Should we grade the severity of intrapulmonary RLS, how should we grade it, and what should we do with a grade 1 RLS? I prefer the quantitative systems described by Gazzaniga et al8 and van Gent et al,3 which use < 20 and < 30 bubbles, respectively, as the upper limit for a grade 1 RLS. Both quantitative systems have a negative predictive value of 100% for the presence of treatable PAVM and good interobserver agreement.3,8 Several authors have pondered whether we should defer chest CT scan in patients with grade 1 RLS unless their grade increases but ultimately concluded that they did not have enough data to recommend a change.3,4 However, Gazzaniga and colleagues8 seemed to favor a change based on their additional data. If one combines the data from the three studies that graded TTCE and used CT scan as a gold standard, there are a total of 144 patients with confirmed or suspected HHT who had a grade 1 RLS by TTCE.3,4,8 Ten patients (6.9%) had PAVM on CT, whereas none had treatable PAVM. When no events are observed in a trial, the “rule of three” states that “we can be 95% confident that the chance of this event is at most three in n.”9 Therefore, the upper limit of finding treatable PAVM with a grade 1 study would be 2.1% (3/144),3,4,8 which is identical to the upper limit of finding a treatable PAVM if the TTCE is completely negative (3/148).3,8 Since we are already comfortable deferring chest CT scan when TTCE is negative,4 the above figures tell us that we should be no less comfortable deferring it with a grade 1 RLS.