Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Effects of therapeutic hypothermia on intracranial pressure and outcome in patients with severe head injury.

Polderman KH, Joe RTT, Peerdeman SM et al.

Intensive Care Med 2002;28: 1563-1573 [abstract]

Reviewed by Satid Thammasitboon, MD, MHPE, Pediatric Critical Care Medicine, Baylor College of Medicine, Houston, TX, and Supat Thammasitboon, MD, Pulmonary and Critical Care Medicine, Tulane University, New Orleans, LA

Review posted March 10, 2003

---

I. What is being studied?:

The study objective:

To investigate the outcomes of patients with severe traumatic brain injury treated with moderate hypothermia using a strict protocol including measures to prevent side effects of hypothermia.

The study design:

Prospective cohort clinical trial at a single institution

The patients included:

All patients (n=136) with severe head injury (Glasgow Coma Scale score ≤ 8 on admission) who were admitted to a tertiary neurosurgical referral center over a five year period. The study also included patients who had GCS score higher than 8 but subsequently deteriorated using sub-analysis for theses two different groups of patients.

The patients excluded:

No exclusion criteria mentioned

The interventions compared:

Standard treatment:

The goals of therapy were stabilization or improvement of the patient's neurologic condition, maintenance of an ICP of 20 mmHg or less and a mean arterial pressure (MAP) 90 or more with vasoactive agents to maintain a cerebral perfusion pressure of 70 mmHg or more. In patients with ICP higher than 20 mmHg initial treatments included appropriate sedation with midazolam, fentanyl, or propofol, as well as treatment with muscle relaxant and administration of mannitol. Neurosurgical intervention was undertaken when necessary to evacuate subdural lesions or large intracerebral lesions. PaCO2 were maintained at 34-40 mmHg; PaO2 levels were maintained above 90 mmHg.

Study protocol:

Control group: Patients who had ICP greater than 20 mmHg after standard treatment and responded to barbiturate coma therapy (pentobarbital bolus 100-200 mg, followed by continuous infusion of 100-300 mg/h until burst suppression on EEG). Response was defined as a significant decrease (20% from baseline value) in ICP within 30 minutes and decrease to and stabilized at levels below 20 mmHg within 60 minutes.

Hypothermia group: Patients who failed to respond to barbiturate coma therapy received artificial cooling using water-circulating blankets. The aim was to lower the body temperature to 32°C. The patient was slowly re-warmed (1°C per 12 h) if ICP remained at 20 mmHg or less for 24 h.

The outcomes evaluated:

Glascow Outcome Scale (GOS) at 6 months

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

No. The patents were not randomized. The study and control groups were classified by their clinical condition after the standard therapy.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Yes. All patients entering the study were followed until they died or 6 months after hospital discharge. Six-months should be long enough for the assessment of neurological outcome.

Were patients analyzed in the groups to which they were randomized?

n/a

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

No. Only examiners who assessed GOS scale were blinded to the treatment group assignments. The clinical management of patients with severe head injury is so critical that keeping clinicians blinded is not practical.

4. Were the groups similar at the start of the trial?

No. More patients in the hypothermia group had severe head injury on CT based assessment (lesion > 25 cc, not surgically evacuated), higher APACHE II scores and lower GCS. More patients in the control group had higher GCS on admission (GCS ≥ 8) and deteriorated during hospitalization. The etiologies of the deteriorations were not described. Since these confounding factors could affect the outcomes of the study, subgroup analysis was used to enhance validity of the results. The hypothermia group was comprised of patients whose ICP failed to respond to the standard therapy. Therefore, they had a worse neurologic condition than those in the control group at the start of the trial.

5. Aside from the experimental intervention, were the groups treated equally?

No. The hypothermia group received the following measures to prevent the occurrence of cooling induced side effects.

• Administration of additional dose of fluid (500-1000 ml of saline in 30 min) upon initiation of artificial cooling and vasoactive medication to maintain CCP at 70 mmHg.
• Electrolyte levels were measured frequently using rapid lab device installed in the ICU to high-normal electrolyte levels (Mg ≥ 1.1 mmol/L, K ≥ 4, P ≥ 1mmol, Ca ≥ 2.3 mmol/L, ionized Ca ≥ 1.2 mmol/L).

These measures not only prevent possible adverse effects of hypothermia but could also decrease the overall morbidity in this group of patients.

III. What were the results?

1. How large was the treatment effect?

The primary outcome measure was the Glasgow Outcome Scale. Those with good recovery (5) or moderate disability (4) were considered to have good outcomes.

All Patients

Outcome	ARR/ABI, 95% CI	NNT, 95% CI
ICU mortality	9.7%, [-6.02%, 25.46%]	10.3, [4, inf]
Good/excellent Neurological outcome	5.9%, [-5.32%, 17.13%]	16.9, [5.8, inf ]

Subgroup analysis: GCS ≥ 7 on admission were excluded.

Outcome	ARR/ABI, 95% CI	NNT, 95% CI
ICU mortality	13.7, [-4.35%, 31.71%]	7.3, [3.2, inf]
Good/Excellent Neurological outcome	10.1, [-3.67%, 23.77%]	9.9, [4.2, inf]

Subgroup analysis: GCS 5-6

Outcome	ARR/ABI, 95% CI	NNT, 95% CI
ICU mortality	24, [-1.76%, 49.76%]	4.2, [2.0, inf]
Good/Excellent Neurological outcome	20, [-2.29%, 42.29%]	5, [2.4, inf]

ARR: Absolute Risk Reduction
ABI: Absolute Benefit Increase
NNT: Number Needed to Treat
CI: Confidence Interval

2. How precise was the estimate of the treatment effect?

The 95% confidence intervals cross zero therefore the treatment effects were not statistically significant (in all except those with GCS scores of 5-6.) This is in conflict with the article which reports p values < 0.05 for overall hospital mortality and good neurologic outcome (page 1569). Based on the above confidence intervals the p values shouldn't have been < 0.05. It is unclear how overall mortality and overall neurologic outcome were statistically analyzed.

3. Did the treatment effect differ in subgroups of patients?

Yes. The benefits of hypothermia were only evident in patients with GCS of 5 or 6 at admission. There was no benefit of hypothermia in patients who had GCS 3 or 4 and those whose GCS were greater than 7 on admission. This finding was consistent with an earlier study (1).

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

Maybe. Children are different from adult in their response to injury. The incidence of diffuse cerebral swelling is 3.5 times more common in children than adults (2). With little knowledge on this phenomenon, using the adult data may not be adequate for assessment of a treatment in pediatric age group. But it may be worth exploring as a therapeutic option. The treatment protocol, however, is very well-described that could be applied to the future study in children. The preventive measures for adverse effects of hypothermia would be particularly useful in the future study.

2. Were all clinically important outcomes considered?

Yes. The 6-month neurological outcome (GOS) and mortality rate are considered adequate. An important aspect we should consider is the quality of life since there is tremendous burden to families and society. The functional outcome measure that is representative to quality of life should be used.

3. Are the likely treatment benefits worth the potential harms and costs?

Although all adverse effects of hypothermia in adult studies could potentially cause secondary brain injury, they are easily prevented. We might anticipate more profound adverse affects on neurological outcome in children. Further studies are needed to determine the safety of hypothermia in pediatric population.

The major limitation of this study is the study design. The effect of the therapeutic intervention was tested on two incomparable groups of patients. The study basically compared the outcomes of patients that responded to a certain level of care with those that required higher level of care. Future studies need to be prospective randomized control trials and excludes patients who have GCS > 7, brain death on neurological examination and severe or refractory hypotension.

References

1. Marion DW, Penrod LE, Kelsey SF, et al: Treatment of traumatic brain injury with moderate hypothermia. N Eng J Med 1997; 336:540-546. [abstract]; PedsCCM EB Journal Club Review
2. Aldrich EF, Eisenberg HM, Saydjri C, et al: Diffuse brain swelling in severely head injured children. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 1992; 76:450-454. [abstract]

Comments

Back to the EB Journal Club Index