Trauma Ultrasound and the 2005 Cochrane Review

Article Outline

• References
• Copyright

To the Editor:

In the March 2007 issue of Annals, Dr. Steven Vance wrote a systematic review abstract summarizing the 2005 Cochrane review of emergency ultrasound-based algorithms for diagnosing blunt abdominal trauma. The authors of the Cochrane review concluded that there was no statistically significant difference in the use of computed tomography, use of diagnostic peritoneal lavage or laparatomy rate between patients who received ultrasound and those who did not.¹ However, the Cochrane review was flawed by statistical errors as well as inclusion of a questionable study.², ³, ⁴, ⁵ When calculating the risk differences in the use of computed tomography and diagnostic peritoneal lavage between subjects who received ultrasound and subjects who did not, the authors of the Cochrane review mistakenly switched the denominators in the Boulanger study. For example, the Cochrane reviewers concluded that although ultrasound reduced the risk of computed tomography use by 46%, it was not statistically significant (95% confidence interval [CI] includes zero). However, with correct numbers from the Boulanger study,² ultrasound significantly reduces the risk of computed tomorgraphy use by 62% (95% CI: (-0.86, -0.39) (Table 1). In addition, the Cochrane analysis included a study by Navarrete-Navarro³ that is not relevant because the authors primarily studied the effect of computed tomography use in trauma. In this study, the majority of subjects (57%) in the “ultrasound group” never received an abdominal ultrasound but the Cochrane review assumed all subjects did. For use of diagnostic peritoneal lavage, with corrected numbers (from the Boulanger study) and the elimination of the Navarrete-Navarro study, ultrasound is associated with a 6% risk reduction (95% CI: (-0.09, -0.02) (Table 2).

Table 1. Comparison of risk difference estimates in use of computed tomography when incorrect versus correct numbers are used.

Correct numbers from the original studies are bolded in the table.

Table 2. Comparison of risk difference estimates in use of diagnostic peritoneal lavage when incorrect versus correct numbers are used.

Correct numbers from the original studies are bolded in the table.

The authors also concluded that the use of trauma ultrasound had no effect on the laparotomy rate but they did not explain the rationale for why it would. The studies analyzed by the Cochrane group², ⁵ were predicting the need of patients for laparotomy. Theoretically, trauma ultrasound may reduce the laparotomy rate by decreasing diagnostic time and thus increasing the number of eligible patients for angioembolization. Unfortunately, the studies providing the laparotomy data in the Cochrane analysis did not have the option of angioembolization for blunt splenic injury.², ⁵ Thus, it does not seem reasonable to expect that ultrasound should decrease the laparotomy rate when looking at the studies in the 2005 Cochrane review. Moreover, the fact that trauma ultrasound significantly reduced the use of computed tomography and diagnostic peritoneal lavage while not increasing the laparotomy rate is never mentioned.

In summary, the Cochrane review of trauma ultrasound was methodologically flawed. While we agree with Dr. Vance that more rigorous studies are needed to determine the full impact of ultrasound on patient outcomes, the data clearly shows significant reductions in both computed tomography and diagnostic peritoneal lavage use when ultrasound is included in trauma algorithms.

Back to Article Outline

References 

1. Stengel D, Bauwens K, Sehouli J, et al. Emergency ultrasound-based algorithms for diagnosing blunt abdominal trauma. The Cochrane Database of Systematic Reviews. 2005;Issue 2. Art. No.: CD00446.pub2.DOI:10.1002/14651858.CD004446.pub2.
   • View In Article
2. Boulanger BR, McLellan BA, Brenneman FD, et al. Prospective evidence of the superiority of a sonography-based algorithm in the assessmant of blunt abdominal injury. Journal of Trauma. 1999;47:632–637
   • View In Article
3. Navarrete-Navarro P, Vazquez G, Bosch JM, et al. Computed tomography vs clinical and multidisciplinary procedures for early evaluation of severe abdomen and chest trauma-a cost analysis approach. Intensive Care Medicine. 1996;22:208–212
   • View In Article
   • MEDLINE
   • CrossRef
4. Arrillaga A, Graham R, York JW, et al. Increased efficiency and cost-effectiveness in the evaluation of the blunt abdominal trauma patient with the use of ultrasound. The American Surgeon. 1999;65:31–35
   • View In Article
   • MEDLINE
5. Rose JS, Levitt MA, Porter J, et al. Does the pressence of ultrasound really affect computed tomographic scan use? (A prospective randomized trial of ultrasound in trauma). Journal of Trauma. 2001;51:545–549
   • View In Article