Headaches in Young People and adults

This is a monster of a guideline from NICE, which is only aimed at people over the age of 12. This email contains the summary, but if you want to read more, then use the link at the bottom.


My main feeling about this guideline, having trawled though the 360 pages of evidence, is that it is more a summary of what we currently know, than a guideline. They exclude commonly used treatments like amitriptyline and pizotifen from the guideline, not because there is evidence that they are ineffective, but because the studies just haven’t been done yet. A lot of the evidence that is available, is of poor quality and is often contradictory.

From my personal point of view, this guideline will not be changing my practice much.

Having said that – there are some useful bits and I will try to pull these out without getting too bogged down in the evidence.

A summary (with my own interpretation, if you don’t want to read through the rest) is:

  • Imaging. Don’t image patients without red flags – it doesn’t reassure them much.
  • Headache Diaries. Get patients to do a headache diary as part of their evaluation.
  • Give patients a positive diagnosis, lots of reassurance and recognise the impact on their lives.
  • The panel felt that chronic tension headache was likely to be unusual. If there any features consistent with migraine, treat the patient as if having migraine.
  • Remember Medication Overuse Headache
  • Episodic Tension Headache – paracetamol, aspirin and NSAIDs all work equally well.
  • Chronic Tension Headache – evidence is very limited. The only recommendation is to use acupuncture.
  • Episodic Migraine – discuss what the patient prefers, there is not a lot to choose between treatments, though combinations of triptans with either NSAID or paracetamol seem to be the most effective. If using aspirin, use 900mg. Antiemetics give added benefit, even if there is no nausea or vomiting.
  • Migraine Prophylaxis – topiramate and propranolol are 1st line. New for 2015 is to consider amitriptyline depending on preferences, co-morbidities and likelihood of adverse effects. Acupuncture course (10 sessions over 5-8 weeks) can be considered if 1st line fails or not suitable. Changed in 2015, gabapentin now not offered as 2nd line. There aren’t any studies looking at pizotifen – so they haven’t been shown not to work. The evidence out there is very contradictory and often of poor quality.
  • Cluster Headaches – Ask a neurologist for advice about whether imaging is required after a first cluster. Only O2 and subcut or nasal triptans are advised for acute treatment. Only verapamil is advised for prophylactic treatment.

Red Flags:

Consider the need for further investigation / referral if:

  • Compromised immunity (eg HIV with CD4 count < 200). There is a risk of cerebral infection.
  • Under 20s with history of malignancy (in one small study up to 14% of patients presenting with headaches had cerebral mets).
  • History of malignancy known to metastise to the brain
  • Vomiting without obvious cause

Evaluate carefully and consider the need for further investigation and referral if:

  • Worsening headache with fever
  • New neurological deficit or new cognitive dysfunction
  • Change in personality
  • Impaired consciousness
  • Recent head trauma (eg last 3m)
  • Triggered by cough / valsalva / sneeze / worse with exercise / orthostatic (changes with posture)
  • Substantial change in characteristics
  • Migraine with atypical aura ( motor weakness, double vision, visual symptoms only affect 1 eye, poor balance, decreased consciousness)
  • Suggestive of GCA / glaucoma / sudden onset reaching maximum within 5 mins

Early morning headaches? Interestingly this isn’t a characteristic included, as the studies looking at this were of poor quality. This obviously doesn’t mean that it is not a concerning feature.

 Is imaging for reassurance useful?


Imaging involves radiation, it is expensive and up to 10% of scans will show up an incidental finding. Less than 1% of scans show up anything serious.

Interestingly, if you feel doing a scan for reassurance works – it doesn’t. It does reduce visits to neurologists and it may reduce visits to the GP for reassurance, but it doesn’t reduce overall future GP visits and doesn’t decrease requests for sick notes, health worries or fear of illness.

You can also be reassured, that of patients diagnosed by the GP as having primary headache (ie not caused by anything else), only 4 in 10,000 go on to be diagnosed as having a brain tumour over the coming year (compared to 1.7 in 10,000 of people without headaches). In the under 18s, there is no increase in tumour rate seen.

How should you evaluate a patient with headache?

There isn’t specific advice on what history and examination should be done (eg is a full CNS exam needed on everyone?).

Headache diaries are, however, worth doing. The patient should keep a diary for 8 weeks, with frequency / duration / severity / associated symptoms / medications used / possible precipitants / menstruation.

 What information should be given to patients?

A common theme is that patients feel fobbed off by their doctors. Therefore the following should be done for all patients:

  • Give a positive diagnosis
  • Explain the cause
  • Reassure that other pathology has been ruled out
  • Explain management options
  • Recognise the impact that it can have on their lives.
  • Give people written advice, including about support groups.

Medication Overuse Headache

Patients are likely to be at risk if they use:

  • Triptans / opioids / ergots on 10d per month or more
  • Paracetamol / aspirin / NSAIDs on 15d per month or more

Patients should be advised to withdraw all the above drugs for 1 month. They should be warned that their headaches are likely to get worse before they get better. Most patients can manage well without prophylactic treatment during this time, but evidence suggests that it does work well if needed.

 Tension Headache – treatment of acute attacks:

Paracetamol, NSAIDs and aspirin all work equally well.

Avoid opiates (including co-codamol). They haven’t been shown to give any added benefit.

 Tension Headache – prophylactic treatment:

The only treatment suggested is acupuncture, which did have significant benefit over sham acupuncture in trials.

The committee say “ there was not enough evidence to recommend pharmacological prophylactic treatment for tension type headaches”.

Amazingly only 1 study looks at amitriptyline use (which would be my normal first line here). It didn’t show any benefit, but was felt to be a low quality study. Interestingly, only 2 other studies into chronic tension headache have been done that were considered and these showed that physiotherapy was better than acupuncture and that ‘manual therapies’ were better than ‘normal care’. As these studies were low quality, the committee felt that they couldn’t include these in the guideline.

Migraine – treatment of acute attacks:

Combination treatments seem to work best, though the evidence isn’t great:

  • Triptan + NSAID
  • Triptan + paracetamol.

If patients want to use single treatment, then that is fine too. Triptans, NSAID, parecetamol and 900mg aspirin have little between them.

Under 18s – oral triptans aren’t licensed, so consider using nasal preparations.

Antiemetics – interestingly antiemetics seem to have an effect even when there is no nausea, so consider adding them in as a treatment option.

Second line treatments: If the above aren’t working, then the treatment with the best evidence is non-oral antiemetics like metoclopramide or prochlorperazine.

Migraine – prophylactic treatment

1st line – topiramate or propranolol

Be aware that topiramate can be teratogenic and can reduce COC effectiveness

2nd line – acupuncture or gabapentin

They also state that Riboflavin 400mg OD may be beneficial in some people

People already on treatment – if people are already on established treatment that works, it is OK to leave them on it.

So what about Amitriptyline and Pizotifen? The studies just haven’t been done. There isn’t much good evidence that they don’t work.

Confused? Just to give an idea of the confusion out there…

Interestingly, ACEi may work. Sodium valproate doesn’t work any better than placebo, but nor does topiramate work any better than sodium valproate. B-blockers seem to be effective up until 6m, but haven’t been shown to be effective beyond that. Acupuncture seems to be as effective as B-blockers. Relaxation therapy and exercise are as effective as topiramate. Yoga and ‘manual therapies’ seem to work.

The problem is that a lot of the studies are of poor quality and so the above therapies aren’t being included in the guideline.

Cluster Headaches:

If a patient presents with a first cluster, ask a specialist for advice about whether imaging may be required. You are looking for vascular abnormalities here really.

Acute treatment:

  • O2
  • Triptan (subcut or nasal)

Just in case you are interested – even air has a strong placebo effect… there aren’t any studies that were considered looking at simple analgesics.


Verapamil up to 960mg per day (need ECG before use, with every dose increase and during treatment).

Oral triptans have been shown not to be effective. There is some evidence for melatonin, but sodium valproate doesn’t seem to work. Topiramate may work.

1 Response

  1. Mairi
    Really useful Louise - thank you!

Leave your comments