Accepting the presence of any pulmonary shunt on TTCE as a clinical Curaçao criterion enhanced the number of positive criteria in 92 of 487 individuals (18.9%), which actually changed the clinical diagnosis in 30 of them (32.6%). The remaining 62 (67.4%) changed from three to four positive criteria, which had no clinical consequences because three Curaçao criteria already suffice for a definite diagnosis of HHT (Table 3). This increased criteria sensitivity from 88.0% (95% CI, 0.84-0.91) to 94.3% (95% CI, 0.91-0.96) at the expense of a decreased specificity from 73.9% (95% CI, 0.66-0.80) to 69.9% (95% CI, 0.62-0.77) (Table 4). This decline in specificity was completely related to the presence of grade 1 pulmonary shunts in individuals without HHT; all nine of the 153 people (5.9%) without HHT incorrectly given a diagnosis of possible or definite HHT showed a pulmonary shunt grade 1 on TTCE. The range of microbubbles counted in the left side of the heart in these nine individuals was between three and 19, which illustrates that there was no doubt about the presence of a pulmonary shunt grade 1 (and not grade 2) on TTCE. In contrast, genetic testing confirmed the presence of HHT in all individuals with a pulmonary shunt grade ≥ 2 on TTCE. Therefore, we also determined the diagnostic accuracy of the clinical Curaçao criteria with the addition of only pulmonary shunt grades ≥ 2 on TTCE as a positive criterion. This strategy enhanced the number of positive criteria in 30 of 487 individuals (6.2%), which correctly changed the clinical diagnosis in seven of 30 (23.3%) and prevented additional incorrect clinical diagnoses of HHT, with no difference between HHT1 and HHT2 (Table 5). Accepting the presence of only pulmonary shunt grades ≥ 2 on TTCE as a positive clinical Curaçao criterion actually improved the overall diagnostic accuracy of the current criteria, resulting in a sensitivity, specificity, PPV, and NPV of 90.1% (95% CI, 0.86-0.93), 73.9% (95% CI, 0.66-0.80), 99.0% (95% CI, 0.97-0.99), and 100% (95% CI, 0.97-1.0), respectively, in diagnosing HHT (Table 6).