Elsevier

The Spine Journal

Volume 4, Issue 2, March–April 2004, Pages 130-137
The Spine Journal

Clinical studies
Mild hypothermia, blood loss and complications in elective spinal surgery

https://doi.org/10.1016/j.spinee.2003.08.027Get rights and content

Abstract

Background context

Spinal surgery carries risks of incidental spinal cord and nerve root injury. Neuroprotection, to minimize the extent of such injuries, is desirable. However, no neuroprotective strategies have been conclusively validated in nonvascular spinal surgery. Mild hypothermia resulting from general anesthesia is a readily achievable potential neuroprotective strategy. Mild hypothermia, however, has been associated with wound infection, increased operative blood loss and other complications. No previous studies have specifically evaluated whether mild hypothermia is associated with an increased risk of these complications in elective spinal surgery.

Purpose

We investigated the association between incidental mild hypothermia, perioperative complications and operative blood loss.

Study design/setting

This is a retrospective study employing cohort analysis, rank analysis and single and multivariate linear regression. The setting was the Veterans Administration Medical Center, a teaching hospital of the University of Miami.

Patient sample

Data on a total of 70 adult veterans aged 23 to 81 years undergoing complex spinal procedures in which passive cooling was employed during surgical decompression.

Outcome measures

The variables measured were temperature, blood loss, mean arterial pressure (MAP) and duration of anesthesia. The outcome measured was the presence or absence of complications.

Methods

After 70 patients had been acquired, regression and rank analyses were performed to test for a link between mild hypothermia and blood loss. In addition, two cohorts, patients who experienced complications, and those who did not experience complications in the perioperative period, were compared for several variables including three measures of exposure to hypothermia. Surgical procedures included 60 cervical, 1 occipitocervical, 1 cervicothoracic, 7 thoracic and 1 thoracolumbar procedure. Hypothermia followed induction of anesthesia; esophageal or bladder temperature was monitored. Cooling was passive; warming utilized a forced air blanket. Temperature data from anesthetic records was used to derive mean intraoperative temperature, nadir intraoperative temperature and the rates of cooling and rewarming. The time course of hypothermia, the overall fluctuation in core temperature and the quantity of subbaseline temperature were determined. Medical and surgical complications were included. Two patients with complications considered irrelevant to hypothermia were removed from further analysis. Patients with and without complications were compared as cohorts for differences in mean values of age, comorbid risk factors, intraoperative MAP, intraoperative blood loss, anesthetic duration and temperature-related measures. Relationships between blood loss, anesthesia duration and temperature parameters were assessed in rank and regression analyses.

Results

Patients with complications (n = 12) had longer mean anesthetic durations (p = .0001) and larger mean surgical blood losses (p = .001) than patients without complications (n = 56). Neither mean nor nadir intraoperative hypothermic temperatures were statistically associated with complications. However, large hypothermic integrals (p = .04) and the total quantity of recorded temperature fluctuation (p = .01) were both associated with complications. Comorbid risk factors, MAP and age were not statistically linked to complications. Finally, no relationship between any of the temperature measures and increased blood loss was found.

Conclusion

Operative blood loss was not linked to any index of the patient's temperature. Longer anesthesia durations were linked to complications and increased blood loss. Regarding mild hypothermia, neither mean nor nadir hypothermic temperatures were linked to complications, but the estimated total quantity of subbaseline temperature was linked, as was total fluctuation in temperature. Lengthy exposure to mild hypothermia appeared to be associated with wound infections. The use of mild hypothermia as a potential neuroprotective strategy during spinal surgery appears to be reasonably safe, but to avoid complications, the duration of hypothermic exposure should be minimized.

Introduction

Strategies to prevent neurological injury during spinal surgery include the use of neurophysiologic monitoring and neuroprotection. No specific studies of the efficacy of neuroprotection in spinal surgery exist, and evidence obtained in other clinical settings has been extrapolated into the practice of spinal surgery. Because methylprednisolone has gained acceptance as a treatment for acute spinal cord injury, many spine surgeons use the Second National Acute Spinal Cord Injury Study (NASCIS II) [1] dosing protocol during “high risk” cases on a prophylactic basis or administer steroids if there is an apparent adverse change in monitored neurophysiologic variables. Using similar reasoning, hypothermia, which has an established neuroprotective efficiency to prolong the ischemic tolerance of the brain or spinal cord [2], [3], [4], is thus hypothesized to provide neuroprotection against potential mechanical injury during spinal surgery. Such injuries might involve accidental contusion, spinal cord or nerve root stretching or compression, or injury associated with the removal of intrinsic tumors. It is hypothesized that the neuroprotective agent, whether methylprednisolone or hypothermia, will reduce the extent of injury mainly by altering mechanisms of secondary cellular damage. However, both methylprednisolone and mild hypothermia have been associated with a risk for increased complications in some series of surgical patients [5], [6], and therefore data regarding risk/benefit should influence their use during elective spinal surgery.

Because general anesthesia impairs thermal regulation, mild hypothermia is commonly observed during surgery. Continuous monitoring of core temperature is a surgical standard and provides data that can be correlated with blood loss and complications. Experimental data have indicated that mild hypothermia (32.5C to 35.5C) can reduce some aspects of secondary spinal cord injury after spinal cord contusion [7], [8], [9], [10]; however, this efficacy of neuroprotective hypothermia is not clearly established in human spinal cord injury. Clinical and experimental studies have indicated that hypothermic neuroprotection from ischemia is markedly more effective if established before injury than if applied after injury as a therapy [11], [12]. Mild hypothermia has been studied as a treatment for brain injury [13], but only a few studies have reported its impact in elective neurosurgery [14], [15], [16].

Deep levels of hypothermia (less than 32.5C), clinically established for ischemic neuroprotection during high-risk cerebrovascular surgery, have been associated with altered hemostasis [17], [18], [19] and cardiac arrhythmia [20]. Mild hypothermia, considered safer, has also been associated with several potential complications, including wound infection [21], increased surgical blood loss [22], [23], perioperative myocardial infarction [24], and altered surgical wound healing [25], [26]. Mild hypothermia is therefore both a potential risk factor for perioperative morbidity [27], [28] and a potential neuroprotective strategy in spinal surgery. Furthermore, the hypothermic temperatures at which risks for complications increase and neuroprotective benefits occur may not be the same. As a risk factor, “mild” hypothermia is considered to exist at any temperature below 37C; however, experimental studies indicate that neuroprotection requires central nervous system temperatures below approximately 35.5C [29]. The threshold for hypothermic neuroprotection in humans is not known with certainty. The relative risks and potential neuroprotective benefits of hypothermia are likely to be related to the timing, duration, depth of hypothermia and rate of rewarming. Because of its neuroprotective potential, apparent simplicity, possible risks and inherent association with general anesthesia, it is important to determine the impact of mild hypothermia in the setting of spinal surgery.

This study assesses the association between incidental mild hypothermia, surgical blood loss and perioperative complications in a series of patients undergoing spinal procedures (ie, cervical or thoracic) that placed the spinal cord at risk for injury. A new measure to quantify subbaseline temperature, the time-temperature integral t1tx (TB−TO), is introduced, where TB = temperature baseline, TO = a patient's documented temperature, t1 = the first time point with a core temperature less than 37°C, and tx = the last time point with a core temperature less than 37°C.

Section snippets

Methods

We acquired data prospectively on 70 consecutive surgical patients in a single surgeon series at the Veterans Affairs Medical Center in Miami, Florida (institutional review board approval number 00242), a teaching hospital of the University of Miami. The population consisted of adult veterans, including patients with chronic and acute spinal cord injury. The surgical procedures included 60 cervical, 1 occipitocervical, 1 cervicothoracic, 7 thoracic and 1 thoracolumbar procedure. Indications for

Results

The surgical procedures performed are summarized in Table 2.

Time-temperature integral

We found that t1∫tx (TB−TO) is useful to represent intraoperative body temperature changes. It is strongly associated with duration of subbaseline temperature, and mean and nadir temperatures. Because t1∫tx (TB−TO) is measured using only the duration between time points when subbaseline temperature (<37C) is actually present, it differs from the mean in which all the temperature measurements are divided by the total anesthetic time. The mean could underestimate the occurrence of a brief but

Conclusion

Brief periods of mild hypothermia during elective spinal surgery were not associated with an increased rate of complications, but extended periods may increase risks of wound infection. Mild hypothermia was not associated with increased surgical blood loss. Studies to determine whether mild hypothermia is definitely neuroprotective for human spinal cord trauma are still needed. A randomized prospective study of the safety of mild hypothermia in elective adult spinal surgery is recommended.

Acknowledgements

The anesthesia staff at the Miami Veteran's Administration hospital, especially Dr. Michael Lewis and Dr. Gerard DeSousa. Elizabeth Owen and Janet Gabrin-Levy performed manuscript editing and figure preparation.

References (42)

  • C.G Yu et al.

    Beneficial effects of modest systemic hypothermia on locomotor function and histopathological damage following contusion-induced spinal cord injury in rats

    J Neurosurg

    (2000)
  • K Chatzipanteli et al.

    Posttraumatic hypothermia reduces polymorphonuclear leukocyte accumulation following spinal cord injury in rats

    J Neurotrauma

    (2000)
  • R.R Hansebout et al.

    Current status of spinal cord cooling in the treatment of acute spinal cord injury

    Spine

    (1984)
  • W.R Yu et al.

    Systemic hypothermia following compression injury of rat spinal cord: reduction of plasma protein extravasation demonstrated by immunohistochemistry

    Acta Neuropathol (Berl)

    (1999)
  • M Kakinohana et al.

    The effect of graded postischemic spinal cord hypothermia on neurological outcome and histopathology after transient spinal ischemia in rat

    Anesthesiology

    (1999)
  • M Marsala et al.

    Effect of graded hypothermia (27 degrees to 34 degrees C) on behavioral function, histopathology, and spinal blood flow after spinal ischemia in rat

    Stroke

    (1994)
  • G.L Clifton et al.

    Lack of effect of induction of hypothermia after acute brain injury

    N Engl J Med

    (2001)
  • P.C Gerszten et al.

    Intraoperative hypothermia and ventricular shunt infections

    Acta Neurochir

    (1998)
  • K.Z Baker et al.

    Deliberate mild intraoperative hypothermia for craniotomy

    Anesthesiology

    (1994)
  • G.L Clifton et al.

    Use of moderate hypothermia during elective craniotomy

    Tex Med

    (1992)
  • C.S MacCarty et al.

    The application of profound hypothermia to the treatment of cerebrovascular disorders

    Res Publ Assoc Res Nerv Ment Dis

    (1966)
  • Cited by (67)

    • Interventions to minimize blood loss and transfusion risk in spine surgery: A narrative review

      2020, Clinical Neurology and Neurosurgery
      Citation Excerpt :

      Collectively, they found decreases in blood loss of up to 40 % and decreases in transfusion rates of up to 45 %. Hypothermia – defined as a core body temperature less than 36 °C – produces a relatively hypocoagulable state, increasing the risk for intraoperative blood loss and the need for transfusion [154,155]. This is thought to occur secondary to inhibition of thromboxane synthetase, resulting in knockdown of thromboxane A2 – a prostaglandin responsible for platelet aggregation under normal conditions [156,157].

    View all citing articles on Scopus

    FDA device/drug status: not applicable.

    Nothing of value received from a commercial entity related to this research.

    View full text