#442 - [Journal Club] - 📌 Does combining EEG and MRI improve neurodevelopmental prognostication in preterm infants?

In this episode of Journal Club, we wrap up a marathon recording session with a deep dive into the world of neonatal neuroprognostication. Daphna reviews a systematic review and meta-analysis from Pediatric Neurology that evaluates whether combining EEG and MRI provides better answers for families of preterm infants. While MRI remains a powerful tool for structural assessment, the data suggests that adding the functional insights of EEG significantly boosts specificity, particularly when predicting severe neurodevelopmental outcomes. We discuss the importance of timing these studies and the clinical value of sleep-wake cycling as a developmental milestone at the bedside.

----

Combined Use of Electroencephalography and Magnetic Resonance Imaging in the Prognostication of Neurodevelopmental Outcomes in Preterm Infants - A Systematic Review and Meta-Analysis. Forrest CD, Biagioni T, Liley HG, Lai MM, Colditz PB, Ware RS, Boyd RN, Roberts JA.Pediatr Neurol. 2026 Feb;175:116-129. doi: 10.1016/j.pediatrneurol.2025.11.005. Epub 2025 Nov 13.PMID: 41337899 

Transcript

Ben Courchia (00:00.546) Hello everybody, welcome back to the Incubator Podcast. We're back for one more episode of Journal Club this week. Daphna, you're ready to wrap this up. We have new news tomorrow, but you're wrapping up Journal Club.

Daphna Yasova Barbeau (00:08.238) I know. You were ragging on me for things going wrong, but sometimes you ask a lot of questions.

Ben Courchia (00:18.464) I'm not blaming you; I am making an assessment of the fact that our Journal Club recording has gone on for quite some time. We started recording about two and a half hours ago.

Daphna Yasova Barbeau (00:27.064) We're not done yet.

Ben Courchia (00:35.062) And we're not done yet, but it's okay. The HIE paper from the American Academy of Pediatrics obviously warranted a prolonged conversation. Where are we going for this Thursday?

Daphna Yasova Barbeau (00:47.788) To end the week, we have a paper coming out of Pediatric Neurology. We are always looking at these interesting journals that we may not always hit as neonatologists. This is entitled, "Combined Use of Electroencephalography and Magnetic Resonance Imaging in the Prognostication of Neurodevelopmental Outcomes in Preterm Infants: A Systematic Review and Meta-Analysis".

Daphna Yasova Barbeau (00:47.788) Coming out of Australia, the goal is to evaluate the predicament we have. We have these wonderful advances in neonatal and perinatal medical care. We've decreased gestation-specific morbidity and mortality associated with preterm birth, but long-term neurologic morbidity remains high. This is clear evidence of the detrimental impact of prematurity on the development of the brain.

Daphna Yasova Barbeau (00:47.788) We really lack good prognostic indicators to provide information to families. The authors write that prognostication of neurodevelopmental outcomes in infants born preterm remains one of the most challenging and simultaneously vital aspects of care in neonatology. The aim of this systematic review is to assess the prognostic potential of a combined multimodal neuro-investigative approach using both structural brain MRI and EEG to identify abnormal brain structure and function in preterm-born infants to predict neurodevelopmental outcomes.

Daphna Yasova Barbeau (00:47.788) Articles were eligible for inclusion if they were retrospective or prospective cohort studies containing a minimum of 20 infants in the final analysis. They did not include case reports, case series, or cohort studies involving fewer than 20 patients. They excluded studies where the full text could not be found, accessed, or translated into English. They included studies that had infants born preterm between 22 weeks and 36 and 6/7 weeks gestation.

Daphna Yasova Barbeau (03:02.41) The studies employed inpatient EEG and MRI with at least one EEG, either amplitude-integrated or video EEG, during the preterm time. MRI could either be completed at preterm or term-equivalent gestation, which is an interesting point for this study. Studies reporting investigations after 44 and 6/7 weeks post-menstrual age were excluded from the review.

Daphna Yasova Barbeau (03:02.41) They also excluded babies if the majority had medical diagnoses outside of prematurity, such as congenital heart disease. The outcome measure for the purposes of the review was neurodevelopment assessed at any time from 12 months corrected age using standard tests. Abnormal developmental outcome was defined as one or more of the following: cognitive, motor, or language scores of more than two standard deviations below the mean on a validated test; a diagnosis of autism spectrum disorder; developmental delay; or developmental coordination disorder.

Daphna Yasova Barbeau (03:02.41) They also included abnormal motor exams identified by clinical exam or standardized testing, a formal diagnosis of a recognized seizure disorder, and severely abnormal neurodevelopment defined as cerebral palsy or severe motor impairment, blindness, or deafness which is not correctable with hearing aids. They also included treatment-resistant epilepsy.

Daphna Yasova Barbeau (03:02.41) First, they assessed the use of EEG and amplitude-integrated EEG. Five studies used aEEG and seven used EEG. All studies performed a combination of qualitative and quantitative assessment to determine normal and abnormal outcomes. Seven of these studies, which included 952 infants, were sufficiently comparable to perform the meta-analysis. For any abnormal neurodevelopmental outcome, EEG prognosticated with an overall sensitivity of 64% and a specificity of 70%.

Daphna Yasova Barbeau (05:19.224) When they were looking specifically at severe neurodevelopmental outcomes only, this led to a similar sensitivity of 68% and specificity of 68%. They also did a sub-analysis; pooling EEG and aEEG did not significantly affect these results. Then they wanted to look at MRI.

Daphna Yasova Barbeau (05:19.224) A meta-analysis assessing the prognostic potential of MRI showed that for any abnormal developmental outcome, MRI had an overall sensitivity of 64% and a specificity of 89%. For severe neurodevelopmental outcome, the sensitivity and specificity were slightly improved at 74% and 86%, respectively. They wanted to look at potential variability in the time points because not all MRIs were obtained at the same time. They performed a subgroup analysis, and this did not result in a significant change in the prognostic variables.

Daphna Yasova Barbeau (05:19.224) The real purpose of the study was to combine the EEG and MRI. Seven of the included studies used a combination to prognosticate neurodevelopmental outcomes. Three of the studies were sufficiently homogeneous to undergo meta-analysis. The meta-analysis showed that for any abnormal developmental outcome, EEG and MRI combined prognosticated with an overall sensitivity of 70% and specificity up to 96%. When prognosticating severe outcomes only, sensitivity was up to 94% and specificity was up to 95%.

Daphna Yasova Barbeau (07:26.072) In conclusion, the systematic review assessed the combined approach of using both EEG and brain MRI to provide prognostication of abnormal neurodevelopment in preterm infants. They used 12 appropriate studies in total. Pooled sensitivity and specificity for the combination were higher than those for either EEG or MRI alone, particularly when prognosticating for severe neurodevelopmental outcomes. I think it is a step in the right direction when thinking of how we can include more information for families.

Ben Courchia (08:09.228) Wrap this up for us. I have a feeling from what you've described in the paper that we have been trying to give prognostic information to families. MRI is the best we have, but it is not perfect. Adding EEG seems to make things better, but it is still not quite there. Am I making that assessment correctly?

Daphna Yasova Barbeau (08:29.282) You nailed it. We are not very good at it yet. I do want to say I think there is potential regarding this concept about EEG versus aEEG and when we obtain those things. That may help move the curve up to improve sensitivity and specificity, but I think we just haven't nailed down what is the right time point for either of those.

Ben Courchia (08:51.744) And doing an amplitude-integrated EEG is much easier to get, whereas a full EEG is more invasive. Do you think tweaking the timing of the EEG could play a bigger role?

Daphna Yasova Barbeau (09:10.05) Absolutely. These studies did have some heterogeneity on when they got them. There are some specified time points by which we know that if you don't have sleep-wake cycling, for example, then that really prognosticates a poor outcome. You could use aEEG in the hands of somebody who really knows how to use it as a developmental milestone.

Daphna Yasova Barbeau (09:39.406) You can say the baby is not meeting this developmental milestone at the time point they should, which gives you different information than the MRI, which is really anatomic. I still think in these very little babies, we have some power to change the findings on MRI with good intervention.

Ben Courchia (09:59.618) For sure. I think it is like BPD as well; if you have to make a prediction, the later you can make that call, the more accurate you will probably be. Moving that to as close to discharge as possible, you probably will be much better at predicting outcomes on a 47-week baby that is about to be discharged than the same baby at 37 weeks.

Ben Courchia (10:29.804) Who was it that we interviewed? Terrie Inder, right? This is what she was saying. The prognostic value of these studies is not great, but it is the best value you can give to families, and they want that information.

Ben Courchia (10:48.13) Absolutely. All right, buddy, that was great. I'll see you tomorrow for new news. Thank you, everybody. See you next time.

Daphna Yasova Barbeau (10:56.687) Bye.