DVT and pulmonary embolism after acute infection.

摘要

At first glance, this study by Liam Smeeth and colleagues (April 1, p 1075)1 was impressive, making use of a database comprising "more than 20 million person-years of observation from 220 general practices". The method used was laid out in an earlier paper on adverse events from vaccines.The authors of that paper reported on both the relative incidence and. attributable risk associated with the vaccine. We were surprised that Smeeth and colleagues chose to report only the incidence risk ratio, despite apparently having data on unexposed cases (used in adjustment for the effects of age) which would give access to attributable risk rates.Surely this was a missed opportunity. A doubling of relative risk might mislead the unwary reader into considering thromboprophylaxis after any lower respiratory tract or urinary tract infection. This would be appropriate if, for example, the risk doubles from 10 in 100 to 20 in 100, but would not make clinical sense if, for example, the risk increased from one in a million to two in a million. Perhaps the existing data set would allow this aspect to be investigated?