Role of Computed tomography brain scan in complex febrile seizure: Number Needed to Screen (NNS)

Every year, millions of young children experience a sudden, frightening event: a seizure triggered by a fever. While most of these episodes are brief and harmless, a significant portion are more complex. These complex cases might involve a seizure that lasts longer than usual, one that affects only one side of the body, or a series of seizures occurring within a single day. When a child arrives at a hospital with such symptoms, doctors face a difficult choice. They must decide whether to order a brain scan to look for hidden dangers, such as a tumor or an infection deep inside the skull. On one hand, missing a serious condition could be catastrophic. On the other hand, these scans often expose young children to radiation and may require sedation, a chemical sleep that carries its own risks. For decades, medical guidelines have been clear for simple cases, but for these complex situations, the path forward has remained murky, leaving families and doctors to wonder if the scan is truly necessary or just a precautionary measure.

A team of researchers at United Christian Hospital in Hong Kong set out to bring clarity to this dilemma by looking back at real-world cases. They reviewed the medical records of children aged six months to five years who had been admitted with complex febrile seizures between 2019 and 2024. The goal was not just to see how many scans showed something wrong, but to understand exactly which children were the ones with those problems. By studying 117 children who had undergone a computed tomography, or CT, scan—a type of imaging that uses X-rays to create detailed pictures of the brain—the team could calculate a specific metric known as the number needed to screen. This figure tells us how many children must be scanned to find just one person with a significant abnormality. Without any special filters, the researchers found that for every 15 children scanned, only one had an abnormal result. This low rate suggests that scanning every single child with a complex seizure is an inefficient use of resources and exposes many children to unnecessary radiation.

However, the story changed dramatically when the researchers looked closer at the details of the children's conditions. They discovered that two specific factors acted as powerful signals for trouble. The first was the length of the seizure; if a child's seizure lasted longer than 30 minutes, the likelihood of finding an abnormality on the scan increased sharply. The second factor was developmental delay, meaning the child had already shown signs of falling behind in their growth or learning milestones. When the researchers applied a model using these two factors to decide who should be scanned, the results were striking: the model successfully identified all eight children with abnormal scans. In this targeted approach, the number of children needed to scan to find one problem dropped from 15 down to just 2. This means that if a doctor sees a child with a complex seizure who also has a prolonged duration or a history of developmental delay, the scan is highly likely to reveal something important. Conversely, if a child has a complex seizure but does not have these specific risk factors, the scan is almost certainly going to be normal.

The study did not find that other common concerns, such as the specific type of virus causing the fever or whether the child had a family history of seizures, were reliable indicators for needing a scan. In fact, the data showed that children with focal neurological signs, where a doctor notices a specific weakness or abnormality during an exam, did not necessarily have abnormal scans in this particular group, though the researchers noted that examining young children can be difficult and such signs are sometimes hard to spot. The researchers concluded that by using these two clear predictors—seizures lasting over 30 minutes and developmental delays—hospitals could avoid scanning over 100 children who would likely have normal results, while still catching every single case of a hidden brain problem. This approach does not eliminate the need for scans, but it refines them, ensuring that the technology is used where it matters most. By focusing on the children who truly need it, medical teams can protect young patients from unnecessary radiation while ensuring that serious conditions are not missed, turning a broad, uncertain practice into a precise, evidence-based strategy.