Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization

Summary

Mild perioperative core hypothermia is common in patients undergoing colon surgery. Hypothermia leads to vasoconstriction and decreased oxygen tension at the wound site, to impaired neutrophil function and also, potentially, to decreased wound strength. This randomized, double-blind study was undertaken to determine whether maintaining perioperative normothermia would decrease infections, improve wound healing and decrease length of stay in patients undergoing colon resection.

Methods

• Subjects: Patients 18-80 years old, undergoing elective colorectal resection for cancer or inflammatory bowel disease.

 
• Exclusions: Chemotherapy, steroid therapy, infection, malnutrition.

 
• Intervention: All patients had standard bowel prep and received IV antibiotics at induction of anesthesia and for about four days thereafter. All patients were hydrated aggressively (to prevent hypovolemia induced vasoconstriction), were given supplemental oxygen postoperatively and pain was treated using a patient-controlled system.

Patients were randomized to normothermia (core temperature maintained at about 36.5 degrees Centigrade) or hypothermia (allowed to drop to about 34.5). In both groups, IV fluids passed through a warming device and forced air was circulated over the patients' upper body, but in the normothermia group the devices were turned on, in the hypothermia group the fluid warmer was not turned on and the air was not heated. The authors state that the surgeons were not able to detect whether the patients were being warmed or not.
 

• Analysis: The analysis examined a number of factors, including pre-operative predictors of infection, the difference in core temperature between the groups, the incidence of wound infections during the first 15 post-operative days, hospital length of stay, and collagen deposition near the wound in a subgroup of patients.

The authors planned to enroll 400 patients but also to perform interim analyses after the enrollment of 200 and 300 patients, with appropriately stringent criteria for early termination of the trial.

• The trial was stopped after 200 patients were enrolled (between July, 1993 and March, 1995) because of a significant decrease in wound infection in the normothermia group.
   
• Baseline patient characteristics, surgical site (colon vs. rectum) and indications and preoperative risk of infection were similar in the two groups.
   
• As would be expected, core temperature in the hypothermia group dropped significantly compared with the normothermia group and remained lower for about 6 hours postoperatively.
   
• There were 6 surgical wound infections out of 104 patients in the normothermia group, compared with 18 out of 96 in the hypothermia group (p=0.009; reason for terminating the study).
   
• Clinical variables were also better in the normothermia group, including days of hospitalization (12.1 vs. 14.7), collagen deposition near the wound in a subgroup of patients, days to first solid food and days to suture removal. Interestingly, these clinical variables were also better when only uninfected patients were looked at.
   
• The WBC count was lower on postop day 1 in the hypothermia group (11,500 vs. 13,400) suggesting an impaired response, but higher on postop day 3 (10,100 vs 8,900) related to infection.
   
• Multivariate analysis yielded five significant risk factors for wound infection: the most potent risk factor was tobacco use (odds ratio 10.5), then hypothermia (4.9), rectum vs. colon surgery (2.7), preoperative National Nosocomial Infection Surveillance System score (2.5 per unit increase) and age (1.6 per decade).

Comment

Similarly, these results apply to colorectal resection surgery. Although it seems logical that they should apply to other types of surgery, some further trials will need to be done to confirm this.

As the authors and the authors of an accompanying editorial note, the risk of infection associated with smoking is even greater than the risk of hypothermia. It would be interesting to determine whether this is due to the vasoconstrictive effect of smoking, in which case discontinuing the habit a day or two prior to surgery would be of significant utility.

These points aside, this is a very well-designed study, with results that are both statistically and clinically significant. It is all the more impressive for its use of a low-cost, low-tech approach. It will be interesting to see how quickly these results will be translated into practice.

May 25, 1996

Reader comments

Date: Sat, 01 Jun 1996
From: Olafur Jakobsson <opj@isl.pp.se>

Once again 'a feeling', but I strongly believe that the principles of adequate hydration, pain relief and normal body temperature have a significant influence on the results of free-flap surgery. The results are very dependent upon peripheral skin circulation, which otherwise can be difficult to maintain.

O P Jakobsson, MD, PhD, plastic surgeon

Makes sense. I wonder, is there significant core hypothermia with this sort of surgery? -- mj

Yes, as free-flap operations tend to take a long time with a large part of the body exposed. We certainly had problems until this was recognized and solved.

Olafur J

Date: 27 Jun 96
From: Daniel Sessler <76735.2602@CompuServe.COM>
To: Michael Jacobson

 Thank you for including our recent NEJM article on your WWW site. Your summary is excellent and I have nothing to add to it. I would be pleased to respond to readers' questions.

Best, Dan.

September 5, 1996

Letters to the Editor about this article from the NEJM website.

December 3, 1996

 From: wengered@chr.mts.kpnw.org.

 HI,

What do you think of using normovolemic hemodilution to lessen the possibility of infection?

E. WENGER MD.
 
 

Acute normovolemic hemodilution involves removal of some of the patient's blood prior to surgery and replacing it with enough fluid (colloid / crystalloid) to maintain euvolemia. When blood loss occurs during surgery, it involves loss of a lesser amount of RBC's. After surgery, the blood that was removed can be re-infused. This is a technique to minimize the need for blood transfusion.

A quick literature search didn't turn up any obvious articles about how this technique affects wound infection rates, and I'm not sure how it would interact with perioperative normo- or hypothermia. I'll forward this question to Dr. Sessler, to see if he has anything to add. -- mj

Date: Mon, 21 Apr 1997 From: Arni Björnsson <arni.bjornsson@swipnet.se>
Arni Bjornsson MD, Dep of Anesthesiology
Univ. Hospital Linkoping
SWEDEN

The question is highly relevant. We have an infection problem in our spinal surgery unit. The patients are hemodiluted to Hb 8.0 mg/dL, the drained volume being replaced with saline and 4% albumin in a ratio 1:2.

Most of the patients have hypothermia (35 deg. Celcius) at the end of surgery. We routinely use plasma sequestration, red cell salvage (Haemonetics AT-1000) and platelet gel to minimize the intraoperative blood loss. Proper antibiotic therapy and skilled surgeons give good results but still the infection rate (20%) and subsequent reoperations should be lower.

The theatre ventilation is being replaced and new routines in sterility are adapted. Still we can't keep the patients warm during these lengthy operations and I'm certain that is a contributing fact to the high infection rate.

Comments are welcome to:

arni.bjornsson@swipnet.se
arni.bjornsson@ane.us.lio.se
 
 

Could the hemodilution contribute to the hypothermia, and thus indirectly to the infection rate? -- mj.

June 12, 1997

The New England Journal of Medicine just published a detailed review of the topic of perioperative hypothermia (N Engl J Med 336:1730-7) by Dr. Daniel Sessler, corresponding author of this study.
 

Date: Fri, 20 Jun 1997
From: Frank Lloyd <flloyd@delphi.com>

Wound infections can be treated on outpatient basis. So if wound infection is the cause of prolonged hosp stay, it may be associated with other factors.

Good study, although the ethics could be questioned, as was done.

Frank Lloyd
 

Date: Tue, 19 Aug 1997
From: Wengeredu@aol.com

I would like to add that hotline fluid warming by itself fails to maintain normothermia. The claim by some physicians that fluid warming alone will maintain normothermia doesn’t agree with much of what is known about perioperative heat balance and seems inconsistent with the laws of thermodynamics.

E. Wenger MD
Anesthesiologist
wengeredu@aol.com
 
 

Thus the importance of blankets and forced-air warming, as described here.

All I remember of the laws of thermodynamics is that entropy is constantly increasing, which the surface of my desk confirms daily. -- mj

From: Alo IPSS [aihuancayo@aloipss.sld.pe]
Sent: Tuesday, February 16, 1999
 

I think the article is very interesting, because for the general surgeon hypothermia may not be so important, but for pediatric surgeon it is very important. Perhaps it´s necesary to use
warm or radiant table in operatory room as in pediatric surgery  for maintenance of normothermia.
 

José Cárdenas, M.D.- Pediatric Surgeon.
Huancayo-Perú-South América
 

Submit a comment about this article
 

Journal Club on the Web main screen