Does timing matter? Unlocking the value of repeated MRI in the early phases after severe traumatic brain injury

Cerebral MRI after severe traumatic brain injury (sTBI) may refine lesion characterization and support prognosis, but the optimal timing within the early course remains uncertain.

We retrospectively analyzed 17 adult sTBI patients (GCS ≤ 8) who underwent two clinically indicated 3 T MRI examinations: an early scan (≤ 72 h) and a subacute scan (day 12-14). Three board-certified neurointensivists, blinded to scan timing, independently rated lesion burden across standard sequences (T1, T2/FLAIR, SWI, DWI) and evaluated (i) comparative clinical utility of early vs subacute MRI, (ii) imaging-only prognosis (expected favorable vs unfavorable functional outcome, dichotomized by Glasgow Outcome Scale), and (iii) whether imaging findings would be expected to prompt a management change. Inter-rater agreement was quantified by ICC (lesion counts) and Fleiss' kappa (categorical ratings). The patient (MRI-pair) was the statistical unit; rater-level ratings were used for reliability estimation, not as independent observations.

At the patient level, early and subacute MRI were judged equally useful in 12/17 (70.6%) cases (Fleiss' kappa for pairwise utility rating: 0.33). Lesion burden demonstrated substantial overlap between time points; paired comparisons of rater-averaged lesion counts did not indicate a systematic difference between early and subacute MRI across T1, T2/FLAIR, SWI, or DWI (all p > 0.05). Monitoring-related artifacts were frequent but rarely reduced overall interpretability.

In this selected cohort able to undergo two MRIs within the early course of sTBI, early and subacute MRI provided largely concordant information. MRI timing should be individualized based on clinical stability and the specific diagnostic question; routine repetition within the first two weeks may not be necessary.