Rossetti, AO; Oddo, M; Logroscino, G; et al.
ANNALS OF NEUROLOGY Volume: 67 Issue: 3 Pages: 301-307 Published: March 2010
Objective: Current American Academy of Neurology (AAN) guidelines for outcome prediction in comatose survivors of cardiac arrest (CA) have been validated before the therapeutic hypothermia era (TH). We undertook this study to verify the prognostic value of clinical and electrophysiological variables in the TH setting.
Methods: A total of 111 consecutive comatose survivors of CA treated with TH were prospectively studied over a 3-year period. Neurological examination, electroencephalography (EEG), and somatosensory evoked potentials (SSEP) were performed immediately after TH, at normothermia and off sedation. Neurological recovery was assessed at 3 to 6 months, using Cerebral Performance Categories (CPC).
Results: Three clinical variables, assessed within 72 hours after CA, showed higher false-positive mortality predictions as compared with the AAN guidelines: incomplete brainstem reflexes recovery (4% vs 0%), myoclonus (7% vs 0%), and absent motor response to pain (24% vs 0%). Furthermore, unreactive EEG background was incompatible with good long-term neurological recovery (CPC 1-2) and strongly associated with in-hospital mortality (adjusted odds ratio for death, 15.4; 95% confidence interval, 3.3-71.9). The presence of at least 2 independent predictors out of 4 (incomplete brainstem reflexes, myoclonus, unreactive EEG, and absent cortical SSEP) accurately predicted poor long-term neurological recovery (positive predictive value = 1.00); EEG reactivity significantly improved the prognostication.
Interpretation: Our data show that TH may modify outcome prediction after CA, implying that some clinical features should be interpreted with more caution in this setting as compared with the AAN guidelines. EEG background reactivity is useful in determining the prognosis after CA treated with TH.
Comment by Gian Luigi Lenzi
I would like to utilise this interesting paper for a more general comment: how many times does our working hypothesis shape the final results? When I was first introduced to experimental neurophysiology my very illustrious Professor taught me that the results of the first experiment were to write the paper and the other experiments were to confirm the first one. This is evidently a paradox, but how many times does our strong belief force the results towards a particular direction? It is an old story that only supports the Galilean needs of external confirmation of every kind of experimental data. Unfortunately in a clinical environment sometimes patients die.