Insulin Therapy in the Medical ICU
http://www.100md.com
《新英格兰医药杂志》
To the Editor: According to the conclusions of Van den Berghe and colleagues (Feb. 2 issue),1 the mortality among patients staying three days or longer in a medical intensive care unit (ICU) was significantly decreased with the use intensive insulin treatment (200 of 381 patients in the group receiving conventional treatment vs. 166 of 386 patients receiving intensive treatment, P=0.009). However, in the subgroup of patients staying in the ICU less than three days, the mortality was higher in the intensive-treatment group than in the conventional-treatment group (56 of 209 patients vs. 42 of 224 ). A German multicenter study of intensive insulin therapy in patients with severe sepsis was recently suspended by the safety monitoring board because of a significant excess risk of severe hypoglycemia without any evidence of improved survival.2 Hypoglycemia is associated with an increased risk of death, and this increased risk may have been related to undiagnosed episodes of hypoglycemia at the earlier phase of the stay in the ICU in patients who were unconscious or comatose. Consequently, it might be important to develop a system for continuous glycemic monitoring before considering intensive insulin therapy for patients in a medical ICU.
Laure Hammer, M.D.
Géraldine Dessertaine, M.D.
Jean-Fran?ois Timsit, M.D., Ph.D.
University Hospital
38043 Grenoble, France
References
Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-461.
ClinicalTrials.gov. Efficacy of volume substitution and insulin therapy in severe sepsis (VISEP trial). (Accessed April 21, 2006, at http://clinicaltrials.gov/show/NCT00135473.)
To the Editor: The main finding of Van den Berghe et al., that intensive insulin therapy does not improve mortality in the medical ICU, is in contrast to the large mortality benefit observed in the surgical ICU in a previous study conducted by the same group.1 This negative result gives credibility to the recommendation that trials stopped early because of benefit should be viewed with skepticism, since multiple interim analyses and questionable early-stopping rules may lead to an exaggeration of positive results.2 In this trial in the medical ICU, improvements in morbidity (secondary end points) in the entire study population were marginally significant, and the post hoc analysis has limited usefulness for clinicians, since patients requiring stays of three days or longer in the medical ICU cannot be accurately identified prospectively. Moreover, the lack of blinding and the failure to adjust for multiple comparisons necessitate a cautious interpretation of the results. Therefore, because the previous trial in the surgical ICU1 was stopped early, the potential benefits of intensive insulin therapy were likely to have been inflated, thereby causing a premature rush to adopt the therapy2,3; the current trial runs a similar risk by overemphasizing post hoc results and secondary end points.
Scott K. Aberegg, M.D., M.P.H.
Ohio State University College of Medicine and Public Health
Columbus, OH 43214
scottaberegg@hotmail.com
References
Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-1367.
Montori VM, Devereaux PJ, Adhikari NKJ, et al. Randomized trials stopped early for benefit: a systematic review. JAMA 2005;294:2203-2209.
Angus DC, Abraham E. Intensive insulin therapy in critical illness. Am J Respir Crit Care Med 2005;172:1358-1359.
To the Editor: The benefits of strict normalization of blood glucose levels reported by Van den Berghe et al. in patients in a medical ICU who have stress hyperglycemia are in accordance with observations in more homogeneous populations of patients in the surgical ICU. The declaration in the accompanying editorial by Malhotra1 that the present study was "negative" obscures the significant reduction in morbidity overall and the decreased mortality in the target population of patients staying in the ICU for three days or longer. The novel arbitrary recommendation1 of aiming for a blood glucose level of 150 mg per deciliter during the first three days in the medical ICU has no foundation in published data and will reduce the benefits of glycemic control shown in this and other studies — notably, the demonstrated overall benefit in terms of morbidity of early intervention with intensive insulin therapy. A recent study provides guidance for the prevention, detection, and treatment of hypoglycemia in the ICU, a condition that is quite manageable in nearly all cases.2 Pending further studies, tight control of blood glucose levels is clearly indicated for all patients in the ICU.
Ronald Tamler, M.D., Ph.D.
Derek LeRoith, M.D., Ph.D.
Mount Sinai School of Medicine
New York, NY 10029
ronald.tamler@mssm.edu
Jesse Roth, M.D.
Albert Einstein College of Medicine
New York, NY 11357
References
Malhotra A. Intensive insulin in intensive care. N Engl J Med 2006;354:516-518.
Vriesendorp TM, van Santen S, DeVries JH, et al. Predisposing factors for hypoglycemia in the intensive care unit. Crit Care Med 2006;34:96-101.
The authors reply: We share the evidence-based viewpoint of Dr. Tamler and colleagues, rather than that of the editorialist, Dr. Malhotra, who suggested delaying intensive insulin therapy until day 3 of intensive care. Our data indicate significant benefits in terms of mortality and morbidity in the target population. Our previous study in a surgical ICU suggested that three days or more of intensive insulin therapy, started on admission to the ICU, are needed to reduce in-hospital mortality. Thus, we believe it was scientifically and ethically correct to power this study, conducted in the medical ICU, statistically for a mortality effect in patients treated for at least three days. As hypothesized, a significant 10 percent absolute reduction in mortality was observed in this target population, along with a 3 percent lower mortality in the intention-to-treat group — numbers remarkably similar to those in our study conducted in the surgical ICU. Further analysis showed that mortality in the intention-to-treat group was, in fact, significantly reduced (odds ratio for death, 0.77; 95 percent confidence interval, 0.60 to 0.99; P=0.04) when corrected for preexisting cancer, end-stage organ failure, and severity of illness on admission to the medical ICU.
Morbidity was significantly reduced in all patients in the study. The benefits of accelerated weaning from mechanical ventilation and shorter stays in the ICU and hospital are clear. Furthermore, these morbidity end points were the goal of earlier important studies of ICU outcomes.1
Although numerically more deaths occurred among medical patients, as compared with surgical patients, receiving intensive insulin therapy for less than three days, the difference was not statistically significant and appeared to be attributable to selection bias. In the group receiving this therapy for less than three days, death was unrelated to hypoglycemia. Among patients who stayed in the ICU for three days or more, kidney failure and liver failure increased both the risk of death and the incidence of hypoglycemia. Thus, hypoglycemia, occurring more often in such patients when treated with insulin, stands out as a statistical risk factor for death.
The German sepsis study (involving 488 patients, with both short and long stays in the ICU), to which Dr. Hammer and colleagues refer, was stopped early, because hypoglycemia developed in 12 percent of patients. That study did not provide information regarding efficacy, because it was underpowered for a mortality effect; morbidity also was not reported. Furthermore, since our studies in the medical and the surgical ICU have shown that survival curves separated only after 30 days or more, studies aiming for an early mortality benefit would predictably fail.
Sustained blood-glucose control in patients with diabetes2 or critical illness prevents the cellular damage inflicted by hyperglycemia, a benefit that outweighs the risk of hypoglycemia. Continuous glucose-monitoring systems are awaited, which could further reduce episodes of hypoglycemia and hyperglycemia. Current data support careful maintenance of normoglycemia in all patients in the ICU, from admission on.
Greet Van den Berghe, M.D., Ph.D.
Alexander Wilmer, M.D., Ph.D.
Roger Bouillon, M.D., Ph.D.
Catholic University of Leuven
B-3000 Leuven, Belgium
greta.vandenberghe@med.kuleuven.be
References
Takala J, Ruokonen E, Webster NR, et al. Increased mortality associated with growth hormone treatment in critically ill adults. N Engl J Med 1999;341:785-792.
Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-986.
The Editorialist replies: Tamler et al. suggest that we should overlook the excess mortality potentially associated with short-term intensive insulin therapy in the study by Van den Berghe et al. The excess early deaths in the intensive-treatment group could have a substantial effect on both the subgroup analyses of patients with long stays in the medical ICU and on morbidity. The assignment to subgroups after randomization is problematic, since patients were considered to have a long stay in the ICU on the basis not just of randomization but also of survival during the first three days in the ICU. Thus, the results for those with long stays may reflect true benefits of intensive insulin therapy or survivor effects before these patients were included in the subgroup (i.e., the sickest patients died before being included). The morbidity data in the intention-to-treat analysis are exciting but difficult to interpret because of the differential time to death — early deaths in the intensive insulin group could effectively protect against the development of renal failure and prolonged use of mechanical ventilation. My opinion regarding target therapeutic values of glucose is based on published international consensus,1 since the data available from randomized trials are equivocal (one positive phase 2 study, one negative phase 2 study, and the reportedly negative multicenter Efficacy of Volume Substitution and Insulin Therapy in Severe Sepsis study2).
Atul Malhotra, M.D.
Harvard Medical School
Boston, MA 02115
References
Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 2004;32:858-873.
Brunkhorst FM, Kuhnt E, Engel C, et al. Intensive insulin therapy in patients with severe sepsis and septic shock is associated with an increased rate of hypoglycemia: results from a randomized multicenter study (VISEP). Infection 2005;33:19-20.
Laure Hammer, M.D.
Géraldine Dessertaine, M.D.
Jean-Fran?ois Timsit, M.D., Ph.D.
University Hospital
38043 Grenoble, France
References
Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449-461.
ClinicalTrials.gov. Efficacy of volume substitution and insulin therapy in severe sepsis (VISEP trial). (Accessed April 21, 2006, at http://clinicaltrials.gov/show/NCT00135473.)
To the Editor: The main finding of Van den Berghe et al., that intensive insulin therapy does not improve mortality in the medical ICU, is in contrast to the large mortality benefit observed in the surgical ICU in a previous study conducted by the same group.1 This negative result gives credibility to the recommendation that trials stopped early because of benefit should be viewed with skepticism, since multiple interim analyses and questionable early-stopping rules may lead to an exaggeration of positive results.2 In this trial in the medical ICU, improvements in morbidity (secondary end points) in the entire study population were marginally significant, and the post hoc analysis has limited usefulness for clinicians, since patients requiring stays of three days or longer in the medical ICU cannot be accurately identified prospectively. Moreover, the lack of blinding and the failure to adjust for multiple comparisons necessitate a cautious interpretation of the results. Therefore, because the previous trial in the surgical ICU1 was stopped early, the potential benefits of intensive insulin therapy were likely to have been inflated, thereby causing a premature rush to adopt the therapy2,3; the current trial runs a similar risk by overemphasizing post hoc results and secondary end points.
Scott K. Aberegg, M.D., M.P.H.
Ohio State University College of Medicine and Public Health
Columbus, OH 43214
scottaberegg@hotmail.com
References
Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359-1367.
Montori VM, Devereaux PJ, Adhikari NKJ, et al. Randomized trials stopped early for benefit: a systematic review. JAMA 2005;294:2203-2209.
Angus DC, Abraham E. Intensive insulin therapy in critical illness. Am J Respir Crit Care Med 2005;172:1358-1359.
To the Editor: The benefits of strict normalization of blood glucose levels reported by Van den Berghe et al. in patients in a medical ICU who have stress hyperglycemia are in accordance with observations in more homogeneous populations of patients in the surgical ICU. The declaration in the accompanying editorial by Malhotra1 that the present study was "negative" obscures the significant reduction in morbidity overall and the decreased mortality in the target population of patients staying in the ICU for three days or longer. The novel arbitrary recommendation1 of aiming for a blood glucose level of 150 mg per deciliter during the first three days in the medical ICU has no foundation in published data and will reduce the benefits of glycemic control shown in this and other studies — notably, the demonstrated overall benefit in terms of morbidity of early intervention with intensive insulin therapy. A recent study provides guidance for the prevention, detection, and treatment of hypoglycemia in the ICU, a condition that is quite manageable in nearly all cases.2 Pending further studies, tight control of blood glucose levels is clearly indicated for all patients in the ICU.
Ronald Tamler, M.D., Ph.D.
Derek LeRoith, M.D., Ph.D.
Mount Sinai School of Medicine
New York, NY 10029
ronald.tamler@mssm.edu
Jesse Roth, M.D.
Albert Einstein College of Medicine
New York, NY 11357
References
Malhotra A. Intensive insulin in intensive care. N Engl J Med 2006;354:516-518.
Vriesendorp TM, van Santen S, DeVries JH, et al. Predisposing factors for hypoglycemia in the intensive care unit. Crit Care Med 2006;34:96-101.
The authors reply: We share the evidence-based viewpoint of Dr. Tamler and colleagues, rather than that of the editorialist, Dr. Malhotra, who suggested delaying intensive insulin therapy until day 3 of intensive care. Our data indicate significant benefits in terms of mortality and morbidity in the target population. Our previous study in a surgical ICU suggested that three days or more of intensive insulin therapy, started on admission to the ICU, are needed to reduce in-hospital mortality. Thus, we believe it was scientifically and ethically correct to power this study, conducted in the medical ICU, statistically for a mortality effect in patients treated for at least three days. As hypothesized, a significant 10 percent absolute reduction in mortality was observed in this target population, along with a 3 percent lower mortality in the intention-to-treat group — numbers remarkably similar to those in our study conducted in the surgical ICU. Further analysis showed that mortality in the intention-to-treat group was, in fact, significantly reduced (odds ratio for death, 0.77; 95 percent confidence interval, 0.60 to 0.99; P=0.04) when corrected for preexisting cancer, end-stage organ failure, and severity of illness on admission to the medical ICU.
Morbidity was significantly reduced in all patients in the study. The benefits of accelerated weaning from mechanical ventilation and shorter stays in the ICU and hospital are clear. Furthermore, these morbidity end points were the goal of earlier important studies of ICU outcomes.1
Although numerically more deaths occurred among medical patients, as compared with surgical patients, receiving intensive insulin therapy for less than three days, the difference was not statistically significant and appeared to be attributable to selection bias. In the group receiving this therapy for less than three days, death was unrelated to hypoglycemia. Among patients who stayed in the ICU for three days or more, kidney failure and liver failure increased both the risk of death and the incidence of hypoglycemia. Thus, hypoglycemia, occurring more often in such patients when treated with insulin, stands out as a statistical risk factor for death.
The German sepsis study (involving 488 patients, with both short and long stays in the ICU), to which Dr. Hammer and colleagues refer, was stopped early, because hypoglycemia developed in 12 percent of patients. That study did not provide information regarding efficacy, because it was underpowered for a mortality effect; morbidity also was not reported. Furthermore, since our studies in the medical and the surgical ICU have shown that survival curves separated only after 30 days or more, studies aiming for an early mortality benefit would predictably fail.
Sustained blood-glucose control in patients with diabetes2 or critical illness prevents the cellular damage inflicted by hyperglycemia, a benefit that outweighs the risk of hypoglycemia. Continuous glucose-monitoring systems are awaited, which could further reduce episodes of hypoglycemia and hyperglycemia. Current data support careful maintenance of normoglycemia in all patients in the ICU, from admission on.
Greet Van den Berghe, M.D., Ph.D.
Alexander Wilmer, M.D., Ph.D.
Roger Bouillon, M.D., Ph.D.
Catholic University of Leuven
B-3000 Leuven, Belgium
greta.vandenberghe@med.kuleuven.be
References
Takala J, Ruokonen E, Webster NR, et al. Increased mortality associated with growth hormone treatment in critically ill adults. N Engl J Med 1999;341:785-792.
Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-986.
The Editorialist replies: Tamler et al. suggest that we should overlook the excess mortality potentially associated with short-term intensive insulin therapy in the study by Van den Berghe et al. The excess early deaths in the intensive-treatment group could have a substantial effect on both the subgroup analyses of patients with long stays in the medical ICU and on morbidity. The assignment to subgroups after randomization is problematic, since patients were considered to have a long stay in the ICU on the basis not just of randomization but also of survival during the first three days in the ICU. Thus, the results for those with long stays may reflect true benefits of intensive insulin therapy or survivor effects before these patients were included in the subgroup (i.e., the sickest patients died before being included). The morbidity data in the intention-to-treat analysis are exciting but difficult to interpret because of the differential time to death — early deaths in the intensive insulin group could effectively protect against the development of renal failure and prolonged use of mechanical ventilation. My opinion regarding target therapeutic values of glucose is based on published international consensus,1 since the data available from randomized trials are equivocal (one positive phase 2 study, one negative phase 2 study, and the reportedly negative multicenter Efficacy of Volume Substitution and Insulin Therapy in Severe Sepsis study2).
Atul Malhotra, M.D.
Harvard Medical School
Boston, MA 02115
References
Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 2004;32:858-873.
Brunkhorst FM, Kuhnt E, Engel C, et al. Intensive insulin therapy in patients with severe sepsis and septic shock is associated with an increased rate of hypoglycemia: results from a randomized multicenter study (VISEP). Infection 2005;33:19-20.