External Cooling in Severe Sepsis May Improve Early Mortality

Resident Focus - Volume 9, Issue 12

Read the Complete DynaMed Resident Focus

Reference: Am J Respir Crit Care Med 2012 May 15;185(10):1088 (level 2 [mid-level] evidence)

Sepsis is an extremely serious and common diagnosis responsible for between 650,000 -– 750,000 admissions per year (Emer Med Cardiac Research and Ed Group. 2006 September 2; 1-8, Crit Care Med. 2001 July;29:1303-1310). Septic shock has an exceptionally high mortality rate of 40-60% (Am J Resp Crit Care Med. 2003;168:165-172, Crit Care Med. 2006;34:344-353). Shock can develop rapidly and prompt recognition and intervention are essential to decrease mortality. These first crucial hours can occur in the emergency department, medical floor, or ICU; in any of these locations, a primary care physician may be involved in care and can have a positive impact on outcomes. Since the advent of early goal directed therapy (EGDT) recommendations have focused on the first few hours of management (see Surviving Sepsis Campaign [Crit Care Med. 2013 Feb;41(2):580-637]). Fever is a common symptom of sepsis, however little research has shown whether this is a beneficial physiologic response to be permitted or a pathologic response to be treated. When fever accompanies severe sepsis, is there benefit to normalizing body temperature with external cooling?

The “Sepsiscool” trial was a multicenter randomized, controlled trial with the primary objective to determine whether controlling fever with external cooling in septic shock can decrease vasopressor requirements by 50% after 48 hours. Secondary outcomes were also measured including shock reversal, length of stay in the ICU and in the hospital as well as mortality rates at two weeks, at ICU discharge and at hospital discharge. Over 19 months, 200 patients with severe septic shock requiring vasopressor infusion as well as fever greater than 38.3°C (100.9°F) were randomized to external cooling or no external cooling. The locations of the patients just prior to ICU admission was similar in both groups with nearly 45% of those cooled being on the medical or surgical floor. Median time of inclusion in the trial was 1 day after ICU admission. All patients were intubated, sedated, and severely ill based on standardized sepsis scoring with the cause of infections similar in both groups, most often pneumonia. External cooling measures were instituted for a period of 48 hours, keeping core body temperatures between 36.5°C and 37°C (97.7°F – 98.6°F). During this period, vasopressor infusions were weaned by nurses based on an algorithm while still maintaining adequate mean arterial pressures.

The number of patients with 50% vasopressor dose decrease was most evident and statistically significant at 12 and 24 hours in the cooling group (P < 0.01). However, the difference was no longer significant at 48 hours. Interestingly, fewer patients needed a vasopressor dose increase during these 48 hours in the cooling group (35% vs. 53%, p = 0.011). Overall shock reversal, defined as absence of need of vasopressor therapy for 24 consecutive hours, was more common in the cooling group (87% vs. 72%, p = 0.02). The overall mortality rate was lower in the cooling group at 14 days (19% vs. 34%, p = 0.013), but survival benefit did not reach statistical significance at the other measured time points including at ICU discharge and at hospital discharge.

This study showed that normalizing body temperature in severely septic patients using external cooling yields some benefits: lower required doses of vasopressor, higher rates of shock reversal, and lower early mortality in the first 14 days. Although later mortality was not statistically different at other timepoints, mortality was not the primary outcome being studied, and the study was not powered to detect these differences.

For more information, see Sepsis treatment in adults in DynaMed.