Poor test characteristics for the digital rectal examination in trauma patients.

ABSTRACT

STUDY OBJECTIVE: Current advanced trauma life support guidelines recommend that a digital rectal examination be performed as part of the initial evaluation of all trauma patients. Our goal is to estimate the test characteristics of the digital rectal examination in trauma patients. METHODS: We conducted a retrospective medical record review study of consecutive trauma patients treated in our emergency department from January 2003 to February 2005 for whom the trauma team was activated and who had a documented digital rectal examination. RESULTS: One thousand four hundred one patients met our selection criteria and were included in the analysis. We estimated the composite sensitivity of the digital rectal examination (any abnormal finding) for detecting any of the index injuries to be 22.9% (95% confidence interval [CI] 16% to 30%) and the specificity to be 94.7% (95% CI 93% to 96%). The calculated sensitivity and specificity for the digital rectal examination were 37% (95% CI 23% to 50%) and 96% (95% CI 95% to 97%), respectively, for detection of spinal cord injury, 5.7% (95% CI 0% to 13%) and 98.9% (95% CI 98% to 99%) for detection of bowel injury, 33.3% (95% CI 0% to 87%) and 99.8% (95% CI 99% to 100%) for detection of rectal injury, 0% and 99.8% (95% CI 99% to 100%) for detection of pelvic fracture, and 20% (95% CI 0% to 55%) and 99% (95% CI 98% to 100%) for detection of urethral disruption. CONCLUSION: The digital rectal examination has poor sensitivity for the diagnosis of spinal cord, bowel, rectal, bony pelvis, and urethral injuries. Our findings suggest that the digital rectal examination should not be used as a screening tool for detecting injuries in trauma patients.