Monitoring ADRs: Need and methodes

Need for ADR monitoring: ADRs are responsible for significant morbidity and mortality and can adversely affect a patient’s quality of life. Studies have shown that ADRs are responsible for a significant number of hospital admissions, with reports ranging from 0.3% to as high as 11%. ADRs also have a significant impact on the healthcare costs. Studies have estimated that the occurrence of an ADR increased the cost of patient care by $2262 per patient.

Another study has estimated the cost of preventable ADRs in a 700 bed hospital to be $2.8 million per annum.

The importance of ADR monitoring has been highlighted time and again by advent of global tragedies with drugs that were initially believed to be safe like thalidomide, chloramphenicol, and the COX-2 inhibitor rofecoxib. ADR monitoring is especially needed because many less common but significant adverse effects of drugs can go unnoticed during clinical trials due to limited number of population and controlled environmental settings. In a prospective spontaneous reporting study conducted to assess the pattern and severity of ADRs in local population in a South Indian teaching hospital, a total of 270 suspected ADRs were reported and evaluated from 164 patients. A total of 3.7% of the hospitalized patients experienced an ADR, 0.7% of the admissions were due to ADRs and 1.8% had a fatal ADR. 

Very few patients are studied in clinical trials and only those adverse reactions with a relatively high incidence will be detected. Furthermore, these adverse events have no background incidence, i.e. they do not occur either as a natural disease or in response to something other than the drug. In a real setting, however, many adverse drug reactions tend to mimic signs and symptoms of non-drug induced diseases. So, in the setting of clinical trials, it becomes even more difficult to detect an adverse drug reaction. Therefore, it is important to devise methods to detect ADRs as quickly as possible after marketing, for confirming that the events detected during clinical trials are truly ADRs and for assessing their overall incidence, to detect other less frequent ADRs and to evaluate the balance of benefit and risk.   

Methods of monitoring ADR: Though different surveillance methods have been devised for ADR monitoring, none has yet proved completely satisfactory. These programs mostly aim for speed of detection of the ADRs, estimation of incidence or prevalence in the population, and the predisposing factors involved. This information can be used to estimate the benefit: risk ratio and also to prevent such reactions as far as possible. There are several methods of surveillance of ADRs, which are discussed in brief here.  

I. Anecdotal reporting: It is reporting of ADRs through anecdotal reports by individual doctors that a patient has suffered some peculiar effect due to a drug. Such reports need to be verified by further studies and they sometimes fail to confirm a problem. Still, the skill of individual observant clinicians is valuable in the detection of ADRs. Some examples of ADRs detected through this system are: jaundice caused on repeated administration of halothane, oculomucocutaneous syndrome due to practolol, etc.

II. Voluntary but organized reporting: In the UK, the ‘yellow card’ is currently used for voluntary reporting of suspected ADRs to the UK Committee on Safety of Medicines (CSM). These cards are found as detachable pages in the back of the British National Formulary (BNF). In case of a suspected ADR, the doctor should complete one of these cards and send it to the CSM. Although doctors are asked to report adverse reactions to any therapeutic agent, the CSM is particularly interested in reports on the following types of reactions:

1. Reactions to drugs that are marked with the symbol of a black inverted triangle in the BNF; these are drugs that have not been on the market for long.

2. Reactions  to any drug when the suspected ADR is serious, including reactions that are fatal, life threatening, disabling, or incapacitating, or that result in a prolonged stay in hospital.

3. Reactions that may represent delayed drug effect.

4. Congenital abnormalities.

5. Reactions to all vaccines.

Though this organized voluntary reporting system has provided extremely useful information on occurrence of ADRs in the national community, it has some disadvantages as well. The major disadvantages are:

1. Difficulty in spotting ADRs that are not known to exist.

2. Tendency of over reporting recently known ADRs and not reporting already established ones.

3. Tendency to under-report, mainly due to lack of time and pressure of other business.

III. Intensive event reporting: It is a hospital based ADR reporting system in which a group of individuals are designated to screen a defined population specifically to detect ADRs and then relate them to specific drugs. These programmes can provide interesting statistics on the occurrence of ADRs in hospitals but are not generally effective in detecting anything new because the population studied is relatively small and each patient is studied for only a short period of time. An example of such programme is the Boston Collaborative Surveillance Program. 

IV. Cohort studies (prospective studies):  In this method, the patients taking a particular drug are identified and events are then recorded. Here also, a relatively small number of patients are studied and there is lack of suitable control group in which to assess the background incidence of any adverse events. Also, such studies are very expensive and it is difficult to justify and organize such studies for every newly marketed drug.  

V. Case-control studies (retrospective studies): Here, patients who present with symptoms or an illness that could be due to an adverse drug reaction are screened to see if they have taken the drug. The prevalence of drug taking in this group is then compared in a reference population who do not have the symptoms or illness. This study is suitable for determining whether a drug causes a given adverse effect once there is some indication that it might, but it is not a suitable method for detecting completely new ADRs. The relationship between maternal stilbestrol ingestion and vaginal adenocarcinoma was confirmed by this method. 

VI. Case-cohort studies: This is a hybrid of prospective and retrospective studies. Here, patients who present with symptoms or an illness that could be due to an ADR are screened to see if they have taken the drug and the results are then compared with the incidence of the symptoms or illness in a prospective cohort of patients who are taking the drug. This technique has been used, for example, to study the risk of breast cancer in women taking oestrogens. 

VII. Use of population statistics: Registers of causes of death, congenital malformations and other information from other sources like hospital and other records, are screened for a change in the pattern of deaths. This might stimulate an investigation of a possible drug- related cause. For example, the increase in deaths among young asthmatics in the early 1960s seemed attributable to the increased use of bronchodilator aerosols containing non- selective adrenoceptor agonists. Subsequently, a similar increase in deaths in young asthmatics in New Zealand was attributed to the chronic use of b₂-adrenoceptor agonists. These findings have probably been a factor in the emphasis on prophylactic use of inhaled steroids, reserving b- adrenoceptor agonists for symptom relief. 

VIII. Record linkage: Here, the idea is to bring together a variety of patient records like:

i. General practice records of illness events

ii. General practice records of prescriptions

iii. Hospital records of illness events  and

iv. Hospital records of prescriptions.

In this way, it may be possible to match illness events with drugs prescribed. An example of the use of record linkage is the prescription- Event Monitoring Scheme, in which all the prescriptions issued by selected practitioners for a particular drug are obtained from the Prescription Pricing Authority. The prescribers are then asked to report any events in the patients taking the drug (whether attributable to ADRs or not) to those running the scheme. This scheme is less expensive and time consuming than other surveillance methods. 

IX. Meta- analysis: If many studies of an ADR have been performed, a meta- analysis of all available data can be done to determine whether there is an association.