November 2015 - Surgical Critical Care

 

November 2015
EAST Monthly Literature Review


"Keeping You Up-to-Date with Current Literature"
Brought to you by the EAST Manuscript and Literature Review Committee

This issue was prepared by EAST Manuscript and Literature Review Committee Member Jason Smith, MD, PhD. MBA, FACS.

In This Issue: Surgical Critical Care

Scroll down to see summaries of these articles

Article 1 reviewed by Jason Smith, MD, PhD, MBA, FACS
Therapeutic Hypothermia in Deceased Organ Donors and Kidney-Graft Function. Niemann CU, Feiner J, Swain S, Bunting S, Friedman M, Cruthfield M, Broglio K, Hirose R, Roberts JP, Malinoski D. N Engl J Med. 2015 Jul 30;373(5):405-14.

Article 2 reviewed by Jason Smith, MD, PhD, MBA, FACS
Age of Transfused Blood in Critically Ill Adults. Lecroix J. et al. New England Journal of Medicine. 2015 Apr 9;372(15):1410-8.

Article 3 reviewed by Jason Smith, MD, PhD, MBA, FACS
Acetaminophen for Fever in Critically Ill Patients with Suspected Infection. Young P, Saxena M, Bellormo R, Freebairn R, et al. New England Journal of Medicine. 2015 Dec 3;373(23):2215-24.

Article 4 reviewed by Jason Smith, MD, PhD, MBA, FACS
Permissive Underfeeding or Standard Enteral Feeding in Critically Ill Adults. Arabi YM, Aldawood AS, Haddad SH, et al. N Engl J Med. 2015 Jun 18;372(25):2398-408.

Article 1
Therapeutic Hypothermia in Deceased Organ Donors and Kidney-Graft Function. Niemann CU, Feiner J, Swain S, Bunting S, Friedman M, Cruthfield M, Broglio K, Hirose R, Roberts JP, Malinoski D. N Engl J Med. 2015 Jul 30;373(5):405-14.

This article by Niemann et al in the NEJM highlights the effects of induced mild hypothermia during deceased organ donor management on renal graft function in the transplant recipient.  It is one of the first prospective studies on donor management that demonstrates a link to graft outcomes in the recipient patient.  Delayed graft function (DGF - defined as dialysis during the first week post kidney transplant) has been reported in up to 50% of recipients of kidney transplants.  It is costly and ultimately is a marker for a decrease in long-term graft survival. This study enrolled deceased organ donors who were randomized to normothermia (36.5 to 37.5oC) or mild hypothermia (34-35oC) and maintained at that temperature throughout donor management until organ recovery. Temperatures were primarily maintained using the Bair Hugger.  The trial was terminated early after the interim data analysis demonstrated efficacy of hypothermia in decreased DGF.  In total there were 370 organ donors enrolled in the two groups (180 in the hypothermia group and 190 in the normothermia group).
 
 The results were promising with an overall reduction of DGF in recipients who received a kidney from a donor in the hypothermia group.  39% of recipients in the normothermia group developed DGF compared with 28% in the hypothermia group (OR, 0.62; 95% CI 0.43-0.92; p = 0.02). While this overall rate of DGF is a bit high compared to some reports, it does fall within the overall reported incidence of up to 50% following renal transplantation.  The authors’ statistical analysis took in to account many donor variables, differences in the organ procurement organizations and cold-ischemic time to control for other variations in the possible outcomes.  The highest level of efficacy for hypothermic treatment was noted in those patients who were deemed expanded-criteria donors at enrollment. This group represents an older and generally sicker patient population in which the DGF rate dropped from 56.5% in the normothermia group to 31% in the hypothermia group. The number of organs transplanted per donor was not significantly different between the two study groups and overall complications were no different. A nice appendix is available here that gives further details about the study groups and numbers of organs transplanted.  Overall, this is a very nice prospective study across a varied patient population in several centers that addresses a real problem for renal transplant recipients; DGF. I would recommend the article to those who care for neurologically devastated patients potentially involved in the organ donation process.The technique is simple, cost-effective and clearly demonstrates a linked outcome from donor management to transplant recipient organ outcome.

Article 2
Age of Transfused Blood in Critically Ill Adults. Lecroix J. et al. New England Journal of Medicine. 2015 Apr 9;372(15):1410-8.

This study, termed the ABLE (Age of Blood Evaluation) trial, was designed to see if transfusions using fresh red blood cells lead to a decreased 90-day mortality in critically ill patients as compared to older blood. The “fresh” RBC group was defined as patients receiving blood that was stored for less than 8 days.  2400 patients were accrued in over 64 centers across Canada and Europe, and randomized.   Inclusion criteria were an age of 18 years or older, admission to an ICU, a RBC transfusion prescribed within 7 days of admission to the ICU, and an expectation for the patient to require at least 48 hours of invasive or noninvasive mechanical ventilation. No other treatment guidelines or transfusion guidelines were mandated in the study.
 
There were 1200 patients in each arm, with similar baseline characteristics. The anemia and red cell transfusion characteristics were also nearly identical in each arm.  The average Hb level before the first transfusion was 7.69 and 7.64 between the fresh group and the standard group respectively. The duration of storage of the RBC’s transfused was 6.1d vs 22d.  The 90-day all cause mortality was similar; 37% for the “fresh” group vs 35% for the “older” group.  In the subgroup analysis, there was also no difference in mortality between the groups on the basis of age, APACHE II score, number of units transfused, or admission category. There were no significant differences in any of the secondary outcomes measured (organ dysfunction, infections, length of hospital and ICU stay, and the duration of respiratory, renal, and hemodynamic support).  Strengths cited include large numbers for an adequately powered study, diverse patient sets to ensure broad applicability of the results, and minimization of bias due to study design.  Limitations include under-representing groups that may be particularly vulnerable to the consequences of prolonged RBC storage, and the applicability of the study to centers practicing a liberal transfusion strategy or those that do not use leukocyte-reduced blood.
 
Article 3
Acetaminophen for Fever in Critically Ill Patients with Suspected Infection. Young P, Saxena M, Bellormo R, Freebairn R, et al. New England Journal of Medicine. 2015 Dec 3;373(23):2215-24.

The pharmacologic treatment of fever in the ICU is a common occurrence and nursing request, though the benefit of doing so is controversial. Proponents suggest that fever should be treated with the rationale that it placed additional physiologic stress on patients who were already seriously ill. Critics argue that fever is a natural response to fight infection and may act to enhance immune-cell function and inhibit bacterial growth. Previous studies have provided support to both viewpoints, but there has not been a large PRCT to provide high-level evidence.

This study was a multi-center study conducted in Australia and New Zealand as a prospective, parallel-group, blinded, randomized controlled trial. It included adult patients with a temperature > 38 and who were receiving antimicrobial therapy for a known or suspected infection. The primary outcome measure was ICU-free days (to day 28), a composite outcome measure of ICU mortality and length of stay. 
             
A total of 700 patients were eventually randomized and included in the trial.  The two groups were similar in baseline characteristics. The study group had statistically significant lower mean daily peak and average body temperatures. There was no difference between the groups with respect to the primary outcome of ICU-free days. ICU length of stay was longer in the study group among patients who ultimately died, with the authors suggesting that this was consistent with a previous study that cooling to normothermia delayed death in mechanically ventilated patients with septic shock (but did not change overall long-term mortality). 

The results of this study suggest that the use of acetaminophen to treat fever in the ICU does not affect ICU-free days or overall mortality.  Some of the limitations of the study include a short duration of drug administration (<= 8 doses) and the allowance of open label use of acetaminophen by the control group after the study period.  Roughly 1/3 of control group patients eventually were treated with acetaminophen during their ICU stay. 

Overall, I think the study does lend some credence that low-grade fever does not need to be pharmacologically treated in the ICU.  However, the limitations of the study and difficulties of investigating a complex and heterogeneous ICU population are likely to lead to continued controversy. The finding of decreased early mortality among non-survivors in the study group is a target for future investigations. 

Article 4
Permissive Underfeeding or Standard Enteral Feeding in Critically Ill Adults. Arabi YM, Aldawood AS, Haddad SH, et al. N Engl J Med. 2015 Jun 18;372(25):2398-408.

Growing controversy has involved the appropriate dosing of nutrition support in the critically ill patient and “full feeding” strategies have recently been challenged. Recently, a large multicenter, prospective, randomized controlled trial was performed in Saudi Arabia and Canada examining different dosing strategies in the ICU from 2009 to 2014. A total of 894 patients were examined with 448 patients in the “permissive underfeeding” group and 446 patients in the “standard feeding” group. Permissive underfeeding (PU) targets were 40-60% of goal and standard feeding (SF) targets were 70-100% goal. Important to note, protein targets were equivalent in both groups. The heterogeneous population examined was comprised of 75% medical, 22% non-operative trauma, and 3% surgical patients.  Nearly all (>95%) of patients were receiving mechanical ventilation and more than half (55%) were on vasopressor medications. The average APACHE II score was 21 and SOFA was 9.9. The PU group received 47% of goal calories and 68% protein goals and the SF group received 71% and 69% of goals with calories and protein respectively successfully achieving the desired differences between the groups.

Outcomes between the groups demonstrated no differences in the primary, secondary, and tertiary endpoints. The primary endpoint was 90-day all cause mortality. Secondary outcomes included serial SOFA scores and 28-day, in-hospital, 180-day, and ICU mortality. Tertiary outcomes included vent-free days, ICU free days, feeding intolerance, ICU infections, electrolyte abnormalities, transfusion requirements, and hospital length of stay. With regard to co-interventions, patients in the SF group received more insulin, had higher blood glucose values, and had more positive fluid balance over the course of their hospitalization.

The authors concluded that PU strategies were equivalent with regard to the above outcomes when compared to SF strategies. An important difference between this trial and other preceding trials remains the standardization of protein delivery between the groups. There is growing sentiment that protein delivery remains the key driver in appropriate nutrition support in the critically ill patient rather than non-protein caloric provision (Allinstrup). Of note, extrapolating the findings of this study to surgical and trauma populations should be done with some caution given that surgical patients represented a very small minority (2-4%) of the study population and only non-operative trauma patients were included. Despite these cautionary statements, this study does give credence to the avoidance of full feeding strategies in intolerant patients and supports the notion that permissive underfeeding may be appropriate in critically ill patients.