Introduction: More than half of ischemic stroke patients present as minor strokes (NIHSS<6). A lack of thrombolysis guidelines for this population leads to untreated strokes and erroneously treated stroke mimics, producing adverse outcomes. Data suggest that whole brain CT perfusion (WB-CTP) improves detection of ischemia, offering a potential method of ameliorating diagnostic uncertainty in these patients.
Hypothesis: WB-CTP can guide clinical decisions by identifying patients with ischemic episodes that would benefit from thrombolysis or early intervention.
Methods: This retrospective chart review enrolled 524 consecutive patients receiving WB-CTP with a Toshiba 320 detector scanner between 08/2008 and 06/2015, for acute stroke less than 6 hours from onset and NIHSS<6, and who showed no evidence of intracranial hemorrhage. Patients were excluded for non-diagnostic (n=25) or unreported (n=8) scans and non-ischemic findings (7). For diagnostic accuracy calculations, the reference standard was the final clinical impression suggesting ischemic events, as only 52% had follow-up imaging. Subgroup analyses were performed in patients receiving follow-up imaging.
Results: A total of 484 patients (age 17-101, 54% men, mean NIHSS 2.47) were included. Follow-up imaging was performed in 251 patients; 150 underwent MRI with diffusion-weighted imaging (DWI). A summary of diagnostic accuracy values is shown in table below. WB-CTP is highly specific with a high positive predictive value in all groups and has moderate to high negative predictive value. Positive and negative likelihood ratios were 21.24 and 0.5 in the whole group analysis.
Conclusions: Positive WB-CTP findings may warrant early intervention, including thrombolysis, while negative findings alone are not a sufficient basis upon which to confidently withhold interventions.