Orthopedics · Part Four · Seeing and Ruling Out
Lesson 38 / 44
When a Headache Is Dangerous: Red Flags, Thunderclap Onset, and the Subarachnoid Decision Rules
Most headache is primary and benign. The clinical skill that matters is the same one a trained clinician applies to every headache. That skill decides whether the pain is the disorder itself or a signal of something rarer that must be caught and referred.
A dangerous headache is a secondary headache, meaning the head pain is produced by another condition such as bleeding, infection, a mass lesion, or an inflamed artery. Most headache is primary and benign. After a new headache in primary care, the one-year risk of a malignant brain tumor was 0.15 percent. The Unified Model of Tone reads screening as assessment doing its highest-value work, because delay becomes its own kind of harm.
One-year risk of a malignant brain tumor after a new undifferentiated headache
0.15 percent
Ottawa subarachnoid hemorrhage rule, sensitivity against specificity
100 percent against 15.3 percent
Thunderclap headaches that proved to be subarachnoid hemorrhage
11.3 percent of 137 patients
Hemorrhage patients not diagnosed at their first medical contact
56 of 482, or 12 percent
Thunderclap headache
A thunderclap headache reaches its worst intensity instantly, which is why a clinician asks how many seconds passed between nothing and unbearable. Emergency research usually enrolls a wider group, meaning any nontraumatic headache that peaks within one hour of onset. Speed of onset, rather than severity, defines the category.
Why an abrupt onset changes the reasoning
Blood escaping into the subarachnoid space spreads across the surface of the brain and irritates the meninges within seconds. That anatomy sits behind the features the decision rules score: neck pain, limited neck flexion on examination, and onset during exertion when arterial pressure surges. The tempo of a headache reports the speed of the event that produced it.
01Primary against secondary headache
Most headache is primary, and screening asks which kind this one is
Most headache is primary and benign, which means the headache is the disorder itself. Tension-type headache, migraine and cluster headache make up the overwhelming majority of what any clinician sees. In these the nervous system carries the tendency to generate the pain.
Secondary headache works the other way. The head pain is a symptom of another condition such as infection, hemorrhage, a mass lesion, or inflammatory disease. A review of the flags used to catch them states that only a minority of headache patients carry one (Do 2019).
The clinical value of a primary-contact clinician lies in knowing this distinction cold, screening every headache against it, and reassuring the large majority whose pattern is unmistakably benign.
What happens to a new headache over the following year
Primary care records size the risk. Kernick and colleagues followed 21,758 primary and 63,921 undifferentiated headaches through the year after presentation (Kernick 2008). The one-year risk of a malignant brain tumor was 0.15 percent after an undifferentiated headache.
After a headache already labeled primary the risk was 0.045 percent. Age moved it. Above 50 the undifferentiated figure reached 0.28 percent. The likelihood ratio for subarachnoid hemorrhage after an undifferentiated headache was 29, and it also rose with age.
Those numbers are why screening reassures. The clinician who runs the screen can hand a patient figures rather than atmosphere.
02Findings
What the research shows
From a primary care cohort, a childhood tumor meta-analysis, and five hospital cohorts.
03The tumor patients fear
A brain tumor almost never presents as an isolated headache
Dangerous secondary causes of headache are uncommon, and a clinician screens for them first precisely because they are rare rather than because they are likely. Forsyth and Posner examined 111 consecutive patients with primary or metastatic brain tumors (Forsyth 1993). Headache was present in 48 percent.
The headache itself was ordinary, resembling tension-type headache in 77 percent of cases and migraine in 9 percent. It was the worst symptom in only 45 percent of patients, and the classic early morning brain tumor headache proved uncommon.
What marks a tumor headache is the company it keeps. Nausea, vomiting, an abnormal neurological examination, or a significant change in a prior headache pattern are what suggest a tumor. Nausea or vomiting accompanied 40 percent, and bending over worsened 32 percent.
Recognizing that the neurology tends to precede the headache is what lets a clinician reassure rather than alarm.
The same principle in children
Children show it with different signs. A systematic review of 74 papers and 4,171 children ranked the presenting features of central nervous system tumors by site (Wilne 2007). For posterior fossa tumors, abnormal gait and coordination appeared in 60 percent and papilledema in 34 percent.
Brainstem tumors were louder in the motor system and quieter in pain. Abnormal gait and coordination appeared in 78 percent and cranial nerve palsies in 52 percent, with headache in only 23 percent. Supratentorial tumors produced seizures in 38 percent.
Symptoms of raised intracranial pressure were absent in more than half of the children. Several signs occurring together is what prompts a clinician to look further.
Periodic episodes of dizziness, abdominal pain or vomiting in an otherwise well child belong elsewhere. Migraine vs Seizure carries the childhood migraine equivalents, including benign paroxysmal vertigo and abdominal migraine, and the time course that separates them from epilepsy.
04Four features that change the plan
Four features move a headache from routine to investigated
A new headache appearing in later adulthood, a sudden change in a familiar pattern, focal neurological signs, and cognitive change are the four features that move a headache from routine to investigated.
Longstanding primary headache carries a history stretching back years, so a genuinely new headache in an older adult is a different object. The SNNOOP10 list scores onset after 65 as an item in its own right, alongside pattern change or the recent onset of a new headache (Do 2019).
Focal signs such as hearing loss, vision change, weakness, coordination difficulty or speech trouble point beyond a typical primary headache. So does cognitive change, meaning confusion, memory disturbance, or a shift in personality that family members often notice first.
Position, straining and the pressure signs
A headache that wakes a person from sleep, or worsens with bending, straining or a change in position, reads as a pressure problem. Raised intracranial pressure behaves that way. Positional headache, headache precipitated by coughing or exercise, papilledema, and progressive headache each sit on the same list as separate flags (Do 2019).
The authors leave open how much any single flag raises the odds on its own, because the low incidence of most secondary headaches has kept the prospective studies small. The list works as a screen, and a clinician uses it to decide who needs a scan.
That decision is reasoned first, then referred out and read alongside the history and the examination. What MRI Is Really For sets out when an image changes what happens next.
The first job with any headache is to decide whether the headache is the disorder itself or a symptom of something else. The reassuring truth is that the dangerous causes are rare and the trained clinician is built to catch them.
05Thunderclap onset
Abrupt onset is a tempo, and the tempo is what triggers the workup
A subarachnoid hemorrhage classically arrives as a thunderclap headache that patients describe as the worst of their life. Speed is the feature that matters. Landtblom and colleagues followed 137 consecutive patients with sudden onset headache through CT, cerebrospinal fluid examination and up to twelve months of follow-up (Landtblom 2002).
The incidence was 43 per 100,000 adults per year, and 11.3 percent of those patients had a subarachnoid hemorrhage. Other findings included cerebral infarction, intracerebral hematoma, aseptic meningitis and sinus thrombosis. No specific cause was found in the majority.
The conclusion of that study is the reason decision rules exist. Attacks caused by a hemorrhage could not be separated from the rest on clinical grounds. Nausea, neck stiffness, occipital location and impaired consciousness were more frequent with hemorrhage, and none occurred in every case.
Recurrence turned out to be reassuring. Repeat thunderclap attacks occurred in 24 percent of the patients without hemorrhage, and none of that group bled later. A first thunderclap attack is the one that earns urgent evaluation.
Why worst headache of my life is a weak discriminator
The phrase is nearly universal in the population being screened. Among 3,132 emergency patients evaluated for possible subarachnoid hemorrhage, 2,571 of them, or 82.1 percent, described the worst headache ever (Perry 2011). Of that whole group, 7.7 percent had a hemorrhage.
Intensity therefore sorts almost nobody. The formal rules score onset, exertion, neck behavior and age instead of how bad the pain feels.
06The subarachnoid decision rules
The Ottawa rule catches every hemorrhage and clears very few patients
The Ottawa subarachnoid hemorrhage rule identifies which alert, neurologically intact adults with sudden headache need no testing at all. Perry and colleagues studied 2,131 such adults whose headache peaked within one hour (Perry 2013). Of those, 132, or 6.2 percent, had a hemorrhage.
The rule scores six features. They are age 40 or older, neck pain or stiffness, witnessed loss of consciousness, onset during exertion, thunderclap headache meaning instantly peaking pain, and limited neck flexion on examination. Any single one calls for investigation.
A four-item version reached 98.5 percent sensitivity. Adding thunderclap onset and limited neck flexion produced 100 percent sensitivity, with a confidence interval of 97.2 to 100. Specificity was 15.3 percent.
That second number is the price of the first. A rule built to miss nothing must send most of the people it screens onward, because a missed hemorrhage costs immeasurably more than an unnecessary scan. The authors also state the boundary: the findings apply only to patients with those clinical characteristics.
What happened when the rule was used prospectively
Implementation across six Canadian emergency departments enrolled 3,672 patients, 188 of whom had a hemorrhage (Perry 2020). The rule stayed 100 percent sensitive, and lumbar puncture use fell from 38.9 percent to 25.9 percent.
Timing changes what a normal scan means. Among 3,132 patients, overall CT sensitivity for hemorrhage was 92.9 percent. Among the 953 scanned within six hours of onset, all 121 hemorrhages were identified (Perry 2011).
Prospective use of that six-hour rule was less perfect. It was 95.5 percent sensitive and missed five hemorrhages (Perry 2020). One was a radiology misread, two were incidental aneurysms, one had a nonaneurysmal cause, and one patient was profoundly anemic.
Both numbers belong here. The six-hour window is powerful and it is not absolute, so the decision stays with the whole clinical picture.
What a delay actually costs
Kowalski and colleagues followed 482 patients with subarachnoid hemorrhage (Kowalski 2004). Fifty-six, or 12 percent, were not diagnosed at their first medical contact, rising to 19 percent among the 221 who arrived with normal mental status.
Migraine or tension headache was the incorrect label in 36 percent of those cases, and failure to obtain a CT scan was the error in 73 percent. Twelve of the misdiagnosed patients rebled before the diagnosis was corrected.
Among those who had looked well at first contact, misdiagnosis was associated with worse quality of life at three months and a higher risk of death or severe disability at twelve. The people most exposed to a delay are the ones who present most reassuringly.
07Vision, vessels and meninges
Three other emergencies each announce themselves in a different way
Temporal arteritis threatens vision, and the window for protecting it is measured in days. It is an inflammatory disease of medium and large arteries, marked by localized pain over the temple, a tender or thickened temporal artery, and an association with polymyalgia rheumatica. Gonzalez-Gay and colleagues studied 239 patients with biopsy-proven giant cell arteritis (Gonzalez-Gay 1998). Visual involvement occurred in 69 of them, and 34 had permanent visual loss.
The interval to involvement of the second eye was 5 days. After adjustment for the corticosteroid regimen, treatment started within the first day of visual loss was the only predictor of improvement in acuity. Transient visual loss and jaw claudication predicted permanent loss.
That is what makes rapid referral protective. In primary care records, temporal arteritis was the most common of the serious outcomes studied after a new undifferentiated headache, at 0.66 percent (Kernick 2008).
Cervical artery dissection arrives as pain first
Cervical artery dissection announces itself with head or neck pain before anything else. Silbert and colleagues described 161 consecutive patients with spontaneous carotid or vertebral dissection (Silbert 1995). Headache was reported by 68 percent of the carotid group and 69 percent of the vertebral group.
When present it was the initial manifestation in 47 percent of carotid and 33 percent of vertebral dissections. The median interval from headache onset to other neurological manifestations was 4 days for carotid and 14.5 hours for vertebral dissection. Neck pain accompanied 26 percent of carotid and 46 percent of vertebral cases.
New head or neck pain in a person developing neurological symptoms is worth taking at face value, because the pain arrives inside a short window. Those figures describe how a dissection presents. The series was not designed to answer what precipitates one.
The meningitis triad is less reliable than its reputation
The classic triad of fever, neck stiffness and altered mental status was present in only 44 percent of 696 episodes of community-acquired bacterial meningitis (van de Beek 2004). Waiting for all three would miss more than half of cases.
Ninety-five percent had at least two of headache, fever, neck stiffness and altered mental status. On admission 33 percent had focal neurological abnormalities. Overall mortality was 21 percent.
A slower tempo carries its own flags. Posttraumatic onset of headache and onset after 65 are both SNNOOP10 items (Do 2019). Together they describe the older adult whose headache and confusion build over weeks after a fall, which is how a subdural hematoma behaves. These are the exceptions a trained eye catches and refers.
08Claims removed from this page
Five claims from the earlier version were removed
The earlier text said that untreated retinal artery involvement in temporal arteritis can cause blindness within twenty-four hours. No source could be found for that figure, so it came off. The measured findings replace it, meaning 34 permanent visual losses among 239 patients and a 5-day interval to the second eye (Gonzalez-Gay 1998).
Three more figures went for the same reason. The claim that most adult tumors sit above the tentorium while childhood tumors sit below it had no source, and the ranked features by site replace it (Wilne 2007). So did the claim that roughly half of childhood dizziness traces to migraine, and the childhood migraine prevalence figures, which belong to Migraine vs Seizure.
The six-month threshold given for what counts as a new headache also had no source. The SNNOOP10 items carry that job instead (Do 2019).
09The model on dangerous headache
What the Unified Model of Tone claims about recognizing a dangerous headache
Everything above is established science, including the five hemorrhages the six-hour rule missed. What follows is this model’s reading, stated as ours rather than drawn from the papers cited.
Tone is the integrated organization through which the body’s interacting processes relate to one another at a given moment. Our model holds that assessment, and not delivery, is the true seat of accuracy. A decision rule is that principle written down.
Our model also holds that a disturbance surfaces wherever a body is least able to absorb it. The pain is localized where capacity ran out, which is not always where the process began. That is why 137 patients with the same abrupt onset could not be separated on clinical grounds (Landtblom 2002).
Why escalation is the correct dose
Our model places every intervention in medicine on one continuous axis of magnitude, running from the lightest sustained touch to the most invasive surgery. Once a distortion has descended past what any surface input can reach, the larger magnitude is the correct one, and delay becomes its own kind of harm.
The misdiagnosis data measure that harm directly (Kowalski 2004). Recognizing the pattern and moving immediately is the portal-of-entry role at full strength.
The reassuring reality holds across ages. Dangerous headache is genuinely rare, most headache is primary, and a clinician who can read an amplified protective state separates the common and benign from the uncommon exception with confidence.
The prediction
From that follows a claim the headache literature does not make. Repeat thunderclap attacks that resolve without bleeding, seen in 24 percent of the patients without hemorrhage, mark a regulatory state rather than a vessel lesion (Landtblom 2002). Our model predicts that state is measurable between attacks.
Four measures recorded together in the same people will share one underlying factor rather than varying independently. They are cerebrovascular reactivity by breath-hold index, resting heart rate variability, pressure pain threshold at the upper trapezius, and recovery time after an orthostatic challenge.
This is a claim about how a dangerous headache is organized rather than a claim about what treatment does. If breath-hold cerebrovascular reactivity, resting heart rate variability, pressure pain threshold and recovery time after an orthostatic challenge are shown to move together, the unification claim is confirmed.
10The tone reading
How a dangerous headache expresses tone
Every topic in this library expresses all of tone. In dangerous headache three aspects carry the signature, because the tempo of the onset and the behavior of the neck carry more information than the severity of the pain.
Time course
Onset time carries the diagnosis. A thunderclap peaks instantly, while carotid dissection pain preceded other neurological signs by a median of four days.
Coupling
Dangerous headache travels with company. In bacterial meningitis, 95 percent of 696 episodes carried at least two of headache, fever, neck stiffness and altered mental status.
Constraint
Blood in the subarachnoid space restricts neck movement. Limited neck flexion on examination is one of the six features the Ottawa rule scores.
The remaining foundations run through dangerous headache as well. Input quality: a decision rule raises the quality of what the examination reads. Prediction: a headache that breaks a familiar pattern violates the body’s own forecast, which is why pattern change scores as a flag. Set point: the level at which a person calls a headache the worst of their life differs between bodies, which is why 82.1 percent of screened patients used the phrase. Gain: meningeal irritation raises the response to ordinary neck movement. Load: exertion raises arterial pressure, and onset during exertion is scored for that reason. Oscillation: attacks that repeat and resolve report a rhythm rather than a rupture. These are readings of one organization rather than separate systems, which is the core claim of the Unified Model of Tone.
11Across the library
How this page relates to the rest of the library
Screening decides which of the neighboring pages a headache actually belongs to.
Cervicogenic headache, tension-type headache and migraine told through trigeminocervical convergence, once the dangerous causes are off the table.
The childhood migraine equivalents and the time course that separates a benign episode from epilepsy.
The same screening logic in the spine, where the accuracy of individual flags has been measured directly.
The other emergency where time to decompression decides the outcome.
Escalation as correct dosing, and the thresholds that make an immediate referral the expert move.
The structures whose signs arrive before the pain when a lesion sits in the posterior fossa.
Headache as a measurable regulatory state, with the instruments used to read it.
12Frequently asked
Questions patients ask about dangerous headaches
When is a headache a warning sign?
A clinician looks for four features. They are a new headache appearing in later adulthood, a sudden change in a familiar pattern, focal neurological signs such as vision or speech trouble, and cognitive change. A headache that peaks instantly, one that wakes a person from sleep, and one that worsens with bending or straining also move a case forward. All of these are uncommon. After a new headache in general practice, the one-year risk of a malignant brain tumor was 0.15 percent.
Are most headaches dangerous?
No. The great majority are primary headaches such as migraine, tension-type headache and cluster headache, and these are benign. A review of the flags used to catch secondary causes states that only a minority of headache patients have one. Even in an emergency department, among 3,132 patients being evaluated specifically for possible bleeding around the brain, 7.7 percent had it. Screening exists to find the rare serious cause early and to let everyone else be reassured with real figures.
What is a thunderclap headache?
A thunderclap headache reaches its worst intensity instantly, so a clinician asks how many seconds passed between nothing and unbearable. It is the classic presentation of a subarachnoid hemorrhage. Speed rather than severity defines it. In a prospective series of 137 consecutive patients with sudden onset headache, 11.3 percent had a hemorrhage, at an incidence of 43 per 100,000 adults per year. Repeat attacks occurred in 24 percent of the patients who had no hemorrhage, and none of them later bled.
What is the Ottawa subarachnoid hemorrhage rule?
It is a screening rule for alert, neurologically intact adults whose headache peaked within one hour. It scores six features: age 40 or older, neck pain or stiffness, witnessed loss of consciousness, onset during exertion, instantly peaking pain, and limited neck flexion on examination. Any one of them calls for investigation. In 2,131 patients the rule was 100 percent sensitive and 15.3 percent specific. It was designed to miss nothing, which is exactly why it clears so few of the people it screens.
Can a normal CT scan rule out bleeding around the brain?
Timing decides how much a normal scan settles. Among 3,132 emergency patients, overall CT sensitivity for subarachnoid hemorrhage was 92.9 percent. Among the 953 scanned within six hours of onset, all 121 hemorrhages were found. When that six-hour rule was later applied prospectively in a cohort of 3,672 patients, sensitivity was 95.5 percent and five hemorrhages were missed. A scan performed early is powerful, and the clinical picture and the timing of the scan still carry part of the decision.
What is temporal arteritis?
Temporal arteritis, also called giant cell arteritis, is inflammation of an artery near the temple that can threaten vision. It is one of the specific causes screened for in older adults with a new headache. Among 239 patients with biopsy-proven disease, 69 had visual involvement and 34 lost vision permanently. The interval to the second eye was 5 days, and corticosteroid treatment begun within the first day of visual loss was the only predictor of improvement in acuity.
What does the Unified Model of Tone say about a dangerous headache?
That assessment, rather than delivery, is the true seat of accuracy, so a decision rule is skilled reading written down. The model places every intervention on one continuous axis of magnitude, and holds that delay becomes its own kind of harm once the picture calls for a larger one. From that the model predicts that cerebrovascular reactivity, resting heart rate variability, pressure pain threshold and recovery time after an orthostatic challenge share one underlying factor between attacks rather than varying independently.
13The sources
References
12 primary sources, each linked to its record. Figures quoted on this page were checked against the published abstract.
Related evidence