Placebo blinding in Randomised trials of acute respiratory Infections in primary care: a Meta-Epidemiological Study

PI Title: Taeko Becque

Lead member: Southampton

Coughs and colds are one of the most common conditions seen by GPs and lead to a lot of antibiotic prescribing.
Randomised controlled trials are one of the best ways of measuring how effective treatments for coughs and colds are. They often compare people in two groups: one group who take an active treatment, for example antibiotics, and another group who don’t take an active treatment. There are two main types of randomised trials:
1) Placebo trial – the group who is not receiving antibiotics receives a placebo which looks, tastes and smells the same as an antibiotic, but does not contain any active ingredients. Patients are not aware whether they are receiving antibiotics or placebo, which controls for the potential psychological effect of taking a medication.
2) Open trial – the group who is not receiving active treatment does not receive any treatment at all, so participants know whether they are taking an antibiotic or not.

The main concern with open trials is that patients may report longer symptom duration if they know that they are receiving no treatment. This would lead us to exaggerate the effectiveness of antibiotics. For example, in a placebo trial, patients may report their symptoms last an average of 7 days in the antibiotics group and 8 days in the placebo group, implying that symptoms are 1 day shorter in the antibiotics group. In an open trial, patients may report their symptoms last an average of 7 days in the antibiotics group and 10 days in the no treatment group, implying that symptoms are 3 days shorter in the antibiotics group.
Producing placebos can be expensive and difficult, and patients are less willing to participate in trials when they don’t know the treatment they are taking. Open trials are cheaper and easier to set up, and mirror what would happen in normal practice, but participants may report their symptoms differently if they are aware of their treatment.

We plan to search for all the relevant studies and combine them to see whether the differences in symptom duration are significantly longer in open trials compared to placebo trials. We will also measure how bad the symptoms are, and whether they need to go to hospital due to their cold or chest infection getting worse.
If treatment effects are found to be significantly larger in open trials, then we would need to interpret open trials with caution and the use of placebos in future trials of coughs and colds would be recommended. If treatment effects are not significantly different, then open trials may be more suitable, leading to lower costs and a wider range of participants.