Reporting Adverse Drug Reactions: A guide for healthcare professionals
May 2006
Methods of reporting in the UK
- Spontaneous reporting
- What should be reported via the Yellow Card Scheme?
- Patient reporting
- How to report an ADR and what happens next
- Communicating information about ADRs
- Reporting other safety problems to the MHRA
- Structured databases
- Other databases
- Limitations of structured databases
The most common way that regulatory bodies collect ADR information for medicines once they are on the market is through voluntary, spontaneous reporting structures. In the UK, the Yellow Card Scheme is run by the MHRA and the Commission on Human Medicines (CHM) and is used to collect information on ADRs from healthcare professionals and members of the public. The original CHM, known as the Committee on Safety of Drugs, was established in 1964 following the thalidomide tragedy and since then, over half a million reports have been collected. Detailed information about the Yellow Card Scheme can be found on the MHRA website at www.mhra.gov.uk.
When the Yellow Card Scheme was first introduced, only doctors and dentists could submit reports. Gradually this has been extended and now all healthcare professionals, including coroners, pharmacists and nurses are able to report ADRs via the scheme. Available data indicate that the introduction of nurse and pharmacist reporting is proving to be very useful (see references 18, 19 and 20). The extension of reporting to other professionals should not, however, lead to complacency by doctors, who should continue to use the Yellow Card Scheme and take responsibility for reporting suspected ADRs to the regulatory authorities.
It is important to appreciate firstly, that the database used by the MHRA can detect duplicate reports. If, therefore, a doctor deems it necessary to submit a Yellow Card they should do so even if there is a possibility that someone else might have done the same. Secondly, different people will include different information when they complete a Yellow Card, all of which is useful in creating a full picture of the reaction that has taken place.
Yellow Cards should be submitted to either the MHRA directly or to one of five regional monitoring centres (RMC). Paper Yellow Cards are available by writing to either the MHRA or one of the RMCs, and can also be found in copies of the BNF, the Nurse Prescribers’ Formulary (NPF), the Monthly Index of Medical Specialties (MIMS) Companion and from the Association of the British Pharmaceutical Industy (ABPI) Compendium of Data Sheets and Summaries of Product Characteristics. Electronic Yellow Cards were introduced in 2002 and can be downloaded from either the MHRA or the RMC websites (www.yellowcard.gov.uk).
The MHRA also receives ADR reports from pharmaceutical companies, which have a statutory obligation to report suspected serious ADRs. If a doctor passes details of an ADR on to the pharmaceutical firm which markets the drug, this information will subsequently be passed on to the MHRA.
What should be reported via the Yellow Card Scheme?
If it is suspected that a patient has experienced an ADR it should be reported using a Yellow Card. ADRs resulting from prescription medicines, herbal remedies and OTC medications can all be reported. If there is any doubt about whether or not an ADR has occurred and should be reported it is always best practice to submit a report. Causality does not need to have been established.
There are some instances where it is important that all suspected ADRs are reported. These are:
Black triangle drugs
When new drugs and vaccines are first marketed they are intensively monitored in order to confirm the risk/benefit profile of the product (see reference 21). Such products are labelled with an inverted black triangle , and healthcare professionals are encouraged to report all suspected ADRs which occur as a result of the use of all black triangle drugs regardless of the seriousness of the reaction. Newly marketed drugs will usually be intensively monitored for a minimum of two years. It should be noted, however, that a black triangle is not always removed after this length of time and any medication can be assigned a black triangle if it is considered that it needs to be intensively monitored.
For example, black triangle status may be re-assigned to an established product if it is granted a new indication or route of administration, if it is marketed as a new combination with another established active ingredient, or if it is targeted towards a new patient population. If a product is a black triangle drug this will be indicated in the BNF, the NPF, MIMS, and in the ABPI Compendium of Data Sheets and Summaries of Product Characteristics. Advertising material and patient information leaflets (PILs) should also include this information.
A full list of all black triangle products and further information about the black triangle scheme can be found on the MHRA website at www.mhra.gov.uk.
Serious reactions
All serious suspected reactions must be reported via the Yellow Card Scheme, regardless of whether a product is a black triangle drug or vaccine. The side effects of an established drug may be well known but if a serious reaction occurs it should always be reported so that rare or delayed effects can be identified, more detailed advice can be given on potential side-effects and comprehensive information can be used to compare the relative safety of medicines in the same therapeutic class.
Reactions which are considered serious include those that are:
- fatal
- life-threatening
- disabling or incapacitating
- result in or prolong hospitalisation
- congenital abnormalities
- medically significant.
Other examples of serious reactions, given by the MHRA, can be found in appendix C.
Areas of particular importance to the MHRA for reporting include ADRs in children and the elderly, delayed drug effects, congenital abnormalities and ADRs associated with herbal products.
ADRs in children
The MHRA asks that all suspected ADRs occurring in children under the age of 18, should be reported regardless of whether the medication is licensed for use in children. Children are often not exposed to medications during clinical trials and many medications are used in children even if they are not licensed for this purpose. This means that monitoring of drug safety is particularly important for this age group.
In 2005 the first BNF for Children (BNFC) was launched which gives ‘practical information to help healthcare professionals who prescribe, monitor, supply, and administer medicines for childhood disorders’ (see reference 25). This is a vital resource which can improve prescribing to children, and as with the BNF, the BNFC contains useful information that can help with the identification of ADRs. The BNFC clearly indicates whether a medication has a black triangle classification and also contains a few pre-paid paper yellow cards that can be used to report suspected ADRs.
This information about when an ADR must be reported is summarised in the flow diagram below.
Source: MHRA, www.mhra.gov.uk/home/idcplg?IdcService=SS_GET_PAGE&nodeId=750 (accessed March 2006)
ADRs in the elderly
Healthcare professionals should be particularly aware that the elderly may be more susceptible to adverse reactions and it is therefore important to monitor drug safety in this age group. Many elderly patients are more likely to be taking multiple medicines and also may metabolise them less effectively or be more sensitive to their effects. A study of hospital admissions published in 2004, showed that the median age of patients admitted with ADRs was 76 years, which was significantly older than those admitted without ADRs (median 66 years) (see reference 9). If detected, ADRs in the elderly should be reported according to the protocol outlined above.
Delayed drug effects
Some reactions may become manifest months or years after exposure. Any suspicion of such an association should always be reported to the MHRA. Examples of delayed reactions that might need to be reported include: lipodystrophy resulting from medications prescribed to HIV-infected patients, or any withdrawal reactions caused after a patient has been taking SSRI antidepressants.
Pregnancy and congenital abnormalities
It is vital that all suspected ADRs experienced by women during pregnancy are reported (see reference 26).
In cases where a baby is born with a congenital abnormality or where a pregnancy results in a malformed or aborted foetus and it is suspected that an ADR may be the cause, a Yellow Card should be submitted and this should include information of all medications taken during pregnancy.
Herbal medicines
The popularity of alternative therapies, including herbal medications, is growing. One study in the USA indicated that 10 per cent of adults use herbal medicines (see reference 27). The ‘natural’ content of herbal therapies means they are often considered to be ‘safe’ by the public who, along with many professionals, fail to recognise the potential potency of many such products (see reference 3). Patients and healthcare professionals should also be aware that interactions between herbal and other, prescribed or OTC, medications can also cause unexpected reactions. A serious example of this is reduced blood concentrations, which can result from an interaction between the herbal preparation St John’s Wort (hypericum perforatum) and some commonly prescribed drugs, including warfarin, ciclosporin, oral contraceptives, digoxin, theophylline, and selective serotonin reuptake inhibitors, leading to a loss of drug effect with potentially serious outcomes (see reference 3).
The MHRA regulates many herbal medicines but there are still remedies that are available on the market that have not been licensed. It is therefore important that any suspected ADR which occurs from a herbal remedy is reported and that as much information about the ingredients and the source of the remedy are included. In 2002 the CSM advised that the herbal ingredient Kava-kava was associated with rare cases of hepatoxicity which could be serious in nature (see reference 28). This safety concern led to use of Kava-kava being prohibited in unlicensed remedies (see reference 29).
Reclassified medicines
The number of medications that have been reclassified from prescription only medicines (POMs) to pharmacy (P); and from P to OTC (also known as general sale list (GSL)) has risen in recent years as the government encourages the wider availability of medicines alongside greater patient choice (see reference 3). Before a change in status is granted pharmaceutical companies have to demonstrate levels of safety dependent on specific criteria and provide appropriate prescribing information.
Examples of recent POM to P switches include chloramphenicol 0.5 per cent eye drops for the treatment of acute bacterial conjunctivitis and Zocor Heart Pro (simvastatin 10mg) to reduce the risk of a first major coronary event in people who are likely to be at a moderate risk of coronary heart disease (CHD). Clotrimazole for the treatment of Candidal vulvovaginitis (thrush) is an example of a P to GSL switch. With wider availability of these reclassified medicines healthcare professionals, in addition to patients, should pay extra vigilance to their use as rare ADRs or new drug interactions may be identified.
Patient reporting
Healthcare professionals need to be aware that patients, parents or carers, are now also able to take part in spontaneous reporting of ADRs. This has been introduced in recognition of the fact that the patient is directly affected by an ADR and they may notice features that otherwise go unreported. Patient reporting has not been introduced to replace reports from healthcare professionals but instead should be seen as a mechanism through which the data held by the MHRA can be enriched with patient experiences. Although this scheme is in its early stages, all GP surgeries and community pharmacists should have received information about patient reporting and they are encouraged to make their staff aware that it has been introduced. If a patient experiences an ADR they should also be informed that they may submit a Yellow Card if they wish and information should be available in the surgery or pharmacy about the reporting process.
Patients report suspected ADRs in the same way that healthcare professionals do except that they fill out a different form. Patients can also report suspected ADRs to the Yellow Card hotline, on freephone 0808 1003352. Yellow Card forms are available from GP surgeries, pharmacies and other outlets across the NHS, as well as from the Yellow Card hotline and the MHRA website. The patient reporting section is clearly signposted and includes straightforward information for patients about how and what to report.
For further information about medicines that are available in the UK, patients should refer to the website www.medicines.org.uk which lists medicines guides that have been developed as part of the Medicines Information Project. This project aims to encourage patients and the public to make informed choices about their health. It is a collaboration of organisations including the DH, the Royal Pharmaceutical society, NHS Direct and various voluntary patient organisations (see reference 30).
Despite the introduction of patient reporting, healthcare professionals should continue to report as usual. The benefit of having patient reports is that different information is included or may be presented in a different way and thus a broader picture of the ADR and its impact on the patient can be established. Another advantage of patient reporting is that ADRs resulting from OTC medications, or those that do not result in the patient visiting a doctor might now be reported. It is very important that if a healthcare professional observes an ADR which is serious or which results from a black triangle medication they submit a Yellow Card regardless of whether the patient has also submitted a report.
How to report an ADR and what happens next
When a Yellow Card is completed by a healthcare professional, information should be included about which drug is suspected to have caused the reaction, the reaction that has occurred, some information about the patient and contact details of the healthcare professional in case further details are required. The patient details which should be included are the sex, age and weight (if known) of the patient. It may be helpful to include the patient’s initials and/or a local identification number. The local identification number is any number or code that identifies the patient to the reporter, but not to the MHRA. It should be noted that by supplying these anonymised details a healthcare professional will not breach the confidentiality agreement they have with the patient set out in the Data Protection Act 1998 (see reference 31).
Although explicit consent from the patient is not required, it is best practice to inform the patient if a report will be submitted and to keep a copy of the Yellow Card on the patient’s notes. Doctors have an ethical responsibility to ensure that patients are generally aware that ADRs are reported in order to protect the public’s health. This could be done, for example, through notices in hospitals and GP surgeries about the Yellow Card Scheme.
Completed Yellow Cards should be submitted either directly to the MHRA or to one of the RMCs, which will then pass the information on to the MHRA. Once a Yellow Card has been received by the MHRA, the person who made the report will be sent an acknowledgement which will quote the unique identification number assigned to the report. Any information which might identify the patient is then removed from the Yellow Card, and the identification number is added. The Yellow Card is then scanned into the MHRA’s pharmacovigilance database. MHRA staff also input the anonymised information on to the database in a structured format.
The database facilitates the monitoring of ADRs and allows rapid analysis of ADR reports. At any point during this process the reporter may be asked by the MHRA to provide clarification or further information about the ADR. Yellow cards that are received from patients are entered on to the MHRA’s medicines safety database and are considered in the context of all other reports received from patients or healthcare professionals for that medicine. Further information for patients about the reporting and follow up processes can be found on the MHRA website (see reference 32).
The pharmacovigilance scientists and physicians at the MHRA use Yellow Card data to detect ‘signals’ of emerging drug safety problems. They assess the causal relationship between the drugs and reported reactions and identify possible risk factors contributing to the reaction. During this assessment data from other sources may also be referred to, for example (see reference 33):
- case reports in the literature
- pre- and post-marketing clinical trials
- epidemiological studies
- record-linkage databases
- data from other drug regulatory authorities.
The MHRA also works closely with other European regulatory authorities on pharmacovigilance matters. Regulatory changes may include restrictions in use, reduction in dosage, special warning and precautions. In some instances, where it is considered that the risks of a medicine outweigh the benefits, that medicine may be withdrawn from the market. A study of all the new active substances that were authorised in the UK between 1972 and 1994 showed that 24 out of 583 substances authorised during this period were withdrawn for reasons relating to quality (one), efficacy (one) or safety (22) (see reference 34). That corresponds to a withdrawal rate of 3.8 per cent on safety grounds.
Communicating information about adverse drug reactions
An independent review of the Yellow Card Scheme was published in 2004 and made recommendations about ways in which the scheme could be strengthened and how access to Yellow Card data could be improved (see reference 29). As a result of this review it is now possible for healthcare professionals and patients to access complete listings of all of the ADRs that are reported to the MHRA. These data are presented in the form of drug analysis prints (DAP) and include information about what reactions have been reported for each drug. All DAPs and guidance for their interpretation can be accessed via the MHRA website.
For access to more detailed ADR data, systems are being set in place by the MHRA to enable ADR data to be accessed for research and education purposes for the benefit of public health. All requests for ADR data not releasable under Freedom of Information legislation are scientifically reviewed by an independent scientific committee (Independent Scientific Advisory Committee for MHRA Data Research (ISAC) which was established in February 2006) with ethical review provided under the established framework of the Central Office for Research Ethics Committees (COREC) system. For all ADR data requests, consent from a reporter and patient would always be required before access to their data are permitted in line with the provisions of the Data Protection Act, 1998. Further guidance on accessing ADR data for research purposes will be made available on the MHRA website within the first quarter of 2006.
As well as the comprehensive data included in the DAPs, the MHRA communicates with healthcare professionals and patients in a number of ways. Statutory patient information leaflets (PILs) and Summaries of Product Characteristics (SCPs) for healthcare professionals are continually updated with new safety information, while all urgent warnings will be communicated in letters sent out to all doctors and pharmacists. The DH also the Public Health Link, an electronic cascade system, to disseminate urgent information about ADRs and defective products to healthcare professionals when there is not sufficient time to organise a hard copy mailing (see reference 35). The MHRA and the CHM also produce a regular drug safety bulletin called Current problems in pharmacovigilance which is sent to all doctors and pharmacists and is available on the MHRA website. In addition to the freely accessible DAPs, staff at the MHRA provides information on the safety of medicines upon request.
Reporting other safety problems to the MHRA
It should be noted that the MHRA does not only collect information about adverse reactions to medicines and vaccines. Other safety matters that should be reported to the MHRA include:
Defective medicines
The MHRA runs the Defective Medicines Report Centre (DMRC) through which complaints and reports of actual and suspected defects in medicines are collected and assessed. Full information about the DMRC and how to submit a report can be found on the MHRA website.
Adverse incidents involving medical devices
The MHRA receives reports of adverse incidents from both manufacturers and users of medical devices. The Manufacturer Online Reporting Environment (MORE) is an online reporting system for medical device manufacturers and suppliers. The MHRA website can also be used to access online adverse incident reporting forms for medical device users, healthcare professionals, carers, and members of the public.
In 2004 8,840 adverse incidents were reported. Recent examples of reported incidents include over-infusion by a syringe pump which could have been potentially life-threatening and a problem with the fixing system for wheelchair backrests which it was found could have caused serious injury. In both instances an investigation found the cause of the fault; consequently a risk alert was issued against use of the pump, and the wheelchair manufacturer made design alterations and upgraded all existing backrests (see reference 36).
Haemovigilance and the serious adverse blood reactions and events (SABRE) reporting system
Since 2005, the MHRA has been required to collect reports about serious adverse events and serious adverse reactions related to blood and blood components. This is done via a secure online reporting system known as SABRE, to which Blood Establishments and Blood Banks/Hospital Transfusion Teams are able to submit reports. This method of haemovigilance does not replace local reporting arrangements for non-serious events and Serious Hazards of Transfusion (SHOT) reporting should also continue, although this can be facilitated within the SABRE reporting system. Full guidance and information about SABRE and other UK haemovigilance activities can be found on the MHRA website.
Structured databases
Drug safety problems are not only assessed through spontaneous reporting but can also be evaluated through the information from large patient databases. Databases which record all events related to the healthcare of a cohort of individual patients or all instances when a particular drug is prescribed are considered to be extremely useful as they represent what happens under normal conditions of clinical practice rather than only recording when a safety problem or adverse reaction occurs (see reference 16). Such databases are important resources for testing hypotheses generated from Yellow Card data.
General Practice Research Database
A multi-disciplinary team at the MHRA is responsible for the operation of the General Practice Research Database (GPRD). The GPRD is the world's largest database of anonymised longitudinal medical records from primary care (see reference 37). Currently over 300 practices across the UK are registered with the GPRD. The database has three million active patient records and a total over 35 million patient years of validated data. Participating practices supply the GPRD with a wide range of information covering all aspects of patient care, including use of medications (see box 2). Any GP practice that wants to contribute to the GPRD can apply to do so.
These real-life clinical practice data can be utilised for a range of applications, including clinical trials, drug safety, outcomes research and clinical epidemiology. The large population covered by the GPRD means that it has significant value for pharmacovigilance. The data available include ADRs, co-prescription, co-morbidity, dosage details, off-label prescription, and patient demographics (see appendix D for further details).
Information collected from GPs by the GPRD
- Demographics, including age and sex.
- Medical symptoms, signs and diagnoses, including comments.
- Therapy (medicines, vaccines, devices).
- Treatment outcomes.
- Events leading to withdrawal of a drug or treatment.
- Referrals to hospitals or specialists.
- Laboratory tests, pathology results.
- Lifestyle factors (height, weight, BMI, smoking and alcohol consumption).
- Patient registration, practice and consultation details.
There are different levels of access to the GPRD which can be purchased by interested parties under licensing arrangements and a number of services are offered. The GPRD has been used internationally by the pharmaceutical industry, regulators and in academic research. So far, the GPRD has been used to inform over 400 clinical reviews and research papers.
Information about the GPRD, including details of what data are collected and how the database can be accessed, are available on the website, www.gprd.com
Prescription-event monitoring
The Drug Safety Research Unit (DSRU) in Southampton uses prescription-event monitoring (PEM) to evaluate the safety of newly marketed drugs intended for use primarily in general practice (see reference 38). The DSRU, which independently of the MHRA and other government offices, uses the hypothesis generating technique of PEM to determine drug safety problems through pharmacoepidemiology.
PEM is an observational cohort technique which collects data on all prescriptions for the first 20,000 to 50,000 patients given a new drug. After a defined period of time, all the doctors who prescribed the new drug are sent a green form on which they are asked to record events reported by the patient subsequent to the prescription. These outcome data then form the basis of the evaluation study. Not all green forms are returned, but the average cohort size is 10,942 for PEM studies. Since the DSRU was established in 1980, PEM studies have been completed for 90 different medications. Whilst those studies that are for regulatory or internal purposes are kept confidential, over 150 scientific publications have so far resulted from the work of the DSRU (see reference 38).
The DSRU website contains information about PEM and also lists the medications that have been monitored to date as well as details of all the work that has been published www.dsru.org.
Other databases
Other databases which are useful for pharmacovigilance and which healthcare professionals should be aware of include:
The Tayside Medicines Monitoring Unit (MEMO) – www.dundee.ac.uk/memo
MEMO is a University of Dundee based research collaboration that undertakes research into the safe, effective and cost effective use of medicines and devices as well as helping to improve the understanding of disease, all using anonymised healthcare data (see reference 39). An important feature of MEMO is that it allows record linkage with other datasets, such as those relating to hospital care, which enables a broader understanding of healthcare events or the benefits and risks of particular treatments.
The Health Improvement Network (THIN) – www.thin-uk.com
THIN is a medical research database of anonymised patient records from information entered by general practices in their ViSion systems. Data are supplied to approved researchers for drug safety and epidemiological studies (see reference 40).
QResearch – www.nottingham.ac.uk/~mczqres
The QResearch database is run by Nottingham University. It contains data from 468 general practices throughout the UK with records for 3.3 million current patients and 4 million past patients (see reference 41).
Limitations of structured databases
Despite the strength of the data collected by the GPRD and the DSRU, limitations exist and they are yet to be used to the fullest. The Academy of Medical Sciences has identified areas of research which could benefit significantly from greater use of these existing databases including maternal drug exposure on pregnancy outcome, neonatal and early childhood health, and the use of medicines in children (see reference 16).
Both GPRD and PEM only collect information from general practice and thus there are significant gaps in collection of data from secondary care, where more complex medications tend to be used and where the potential for serious ADRs to occur is therefore greater. A lack of financial incentive results in at least 30% of GPs choosing not to return the green cards used in PEM which creates the potential for bias in the data. The Academy of Medical Science notes that the effects of this bias are unknown (see reference 16).
The high price of using the GPRD is likely to be prohibitive to some research groups. Access to these data ranges from between £7,000 and £60,000 for datasets to cover a single research question to between £27,500 and £300,000 per year for varying degrees of online access (see reference 42). It is worth noting, however, that the Medical Research Council (MRC) has now acquired access to the GPRD for the benefit of the UK research community (see reference 43). The MRC hopes to encourage high quality research in public health and primary care by providing an opportunity for UK-based researchers to access to the GPRD datasets. Information about the MRC, the conditions of their GPRD license and the eligibility requirements for researchers wishing to access the GPRD in this way, can be found on the MRC website at www.mrc.ac.uk.