Neuroimaging for the evaluation of chronic headaches: an evidence-based analysis
Medical Advisory Secretariat
Abstract:Objective: The objectives of this evidence based review are: i) To determine the effectiveness of computed tomography (CT) and magnetic resonance imaging (MRI) scans in the evaluation of persons with a chronic headache and a normal neurological examination.ii) To determine the comparative effectiveness of CT and MRI scans for detecting significant intracranial abnormalities in persons with chronic headache and a normal neurological exam.iii) To determine the budget impact of CT and MRI scans for persons with a chronic headache and a normal neurological exam. Clinical need: CONDITION AND TARGET POPULATION Headaches disorders are generally classified as either primary or secondary with further sub-classifications into specific headache types. Primary headaches are those not caused by a disease or medical condition and include i) tension-type headache, ii) migraine, iii) cluster headache and, iv) other primary headaches, such as hemicrania continua and new daily persistent headache. Secondary headaches include those headaches caused by an underlying medical condition. While primary headaches disorders are far more frequent than secondary headache disorders, there is an urge to carry out neuroimaging studies (CT and/or MRI scans) out of fear of missing uncommon secondary causes and often to relieve patient anxiety. Tension type headaches are the most common primary headache disorder and migraines are the most common severe primary headache disorder. Cluster headaches are a type of trigeminal autonomic cephalalgia and are less common than migraines and tension type headaches. Chronic headaches are defined as headaches present for at least 3 months and lasting greater than or equal to 15 days per month. The International Classification of Headache Disorders states that for most secondary headaches the characteristics of the headache are poorly described in the literature and for those headache disorders where it is well described there are few diagnostically important features. The global prevalence of headache in general in the adult population is estimated at 46%, for tension-type headache it is 42% and 11% for migraine headache. The estimated prevalence of cluster headaches is 0.1% or 1 in 1000 persons. The prevalence of chronic daily headache is estimated at 3%. Neuroimaging: COMPUTED TOMOGRAPHY: Computed tomography (CT) is a medical imaging technique used to aid diagnosis and to guide interventional and therapeutic procedures. It allows rapid acquisition of high-resolution three-dimensional images, providing radiologists and other physicians with cross-sectional views of a person's anatomy. CT scanning poses risk of radiation exposure. The radiation exposure from a conventional CT scanner may emit effective doses of 2-4mSv for a typical head CT. MAGNETIC RESONANCE IMAGING: Magnetic resonance imaging (MRI) is a medical imaging technique used to aid diagnosis but unlike CT it does not use ionizing radiation. Instead, it uses a strong magnetic field to image a person's anatomy. Compared to CT, MRI can provide increased contrast between the soft tissues of the body. Because of the persistent magnetic field, extra care is required in the magnetic resonance environment to ensure that injury or harm does not come to any personnel while in the environment. Research questions: What is the effectiveness of CT and MRI scanning in the evaluation of persons with a chronic headache and a normal neurological examination?What is the comparative effectiveness of CT and MRI scanning for detecting significant intracranial abnormality in persons with chronic headache and a normal neurological exam?What is the budget impact of CT and MRI scans for persons with a chronic headache and a normal neurological exam. Search strategy: A literature search was performed on February 18, 2010 using OVID MEDLINE, MEDLINE In-Process and Other Non-Indexed Citations, EMBASE, the Cumulative Index to Nursing & Allied Health Literature (CINAHL), the Cochrane Library, and the International Agency for Health Technology Assessment (INAHTA) for studies published from January, 2005 to February, 2010. Abstracts were reviewed by a single reviewer and, for those studies meeting the eligibility criteria full-text articles were obtained. Reference lists were also examined for any additional relevant studies not identified through the search. Articles with an unknown eligibility were reviewed with a second clinical epidemiologist and then a group of epidemiologists until consensus was established. Inclusion criteria: Systematic reviews, randomized controlled trials, observational studiesOutpatient adult population with chronic headache and normal neurological examStudies reporting likelihood ratio of clinical variables for a significant intracranial abnormalityEnglish language studies2005-present Exclusion criteria: Studies which report outcomes for persons with seizures, focal symptoms, recent/new onset headache, change in presentation, thunderclap headache, and headache due to traumaPersons with abnormal neurological examinationCase reports Primary outcome: Probability for intracranial abnormality SECONDARY OUTCOME: Patient relief from anxietySystem service useSystem costsDetection rates for significant abnormalities in MRI and CT scans Effectiveness: One systematic review, 1 small RCT, and 1 observational study met the inclusion and exclusion criteria. The systematic review completed by Detsky, et al. reported the likelihood ratios of specific clinical variables to predict significant intracranial abnormalities. The RCT completed by Howard et al., evaluated whether neuroimaging persons with chronic headache increased or reduced patient anxiety. The prospective observational study by Sempere et al., provided evidence for the pre-test probability of intracranial abnormalities in persons with chronic headache as well as minimal data on the comparative effectiveness of CT and MRI to detect intracranial abnormalities. OUTCOME 1: PRE-TEST PROBABILITY. The pre-test probability is usually related to the prevalence of the disease and can be adjusted depending on the characteristics of the population. The study by Sempere et al. determined the pre-test probability (prevalence) of significant intracranial abnormalities in persons with chronic headaches defined as headache experienced for at least a 4 week duration with a normal neurological exam. There is a pre-test probability of 0.9% (95% CI 0.5, 1.4) in persons with chronic headache and normal neurological exam. The highest pre-test probability of 5 found in persons with cluster headaches. The second highest, that of 3.7, was reported in persons with indeterminate type headache. There was a 0.75% rate of incidental findings. LIKELIHOOD RATIOS FOR DETECTING A SIGNIFICANT ABNORMALITY: Clinical findings from the history and physical may be used as screening test to predict abnormalities on neuroimaging. The extent to which the clinical variable may be a good predictive variable can be captured by reporting its likelihood ratio. The likelihood ratio provides an estimate of how much a test result will change the odds of having a disease or condition. The positive likelihood ratio (LR+) tells you how much the odds of having the disease increases when a test is positive. The negative likelihood ratio (LR-) tells you how much the odds of having the disease decreases when the test is negative. Detsky et al., determined the likelihood ratio for specific clinical variable from 11 studies. There were 4 clinical variables with both statistically significant positive and negative likelihood ratios. These included: abnormal neurological exam (LR+ 5.3, LR- 0.72), undefined headache (LR+ 3.8, LR- 0.66), headache aggravated by exertion or valsalva (LR+ 2.3, LR- 0.70), and headache with vomiting (LR+ 1.8, and LR- 0.47). There were two clinical variables with a statistically significant positive likelihood ratio and non significant negative likelihood ratio. These included: cluster-type headache (LR+ 11, LR- 0.95), and headache with aura (LR+ 12.9, LR- 0.52). Finally, there were 8 clinical variables with both statistically non significant positive and negative likelihood ratios. These included: headache with focal symptoms, new onset headache, quick onset headache, worsening headache, male gender, headache with nausea, increased headache severity, and migraine type headache. OUTCOME 2: RELIEF FROM ANXIETY Howard et al. completed an RCT of 150 persons to determine if neuroimaging for headaches was anxiolytic or anxiogenic. Persons were randomized to receiving either an MRI scan or no scan for investigation of their headache. The study population was stratified into those persons with a Hospital Anxiety and Depression scale (HADS) > 11 (the high anxiety and depression group) and those < 11 (the low anxiety and depression) so that there were 4 groups: Group 1: High anxiety and depression, no scan group Group 2: High anxiety and depression, scan group Group 3: Low anxiety and depression, no scan group Group 4: Low anxiety and depression, scan group ANXIETY: There was no evidence for any overall reduction in anxiety at 1 year as measured by a visual analogue scale of 'level of worry' when analysed by whether the person received a scan or not. Similarly, there was no interaction between anxiety and depression status and whether a scan was offered or not on patient anxiety. Anxiety did not decrease at 1 year to any statistically significant degree in the high anxiety and depression group (HADS positive) compared with the low anxiety and depression group (HADS negative). There are serious methodological limitations in this study design which may have contributed to these negative results. First, when considering the comparison of 'scan' vs. (ABSTRACT TRUNCATED)