in
1,000 and 1 in 10,000 people are affected.
• Very rare means that fewer than
1 in 10,000 people are affected.
Remember, if a side effect has a risk of
1 in 10,000, then 9,999 out of every 10,000 people taking the medicine are not
expected to experience that side effect.
4. Does a high dose increase the risk of
side effects?
In general, a high dose of a medicine is
more likely to cause side effects. However, high doses are sometimes needed to
ensure maximum benefit. To get the
maximum benefit from your medicine you need to take the recommended dose for
you.
• For medicines you have bought yourself,
the dose is written on the carton or container label and in the patient
information leaflet.
• For medicines that have been
prescribed by your doctor, the dose will be on the pharmacy label. The doctor
will have prescribed a dose for you that takes into account your age, weight,
how ill you are and any other medicines you may be taking. Only change your dose if you have discussed
it with your doctor first.
• With some medicines, you will start on
a low dose that will gradually be increased over a period of weeks (or months).
With other medicines you will stay on the same dose throughout your course of
treatment.
• Sometimes, when you need to stop
taking a medicine, your doctor will gradually reduce the dose to avoid
unpleasant withdrawal effects. You should not increase or decrease the dose
prescribed by your doctor unless you have discussed it with him/her first.
Important: check the patient information leaflet
and speak to your doctor or pharmacist if you feel unwell after your dose has
been increased.
5. How can I reduce the risk of side
effects?
• Take your medicine as your doctor or
pharmacist has advised you.
• Be careful about mixing medicines.
Some medicines should not be taken together. Before taking a new medicine, it
is important to tell your doctor or pharmacist about any other medicines you
are already taking, including herbal remedies or any nonprescription medicines
you may have bought for yourself in a pharmacy or supermarket.
• Understand about "risk
factors". Sometimes risk factors increase the chance of your medicine
causing side effects. These factors will vary depending on what medicine you
are taking. For example, you may be able to lower your risk of side effects by
not drinking alcohol or not eating certain foods during your course of
treatment. Your patient information leaflet will tell you about any known risk
and what you can do to reduce the chance of side effects.
6. Do side effects always come on
straight away?
It depends on the medicine and the
person.
• Some side effects can happen
immediately - for example an allergic reaction.
• Others might not start for several
days or weeks - for example skin rashes - or even longer - for example stomach
problems with some painkillers.
• In general, side effects are most likely
when you start a new medicine or after your
dose has been increased.
• Quite often, mild side effects will go
away as your body adjusts to the new medicine or dose.
7. What should I do if I feel unwell
after taking my medicine?
• Check your patient information
leaflet: it may contain all the advice you need.
• If in doubt, speak to your doctor,
nurse or pharmacist or call NHS Direct, NHS Direct Wales/Galw lechyd Cymru or
NHS 24 on the numbers given at the beginning of this leaflet.
• For worrying or serious effects you
may have to stop taking the medicine, or need other treatment.
• For less serious side effects, you may
be advised to continue with your medicine, or change the dose.
• You or your health adviser can report
suspected side effects to the drug safety watchdog (MHRA). Telephone 020 7084
2000 to find out more.
8. Will my medicine affect my lifestyle?
Although most medicines will not affect
your lifestyle, some can. Examples are:
• Some medicines may affect your vision
or co-ordination or make you sleepy. This may affect your ability to drive,
ride a bicycle or perform skilled tasks safely.
• Some medicines may affect your sex
drive.
• You may need to stop drinking alcohol
or eating certain foods while taking some medicines.
Important: your patient information leaflet will
tell you about any lifestyle issues and advise you about things you should
avoid.
Patient groups were consulted in
developing draft proposals for the leaflet about risks and benefits. The
leaflet was also pilot tested at the MHRA. Annex 9 provides the protocol and
results of the pilot testing.
The Group advised that this leaflet
could be presented in a variety of different formats. Options include paper
leaflets for doctors’ surgeries or pharmacies, or electronic information for
websites (e.g. MHRA and NHS Direct Online).
Electronic formats could also be made
available for health care professionals (e.g., in the National Electronic
Library for Health), or for patients (e.g., ‘My HealthSpace’ – a secure place
on the internet for patients to store information relevant to them).
4.6 CONCLUSION
There are no EU guidelines specifically
dedicated to best practice in risk communication, and other existing guidelines
to aid with the preparation of PILs contain relatively little advice on this
subject. As a result, current practice in risk communication is highly variable
and the outcome is often poor. There is scope within the current legal
framework on PILs to develop risk communication guidelines and the future
changes to the law will give further opportunities to assist with better risk
expression in the leaflet. A number of options for consideration have been
identified or derived from the literature. Discussion with patient
representatives has highlighted a number of concerns which could be addressed
in guidance. Selection and prioritisation of key principles that can be
practically implemented in PILs were the first stages considered in the
development of new guidance. The guidelines which have been developed can be
found at Annex 10.
It is likely that the current variable
standards of PILs in the
Even with this guidance the Group
recognises that it will only be when the final patient information leaflet is
tested with patient groups that the full impact of the principles will be realised.
The guidelines are living documents and as new evidence emerges from, for
example user testing, the principles will be updated to reflect this knowledge.
References
1. Eisner, Richard J. Communication and
interpretation of risk. British Medical Bulletin 1998; 54 (No 4): 779- 790.
2. Gray G M & Ropeik D P. Dealing
with the Dangers of Fear: The Role of Risk Communication. Health Affairs 2002;
21 (No 6): 106-116.
3. Department of Health. Communicating
about Risks to Public Health – Pointers to Good Practice. 1997.
4. Breakwell G M. Risk Communication:
factors affecting impact. British Medical Bulletin 2000: 56 (No1) 110-120.
5. Reagan R T, Mosteller F & Youtz
C. Quantitative meanings of verbal probability expressions. Journal of Applied
Psychology 1989; 74: 433-442
6. Berry D C, Raynor D K & Knapp P.
Communicating risk of medication side effects: an empirical evaluation of EU
recommended terminology. Psychology, Health & Medicine 2003; 8 (No 3);
451-263.
7. Calman K C & Royston G H D. Risk,
Language and Dialect. BMJ 1997; 315: 939-942
8. Paling J. Strategies to help patients
understand risks. BMJ 2003; 237: 745-748
9. Paling J & Paling S. Up to your
Armpits in Alligators.
10. Stewart FH, Shields W C & Hwang
A C. Presenting health risks honestly: mifepristone, a case in point.
Contraception 2004; 69: 177-178.
11. Buetow S, Cantrill J & Sibbald
B. Risk Communication in the Patient-Health Professional Relationship. Health
Care Analysis 1998; 6: 261-270
12. Schapira M M, Nattinger AB &
McHorney C A. Frequency or probability? A qualitative Study of Risk
Communication Formats Used in Health Care. Medical Decision Making 2001;
Nov-Dec: 459-467
13. Knapp P & Raynor