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

Jugular venous oxygen saturation or arteriovenous difference of lactate content and outcome in children with severe traumatic brain injury.

Pérez A, Minces PG, Schnitzler EJ, Agosta GE, Medina SA, Ciraolo CA.

Pediatr Crit Care Med. 2003 Jan;4(1):33-8 [abstract]

Reviewed by Jeffrey Alten MD, Michele Mariscalco MD, Baylor College of Medicine, Texas Children's Hospital, Houston TX

Review posted March 19, 2004

I. What is being studied?

Study objective:

To assess the association between neurologic outcome and jugular venous oxygen saturation (SjvO2) or arteriovenous difference of lactate content (AVDL) in children with severe traumatic brain injury.

Study design

Prospective observational cohort study.

II. Are the results of the study valid?

Note: These questions follow from Randolph AG et al. Understanding articles describing clinical prediction tools. Crit Care Med 1998;26:1603-1612. [abstract]

1. Was a representative group of patients completely followed up?

The group of patients studied included 27 pediatric patients recruited over a 5 year period that met the following inclusion criteria: Ages 1-16, admission to the PICU within 36 hours of a severe traumatic brain injury (STBI), and Glasgow Coma Scale (GCS) < 9 after cardiorespiratory stabilization. Exclusion criteria included those patients with less than 24 hours of follow-up after data collection was begun (presumably due to death or very quick recovery).

It is not clear how many patients with STBI were excluded secondary to less than 24 hours data collection or admission > 36 hours after injury to the PICU, but overall the study likely represented most of the children admitted to the study hospital with STBI over the 5 year period studied.

2. Was follow-up sufficiently long and complete?

Maybe. All patients who had not died (23 of 27 patients) were evaluated at 3 months according to the Pediatric Cerebral Performance Category (PCPC) by a staff neurologist blinded to the results of the SjvO2 and AVDL monitoring. The PCPC was further categorized by dividing patients into favorable (normal to moderate disability) and unfavorable (severe disability to death) groups. Despite this follow-up at 3 months, many researchers believe that children may continue to have cognitive improvement after STBI beyond 3 months - even up to a year. There was only one living child in the unfavorable group at three months - so despite the theoretical potential for cognitive improvement to the more favorable group after 3 months; this represented only one child in this study. This child's improvement would not likely have major impacts on the results.

3. Were all potential predictors included?

No. The authors, though, stated from the outset that they were studying only two predictors: SjvO2 and AVDL. Therefore, other commonly used predictors of outcome in STBI such as age, type of injury, CPP, ICP, MAP and temperature were not included in the analysis as predictors of outcome - even though most of these values were monitored during the study.

The authors did perform a regression analysis in attempt to establish a relationship among intracranial pressure (ICP), cerebral perfusion pressure (CPP), AVDL, and SjvO2 at 24, 48 and 72 hours after placing both ICP and SjvO2 monitors. A relatively strong correlation was found only between AVDL and ICP after 48 hours of monitoring.

The authors also categorized the injuries based on CT findings. It would have been interesting to see if different CT findings had value as an outcome predictor in this study.

4. Did the investigators test the independent contribution of each predictor variable?

The authors performed a univariable analysis (by calculating the area under the receiver operating characteristic curve) and determined there were specific "cut off values" for frequency of abnormal episodes of both AVDL and SjvO2 that were associated with poor outcome in STBI. However, they did not perform a multivariable regression analysis to see if each would independently be associated with poor outcome. In fact, physiologically, it is likely that these two predictors are related.

5. Were outcome variables clearly and objectively defined?

Yes. The outcome assessed in this study was neurologic outcome at 3 months as measured by the PCPC. The PCPC score ranges from one to six (see appendix in the article for explanation of the scoring). The authors further categorized PCPC scores of 1-3 as a "favorable" neurologic outcome and scores of 4-6 an "unfavorable" outcome. The PCPC is a recognized and valid measure of pediatric cognitive and functional disability (1).

III. What are the results?

1. What is(are) the prediction tool(s)?

The authors did not develop a prediction tool. To develop a prediction tool, they would have needed to include other variables, which as stated above, they did not. Instead, they determined "cut off values" for number of episodes of abnormal AVDL and SjvO2 measurements that were associated with increased risk of poor outcome with STBI.

2. How well does the model categorize patients into different levels of risk?

The investigators first determined what would be abnormal values for AVDL (extrapolated from normal adult values, an AVDL > 2 SD was considered pathologic) and SjvO2 (again based on adult values, abnormal measurements were > 74% and < 56%). Using these values, the authors determined via area under the receiver operating characteristic curve, a cutoff point for how many abnormal measurements of each SjvO2 > 74, SjvO2 < 56, and AVDL > 2 SD were associated with an unfavorable neurologic outcome. (In the study AVDL and SjvO2 were simultaneously measured every 6 hours). Using this statistical analysis, it was determined that two or more independent abnormal measurements of AVDL and SjvO2 were associated with unfavorable neurologic outcome in pediatric STBI patients.

The strength of this association was determined by calculating the relative risk of an unfavorable neurologic outcome when a patient had 2 or more abnormal values for SjVo2 or AVDL. They determined that the risk of developing an unfavorable neurologic outcome was 6.6 times higher when the patient had 2 or more SjvO2 desaturation episodes; and 17.6 times higher with 2 episodes of abnormal AVDL measurements.

3. How confident are you in the estimates of the risk?

The strength of association between having 2 or more abnormal values of AVDL or SjvO2 and unfavorable neurologic outcome was determined by calculating relative risk and the 95% confidence intervals. Again, neurologic outcome was defined dichotomously as favorable or unfavorable. (See above) The relative risk of a patient having an unfavorable outcome with 2 or more episodes of SjvO2 < 56 was 6.6 with a fairly wide 95% CI (1.5 to 29.7). The p value was 0.03. The relative risk for having an unfavorable neurologic outcome with 2 or more episodes of pathologic measurements of AVDL was 17.6 with even wider 95% CI (2.5 to 122.5). The p value was 0.001. The risk of having a poor neurologic outcome and 2 or more episodes of SjvO2 > 74 was not found to be statistically significant.

One must take into account the small sample size and the high prevalence of favorable neurologic outcomes vs. unfavorable outcomes (22 vs. 5) while interpreting these results.

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

1. Does the tool maintain its prediction power in a new sample of patients?

Again, the researchers did not develop a prediction tool to study in a new group of patients. The study would likely be valid in this same tertiary pediatric hospital setting, as the authors performed a comprehensive study, accepting all pediatric patients with severe traumatic brain injury with GCS < 9. A repeat study would then presumably include the same patient population. However, particularly given the small sample size, such a validation study is certainly indicated.

The authors mentioned strict management of ICP and CPP as per their institutional protocol. This protocol may contain minor yet potentially significant differences when compared to other pediatric intensive care units' management of elevated ICP. These management differences may have an impact on the outcomes of pediatric STBI and thus reproducibility of this study in other institutions (3,4). In addition even at the authors' hospital, the protocol has changed during the five years of this study, which may have an impact on the outcome of the study if it were to be repeated today.

2. Are your patients similar to those patients used in deriving and validating the tool(s)?

Yes, the patient population in this study should be similar to any tertiary pediatric hospital with an intensive care unit and an active neurosurgical program.

3. Will the results lead directly to selecting or avoiding therapy?

No. The authors presented the AVDL and SjvO2 as possible prognostic tools, not as values to base management decisions upon. There is not enough compelling evidence at this time to adopt the practice of putting in SjvO2 monitors in all children with an ICP monitor. I don't believe that there is enough evidence to suggest that these monitors add any additional prognostic benefit over current commonly measured variables such as ICP, CPP, and MAP. It remains to be seen if studies demonstrate that treating low and or high SjvO2 and AVDL will have an impact on mortality or morbidity - or merely a prognostic tool as shown in adult studies (4). In addition, a true cost-benefit analysis - to include potential harm (thrombosis, infection, etc.) of SjvO2 catheters in children needs to be performed before adopting the practice of routinely using them (5).

4. Are the results useful for reassuring or counseling patients?

No. At this point having the additional information of AVDL or SjvO2 would not change my discussion of prognosis for STBI children. There is a tremendous variability in the outcomes of children following severe traumatic brain injury. If indeed these variables do prove to demonstrate reproducible prognostic value, a clinician would still need to judiciously discuss all possible outcomes with the family given the unpredictability of a child's recovery from head injury

References:

1. Fiser DH, Long N, Roberson PK, Hefley G, Zolten K, Brodie-Fowler M. Relationship of pediatric overall performance category and pediatric cerebral performance category scores at pediatric intensive care unit discharge with outcome measures collected at hospital discharge and 1- and 6-month follow-up assessments. Crit Care Med. 2000 Jul;28(7):2616-20. [abstract]
2. Hackbarth RM, Rzeszutko KM, Sturm G, Donders J, Kuldanek AS, Sanfilippo DJ. Survival and functional outcome in pediatric traumatic brain injury: a retrospective review and analysis of predictive factors. Crit Care Med. 2002 Jul;30(7):1630-5. [abstract]
3. Tilford JM, Simpson PM, Yeh TS, Lensing S, Aitken ME, Green JW, Harr J, Fiser DH. Variation in therapy and outcome for pediatric head trauma patients. Crit Care Med. 2001 May;29(5):1056-61. [abstract]
4. Gopinath SP, Robertson CS, Contant CF, Hayes C, Feldman Z, Narayan RK, Grossman RG. Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry. 1994 Jun;57(6):717-23. [abstract]
5. Coplin WM, O'Keefe GE, Grady MS, Grant GA, March KS, Winn HR, Lam AM. Thrombotic, infectious, and procedural complications of the jugular bulb catheter in the intensive care unit. Neurosurgery. 1997 Jul;41(1):101-7; discussion 107-9. [abstract]

 

Comments

Back to the EB Journal Club Index