Maayke Hunfeld

34 Chapter 2 1 up to 14 d post CA) (30, 32-34). If MRI is obtained too soon after CA (within 2 d after CA), DWI changes might not be visible (31, 35). In addition if imaging is obtained after 7 days, pseudo-normalization can occur (35). Neuroimaging outside this time window might have influenced DWI results. Head CT Two retrospective studies were included in this review including children after CA (n=78 and n=64) (Supplemental Table 6, Supplemental Digital Content 2, http://links. lww.com/PCC/B352 ) (36, 37), concluding that patients with a normal early CT (median time 3.3 hr post ROSC, IQR 1-6 hr) survived with good outcome (36). Loss of gray-white matter differentiation, in particular in the basal ganglia and effacement of the basilar cistern, ambient cistern and sulcal effacement were significantly associated with poor outcome (36, 37). However, both studies were single site studies and retrospective. Similar to most MRI studies, CTs were obtained if clinically indicated, creating a selection bias, and the results might have influenced the medical team in the decision to withdraw or continue treatment. Finally PCPC was used as outcome scale in both studies, determined at hospital discharge. Evoked potentials This review contains two observational prospective studies (Supplemental Table 7 , Supplemental Digital Content 2, http://links.lww.com/PCC/B352 ). Carter et al described SEP findings in a cohort of 105 children with severe brain injury (HIE n=38, unclear how many children after CA) performed within 7 days after the injury (38). At 5 years follow-up using Glasgow outcome scale (GOS) and Health Utilities Index Mark 1 (HUI:1, a questionnaire to assess quality of life ) as outcome measurements, a normal SEP had a PPV for a good outcome of respectively 85.4% (sensitivity 61.2%, specificity 87.8%) and 85.4% (sensitivity 68.6%, specificity 88.9%); A bilateral absent SEP had a PPV of respectively 90.9% (sensitivity 61.2%, specificity 94.6%) and 93.9% (sensitivity 57.4%, specificity 96.1%). Mandel et al performed a SEP in 57 HIE patients within 5 days after event. All children with a bilateral absent SEP had a poor outcome (severe disability, persistent vegetative state or death in the third year after event) (p=0.001). Of 13 patients with a normal SEP, three had poor outcome. BAEP did not have a predictive value (13). Both studies were single center and had heterogeneous inclusion criteria, including children after CA. However, the exact number of CA patients was unclear and no subanalysis was done for this subgroup. Although the strength in Carters cohort was