Imaging for Neuroprognostication After Cardiac Arrest: Systematic Review and Meta-analysis
Predicting neurological outcome in comatose survivors of cardiac arrest relies on clinical findings, radiological and neurophysiological test results. To evaluate the predictive accuracy of brain computed tomography (CT) and magnetic resonance imaging (MRI) for prognostication of neurological outcomes after cardiac arrest.
We searched MEDLINE (database inception to August 2018) and included all observational cohort studies or randomized controlled trials including adult (16 years or older) survivors of cardiac arrest which evaluated the diagnostic accuracy of CT or MRI for predicting neurologic outcome or mortality. Study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 tool. All review stages were conducted independently by 2 reviewers, and where possible data were pooled using bivariate meta-analysis. The main outcome was to evaluate the of accuracy of CT and MRI in neuroprognostication of patients after cardiac arrest.
We included 44 studies that examined brain CT (n = 24) or MRI (n = 21) in 4008 (n per study, 9-398) patients. Decreased grey to white matter ratio on CT (20 studies) was useful for predicting poor neurological outcome (sensitivity 0.44, 95% CI 0.29-0.60; specificity 0.97, 95% CI 0.93-0.99; positive likelihood ratio [LR+] 13.8, 95% CI 6.9-27.7). Similarly, diffusion-weighted imaging (DWI) on MRI (16 studies; sensitivity 0.77, 95% CI 0.65-0.85; specificity 0.92, 95% CI 0.85-0.96; LR+ 9.2, 95% CI 5.2-16.4) and DWI and fluid-attenuated inversion recovery (FLAIR) MRI (4 studies, sensitivity 0.70, 95% CI 0.43-0.88; specificity 0.95, 95% CI 0.79-0.99; LR+ 13.4, 95% CI 3.5-51.2) were useful for predicting poor neurological outcomes. We found marked heterogeneity in timing of radiological examinations and neurological assessments relative to the cardiac arrest.
Decreased grey to white matter ratio on CT and DWI or DWI and FLAIR on MRI are useful adjuncts for predicting poor early neurological outcome after cardiac arrest.
KeywordsCardiac arrest Computed tomography imaging Magnetic resonance imaging Brain anoxia Prognosis
CLS, LD, CCH, AK, RP, NKJA, DCS contributed to the concept and design of study. CLS and LD contributed to the acquisition of data. CLS, LD, CCH, AK, RP, NKJA, DCS contributed to the analyses and interpretation of data. CLS, LD, CCH, AK, RP, NKJA, DCS contributed to the drafting and critical revision of manuscript. NKJA and DCS contributed to the supervision of study. NKJA and DCS contributed equally to this study.
Source of Support
No source of support to declare.
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval and Informed consent
No ethical approval or informed consent needed for this study.
- 21.Macaskill PGC, Deeks JJ, Harbord RM, Takwoingi Y. Chapter 10: Analysing and presenting results. In: Deeks JJ BP, Gatsonis C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 10. The Cochrane Collaboration 2010. http://srdta.cochrane.org/.
- 22.Takwoingi Y. MetaDAS: a SAS macro for meta-analysis of diagnostic accuracy studies. Quick reference and worked example Version 1.3. 2010 July. http://srdta.cochrane.org.
- 43.Lee BK, Jeung KW, Lee HY, Jung YH, Lee DH. Combining brain computed tomography and serum neuron specific enolase improves the prognostic performance compared to either alone in comatose cardiac arrest survivors treated with therapeutic hypothermia. Resuscitation. 2013;84:1387–92.CrossRefGoogle Scholar