Patterns of Care and Outcomes After Computed Tomography Scans for Headache

Transcription

1 CLINICAL RESEARCH STUDY Patterns of Care and Outcomes After Computed Tomography Scans for Headache John J. You, MD, MSc, a,b Jonathan Gladstone, MD, c,d Sean Symons, MD, MPH, e,f Dalia Rotstein, MD, g Andreas Laupacis, MD, MSc, b,g,h,i Chaim M. Bell, MD, PhD b,g,h,i a Departments of Medicine and of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada; b Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; c Cleveland Clinic Canada, Toronto, Ontario, Canada; d Departments of Neurology, and e Medical Imaging, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; Departments of f Medical Imaging, g Medicine, and h Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada; i Keenan Research Centre of the Li Ka Shing Knowledge Institute, St Michael s Hospital, Toronto, Ontario, Canada. ABSTRACT BACKGROUND: Concerns exist about potential overuse of computed tomography (CT) scans for headache in ambulatory care. METHODS: We sought to examine health services use, brain tumor diagnosis, and death during the year after CT scanning for headache by linking records of an audit of 3930 outpatient CT brain scans performed in 2005 in Ontario, Canada, to administrative databases. RESULTS: Of 623 patients receiving CT scans for a sole indication of headache, few (2.1%) scans contained findings potentially causing their headache. For most patients, the index CT scan was the only one received over an 11-year period. However, 28.4% of patients received 1 or more CT brain scans during the preceding decade and 6.7% received 1 or more CT brain scans during the subsequent year. Of the 473 patients (75.9%) whose index scan was ordered by a primary care physician, most (80.3%) did not see a specialist during follow-up. One patient with an indeterminate finding on the index scan was diagnosed with a malignant brain tumor (0.2%), and 6 patients (1.0%) died during follow-up. Among the 4 deaths in which the cause could be determined, none were due to central nervous system causes. CONCLUSION: Because of the potential risk of cancer from exposure to ionizing radiation, efforts should be made to avoid CT scanning for headache when the likelihood of serious illness is low. Evidence-based decision rules that identify which patients with headache do not require neuroimaging may decrease the use of CT scans in situations of little benefit Elsevier Inc. All rights reserved. The American Journal of Medicine (2011) 124, KEYWORDS: Computed tomography; Headache; Health services research Headache is exceedingly common, with a lifetime prevalence of greater than 90% in the general population. 1 Headache is also one of the most common reasons patients visit their primary care physician, visit the emergency department, or consult with a neurologist. 2-4 Although the overwhelming majority of patients will not have a serious or life-threatening underlying cause for their headache, a small minority of patients will, making accurate clinical assessment of the patient with headache critical. Physicians often feel compelled to use neuroimaging to investigate patients with headache because: 1) misdiagnosis (or delay in diagnosis) can result in mortality or lifelong dis- Funding: Dr You is supported by a Career Scientist Award from the Ontario Ministry of Health and Long-Term Care. Dr Bell is supported by a Canadian Institutes of Health Research and Canadian Patient Safety Institute Chair in Patient Safety and Continuity of Care. This study was conducted at the Institute for Clinical Evaluative Sciences, which is partially funded by the Ontario Ministry of Health and Long-Term Care. Conflict of Interest: None of the authors have any conflicts of interest associated with the work presented in this manuscript. Authorship: All authors had access to the data and played a role in writing this manuscript. All authors contributed to the concept and design of the study. Drs You, Gladstone, and Bell drafted the manuscript. All authors contributed to the analysis and interpretation of data, critically revised the manuscript for important intellectual content, and approved the final version. Requests for reprints should be addressed to John J. You, MD, MSc, 1200 Main Street West, HSC-3V51B, Hamilton Ontario, Canada, L8N 3Z5. address: jyou@mcmaster.ca /$ -see front matter 2011 Elsevier Inc. All rights reserved. doi: /j.amjmed

2 You et al Patterns of Care and Outcomes after CT for Headache 59 ability, and 2) patient anxiety about headache as a presenting symptom of serious illness (eg, brain tumor) is often very high because of anecdotal history from friends and family, as well as information from mainstream television and movies. 5 There are surprisingly few guidelines to assist physicians in the use of neuroimaging to evaluate patients with headache. Existing recommendations focus on when imaging is and is not warranted in suspected migraine and, unfortunately, do not provide guidance about the use of neuroimaging for other common headache presentations. 6,7 Cross-sectional studies have shown that computed tomography (CT) scans of the head performed in outpatients for evaluation of headache have low diagnostic yield, with potentially clinically relevant findings uncovered in 0.2% to 2% of scans Such findings have raised concerns that neuroimaging is being overused in the assessment of patients with headache, leading to added strain on already limited healthcare resources and unnecessary exposure to radiation. 13,14 To date, little is known about health services use and outcomes after CT scans performed for headache. Such information may provide greater insights into the appropriateness of current patterns of CT imaging for headache. Accordingly, we conducted a retrospective cohort study in which data from a cross-sectional audit of outpatient CT brain scans for headache were linked to provincial administrative health data to evaluate health services use and patient outcomes after CT imaging for headache in ambulatory care. CLINICAL SIGNIFICANCE MATERIAL AND METHODS Study Sample We used data from a chart abstraction of CT and magnetic resonance imaging (MRI) scans that included a representative sample of outpatient CT scans of the brain performed at 20 randomly selected hospitals offering CT services in Ontario, Canada. Hospital selection was stratified by region (north and south) and teaching status (teaching hospital and community hospital). Teaching status was defined by membership in the Council of Academic Hospitals of Ontario. For the CT brain stratum of the study, the target sample was a consecutive series of 200 outpatient CT scans of the brain performed at each of the 20 hospitals, on or after January 1, CT scans performed on patients in the emergency department were not eligible for study. Data on 3930 outpatient CT scans of the brain were collected. 8 Headache is the most common reason for requesting CT scans of the brain in the ambulatory care setting in Ontario, accounting for 27% (1055/3930) of outpatient CT brain scans. CT scans in ambulatory patients with a sole indication of headache are of low diagnostic yield (2.1%) and infrequently lead to brain tumor diagnosis (0.2%). Because of the potential risk of cancer from exposure to ionizing radiation, clinicians should make every effort to avoid CT scanning for headache when the likelihood of serious illness is low. By using a standardized electronic case report form, trained study nurses abstracted data from physician requisitions and radiologists CT scan reports regarding the indications for and results of the scans. Abstractors categorized each imaging finding described in a given CT scan report as incidental (clinically insignificant), indeterminate (the radiologist could not conclusively identify the presence or describe the nature of an imaging finding), or abnormal. Chance-corrected inter-rater agreement was substantial for the presence of an incidental finding (kappa 0.71; 95% confidence interval, ) and the presence of an indeterminate finding (kappa 0.69; 95% confidence interval, ) in the CT brain reports. 8,15 This study included CT scans of the brain from the original chart abstraction study that were ordered specifically and solely for the evaluation of headache. We chose to focus on this population because individuals with other issues identified on referral constitute a different group of patients for whom there are likely to be fewer concerns about the potential overuse of CT imaging. Scans were thus excluded if there were additional indications for the scan in addition to headache. Scans were also excluded if the record could not be linked to provincial administrative health databases. Data Sources The province of Ontario provides its residents with universal coverage for physician services, hospitalizations, and diagnostic tests. For this study, we linked data from the original chart abstraction study to 4 administrative health databases using an encrypted unique patient number: 1) the Ontario Health Insurance Plan Database, which contains data regarding all physician billing claims; 2) the Ontario Cancer Registry, which contains data regarding incident cases of cancer in Ontario; 3) the Canadian Institute for Health Information Discharge Abstract Database, which contains data regarding all acute care hospitalizations in Ontario; and 4) the Registered Persons Database, which contains vital statistics data for Ontario residents. These linked datasets provide the ability to track, on a provincewide basis, subsequent health services and outcomes for patients with an index CT scan for headache included in the chart abstraction study. Coding within these administrative databases has good agreement with actual procedures and services provided, 16 and the Ontario Cancer Registry captures greater than 95% of incident cases of cancer across the province. 17 Moreover, 95% of CT and MRI scans reviewed in the original chart abstraction study are identifiable in the

3 60 The American Journal of Medicine, Vol 124, No 1, January 2011 Ontario Health Insurance Plan Database, indicating that these administrative databases identify imaging studies with a high degree of accuracy. Outcomes By using the administrative databases described above, we identified CT and MRI scans of the brain performed during the 10-year period before the index CT scan and the 1-year period after the index CT scan. We also identified neurosurgical procedures, new diagnoses of cancer, and deaths occurring during the 1-year period after the index scan. For index CT scans that were ordered by primary care physicians, we identified consultations with neurologists, internists (general internists or subspecialists), and neurosurgeons during follow-up, as well as any follow-up visits with the consultant during the 1-year period after the initial consultation. Statistical Analysis We described baseline characteristics of the study sample, health services use, and outcomes using the mean and standard deviation or median and interquartile range for continuous data, and proportions for categoric data. This study was approved by the Sunnybrook Health Sciences Centre Research Ethics Board. RESULTS Characteristics of the Study Cohort Headache was the most common reason for requesting a CT scan of the brain in the ambulatory care setting. 8 Of the initial 3930 outpatient CT scans of the brain, 1055 were performed for headache. Of these 1055 records, 421 were excluded because the scans were ordered for other indications in addition to headache (Appendix, online), and 11 were excluded because they could not be linked to administrative data, resulting in a final study sample of 623 outpatient CT scans of the brain performed for the sole indication of headache. The study sample was comprised predominantly of younger women. Two thirds of the cohort were female, and two thirds were aged less than 50 years. Most scans (578/623, 92.8%) were completely normal or revealed incidental, clinically insignificant findings. A small proportion of scans revealed indeterminate findings (16/ 623, 2.6%) or abnormal findings (29/623, 4.7%). Thirteen scans (13/623, 2.1%) contained indeterminate or abnormal findings that could potentially have been the cause of the patient s headache (Table 1). Prior and Subsequent Use of Computed Tomography and Magnetic Resonance Imaging Scans of the Brain More than one quarter of patients (28.4%) had at least 1 CT scan of the brain in the preceding 10 years, with the most recent being a median of 2.7 years before the index scan. Table 1 Baseline characteristics of study sample CT scan for headache Characteristic (n 623) Age, mean (SD), y 45.3 (15.3) Age 50 y 406 (65.2) Female 411 (66.0) History of cancer 19 (3.0) Ordering physician Primary care physician 473 (75.9) Neurologist 97 (15.6) Other specialist 53 (8.5) CT scan result Normal 442 (70.9) Incidental* 136 (21.8) Indeterminate 16 (2.6) Unlikely related to headache Hypodensity of uncertain cause 9 (1.4) Colloid cyst vs calcification 1 (0.2) Prominent right posterior cerebral 1 (0.2) artery of questionable significance Sinus thrombosis cannot be excluded 1 (0.2) Circle of Willis aneurysm vs artifact 1 (0.2) Potentially related to headache Acoustic neuroma vs meningioma 1 (0.2) Possible low-grade glioma 1 (0.2) Possible meningioma 1 (0.2) Abnormal 29 (4.7) Unlikely related to headache Old stroke 19 (3.0) Premature atrophy 1 (0.2) Encephalomalacia 1 (0.2) Dilated ventricles 1 (0.2) Other 1 (0.2) Potentially related to headache Hydrocephalus 3 (0.5) Subdural hematoma or hygroma 3 (0.5) Mass effect 2 (0.3) Ruptured lipoma with chemical 1 (0.2) meningitis Acute stroke 1 (0.2) CT computed tomography; SD standard deviation. All data are presented as number (percent) unless otherwise specified. *Scans with only incidental, clinically insignificant finding(s) present. Scans in which the radiologist could not conclusively identify the presence or describe the nature of an imaging finding (an incidental finding may also have been present). Individual abnormalities total 29 because 1 abnormality was present on some of the scans. Few patients (5.5%) had an MRI scan of the brain in the preceding 10 years. During 1 year of follow-up, subsequent scanning with CT and MRI occurred in 42 patients (6.7%) and 41 patients (6.6%), respectively. Patients were more likely to receive a subsequent CT or MRI scan if their index CT scan showed abnormal or indeterminate results than if it was normal or revealed only incidental findings (28.9% vs

4 You et al Patterns of Care and Outcomes after CT for Headache 61 Table 2 Use of computed tomography and magnetic resonance imaging scans of the brain before and after an index computed tomography scan for headache Index CT scan for headache (n 623) No. of CT brain scans in preceding 10 y (21.2) 2 45 (7.2) No. of days since most recent CT brain 977 ( ) scan, median (IQR)* No. of CT brain scans in subsequent year 1 38 (6.1) 2 4 (0.6) No. of days to follow-up CT scan, 106 (71-238) median (IQR)* No. of MRI brain scans in preceding 10 y 1 32 (5.3) 2 1 (0.2) No. of days since most recent MRI 1513 ( ) brain scan, median (IQR)* No. of MRI brain scans during follow-up 1 37 (5.9) 2 4 (0.6) No. of days to follow-up MRI scan, 101 (54-209) median (IQR)) CT computed tomography; IQR interquartile range; MRI magnetic resonance imaging.all data are presented as number (percent) unless otherwise specified. *Among patients who had a prior or subsequent CT/MRI scan of the brain. 5.1% for subsequent CT scan; 33.3% vs 4.5% for subsequent MRI scan). Subsequent CT and MRI scans of the brain typically occurred within 4 months of the index scan (Table 2). Of the 42 CT scans that were ordered during the year after the index CT scan, 32 (76.2%) were ordered by a physician other than the one who ordered the index scan. Of the 41 MRI scans that were ordered during the year after the index CT scan, 25 (61.0%) were ordered by a physician other than the one who ordered the index scan. Health Services Use and Outcomes after Outpatient Computed Tomography Scans for Headache Three quarters (473/623, 75.9%) of the index scans were ordered by primary care physicians. During 1 year of follow-up, approximately one fifth of these patients (93/473, 19.7%) had a total of 108 specialist consultations with a neurologist, internist, or neurosurgeon. (Internists in Ontario function as consultants and are not primary care providers.) Referral was more likely if the index scan result was abnormal or indeterminate (42.5% of these patients were referred) than if it was normal or incidental (17.6% of these patients were referred). Approximately one half of patients referred to a specialist returned for at least 1 follow-up visit with the specialist during the year after the initial consultation (follow-up with specialist occurred after 51 of the 108 initial consultations, 47.2% of cases). Of the overall study cohort, few patients (12/623, 1.9%) were hospitalized for neuropsychiatric causes during the year after the index CT scan for headache, with depression being the most common reason for admission (8 patients were admitted a total of 11 times for depression). Neurosurgical procedures, malignant brain tumor diagnosis, and death were infrequent during follow-up. Only 1 patient (0.2%), who had an indeterminate finding of a possible low-grade glioma on the index CT scan, was diagnosed with a malignant brain tumor during follow-up. Six patients (1.0%) died during follow-up, and among the 4 in whom the cause of death could be determined, no deaths were due to central nervous system causes (Table 3). DISCUSSION Our study of outpatient CT scans of the brain for headache found that most scans were ordered by primary care physicians, few revealed abnormalities that were potentially causing the headache, and most patients who received these scans were not referred to a specialist. Our study extends the findings of prior cross-sectional studies because it is the first, to our knowledge, to report longitudinal data regarding health services use and patient outcomes after CT scanning for headache. Our finding that only 2% of outpatient CT scans for headache revealed abnormalities potentially causing the patient s headaches is consistent with previous studies documenting a low diagnostic yield of CT in this setting During 1 year of follow-up, we also found that a new Table 3 Health services use and outcomes during the year after an index computed tomography scan for headache Index CT scan for headache (n 623) Hospital admission Non-central nervous system cause 48 (7.7)* Neuropsychiatric cause 12 (1.9) Depression 8 (1.9) Malignant brain tumor 1 (0.2) Acute stroke 1 (0.2) Transient ischemic attack 1 (0.2) Benign neoplasm of cranial 1 (0.2) nerves Neurosurgical procedure 2 (0.3) New diagnosis of malignant brain tumor 1 (0.2) Death 6 (1.0) CT computed tomography. All data are presented as number of patients (percent) unless otherwise specified. *These 48 patients had 61 admissions for non-central nervous system causes. These 12 patients had 17 admissions for neuropsychiatric causes.

5 62 The American Journal of Medicine, Vol 124, No 1, January 2011 malignant brain tumor diagnosis and neurosurgical procedures were rare events. Of concern was the finding that, for some patients in our study, the index CT scan for headache was not a single event but one of several CT scans of the brain they received, often ordered by someone other than the original ordering physician and with a median time to repeat scanning of only 106 days. There are increasing concerns that radiation exposure during CT scanning may pose a public health risk. A single CT scan of the head is associated with an average radiation dose of 1 to 2 millisieverts (equivalent to chest x-rays). The estimated lifetime risk of cancer attributable to a single CT scan of the head is 0.005% (1/20,000) for a 45-year-old individual. 18 The risk is greater among younger patients and among those receiving higher cumulative doses of radiation as a result of multiple radiodiagnostic tests. 14,18-20 Our study cohort had a mean age of 45 years, highlighting that patients receiving CT scans for headache tend to be young patients at greater risk of cancer from CT-associated radiation, and with more years of life remaining during which to accumulate greater cumulative doses of radiation. Thus, clinicians should make every effort to minimize the use of CT scans when the likelihood of finding a secondary cause of headache is low. Unfortunately, work to date has not identified clinical features that are useful in ruling out serious causes of headache. 21 The development of highly sensitive, evidencebased clinical decision rules that effectively identify patients who do not require neuroimaging may help to minimize the use of CT scans in situations of little benefit. We found that three quarters of CT scans for headache in ambulatory practice were ordered by primary care physicians. Although this may be largely explained by the fact that there are many more primary care physicians than neurologists in the province of Ontario, there may be additional potential explanations for this finding. First, given long waits for neurologist consultation in Ontario, it is possible that primary care physicians are using the CT scan information to reassure patients that they do not have a serious illness, thus eliminating the need to refer their patient for formal evaluation by a specialist. Indeed, one randomized clinical trial found that anxiety was decreased at 3 months (not sustained at 1 year) among patients with headache who received an offer for neuroimaging compared with patients receiving usual care and that health resource use was lower among the subset of patients with high levels of anxiety or depression at baseline. 22 Second, primary care physicians may be trying to expedite care by requesting that the CT scan be completed in advance of the visit with the specialist, so that imaging findings are available at the time of consultation. Indeed, interviews with specialists in Ontario suggest that some specialists will not see patients for a new consultation unless the referral is accompanied by a CT or MRI scan report. 23 Strengths of our study include the use of a sample of outpatient CT scans from an entire jurisdiction and, by nature of Ontario s single-payer system, the ability to track subsequent health services use and outcomes in all patients using administrative data. A unique aspect of our study is that it reflects as pure a headache sample as possible by specifically excluding patients who had any concomitant symptoms or clinical histories listed on the imaging requisition. We chose to focus on this population because patients with other issues identified on the imaging referral (eg, localizing signs or symptoms, or a history of cancer) represent a different group of patients for whom there are likely to be fewer concerns about the potential overuse of CT scanning. Our study also has some limitations. First, our study was restricted to patients who received CT imaging for headache, and we are not able to compare our findings with a comparison group of patients with headache who did not undergo CT scanning. Second, because of the limited amount of information typically provided by referring physicians on CT scan requisitions, we did not have detailed clinical information about the nature of the headaches, such as their temporal pattern (sudden thunderclap headache vs episodic headache vs chronic daily headache), which could have provided further insights about the appropriateness of the patterns of care we observed. Finally, administrative health claims data in Ontario do not contain information about the reasons for ordering imaging tests. Therefore, we are not able to determine the indications for previous and subsequent CT and MRI scans of the brain that we identified in our study. However, given that we have identified headache as the most common reason for ordering neuroimaging in Ontario, it is likely that many of these previous and subsequent imaging studies of the brain were indeed performed to evaluate headache. CONCLUSIONS Our findings of low diagnostic yield, repeated CT scanning in some patients, and a favorable clinical course for the majority of patients receiving CT scans for headache are likely to continue to fuel concerns, particularly among payers, that CT scans are being overused for this indication. Given concerns about the risk of cancer from radiation exposure during CT scanning, clinicians should make every effort to ensure that CT scans are not ordered in patients who are unlikely to benefit. Development of highly sensitive clinical prediction rules that can effectively rule out the need for imaging in patients with headache would provide useful guidance for primary care physicians who provide the bulk of care for patients with this exceedingly common condition. ACKNOWLEDGMENTS The authors thank Alice Newman and Lingsong Yun for conducting linkage analyses for this study. References 1. Rasmussen BK, Jensen R, Schroll M, Olesen J. Epidemiology of headache in a general population a prevalence study. J Clin Epidemiol. 1991;44:

6 You et al Patterns of Care and Outcomes after CT for Headache Cherry DK, Hing E, Woodwell DA, Rechtsteiner EA. National Ambulatory Medical Care Survey: 2006 summary. Natl Health Stat Report. 2008;: Pitts SR, Niska RW, Xu J, Burt CW. National Hospital Ambulatory Medical Care Survey: 2006 emergency department summary. Natl Health Stat Report. 2008;: Hopkins A, Menken M, DeFriese G. A record of patient encounters in neurological practice in the United Kingdom. J Neurol Neurosurg Psychiatry. 1989;52: Vargas B. Migraine in modern media. Presented at: The International Headache Conference, September 12, 2009, Philadelphia, PA. 6. Pryse-Phillips WE, Dodick DW, Edmeads JG, et al. Guidelines for the diagnosis and management of migraine in clinical practice. Canadian Headache Society. CMAJ. 1997;156: Frishberg BM, Rosenberg JH, Matchar DB, et al. Evidence-based guidelines in the primary care setting: neuroimaging in patients with nonacute headache. American Academy of Neurology. Available at: Accessed April 27, You JJ, Purdy I, Rothwell D, Przybysz R, Fang J, Laupacis A. Indications for and results of outpatient computed tomography and magnetic resonance imaging in Ontario: a population-based study. Can Assoc Radiol J. 2008;59: Dumas MD, Pexman JH, Kreeft JH. Computed tomography evaluation of patients with chronic headache. CMAJ. 1994;151: Weingarten S, Kleinman M, Elperin L, Larson EB. The effectiveness of cerebral imaging in the diagnosis of chronic headache. Arch Intern Med. 1992;152: Akpek S, Arac M, Atilla S, Onal B, Yucel C, Isik S. Cost-effectiveness of computed tomography in the evaluation of patients with headache. Headache. 1995;35: Sempere AP, Porta-Etessam J, Medrano V, et al. Neuroimaging in the evaluation of patients with non-acute headache. Cephalalgia. 2005;25: Picard A. Scan use is up, but the value is questionable. Globe & Mail April 16, 2009; L Fazel R, Krumholz HM, Wang Y, et al. Exposure to low-dose ionizing radiation from medical imaging procedures. N Engl J Med. 2009;361: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33: Juurlink D, Preyra C, Croxford R, et al. Canadian Institute for Health Information Discharge Abstract Database: A Validation Study. Toronto: Institute for Clinical Evaluative Sciences; Robles SC, Marrett LD, Clarke EA, Risch HA. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol. 1988;41: Brenner DJ, Hall EJ. Computed tomography an increasing source of radiation exposure. N Engl J Med. 2007;357: Smith-Bindman R, Lipson J, Marcus R, et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med. 2009;169: Berrington de GA, Mahesh M, Kim KP, et al. Projected cancer risks from computed tomographic scans performed in the United States in Arch Intern Med. 2009;169: Detsky ME, McDonald DR, Baerlocher MO, Tomlinson GA, McCrory DC, Booth CM. Does this patient with headache have a migraine or need neuroimaging? JAMA. 2006;296: Howard L, Wessely S, Leese M, et al. Are investigations anxiolytic or anxiogenic? A randomized controlled trial of neuroimaging to provide reassurance in chronic daily headache. J Neurol Neurosurg Psychiatry. 2005;76: You JJ, Levinson W, Laupacis A. Attitudes of family physicians, specialists and radiologists about the use of computed tomography and magnetic resonance imaging in Ontario. Healthc Policy. 2009;5:54-65.

7 63.e1 The American Journal of Medicine, Vol 124, No 1, January 2011 Appendix Scans excluded because of additional indications for computed tomography scan in addition to headache Additional indications for CT head Excluded scans (n 421)* Dizziness or lightheadedness 70 (16.6) Suspected cancer 56 (13.3) Suspected intracranial mass 49 (11.6) Posterior fossa signs (eg, vertigo, diplopia) 49 (11.6) Suspected stroke or transient ischemic attack 49 (11.6) Visual disturbances 45 (10.7) Cancer surveillance (for progression or 46 (10.9) recurrence) Other 43 (10.2) Dementia, memory loss, or non-acute cognitive 31 (7.4) deficit Head injury 30 (7.1) Ataxia, impaired balance, or loss of 17 (4.0) coordination Seizure or epilepsy 11 (2.6) Syncope 11 (2.6) Focal sensory deficit 11 (2.6) Suspected cerebral aneurysm 10 (2.4) Focal motor deficit 9 (2.1) Routine postoperative follow-up 5 (1.2) Suspected subdural hematoma 5 (1.2) Pituitary disease 4 (1.0) Suspected multiple sclerosis or demyelination 4 (1.0) Suspected postoperative complication 4 (1.0) Depression 4 (1.0) Falls 4 (1.0) Restaging after cancer treatment 3 (0.7) Cancer surveillance during active treatment 3 (0.7) Tinnitus 3 (0.7) Cancer staging 2 (0.5) Hearing loss 2 (0.5) Follow-up after intracranial hemorrhage 2 (0.5) Follow-up after stroke 2 (0.5) Other cancer-related indication 1 (0.2) Aphasia 1 (0.2) Follow-up after head injury 1 (0.2) Suspected intracranial hemorrhage 1 (0.2) Hemifacial spasm 1 (0.2) CT computed tomography. *Individual indications total 421 because 1 indication was present for some of the scans.