For the video version, click here.
Much of clinical research is problem-solving. You find an issue out there in the medical world – a problem – and then test an idea to fix it. The problem we face this week is the amount of antibiotics received by individuals with asthma. As researchers from Imperial College, London, point out in the Annals of Family Medicine, the vast majority of upper respiratory infections in those with asthma are of viral origin. Antibiotics would not be expected to confer much benefit in this group.
Some small studies, mostly in children with asthma, have suggested that the use of probiotics –
bacterial cocktails that may affect the immune response – carries some benefit in terms of URI rates and antibiotic usage.
The researchers wanted to determine whether that affect translated into a large trial.
Now – how would you do such a trial? The obvious answer would be to take a big group of people with asthma, give half of them probiotics and half placebo and see how they do. That is not what was done here. The researchers performed what’s known as a pragmatic trial.
Pragmatic trials focus on the real-world effectiveness of an intervention, as opposed to an explanatory trial which examines a more idealized efficacy. For an explanatory trial, patients are specifically recruited and extensively screened. They are monitored carefully. Researchers ensure they are receiving, and taking, their study medication.
A pragmatic trial is more real-world. In the case of this trial, they randomized a group of individuals with asthma to receive a standard mailing (encouraging flu vaccine) versus the same mailing plus some information about probiotics and free coupons to receive them.
So the probiotics weren’t sent. The patients had to a) see the mailing, b) order the probiotics on their own initiative and c) take them.
Trusting initiative may be a recipe for disaster. Of the 652 people sent the probiotic coupons, 121 ordered a one-month supply of probiotics. 86 ordered two or more month’s worth. So just 200 people in the intervention group actually received the intervention at all.
Against that backdrop, there was no statistical difference in antibiotic prescriptions when the study was analyzed using all patients. Focusing just on those who actually took probiotics (a subgroup that might have been sicker over all or more attentive to their own health needs), we still see no difference in antibiotic usage compared to the control group.
So are probiotics dead in the water? Not necessarily. Aside from the fact that most people in the intervention group of the study didn’t get the intervention, the decision to link probiotic use to antibiotic prescriptions seems premature. Shouldn’t we first see if probiotics reduce the risk of upper respiratory infections, then worry about antibiotics?
I’d like to suggest that the next study in this space should be a more mechanistic, traditional randomized trial. But I won’t hold my breath.