The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Should We Use Intensive Insulin Therapy to Prevent Critical Illness Polyneuropathy?
Abstract & Commentary
By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle, is Associate Editor for Critical Care Alert.
Synopsis: A prospectively planned sub-analysis of data from a randomized trial of intensive insulin therapy in the medical ICU reveals that in patients in the ICU for greater than 7 days, intensive insulin therapy decreases both the incidence of critical illness polyneuropathy/myopathy and the frequency of prolonged mechanical ventilation.
Source: Hermans G, et al. Am J Respir Crit Care Med. 2006;175:480-489.
Van den Berghe et al have recently reported the results of intensive insulin therapy (IIT) on morbidity and mortality in medical intensive care unit patients.1 In the study discussed below, Hermans and colleagues performed a prospectively planned sub-analysis on data from that trial to determine whether IIT had any effect on the incidence of critical illness polyneuropathy/myopathy (CIP/CIM) and the frequency of treatment with prolonged mechanical ventilation in medical patients who were in the ICU for 7 or more days.
A total of 443 patients met criteria inclusion in this sub-analysis. Out of the 420 patients for whom complete data was available, 212 had been randomly assigned to receive conventional insulin therapy (CIT) while the remaining 208 patients were randomized to the IIT arm. IIT was aimed at blood glucose levels between 80 and 110 mg/dL while the CIT protocol mandated that insulin be given when blood glucose levels were above 215 mg/dL and that insulin doses be decreased or held if the blood glucose fell below 180 mg/dL. The diagnosis of CIP/CIM was made using electroneuromyography (ENMG) while prolonged mechanical ventilation was defined as greater than or equal to 14 days on the ventilator.
Mean glucose levels in the IIT group were 102 mg/dL vs 159 mg/dL in the CIT group. Within the IIT group, 81 out of the 208 patients (38.9%) met ENMG criteria for CIP/CIM compared to 107 out of 212 patients in the CIT group (50.5%). The authors also found that IIT reduced the incidence of treatment with prolonged mechanical ventilation from 99 of 212 patients in the CIT group (46.7%) to 72 of 208 patients (34.6%) in the IIT group. Furthermore, IIT increased the cumulative chance of successful weaning from mechanical ventilation and decreased the number of days on the mechanical ventilation from a median of 14 (interquartile range 9-22) in the CIT group to 12 days (inter-quartile range, 8-20 days) in the IIT group.
In their analysis of risk factors for the development of CIP/CIM, the authors report that while neuromuscular blocking agents increased the risk of CIP/CIM, corticosteroids actually had a protective effect with regard to this outcome, as did increasing patient age. When risk factors for prolonged mechanical ventilation were assessed, however, the number of days of treatment with corticosteroids was found to increase the duration of treatment with mechanical ventilation.
Debate continues as to whether intensive insulin therapy should be standard practice in the ICU. Herman and colleagues make a useful contribution to the growing literature on this issue by providing data on important endpoints—the incidence of critical illness polyneuropathy and duration of mechanical ventilation—which affect the clinical outcomes for critically ill patients. Their additional finding that corticosteroids may exert a protective effect against CIP/CIM is particularly intriguing and runs counter to much previous work which has demonstrated that corticosteroids exerted a deleterious effect in this regard. Unfortunately, their results do not provide any insight into the potential mechanism for this finding beyond the fact that better control of blood glucose levels in the setting of steroid use may be the key factor.
While the article provides useful data in this important debate, it is not clear that these results are enough to convince those on the fence about the usefulness of IIT to switch their practice and incorporate it into ICU protocols. This reservation derives from several methodological issues with the study. First, although the two groups were well-matched with respect to most variables, maximum sequential organ failure assessment (SOFA) scores were lower in the IIT group than in the CIT group. To the extent that severity of illness affects the likelihood of developing CIP/CIM or prolonged mechanical ventilation, this mismatch is potentially of great significance. Second, CIP/CIM was diagnosed solely on the basis of ENMG and it is not clear that this measure correlates well with other markers of patient strength or neurologic function.
A more important concern, however, is the fact that the authors only included patients who had been in the intensive care unit for seven or more days. It is not clear why patients with shorter ICU duration were not included and it is worthwhile to question whether the inclusion of these patients' stay may have, in fact, altered the observed results of the study. This omission is also of importance for clinical practice. If IIT is only of benefit in patients treated in the ICU for greater than seven days, in the absence of useful prognostic factors to identify a priori which patients will fall into that category, clinicians are left wondering in whom and when they should start IIT following ICU admission.
A final concern in considering whether to implement this therapy is the fact that the only positive trials for IIT1,2 have come from this one research group conducting their studies out of a single center; at the same time, several large studies from other centers have been stopped early due to a lack of efficacy or concerns about the safety of IIT. Until positive results emerge from other multi-center trials, it is reasonable to resist the urge of Hermans and colleague's intriguing data and hold off on implementing such protocols until the results of further studies become available.