Subscribe to eTOC

Prognostic Value of SSEP in Hypoxic Coma Limited, New Analysis Finds

ARTICLE IN BRIEF

Investigators reported that the currently accepted superiority of somatosensory evoked potentials testing over other tests barely holds up to the collective evidence, and only in the first day or so after an anoxic-ischemic event.

Despite widespread use of electrophysiological markers in predicting neurological outcomes in hypoxic coma patients, their accuracy is only marginally better than clinical signs after the first day, according to a review of the medical literature in the Feb. 16 Neurology.

Researchers in Australia reviewed 26 studies published in three languages over a 40-year period, ending in 2006. They discovered that the currently accepted superiority of somatosensory evoked potentials (SSEP) testing over other tests barely holds up to the collective evidence, and only in the first day or so after an anoxic-ischemic event.

Led by Thanh G. Phan, MD, professor of neurology at Monash Medical Center in Victoria, Australia, researchers evaluated outcome accuracy of motor and visual SSEP, electroencephalogram (EEG), electromyogram (EMG), M1 (absence of motor response), as well as decorticate (M3) and decerebrate (M2) posturing.

“We provided a formal statistical evaluation of prognostic tests for predicting poor outcome after hypoxic coma, and separated the analysis into two separate time points for comparison,” Dr. Phan told Neurology Today in an e-mail.

“What we found was that the diagnostic accuracy of SSEP is only marginally better than either M1 or pupillary response, especially in the first 24 hours after coma onset,” he said. “The advantage of SSEP over motor signs is diminished when M3 (decorticate posturing) is used as a marker and, at days two and three post-coma onset, there is no statistical difference between the diagnostic accuracy of SSEP and M1.”

Figure

DR. THANH G. PHAN: “What we found was that the diagnostic accuracy of SSEP is only marginally better than either M1 or pupillary response, especially in the first 24 hours after coma onset.”

How these findings translate into clinical practice warrants further discussion, Dr. Phan said. The superiority of SSEP over M1 within 24 hours post-coma onset is small. The difference is statistically significant but this may not necessarily imply clinical significance, but the use of M3 rather than M1 improves the diagnostic performance of motor signs, but diminishes the advantage of SSEP on day one.

“The use of SSEP in clinical practice should take this into consideration,” Dr. Phan said. “This is especially relevant where SSEP is not widely available such as in a rural setting or in an underdeveloped nation. In those contexts, a careful clinical examination provides a reasonable compromise.”

OTHER EVALUATION TOOLS

The most widely used prognostic method is based on eye and motor responses and was developed by David E. Levy, MD, and colleagues, and published in 1985 in the Journal of the American Medical Association.

However, because it was based on a single group of patients, it resulted in “substantial statistical uncertainty,” according to Dr. Phan, and a number of other electrophysiological tests have since been introduced that needed evaluation within the context of the overall evidence in the literature.

In 2006, the AAN Quality Standards Subcommittee conducted a comprehensive literature review to address many of the same questions. In a practice parameter published that year in Neurology, the panel concluded that a poor prognosis can be made in patients lacking any pupillary or corneal reflexes, and in those without extensor motor responses after three days. Bilateral absence of cortical SSE within one to three days also predicts poor outcome, according to the review, a slightly wider window than that found by the Australian team.

G. Bryan Young, MD, professor of neurology and critical care medicine at the University of Western Ontario in London, Ontario, was one of the authors of the 2006 AAN guides. In an e-mail to Neurology Today, he commented that the study would have been better if it included only cases of coma after cardiac arrest. “To pick other etiologies creates inhomogeneity in an otherwise well-designed study. When varied mechanisms are operative, I am not sure it is valid to lump patients together,” he noted.

Dr. Young also pointed out that many patients are now treated with hypothermia after coma in order to reduce the rate of injury to the brain, and that it might have been helpful to have included hypothermia in the analysis.

Eelco F. M. Wijdicks, MD, PhD, professor of neurology in the Division of Critical Care Neurology at Mayo Clinic College of Medicine, Rochester, MN, also served on the 2006 AAN review panel. “I would hope that clinicians would not base their prognostication on the bases of electrophysiology alone, and if the authors are correct, certainly not on a the basis of a single EEG,” he said in an e-mail.

Figure

DR. EELCO F. M. WIJDICKS: “I would hope that clinicians would not base their prognostication on the bases of electrophysiology alone, and if the authors are correct, certainly not on a the basis of a single EEG.”

References

Lee YC, Phan TG, Jolley DJ, et al. Accuracy of clinical signs, SEP and EEG in predicting outcome of hypoxic coma: a meta-analysis. Neurology2010;74:572–580.
Wijdicks EF, Hijdra A, Wiebe S, et al. Practice parameter: Prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review): Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology2006;67:203–210.
Levy DE, Caronna JJ, Plum F, et al. Predicting outcome from hypoxic-ischaemic coma. JAMA1985;253:1420–1426.
Guerit JM, de Tourtchaninoff M, Mahieu P, et al. The prognostic value of three modality evoked potentials in anoxic and traumatic comas. Neurophysiol Clin 1993;23:209–226.