Correspondence (letter to the editor): Prognostic Assessment Always Requires Several Parameters

An important point of criticism regarding the cited studies is the fact that the prognostic parameters under investigation are further used in decisions about stopping therapy. This entails the possibility of a self-fulfilling prophecy. In the study by Bouwes et al., for example, therapy was immediately ended or limited in 40 out of 42 patients with lacking somatosensory evoked potentials (SEP) (1).

We have repeatedly cared for patients who woke up again in spite of concentrations of neuron-specific enolase (NSE) >97 µg/L. The biggest impression was left by the case of a patient with a neuroendocrine tumor, whose NSE rose to >1000 µg/L. In assessing the NSE, malignant tumors should be regarded as confounders. In our opinion, the key message “Elevated serum concentrations of neuron-specific enolase have to be above 97 µg/L to serve as a safe indicator of an unfavorable prognosis” is therefore problematic in this abbreviated form. We think that the factually equating the bilateral absence of the pupillary light response or the corneal reflex with reliable indicators of a poor prognosis is equally problematic, because of possible examiner bias. Especially absent corneal reflexes should be assessed with a great deal of caution. Bouwes and Samaniego described 2 of 23, and 2 of 22, patients whose outcome was good in spite of absent corneal reflexes (1, 3).

A high degree of certainty in the prognostic assessment is possible only by basing it on several parameters, rather than a single one. We treat most patients for seven days before we limit their treatment according to an interdisciplinary prognostic algorithm, on the basis of several consistent parameters for a poor prognosis (2).