Skip to main content
. 2021 Mar 18;4(3):e210994. doi: 10.1001/jamanetworkopen.2021.0994

Table 2. Comparison of Lesions Visible on MR1 vs MR2a.

Abnormality Scan finding, No. (%) of patients Raw P value
MR1 positive MR1 negative
MR1 positive MR2 MR1 negative MR2
Lesion persists Lesion resolved Remains negative Shows new lesion
Any 34 (100) 31 (91) 3 (9) 47 (100) 47 (100) 0 .25
Mass effect
Mass >25 mL 2 (100) 2 (100) 0 79 (100) 79 (100) 0 NC
Midline shift 1 (100) 1 (100) 0 80 (100) 80 (100) 0 NC
Cisternal compression 1 (100) 1 (100) 0 80 (100) 80 (100) 0 NC
Intra-axial
Contusion 20 (100) 20 (100) 0 61 (100) 61 (100) 0 NC
Traumatic axonal injury 21 (100) 20 (95) 1 (5) 60 (100) 60 (100) 0 >.99
Extra-axial
Hemorrhage
Epidural 3 (100) 3 (100) 0 78 (100) 78 (100) 0 NC
Subdural 8 (100) 7 (88) 1 (12) 73 (100) 71 (97) 2 (3) >.99
Subarachnoid 14 (100) 5 (36) 9 (64) 67 (100) 66 (99) 1 (1) .03
Other
Skull fracture 0 0 0 81 (100) 81 (100) 0 NC
Intraventricular hemorrhage 10 (100) 2 (20) 8 (80) 71 (100) 71 (100) 0 .01

Abbreviations: MR1, first magnetic resonance scan after injury; MR2, second magnetic resonance scan after injury; NC, not calculated.

a

Eighty-one patients received a magnetic resonance scan within 72 hours (MR1) and at 2 to 3 weeks after injury (MR2). Lesions visible on the 2 scans were compared using the McNemar test for paired categorical data. Where reports were identical, no P value is shown. None of the P values was significant, assuming a 5% false discovery rate.