Seemingly vague, but potentially dangerous... common, but possibly with consequences... ...or maybe just plain frustrating. Let's talk risk stratification, diagnosis, and management. Primary or Secondary? We can make headache as easy or as complicated as we like, but let's break it down to what we need to know now, and what the parents need to know when they go home. Primary headaches: headaches with no sinister secondary cause – like tension or migraine – are of course diagnoses of exclusion (cluster headache is exceedingly rare in children). Secondary headaches: headaches due to some underlying cause -- are what we need to focus on first. The list of etiologies is vast; here is just a sampling: How do I sort this out? Ask yourself three main questions: Is it a tumor? Is it an infection? Is it a bleed? Is it a tumor? Some historical features are high-yield in screening for signs or symptoms consistent with a space occupying lesion. Progression and worsening of symptoms over time Associated vomiting Pain only in the occiput Headache that is worse with Valsalva – ask if coughing, urinating, or defecating affects the headache Does this headache wake the child from sleep? Is it worse in the morning just after getting up? Conversely, the absence of some historical features may increase suspicion of a space-occupying lesion No family history of migraine No associated aura with the headache. Who needs neuroimaging? The short answer is, if the child has an abnormal exam finding, then obtain a non-contrast head CT in the ED.  If you’re worried enough to get imaging, then you should not feel great about sending him to an expedition to MRI. The reassuring point is that for a child with a normal neuro exam, we have time to figure this out. For the recurrent headache, outpatient MRI really is the way to go if at all possible – not only do we forgo unnecessary radiation, but MRI is more likely to reveal the cause – or rule out the concern. Medina et al. in Pediatrics reported on children with headache suspected of having a brain tumor. They stratified patients into low, intermediate, and high risk, based on clinical predictors from the history and physical. All had imaging. They then calculated probability of tumor in each group. The low risk group had a 0.01% probability of tumor. The intermediate group 0.4%, and the high-risk group had only a 4% probability of tumor. The take-home message is that in the stable patient with a normal neurologic exam and no red flags, time is on our side. The American Academy of Neurology's most recent guidelines, published first in 1994 and revised in 2004. 1. Neuroimaging on a routine basis is not indicated with recurrent headaches and a normal neurologic exam 2. Neuroimaging should be considered in children with an abnormal exam. 3. Neuroimaging should be considered in children with recent onset of severe headache, change in the type of headache, or associated features that suggest neurologic dysfunction Is it an infection? This is nothing new: if you think you need to perform a lumbar puncture, then you’re right. Go after the diagnosis when it meets your threshold for testing. The difficulty is in the child who just has a headache, plus or minus symptoms that may be viral syndrome. Dr Curtis et al. in Pediatrics did a systematic review of Clinical Features Suggestive of Meningitis in Children.  In the history, only obvious features were helpful in this study: bulging fontanel in the infant or neck stiffness in the older child.  Both increased the likelihood of meningitis by 8-fold. In the physical examination, the only reliable predictors in this study were poor general appearance or a change in behavior. You will catch those cases, because you would have tuned into meningitis early on -- especially in the unvaccinated. What about all-comers with fever and headache? The presence of a high fever (so greater than 40 °C) only conferred a positive likelihood ratio of 2.9, only marginally predictiv