Is MRI of the brain indicated for a patient with a typical primary headache (migraine or tension‑type) who is stable and has no neurological signs or red‑flag features?

MRI for Typical Primary Headache Without Red Flags

MRI is not indicated for patients with typical migraine or tension-type headache who have a normal neurological examination and no red-flag features. 1, 2, 3

Evidence-Based Rationale

The yield of clinically significant intracranial pathology in migraine patients with normal neurological examinations is only 0.2% (2/1,086 patients)—identical to the 0.4% rate found in completely asymptomatic volunteers. 1, 2, 3 For tension-type headache with normal examination, the rate is 0% (0/83 patients). 3 This means routine imaging adds no diagnostic value and exposes patients to the substantial risk of false-positive findings that trigger harmful cascades of unnecessary procedures. 3

Documented Harms of Low-Yield Imaging

• False positives vastly outnumber true positives in low-risk populations, leading to invasive follow-up procedures. 3
• One documented case: an incidental asymptomatic aneurysm discovered during imaging for tension-type headache led to endovascular treatment that caused permanent hemiparesis. 3
• White matter abnormalities appear in 33.3% of chronic headache patients versus 7.4% of controls on MRI, but these lesions have no proven causal relationship to headache and do not alter management. 4

When MRI Is Indicated: Red-Flag Assessment

Perform MRI brain without contrast if any of the following are present:

High-Priority Red Flags (Immediate Imaging)

• Abnormal neurological examination (likelihood ratio [LR] = 5.3) 1, 3
• Headache awakening patient from sleep (LR ≈ 3.0) 1, 2, 3
• Thunderclap headache (maximal intensity within seconds) 1, 5

Moderate-Priority Red Flags (Imaging Strongly Considered)

• Headache worsened by Valsalva maneuver, coughing, or exertion (LR = 2.3) 1, 3
• Progressively worsening headache over weeks to months 1, 2, 3
• New-onset headache after age 50 1, 2, 3
• Atypical headache pattern (LR = 3.8; abnormality rate 14.1%) 3
• Rapid increase in headache frequency 1, 2
• Recent head or neck trauma 1
• Focal neurological symptoms or signs 1, 3, 5
• Unexplained fever with neck stiffness (meningitis) 1, 5
• Altered consciousness, memory, or personality 1

Clinical Decision Algorithm

Step 1: Perform a thorough neurological examination, including fundoscopy for papilledema. 3

Step 2: Systematically screen for red-flag features using the history above. 3

Step 3: If examination is normal and no red flags are present:

• Do not order neuroimaging. 1, 2, 3
• Proceed with standard migraine or tension-type headache management. 1, 2
• Document clinical reasoning to mitigate medicolegal concerns. 3

Step 4: If any red flag is present or examination is abnormal:

• Order MRI brain without IV contrast as the first-line study. 2, 3
• MRI is superior to CT for detecting masses, inflammatory lesions, demyelination, and small infarcts. 2, 3
• Standard sequences: T1-weighted, T2-weighted, FLAIR, and diffusion-weighted imaging. 2

Step 5: Reserve non-contrast CT only for:

• Suspected subarachnoid hemorrhage presenting within 6 hours (sensitivity 95% on day 0). 1, 2
• Acute trauma or when MRI is unavailable in emergency settings. 2

Common Pitfalls to Avoid

• Do not image "sinus headache"—this is a common misdiagnosis of migraine and does not warrant neuroimaging. 2
• Do not dismiss medication-overuse headache: ≥15 headache days/month with acute medication use ≥10–15 days/month for >3 months mimics chronic primary headache but requires withdrawal, not imaging. 1
• Do not order CT when MRI is appropriate—CT misses the majority of structural lesions responsible for secondary headache. 2
• Do not skip the neurological examination: 94% of patients with brain tumors causing headache have abnormal neurological findings at diagnosis. 2

Strength of Evidence

These recommendations are Grade B (American Academy of Family Physicians, American Academy of Neurology) and are derived from large cohort studies demonstrating that imaging in low-risk headache patients yields no benefit over clinical assessment alone. 1, 2, 3