Adverse nondrug reactions
Adverse events is an interesting topic, particularly in relation to clinical trials, and how we interpret the results of clinical trials for clinical practice. Adverse events described in patient information leaflets with prescribed drugs can lead to some interesting discussions with patients, especially when comments of frequency or severity of the adverse event are missing, limited, or misunderstood.
In single dose trials, the way in which adverse event data are collected in clinical trials can affect how many adverse events are recorded, and the extent of any differences between active treatment and placebo . In long duration trials, almost every patient will record at least one adverse event, and making sense of even common, reversible and minor adverse events is difficult. What we most want to know is how much more common an adverse event is with treatment.
When placebo is used in clinical trials and adverse events occur, all sorts of convoluted discussions take place about placebo “causing” adverse events. What we really need is some information about adverse event frequency in people not taking part in a trial and not on any drugs, as a background to inform our discussions and thinking. Despite this being obvious and important, Bandolier, prompted by a reader, could find very little information. Only two [2, 3] studies seem to have measured adverse events in a population not in a trial and not on any drugs.
Both studies were retrospective questionnaires about the presence of symptoms often listed as adverse events of drugs. The first was conducted in medical and nonmedical people in Philadelphia in the late 1960s, and the second reproduced the study in medical students in Magdeburg thirty years later.
In both studies, subjects initially recruited were screened for any diseases or medicines being taken, and only those without disease and taking no medicines were asked to complete the symptom questionnaire. The exception was that oral contraceptives were allowed in the German study. Subjects were asked to record whether they had any of the listed symptoms in the three days before questioning.
Most people in the surveys were young and in their 20s. Adverse event reporting was common, and the list of adverse events and the frequency with which they were reported is shown in Table 1. Fatigue was the most common symptom noted in 40% or more, with nasal congestion, sleepiness, irritability, headaches and pain in muscles and joints occurring in 10% or more (Table 1). There were no major differences between medical and nonmedical subjects, nor between studies conducted 30 years apart.
Table 1: Presence of symptoms often listed as adverse events over previous three days in healthy, young, individuals in USA and Germany, 30 years apart
|Inability to concentrate|
|Bleeding from gums after brushing teeth|
|Pain in muscles|
|Pain in joints|
|Faintness of dizziness when first standing up|
|Loss of appetite|
|Bleeding or bruising|
|Giddiness or weakness|
|Excessive bleeding from gums after brushing teeth|
Over the 544 subjects in all three groups, 83% reported at least one symptom over the preceding three days, and only 17% reported none of these symptoms. Large percentages of subjects reported multiple symptoms (Figure 1).
Figure 1: Frequency of reported symptoms
What are we to make of this high level of adverse event reporting in young, healthy individuals? The results were consistent, but both studies used retrospective rather than prospective questioning. It might be conservative, since people taking analgesics or medicines for allergies were omitted.
Certainly it makes us look at clinical trial results with high levels of adverse events, but no difference between active and placebo, in a new light. It is what we should expect, and not necessarily a product of being in a trial. We might also wish to speculate about the results we would find in older, but fit populations, or populations that include people who do have disorders being treated.
Why did Bandolier find only two studies? It may be that our searching was badly flawed, and in that case we would love to be put right by informed readers telling us where we can find out more. But it may just be that there is not much there, which would make this a fertile research territory for young, aspiring, professionals.
- JE Edwards. Reporting of adverse effects in clinical trials should be improved. Lessons from acute postoperative pain. Journal of Pain and Symptom Management 1999 18:427-437.
- MM Reidenberg, DT Lowenthal. Adverse nondrug reactions. New England Journal of Medicine 1968 279: 678-679.
- FP Meyer et al. Adverse nondrug reactions: an update. Clinical Pharmacology and Therapeutics 1996 60: 347-352.