Journal Logo


Utility of transcranial doppler ultrasonography for confirmatory diagnosis of brain death: two sides of the coin

Dosemeci, Levent1 3; Dora, Babur2; Yilmaz, Murat1; Cengiz, Melİke1; Balkan, Sevin2; Ramazanoglu, Atilla1

Author Information
doi: 10.1097/01.TP.0000092305.00155.72


Since the advent of cardiopulmonary support for severely brain-injured patients, the diagnosis of brain death has become a medical, legal, and social issue. Brain death is now generally accepted both legally and medically as a criterion for death. Clinical-examination criteria have been established to determine brain death, and subsidiary investigations such as electroencephalography (EEG), arteriography, and radionuclide scans are used either to determine the arrest of cerebral circulation or to confirm the diagnosis of brain death (1–5).

The growing demand for organ-transplant donors requires quick identification of candidates before the onset of multiple-organ failure. Although the confirmatory tests listed above provide the clinician with objective evidence in addition to clinical examination for brain death, these tests have certain disadvantages. Imaging tests are expensive, some of them are invasive, often not safe for patients, and commonly require the transportation of critically ill patients. Although more portable and noninvasive, EEG is time consuming, frequently occupying a technologist and EEG machine for more than 2 hours because of the required setup time and strict technical standards. When support staff is not present, such as at night or during the weekends, these methods of confirmation pose difficult logistic problems. Furthermore, the use of sedative drugs renders EEG unreliable (6).

The ideal method to assist in the diagnosis of brain death and show the absence of cerebral blood flow should be noninvasive, easy for the intensivist to perform at the patient’s bedside, inexpensive, safe, and specific. Transcranial Doppler ultrasonography (TCD) fulfils these criteria and has been demonstrated to be highly sensitive in documenting the arrest of cerebral circulation in brain-dead patients in many studies (7–11). Waveform abnormalities documented by TCD occur when intracranial pressure increases above the mean arterial pressure and are indicative of arrest or near arrest of cerebral circulation (8,12,13).

Although TCD seems to be the ideal method for confirmation of brain death, the cessation of cerebral blood flow and clinical brain death may not occur at the same time. Such cases may pose a difficult problem to the declaration of brain death. Should the TCD be repeated until cerebral circulatory arrest is demonstrated, thereby putting potential donors at risk of organ dysfunction, or should the patient be declared only on the basis of clinical findings? Studies investigating TCD for the diagnosis of brain death so far have not focused on this problem. The aim of the current study was to assess the clinical utility of TCD for confirmation of brain death and to question whether TCD should be an integral part of the protocol for the diagnosis of brain death.


Patients and Clinical Diagnosis of Brain Death

One hundred and nineteen consecutive patients with Glasgow Coma Scale score less than 5 were included in the study. In all patients, neurologic examination was performed and recorded by intensivists just before TCD examination, which was performed by a neurologist who was unaware of the clinical diagnosis. Three patients were excluded from the study because apnea tests of these patients could not be performed because of desaturation or arrhythmia. Seventy-six patients were diagnosed as clinically brain dead, and 40 patients were not brain dead during TCD examinations. Sixteen patients were excluded from the study because no flow was detected at the initial TCD examination (absence of bone window). Finally, the study population consisted of 61 patients with clinical brain death and 39 patients with no brain death. Fifty-nine patients were male and 41 female, and the mean age was 42.3±21.6 (2–80) years. Comatose state was caused by various pathologic conditions (Table 1).

Table 1
Table 1:
Cause of comatose state in 100 patients in the present series

Clinical Diagnosis of Brain Death

Clinical brain death was diagnosed according to national guidelines, which require the confirmation of irreversible coma, absence of brainstem reflexes, and a positive apnea test in a normothermic, nondrugged patient (14). Confirmatory tests are optional according to these criteria, but the minimum observation period during which the patient’s status should not change is 12 hours if the patient became irreversibly comatose because of a known reason and 24 hours for coma of unknown cause. When clinical diagnosis of brain death was confirmed by a committee consisting of a neurologist, a cardiologist, a neurosurgeon, and an anesthesiologist, patients were declared to the local center for organ donation and coordination. Although confirmatory tests are optional according to our law, we usually prefer having EEG or TCD results of brain-dead patients as a documentary record.

Confirmation of Brain Death Diagnosis by TCD

TCD examinations were performed with a Multidop X-4 TCD instrument (DWL Elektronische Systeme GmbH, Sipplingen, Germany) by using a 2 MHz transducer placed on the temporal bone window. Bilateral middle cerebral arteries, anterior cerebral arteries, and internal carotid arteries were insolated. A stable arterial blood pressure of not less than 90 over 50 mm Hg was maintained throughout the TCD examination.

The following flow patterns were regarded as consistent with the TCD diagnosis of brain death: (1) brief systolic forward flow or systolic spikes and diastolic reverse flow (Fig. 1A), (2) brief systolic forward flow or systolic spikes and no diastolic flow (Fig. 1B), or (3) no demonstrable flow in a patient in whom flow had been clearly documented in a previous TCD examination. The above TCD findings were only accepted as confirmatory of brain death when they were found bilaterally or in at least three different arteries for at least 3 minutes within the same examination. In patients with a unilateral absent temporal bone window, the finding of typical TCD signals in three different arterial segments on one side was accepted as confirmatory. The finding of diastolic back-and-forward flow (to-and-fro flow) (Fig. 2) was not considered as confirmatory of brain death, and the examination was repeated until brain death was confirmed ultrasonographically. Absence of flow in all arterial segments during the first TCD examination can be caused by total cerebral circulatory arrest or possibly by the absence of an adequate temporal bone window. We accepted all cases in which no flow could be detected at the initial TCD examination as having an “absent bone window.”

Figure 1
Figure 1:
(A) Brief systolic forward flow and diastolic backward flow, which confirms brain death. (B) Brief systolic forward flow and no diastolic flow, which confirms brain death.
Figure 2
Figure 2:
Systolic forward flow and diastolic to-and-fro flow as demonstrated by transcranial Doppler ultrasonography (TCD). This flow pattern was not accepted as confirmatory of brain death, and examinations were repeated.

In 15 (19.7%) of the 76 patients with clinical signs of brain death and in 1 (2.5%) of the 40 controls, no flow was detected at the initial TCD examination, and these cases were excluded from the analysis. In clinically brain-dead patients, TCD was repeated if previous TCD examinations did not show a flow pattern consistent with brain death. Forty-three of the 61 clinically brain-dead patients underwent a single TCD examination, and 18 patients were evaluated twice or more (Fig. 3).

Figure 3
Figure 3:
Course of the 119 patients included in the study. GCS, Glasgow Coma Score.


One hundred patients were included in the study. Sixty-one of these patients were clinically brain dead, and 39 were not brain dead when TCD examinations were being performed. In the brain-dead patients, the average time between admission of the patients to our intensive care unit (ICU) and the clinical diagnosis of brain death was 70.5±110.0 hours.

Of the 39 controls with an adequate bone window, 38 showed systolo-diastolic forward flow during TCD examination. One patient still had weak respiratory movements in response to the apnea test, which persisted for 24 hours despite TCD signs of cerebral circulatory arrest.

The initial TCD examination of the 61 patients clinically diagnosed as brain dead and having an adequate bone window revealed flow patterns confirmatory of brain death in 43 patients, but forward systolo-diastolic flow or a diastolic to-and-fro flow pattern was detected in 18 patients. The mean duration between first TCD and clinical examination was 1.8±1.2 (0.5–4) hours. The sensitivity and specificity of the first TCD examination were 70.5% and 97.4%, respectively.

Two of the 18 patients with forward flow died before the second TCD examination could be performed, but they were unsuitable for organ transplantation because of chronic organ failures. The remaining 16 patients underwent a second TCD examination 12.6±8.3 hours after clinical brain death. Brain death was confirmed ultrasonographically in 12 patients, but 4 patients still had systolo-diastolic forward flow or a diastolic to-and-fro flow pattern. The sensitivity after the second TCD examination was 93.2%.

One of the remaining four patients died before a third TCD could be performed. Two of the three patients who underwent a third TCD examination 36 and 48 hours after clinical diagnosis of brain death showed confirmatory TCD signs of brain death, but one patient still had marked systolo-diastolic forward flow. The sensitivity after the third TCD examination was 98.3%. The remaining patient underwent a fourth TCD examination 96 hours after clinical confirmation of brain death, which finally confirmed the clinical diagnosis. This case is reported below. The sensitivity after the fourth TCD examination was 100%. Results are illustrated in Figure 3. Thirty of the 58 (51.7%) patients declared brain dead became organ donors.


A 32-year-old women who suffered from headache collapsed while being examined by a neurologist. We admitted the patient to our ICU in a comatose state after she had been resuscitated successfully within 5 minutes. She did not have any motor response to painful stimuli, and there was no pupillary reaction to light on ICU admission. However, she had weak respiratory movements. We detected subarachnoid hemorrhage caused by aneurysm rupture and brain oedema on computed tomography.

Clinical brain death was diagnosed 3 days after admission. The first TCD performed 4 hours later showed almost normal TCD findings. TCD examinations repeated after 24, 48, and 72 hours of clinical diagnosis of brain death did not show waveform patterns confirmatory of brain death. We repeated the clinical tests for brain death every day before TCD examination, and there was no change in the clinical status. After 96 hours of clinical brain-death diagnosis, TCD showed brief systolic forward flow and diastolic reverse flow, which is confirmatory for brain death.


Many authors have used TCD to investigate cerebral blood flow in brain-dead subjects. A great number of these studies have presented very similar results concerning the specificity and sensitivity of this method for confirmation of the diagnosis of brain death, and the specificity of TCD has usually been 100% in all of these reports (8,9,15–17).

Kirkham et al. (12) studied 23 comatose patients with severe intracranial hypertension. In the 17 fatal cases (demonstrated by EEG and angiography), the authors observed two TCD waveforms, diastolic reverse flow, or brief systolic spikes. In some patients, no Doppler signal could be obtained. The authors confirmed that TCD was a useful technique in monitoring unconscious intensive care patients.

Hassler et al. (13) used TCD to monitor 71 patients with intracranial hypertension who were subsequently declared brain dead. With increasing intracranial pressure, TCD waveforms exhibited different characteristic high-resistance profiles, first with low forward diastolic flow, then no diastolic flow, and then reversed diastolic flow, depending on the relationship between intracranial pressure and cerebral perfusion pressure. As cerebral perfusion pressure finally became zero, three types of TCD waveforms were observed: (1) oscillating flow, (2) very small systolic spikes, and (3) no signal. In the series by Hassler et al. (13), no survival was registered for head-injured patients who developed one of these Doppler waveform patterns.

Petty et al. (8) showed that the finding of either absent or reversed diastolic flow or small early systolic spikes in more than one intracranial artery on TCD in a comatose patient was highly specific (100%) and sensitive (91.3%) for brain death. More recently, Shiogai et al. (18), who studied 58 cases of brain death, showed only two TCD waveforms: brief systolic forward flow and undetectable flow.

It should be emphasized that intracranial circulatory arrest and brain death may not be identical conditions (10,19,20) and that, between the development of cerebral circulatory arrest and total loss of brain function, there may be a time lag not exceeding 24 hours (20). In one case, we detected weak respiratory movements in response to an apnea test persisting for some hours after the first TCD signs of cerebral circulatory arrest. Similar findings of preserved weak spontaneous respiration for a short period of time (about 1 hour) were also reported by Kirkham et al. (12) and Newell et al. (21) in patients with circulatory arrest. In neither of these studies were these results regarded as false-positive because the signs of circulatory arrest proved to be irreversible, and the patients soon became apneic. Exact timing in these conditions is problematic, and additional factors such as initial mild hypothermia may have played a role in the delayed development of brain death. If TCD use is restricted to protocols that require patients to meet clinical criteria of brain death before TCD confirmation, false-positive tests will not occur. In addition, TCD of both middle cerebral and the basilar arteries could be an obligatory minimal examination, which allows one to avoid false-positive results (6,10). But the examination of the posterior circulation by way of the transformational bone window is extremely difficult and sometimes impossible in a bedridden patient who is intubated.

A number of reports show false-negative TCD results in patients with clinical and electroencephalographic signs of brain death. In these patients, preservation of forward flow throughout diastole was thought to be caused by ventricular drains or other reasons (1,11). TCD findings of preserved, positive diastolic flow were more frequent in neonates (22,23). In the study performed by Hadani et al. (10), the preservation of almost normal flow in all major intracranial vessels was observed in one case of brain death resulting from anoxic brain damage. Petty et al. (8) described the same observation. Anoxic brain injury does not always imply severe structural damage of the entire brain, and in some cases, irreversible failure of all vital functions of the brain occurs without persistent elevation of intracranial pressure, which causes total cerebral circulatory arrest. In our patients, the false-negative results cannot be explained by the presence of a ventricular drain or anoxic brain damage because both conditions were also present in some cases that did show cessation of flow. However, in the present series, the reason for coma was anoxic brain injury in a patient (see Case Report) who had preserved, positive flow in TCD for 4 days, although she was already diagnosed as clinically brain dead.

The results of our study show that all patients who had obvious flow, even at a time when they were clinically brain dead, finally developed cerebral circulatory arrest, giving a final sensitivity of 100% for TCD as a confirmatory tool in brain death. Although this finding supports the medical evidence that TCD is highly sensitive in confirming brain death, it raises questions about the necessity of its use. If all patients who are clinically brain dead finally develop cerebral circulatory arrest, then a TCD showing positive flow may delay declaration of brain death and may lead to the loss of possible organ donors. Three of our 18 clinically brain-dead patients died before cerebral circulatory arrest could be demonstrated. Our results may suggest that cerebral circulatory arrest is not necessary for the diagnosis of brain death but that it eventually develops even some time after brain functions have already stopped. However, in cases in which a complete clinical examination is not possible or the results are dubious, confirmatory tests are obligatory for the diagnosis of brain death.

On the other hand, in many countries, an observation time between 6 and 72 hours is required after the first examination showing brain death to the declaration of the patient as brain dead (24). This time could be shortened in patients in whom both clinical evidence of brain death and TCD confirmation of cerebral circulatory arrest can be demonstrated. Such patients could be declared brain dead without a second examination after a period of observation if appropriate legal measurements are taken. Only a few countries allow confirmatory tests to shorten the time to declaration (25).

A patient who is clinically brain dead and shows signs of cerebral circulatory arrest by TCD or any other method is easily declared brain dead. In addition, confirmation of cerebral circulatory arrest by TCD just after the initial clinical diagnosis of brain death may make a second examination 12 to 24 hours after the first one unnecessary. But patients who are clinically brain dead and still have obvious forward systolo-diastolic flow pose a medico-legal problem. The questions are as follows. Can a patient be regarded as brain dead without confirmation of cessation of cerebral blood flow just by clinical examination? When should a patient be declared brain dead? When the brain functions stop or when the cerebral blood flow ceases or both? Are confirmatory tests really necessary in patients with clinical diagnosis of brain death, or do they just delay the declaration of brain death? There is still no global consensus on these questions (24).

We suggest that TCD is a highly sensitive tool in the diagnosis of brain death but may shorten or prolong the time to declaration. Its sensitivity is increased with repeat examinations and should be repeated in cases in which neurologic examination is suspicious of brain death but that cannot be done properly and in which flow is demonstrated after the first TCD. The necessity of demonstrating cerebral circulatory arrest in the confirmation of brain death is debatable.


1. Paolin A, Manuali A, Di Paola F, et al. Reliability in diagnosis of brain death. Intensive Care Med 1995; 21: 657.
2. Ducrocq X, Braun M, Debouverie M, et al. Brain death and transcranial Doppler: experience in 130 cases of brain dead patients. J Neurol Sci 1998; 160: 41.
3. Braun M, Ducrocq X, Huot JC, et al. Intravenous angiography in brain death: report of 140 patients. Neuroradiology 1997; 39: 400.
4. Buchner H, Schuchardt V. Reliability of electroencephalogram in the diagnosis of brain death. Eur Neurol 1990; 30: 138.
5. Larar GN, Nagel JS. Technetium-99 m-HMPAO cerebral perfusion scintigraphy: consideration for timely brain death declaration. J Nucl Med 1992; 33: 2209.
6. Ducrocq X, Hassler W, Moritake K, et al. Consensus on diagnosis of cerebral circulatory arrest using Doppler-sonography: task force group on cerebral death of the Neurosonology Research Group of the World Federation of Neurology. J Neurol Sci 1998; 159: 145.
7. Valentin A, Karnik R, Winkler WB, et al. Transcranial Doppler for each identification of potential organ transplant donors. Wien Klin Wochenschr 1997; 109: 836.
8. Petty GW, Mohr JP, Pedley TA, et al. The role of transcranial Doppler in confirming brain death: sensitivity, specificity and suggestions for performance and interpretation. Neurology 1990; 40: 300.
9. Feri M, Ralli L, Felici M, et al. Transcranial Doppler and brain death diagnosis. Crit Care Med 1994; 22: 1120.
10. Hadani M, Bruk B, Ram Z, et al. Application of transcranial Doppler ultrasonography for the diagnosis of brain death. Intensive Care Med 1999; 25: 822.
11. Zurynski Y, Dorsch N, Pearson I, et al. Transcranial Doppler ultrasound in brain death: experience in 140 patients. Neurol Res 1991; 13: 248.
12. Kirkham FJ, Levin SD, Padaychee TS. Transcranial pulsed Doppler ultrasound findings in brain stem death. J Neurol Neurosurg Psychiatry 1987; 50: 1504.
13. Hassler W, Steinmetz H, Gawlowski J. Transcranial Doppler ultrasonography in raised intracranial pressure and in intracranial circulatory arrest. J Neurosurg 1988; 68: 745.
14. Law #2238 and Organ and Tissue Transplantation Services Regulation. Turkey. Enacted 1993.
15. Ropper AH, Kehne SM, Wechsler L. Transcranial Doppler in brain death. Neurology 1987; 37: 1733.
16. Powers AD, Graeber MC, Smith RR. Transcranial Doppler ultrasonography in the determination of brain death. Neurosurgery 1989; 24: 884.
17. Davalos A, Rodriguez Rago A, Mate G, et al. Value of the transcranial Doppler examination in the diagnosis of brain death. Med Clin 1993; 100: 249.
18. Shiogai T, Sato E, Tokitsu M, et al. Transcranial Doppler monitoring in severe brain damage: relationships between intracranial hemodynamics, brain dysfunction and outcome. Neurol Res 1990; 12: 205.
19. Wijdicks EF. Determining brain death in adults. Neurology 1995; 45: 1003.
20. Shiogai T, Takeuchi K. Relationship between cerebral circulatory arrest and loss of brain functions-analysis of patients in a state of impending brain death [abstract]. Rinsho Shinkeigaku 1993; 33: 1328.
21. Newell DW, Grady MS, Sirotta P, et al. Evaluation of brain death using transcranial Doppler. Neurosurgery 1989; 24: 509.
22. Glasier CM, Seibert JJ, Chadduck WM, et al. Brain death in infants: evaluation with Doppler in US. Radiology 1989; 172: 377.
23. Bode H, Sauer M, Pringsheim W. Diagnosis of brain death by Transcranial Doppler sonography. Arch Dis Child 1988; 63: 1474.
24. Wijdicks EF. Brain death worldwide: accepted fact but no global consensus in diagnostic criteria. Neurology 2002; 58: 20.
25. Haupt WF, Rudolf J. European brain death codes: a comparison of national guidelines. J Neurol 1999; 246: 432.
© 2004 Lippincott Williams & Wilkins, Inc.