Assessing causality of adverse effects

Last updated: Sunday, July 05, 2015

When assessing whether a causal relationship between a medicine and an event exists, the enquirer often expects an opinion or advice on the likelihood that a particular problem or symptom is drug-induced. Sometimes it is relatively easy to do this, if, for example, the patient is taking only one medicine and no evidence is found to implicate the drug as the cause of the suspected reaction. However, when asked about a patient taking more than one medicine and who has several concurrent medical conditions, it can be very difficult to distinguish which, if any, of the medicines is causative.

The TRIP system can be used to help determine whether an ADR has occurred. TRIP stands for Timing, Recovery, Independent evidence, Predictability:

1. Timing

Depending on the type of reaction, time to onset can vary from minutes to months. For example anaphylaxis to IV drugs can begin almost immediately after administration, whereas drug-induced Parkinsonism may take several months to develop. An understanding of the pharmacokinetics and pharmacodynamics of the drug is required, as well as the mechanism of the ADR, when considering the time interval between drug exposure and onset of symptoms. For withdrawal reactions, the duration of administration and the length of the interval since stopping therapy are important timing considerations.

Courtesy of Simon Wills

2. Recovery

Do symptoms improve/resolve when the suspect drug is stopped? This is called dechallenge. Many reactions will resolve on stopping the drug; this supports cause and effect but could be coincidental. Does the reaction improve when a specific treatment is given? (e.g. procyclidine for oculogyric crisis caused by metoclopramide).

3. Independent evidence

Is there another possible non-drug cause for the problem? The clinical picture may be a manifestation of an underlying disease. For example, before attributing disturbed liver function to drug therapy, other causes must be excluded such as viral hepatitis or biliary obstruction. Is there objective evidence of a link with drug therapy? A rechallenge will cause symptoms to recur if the suspect drug is restarted. Positive rechallenge is strongly suggestive that the drug was responsible. Although deliberate rechallenge is seldom justifiable, in clinical practice it can occur by mistake (e.g. patient with known penicillin allergy given co-amoxiclav). There are other types of independent evidence such as plasma levels of drugs or measures of drug activity (e.g. INR).

Courtesy of Simon Wills

4. Predictability

Are the characteristics of the adverse event consistent or inconsistent with the known pharmacology of the drug? For example, many adverse reactions are dose-related and may be avoidable by using a reduced dose. Have similar events been reported previously with the drug or a related drug? (e.g. in textbooks, clinical trials, spontaneous reporting schemes).

These criteria provide useful general guidance but they have some drawbacks. For example, with new medicines there may be only very limited data available. Due to the size and controlled nature of pre-marketing clinical trials, only the most common adverse events will be observed and subsequently listed in the SPC at the time of approval. Rare events, if described, may only be encountered by chance. In this situation you will have less information for the 'predictability' element above.

A positive dechallenge (resolution of symptoms on stopping the suspect drug) is usually a strong indication of iatrogenic disease but coincidences can occur (e.g. rash in a child taking antibiotics may be viral). Failure of symptoms to resolve quickly may not always count against the possibility of an ADR (e.g. regrowth of hair following drug-induced alopecia will take weeks or months).