Intensive Insulin Therapy and Colloid Resuscitation in Severe Sepsis
Abstract & Commentary
By Richard J. Wall, MD, MPH, Pulmonary, Critical Care, & Sleep Disorders Medicine, Southlake Clinic, Valley Medical Center, Renton, WA, is Associate Editor for Critical Care Alert.
Dr. Wall reports no financial relationship to this field of study.
Synopsis: This study of severe sepsis patients showed that intensive insulin therapy does not improve mortality or organ dysfunction, and resuscitating patients with the colloid hydroxyethyl starch is harmful.
Source: Brunkhorst FM, et al. N Engl J Med. 2008; 358(2): 125-139.
This multicenter randomized trial—nicknamed the VISEP study—used a two-by-two factorial design to test two independent treatments in severe sepsis patients. The study included 537 adult patients from 18 academic tertiary hospitals in Germany. Patients had severe sepsis or septic shock (defined using ACCP-SCCM consensus criteria). The first study compared intensive insulin therapy with conventional insulin therapy. The second study compared resuscitation with the colloid hydroxyethyl starch (HES) against resuscitation with traditional lactated Ringers solution (LR). The primary outcomes for both studies were 28-day mortality and morbidity rates. The latter was measured using the Sequential Organ Failure Assessment scale.
After the first safety analysis, the insulin study was prematurely stopped by the data and safety monitoring board because there was a higher number of hypoglycemic events in the intensive insulin arm, as compared with the conventional therapy group. The study comparing HES and LR was continued, but all subsequent patients received only conventional insulin therapy. However, at the planned interim analysis, the fluid resuscitation study was also aborted because the colloid HES group had a significantly higher incidence of renal failure and a trend toward higher mortality.
In the insulin trial, a continuous insulin infusion maintained blood sugars between either 80-110 mg/dL (intensive group) or 180-200 mg/dL (conventional group). Overall, there was no difference between groups for mortality or morbidity. However, the intensive group had significantly more severe hypoglycemic events, defined as blood sugar ≤ 40 mg/dL (17% vs 4.1%; p <0.001). In the intensive group, the hypoglycemic events were more likely to be life-threatening (5.3% vs 2.1%; p = 0.05) and to prolong the hospitalization (2.4% vs 0.3%; p = 0.05).
In the fluid trial, a mandatory protocol achieved a central venous pressure (CVP) of 8 mm Hg during the first 96 hours. If the mean arterial pressure remained <70 mm Hg or central venous oxygen saturation (ScVO2) remained <70%, the treating physician could implement further measures to achieve these goals (fluids, vasopressors, and/or inotropes). HES could not be used as a maintenance fluid. Overall, there was no difference between groups for mortality or morbidity. As expected, the LR group ended up receiving more fluid volume than the HES group.
The HES patients had a higher median CVP (11.8 vs 10.7 mmHg), a higher median ScVO2 (73.6% vs 72.4%), and achieved targeted hemodynamic values faster than the LR group. Despite these apparent advantages, the incidence of acute renal failure (defined as doubling of serum creatinine level or the need for renal-replacement therapy) was much higher in the HES group (34.9% vs 22.8%; p = 0.002). There was a direct correlation between cumulative HES dose and need for renal-replacement therapy. Of note, the HES group received more units of packed red cells than the LR group (p <0.001).
Commentary
This paper essentially tells the story of two different randomized controlled trials. Because there is no reason to expect an interaction between the two tested strategies, let's discuss the results separately.
Insulin Study
In 2001, Van Den Berghe, et al. published a landmark study showing that intensive insulin therapy lowered in-hospital mortality from 10.9% to 7.2% in critically ill (mostly cardiac) surgical patients.1 A follow-up study in medical ICU patients, however, showed no beneficial effect.2 Furthermore, use of intensive insulin therapy in the latter study was associated with a 5-fold increase in hypoglycemic events. The current study in severe sepsis patients supports the second study's findings-intensive insulin is not always beneficial in medical ICU patients, and it may actually be dangerous because it increases risk of severe hypoglycemia.
As the pendulum swings yet again, what should ICU clinicians do? First, do no harm. When implementing a protocol in the ICU, it is essential that you also design a system to monitor for unintended consequences. If Brunkhorst et al had only measured their insulin protocol's success, they would never have detected the excessive rate of hypoglycemic episodes. Second, realize that "critically ill" does not describe a patient population. The ICU is merely a physical location where patients receive care for a cardiac, neurological, surgical, or medical condition. Being admitted to the ICU does not mean a patient will necessarily benefit from a protocol that was validated in another ICU population.
In other words, don't just read the "Discussion" section in a critical care article. Scrutinize the study population and ask yourself: "Do I care for similar patients? Are these results applicable to my hospital?" For example, despite "negative" results in the second study, Van Den Berghe later showed that intensive insulin therapy reduces critical illness polyneuropathy (51% vs 39%; p = 0.02) and need for prolonged mechanical ventilation (46.7% vs 34.6%; p = 0.01).3
Fluid Study
There are few data regarding whether colloids or crystalloids are better for resuscitating patients with severe sepsis. Past studies suggested that HES adversely affected renal function in critically ill patients. The current study found similar results in a population of severe sepsis patients. Unless a future study shows that HES is beneficial in sepsis, don't use it in this population. However, this does not mean that every colloid is harmful in sepsis patients. Nor does it mean that HES is harmful in every ICU patient. It simply means that the tested strategy (HES) was more dangerous than the control therapy (LR) in this population (severe sepsis patients). Given the large number of other fluids available for resuscitating sepsis patients, however, I recommend that clinicians use a fluid that has not been shown to be harmful. And stay tuned-this will certainly be a hot research topic in years to come!
References
- Van den Berghe G, et al. N Engl J Med. 2001;345:1359-1367.
- Van den Berghe G, et al. N Engl J Med. 2006;354:449-461.
- Hermans G, et al. Am J Respir Crit Care Med. 2007;175(5):480-489.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content