The Rotterdam CT score of traumatic brain injury is a classification aimed at improving prognostic evaluation of patients admitted with moderate or severe traumatic brain injuries. 

It was published in 2006 ¹ and is gaining in popularity. Along with the Marshall classification system, at the time of writing (June 2016), it is the most commonly used system to classify CT brain in traumatic brain injury. 

It should not be confused with the Rotterdam criteria for polycystic ovarian syndrome. 

Structure

The Rotterdam classification includes four independently scored elements. Like the Marshall system, it includes 1) degree of basal cistern compression and 2) degree of midline shift. It does not, however, include contusions, but rather restricts mass lesions to 3) epidural hematomas, and adds 4) intraventricular and/or subarachnoid blood ¹. 

Each of these is given a score, and these scores are tallied, with the addition of 1 to the total. In other words, a completely normal appearing scan has a Rotterdam score of 1 and the worst possible score is 6.  

Classification

• basal cisterns

  • 0: normal

  • 1: compressed

  • 2: absent

• midline shift

  • 0: no shift or <= 5 mm

  • 1: shift > 5 mm

• epidural mass lesion

  • 0: present

  • 1: absent

• intraventricular blood or traumatic SAH

  • 0: absent

  • 1: present

Prognosis

In adults the mortality at six months increases with the score ¹: 

• score 1: 0%

• score 2: 7%

• score 3: 16%

• score 4: 26%

• score 5: 53%

• score 6: 61%

Note that the initial CT scan (done in the first 4 hours after admission) and not the one with the worst findings was used for estimating prognosis in the original study. Furthermore, patients with mild head injury were not included in the study ³.

Children have lower mortality in lower Rotterdam scores (scores 2 and 3), and higher mortality at higher scores (scores 4 to 6) ².