The Interventional Effect Cannot Be Interpreted
Schöberl et al. reported the therapeutic effect of interventional patent foramen ovale (PFO) closure on the risk of ischemic cerebral insults: “However, the number needed to treat by PFO closure to prevent another stroke is 67 (e9).” (1)
Although it is desirable that a therapeutic effect should be summarized in a single number, this extremely brief reporting style is bound to result in a situation where readers cannot interpret the therapeutic effect anymore (2). The NNT is calculated by inverting the risk reduction (1/risk reduction). Without answers to the following questions, the NNT cannot be interpreted:
1. Which treatment alternative (comparator) was interventional PFO closure compared with? And:
2. What was the time frame for which the risk of ischemic cerebral insults was determined?
The risk differences determined by studies, and the NNTs crucially depend on these two factors. The meta-analysis cited by Schöberl et al. (3) included three studies with varying comparators: platelet aggregation inhibition in isolation or combined with anticoagulation. In the PFO closure group, medication treatment (platelet aggregation inhibitors) was also given; and durations of treatment differed (3).
The phrase “to prevent another stroke” is incorrect. The NNT of 67 means that if 67 patients are treated by PFO closure and 67 patients are treated with the comparator treatment, one additional ischemic cerebral insult can be expected to be prevented in the PFO group over an observation period of 2.5 years. Schöberl et al. are not referring to an additional case, but one case. I have my doubts whether the NNT was a helpful measure for doctors in terms of risk communication.
Prof. Dr. med. Andreas Stang, MPH
Institut für Medizinische Informatik, Biometrie & Epidemiologie (IMIBE)
Conflict of interest statement
The author declares that no conflict of interest exists.
|1.||Schöberl F, Ringleb PA, Wakili R, Poli S, Wollenweber FA, Kellert L: Juvenile stroke—a practice-oriented overview. Dtsch Arztebl Int 2017; 114: 527–34 VOLLTEXT|
|2.||Stang A, Poole C, Bender R: Common problems related to the use of number needed to treat. J ClinEpidemiol 2010; 63: 820–5 CrossRef MEDLINE|
|3.||Kent DM, Dahabreh IJ, Ruthazer R, et al.: Device closure of patent foramen ovale after stroke: pooled analysis of completed randomized trials. J Am Coll Cardiol 2016; 67: 907–17 CrossRef MEDLINE PubMed Central|