Variability in occurrence of embolic signals in long term transcranial Doppler recordings

Albeit still unproven, it is supposed that the presence and number of asymptomatic circulating cerebral microemboli detected by transcranial Doppler ultrasound (TCD) may be an indicator of stroke risk. Little research has been done to assess the reproducibility of these data and the required time for recording. We examined one middle cerebral artery in 7 subjects with carotid, aortic or cardiac embolic source for 24 h by TCD. Analysis for embolic signals was done off line completely blinded to the diagnosis and the time of the day during the recording. Embolic signals were found in all 7 subjects varying from 0 per h to 13 per h. Embolic signals occurred throughout the day with a nonsignificant tendency towards higher values in the early morning hours. Half an hour recording would not have been suitable to rule out or to confirm the presence of embolic signals as there were too many gaps between embolic signals of more than 1/2 h. One hour is the required minimum. Concerning the number of embolic signals, even a recording of four hours yielded variabilities of 0.25 and 8 embolic signals per hour in the same person. We recommend to perform follow-up studies the same time of the day. In patients with a low number of embolic signals longer recordings or a lower detection threshold with a higher number of detected embolic signals are necessary to compare frequencies of embolic signals.