Long-term outcome of patients hospitalized in intensive care units with central or extrapontine myelinolysis.
Crit Care Med. 2011 Oct 27;
Authors: Louis G, Megarbane B, Lavoué S, Lassalle V, Argaud L, Poussel JF, Georges H, Bollaert PE
OBJECTIVE:: Very few data are available for critically ill patients with central or extrapontine myelinolysis and according to available evidence, the prognosis seems to be poor. We aimed to describe the baseline characteristics, the management, the long-term prognosis, and the prognostic factors in central or extrapontine myelinolysis. DESIGN:: Retrospective observational study considering modified Rankin Scale score >3 or death as an unfavorable outcome. SETTING:: Forty-six French intensive care units. PATIENTS:: Thirty-six patients with central or extrapontine myelinolysis treated in 2000-2010. INTERVENTIONS:: None. MEASUREMENTS AND MAIN RESULTS:: At baseline, 31 (86%) patients were alcoholics and 33 (92%) presented with hyponatremia. Mechanical ventilation was required in 32 (89%) patients. At 1-yr follow-up, 11 (31%) patients have died, whereas 14 (56%) survivors have returned to a Rankin score ?1. Life-supporting therapies were withheld in 11 (31%) patients. Severe cerebral motor disability was the most frequently cited reason. However, five of them were still alive at 1 yr with Rankin score ?1 for four of them. We found no statistical difference between the 18 (50%) patients with a favorable outcome and the 18 (50%) patients with an unfavorable outcome with regard to severity of illness, suggesting that recovery is possible and unpredictable on the basis of clinical presentation. Chronic alcoholism was less frequent in patients with a favorable outcome as compared with patient with an unfavorable outcome (13 [72%] vs. 18 [100%], p = .04). CONCLUSIONS:: The prognosis of critically ill patients with central or extrapontine myelinolysis is better than thus far thought despite initial severe clinical manifestations. Regarding the high rate of decisions to withhold life-supporting therapies, the probability of a favorable outcome might be underestimated by intensivists.
PMID: 22036854 [PubMed - as supplied by publisher]