The determination of brain death is essential for medical decision making in comatose patients. However, how to determinate brain death accurately with appropriate tests has been an ethical and academical focus for long time. Currently, there are some disconformities across countries on the procedures of brain death determination. Our study reported the variability on criteria of brain death in China, US, UK, Germany, and France, and investigated the diagnostic performance of ancillary tests according to criteria of different countries.
Apnea test is a mandatory technique to confirm the absence of spontaneous respiration in criteria of all the countries. The positive result of apnea test is arterial PCO2 ≥ 60 mm Hg (or 20 mm Hg increase from the baseline level) and the absent respiratory movements6, 7. However, the duration of apnea test was not the same across countries. Short duration may lead to false negative, whereas long duration may induce hypotension, severe hypoxemia, and serious arrhythmia which cause secondary insults10. The duration of apnea test is five minutes in UK, eight to ten minutes in US and China, and unspecified in German and French criteria. Identifying the optimum duration of apnea test is a crucial issue that should be investigated in future studies. Moreover, apnea test with continuous positive airway pressure was proposed to be a safer method than the classic oxygen insufflation method11–13. More related evidence may lead to revisions in criteria of brain death determination in the future.
EEG is used to confirm the electrical inactivity of cortical layers in brain death determination. Very few studies investigated the validity of EEG in determination of brain death. A study14 in 1995 (n = 15) showed that EEG only had a sensitivity of 53%, and another study15 in 1987 (n = 56) reported a sensitivity of 80.6%. These two early studies compared the EEG with clinical criteria of brain death and used paper EEG recordings which might contain substantial artifacts. A recent German study16 (n = 67) and a Chinese study (n = 37)17 indicated that EEG had a high sensitivity of 94%-100% which was consistent with our findings.
Somatosensory evoked potentials (SEPs) examines lemniscal pathways in the hemispheres and brainstem, and was reported to be a reliable test for brain death determination with a sensitivity of 87.4%-100%18–20. The examination of cerebral circulation provides an alternative evaluation of brain function. TCD is recognized as one of the three confirmatory tests in Chinese criteria of brain death. However, TCD has a lower sensitivity (73%-78%) than EEG and SEPs19, 21. Therefore, TCD was used less often than the other two ancillary tests in our study and was confirmed to have a medium sensitivity (84.3%-86.0%). Besides, TCD is limited by the penetrability of transtemporal windows and interexaminer variability. MR and CT angiography were proposed to assist brain death determination22, 23. However, angiography requires higher cost and the transfer of patients out of intensive care unit which generates potential safety hazards.
Among patients who had negative results of confirmatory tests (3 in EEG, 2 in SLSEP, and 8 in TCD), few were not diagnosed as brain death. Therefore, the specificity and NPV of confirmatory tests were very low in our study. The positive results of confirmatory tests suggest a high likelihood of brain death, whereas the negative results do not necessarily mean not of brain death. Among 135 (67.8%) patients who met the clinical criteria of brain death in our study, there were three patients who did not reach the brain death criteria of confirmatory tests, and they were not diagnosed as brain death in China. These three patients had no recovery of consciousness until the withdrawal of medical support. So far there are no reports of neurologic function recovery in patients who met clinical diagnosis of brain death. The routine and repeat use of ancillary tests in brain death determination need to be further justified. However, the confirmation of ancillary tests may avoid some possible human errors in clinical evaluations of brain death. Moreover, the raters of brain death should be trained and certified by national medical societies in order to minimize potential variability in the practice.
There are some limitations in this study. Apnea tests were conducted for eight minutes in our patients according to Chinese criteria. Therefore, we are unable to investigate the positive rate and safety of different test durations. There is a possibility that the positive rate of apnea test was slightly overestimated and the completion rate was mildly underestimated in patients evaluated according to UK criteria. Moreover, as EEG, SLSEP, and TCD are the only ancillary tests recognized in China, other ancillary tests, such as angiography, bispectral index, and positron emission tomography, were not conducted and investigated in this study.