Prevalence of normal head CT and positive CT findings in a large cohort of patients with chronic headaches

Abstract

This study was carried out to ascertain the frequency of normal head computed tomography (CT) scans and positive CT scan findings in patients having chronic headache as chief complaint. Head CT scans done over a period of two years were retrospectively evaluated. On the basis of CT reports, the patients were divided into two groups: Group A, having headache as the only complaint, and Group B, having headache and additional neurological signs or symptoms. A total of 2498 patient reports were evaluated. There were 1772 patients in Group A and 726 patients in Group B. In Group A, 82% (n = 1453) patients had normal head CT, whereas in Group B 74.5% (n = 541) patients had a normal CT scan. There were 13.22% head CT scans showing significant findings in Group B, as compared to 6.2% in Group A. Both these differences were found to be statistically significant. CT findings such as infections, neoplasm, hydrocephalus, and extra-axial collections were higher in Group B when compared to Group A. CT examination in patients with isolated chronic headache is normal in high percentage of patients. The frequency and distribution of various CT findings over different age groups in a large cohort of patients presenting with chronic headache are discussed.

Introduction

Chronic headache is defined as “headache occurring for 15 or more days in a month for at least three months.”¹ It is one of the common complaints encountered in day-to-day neurological practice. However, when the headache becomes recurring or chronic, physician consultation is often sought. In a document on the global burden of headache, a prevalence of 50% has been reported in Asia, Australia, Europe, and North America.² The incidence and prevalence of headache in the Indian subcontinent have not been adequately researched. Recently, a population-based survey on the prevalence of headache in the state of Karnataka has shown that tension-type headache and migraine are the two most common causes of headache.³ Apart from these common causes, there are multiple other causes such as trauma, vascular disorders, infections, tumors, substance abuse, etc. Some forms of headache like ophthalmoplegic migraine have a typical clinical presentation and imaging may or may not be informative,⁴ while in others like neoplasms imaging offers an early diagnosis and a chance for potential treatment.

The two main neuroimaging modalities available for evaluation of the various causes of headache are computed tomography (CT) and magnetic resonance imaging (MRI). CT, by virtue of its ubiquity, ease of use, speed, and lower financial footprint, is usually the first investigation to be asked for in cases of headache. Multiple earlier studies have shown that in the vast majority of patients with chronic headache, CT scan may be normal.^5–8 This has led to formation of guidelines as to when to order neuroimaging in these patients.⁹ However, not much work has been published on the frequency of positive CT findings in these patients and whether they are significant or just incidental.

The objective of this study was to assess the prevalence of normal CT scans in patients with chronic headache as main presentation and to study the distribution and frequency of positive CT findings in this group of patients.

Materials and methods

This was a retrospective analysis of all the CT head examination records done in the department. We included only those head CT scans of patients who were seen in the different outpatient clinics of our tertiary care hospital and referred for CT-head examination to our department. Patients who were admitted in the hospital at the time of imaging were not included for this study. The CT head examinations that were performed between 2011 and 2013 were retrieved from the database and evaluated. All CT examinations in which the reason for referral was chronic headache as the chief complaint of the patient were included in the study. The CT scans were carried out on 4-slice and 16-slice spiral CT scanners; 10 mm contiguous slices were taken from foramen magnum to the vertex, which were then reconstructed into 3 mm slices both in soft tissue and bone window.

The reports were segregated on the basis of presentation into two groups: Group A, in which chronic headache was the only presenting complaint, and Group B, in which chronic headache was associated with other complaints like dizziness, tinnitus, vomiting, focal weakness, or any neurological signs. The patients in both these groups were then divided in age groups from 0 to 90 years of age at increments of ten.

In both groups, the CT findings were grouped under 11 categories: normal scan, cerebral atrophy consistent with age, basal ganglia calcification, calcified granulomas without edema, infections, focal gliosis/infarcts, neoplasms, extra-axial collections, lobar atrophy, and miscellaneous groups (see Table 1). In the miscellaneous group, conditions like contusions, white matter edema without clear cause, gyral swelling, and indeterminate lesions which needed further investigation were kept. The first four groups of CT findings (normal study, cerebral atrophy, basal ganglia calcification, and calcified granuloma) were taken as normal scan/insignificant CT findings for study purposes as the majority of these findings are seen as incidental findings on CT. The remaining seven groups of CT findings constituted positive CTs/significant CT findings as all these conditions could potentially give rise to chronic headache.

Table 1.

CT head findings distribution over Group A and Group B.

Category	CT findings	Number in Group A (%)	Number in Group B (%)	p-value
Normal + clinically insignificant CT findings
1	Normal study	1453 (81.9)	541 (74.5)	0.000
2	Cerebral atrophy	57 (3.2)	29 (4.0)	0.201
3	Basal ganglia calcification	15 (0.8)	4 (0.5)	0.310
4	Calcified granulomas	92 (5.2)	56 (7.7)	0.012
Clinically significant CT findings
5	Infections	12 (0.67)	18 (2.47)	0.000
6	Chronic infarcts or gliosis	52 (2.9)	19 (2.6)	0.384
7	Neoplasms	24 (1.35)	25 (3.4)	0.001
8	Extra-axial collections	16 (0.9)	12 (1.65)	0.084
9	Hydrocephalus	13 (0.7)	10 (1.37)	0.101
10	Lobar atrophy	8 (0.45)	1 (0.13)	0.212
11	Miscellaneous	30 (1.7)	11 (1.51)	0.449
	Total	1772 (100)	726 (100)

Three experienced neuroradiologists with 3–10 years of experience in neuroradiology reviewed all the scans. The findings were tabulated and analyzed using Statistical Package for the Social Sciences, Version 16.0 (SPSS, Chicago, IL). Difference of proportions was analyzed using Pearson’s chi square test and a value of p < 0.05 was taken as a statistically significant difference.

Results

A total of 2498 patient CT head scans were included. There were 58.56% (n = 1463) females and 41.43% males (n = 1035) in the study. The mean age was 36.62 years with standard deviation of 14.63. Median age was 35.

Of these 2498 patients, Group A had 1772 (70.93%) patients and Group B had 726 (29.06%) patients. Most patients (48%, n = 1202) were between 20 and 39 years of age, whereas age groups of 0–9 years and 80–89 years had only 7 and 11 patients, respectively.

In total, 79.82% (n = 1994/2498) of the head CTs were normal. In Group A, 82% (n = 1453/1772) of the patients had normal head CT as compared to 74.5% (n = 541/746) of the patients in Group B.

The combined number of patients who had either normal head CT or CT findings not significant for patients with chronic headache (categories 1–4) was 91.25% (n = 1617/1772) in Group A versus 86.77% (n = 630/726) in Group B (Table 1).

The findings, which could potentially cause headache, were further divided into seven categories, as enumerated in Table 1. The total percentage of patients with significant findings on CT was 6.20% (155 of 1772 patients) in Group A and 13.22% (96 out of 726 patients) in Group B (Table 1).

The distribution of significant and insignificant CT findings was also studied over various age groups from 0–90 and the results are given in Table 2.

Table 2.

Distribution of CT findings with age.

Age	No.	Normal CT (%)	Cerebral atrophy (%)	Basal-ganglia calcification (%)	Calcified granuloma (%)	Infection (%)	Gliosis (%)	Neoplasm (%)	Collection (%)	Hydrocephalus (%)	Lobar atrophy (%)	Miscellaneous (%)
0–9	7	3 (42.85)	0 (0)	0 (0)	0 (0)	1 (12.28)	0 (0)	1 (12.28)	0 (0)	2 (28.57)	0 (0)	0 (0)
10–19	292	242 (82.87)	0 (0)	0 (0)	20 (6.89)	13 (4.45)	8 (2.73)	2 (0.68)	1 (0.34)	5 (1.71)	0 (0)	1 (0.34)
20–29	555	474 (85.40)	2 (0.36)	2 (0.36)	37 (6.66)	3 (0.54)	13 (2.34)	9 (1.62)	3 (0.54)	3 (0.54)	0 (0)	9 (1.62)
30–39	647	538 (83.15)	4 (0.61)	8 (1.23)	41 (6.33)	3 (0.46)	15 (2.31)	14 (2.16)	8 (1.23)	4 (0.61)	2 (.30)	10 (1.54)
40–49	506	425 (84.18)	2 (0.39)	4 (0.8)	28 (5.53)	6 (1.18)	13 (2.56)	7 (1.38)	5 (0.98)	5 (0.98)	3 (0.59)	7 (1.38)
50–59	262	186 (71.37)	19 (7.25)	3 (1.14)	12 (4.58)	3 (1.14)	7 (2.67)	11 (4.19)	6 (2.29)	4 (1.52)	2 (0.76)	8 (3.04)
60–69	162	104 (64.19)	31 (19.13)	1 (1.61)	9 (5.55)	1 (1.61)	5 (3.08)	3 (1.85)	2 (1.23)	0 (0)	2 (1.23)	4 (2.46)
70–79	58	21 (36.20)	23 (39.65)	1 (1.72)	1 (1.72)	0 (0)	7 (12.06)	2 (3.44)	2 (3.44)	0 (0)	0 (0)	1 (1.72)
80–89	11	1 (9.09)	5 (45.45)	0 (0)	0 (0)	0 (0)	3 (27.27)	0 (0)	1 (9.09)	0 (0)	0 (0)	1 (9.09)
	2498	1994	86	19	148	30	71	49	28	23	9	41

The difference between number of patients having normal CT scan and of positive CT scan in both the groups was statistically significant (p < 0.05).

Discussion

Headache is one of the common symptoms for which patients seek a consultation with a healthcare provider. In the present era of advanced diagnostic imaging, CT is easily available and is more and more utilized as screening modality for these headaches. It has led to their indiscriminate use, leading to increasing radiation exposure to the patients and increasing healthcare cost.

The other side of the issue is the ethical aspect of whether it is correct to deny a patient an investigation which can potentially lead to diagnosis of a life-threatening disease, e.g. a brain tumor. The US headache consortium has given recommendations in doing neuroimaging for chronic headache patients. According to these, neuroimaging should be considered in patients who have a non-acute headache associated with abnormal findings on neurological examination. (Grade B recommendation).⁹

We evaluated the head CT of patients referred to our imaging department to see the percentage of normal scans and to study the positive findings revealed by these scans.

In 1993, Mitchell et al. evaluated whether routine CT evaluation is necessary in patients of headache irrespective of presence or absence of neurological findings.⁵ They studied 350 patients of which only 2% had CT findings which were clinically significant. They found that an additional 7% of the patients had positive CT findings that were clinically insignificant. More importantly, all of the patients in their study who had significant CT findings had some neurological finding or abnormal symptom.⁵ In comparison, we found 79.8% (n = 1994) of the patients with normal head CT (in both groups combined) and an additional 10% (n = 253) of patients having incidental findings on CT examination. In the present study, 10% (n = 251) of the patients had significant CT findings, which is significantly more than reported in the study by Mitchell et al.⁵

In another review by Dumas et al. of 373 patients with chronic headache involving 402 CT scans,⁶ approximately 4.47% of the CT scans were positive for any finding but only 0.98% (4 out of 402) had any significant CT finding. In these four patients, diagnosis of aneurysm (n = 1), osteomas (n = 2), and low-grade glioma (n = 1) were given.

A few other studies, such as those by Frishberg and Thomas et al,^8,10 support these findings. These studies showed that a routine CT examination for a chronic headache in the absence of focal neurological signs or any abnormal symptoms is less likely to have a positive CT scan.

Recently, a study by Simpson et al. found CT positivity of 10.5% in 4404 CT scans referred by general physicians between 1999 and 2007.⁷ Out of this, 1.4% of the CT findings were thought of as having abnormalities which could potentially be attributed as a cause of headache and 9.1% of the abnormalities were thought to be incidental in nature.⁷

In comparison, our results show that in a group of patients presenting with chronic headache as the only symptom, 6.2% of the scans showed a significant finding which increased to 13.22% in a group of patients having some neurological sign or symptom in addition to the complaint of chronic headache and was statistically significant (p < 0.005). Also, there were 82% absolutely normal head CTs in Group A as compared to 74.5% in Group B, a difference of 7.5%, which again was statistically significant (p < 0.005). This highlights the facts that in patients presenting with chronic headache as the only presenting feature CT head is normal in majority and if chronic headache is combined with the presence of additional neurological sign or symptom the number of normal CT head reduces significantly (74.5% in the present study).

The present study detected much more positive head CT scans as compared to previous studies. There can be multiple reasons for that. The most obvious is the referral bias; the patients coming to our hospital are those who are usually referred from primary or secondary care centers, which means the significant number of normal patients will be reduced at that level. The present CT scanners have higher resolution and they acquire volumetric data instead of axial slices with inter-slice gap as used in the older generation scanners. This could potentially be another reason. Also, all the scans were reviewed by neuroradiologists instead of general radiologists, which could also explain the higher number.

Among the positive CT findings in our study population, the calcified granulomas were seen in 148 patients head CT and constituted approximately 6% of positive CT findings. This is not surprising as tuberculosis and cysticercosis, the two most common causes of calcified granulomas, are endemic in the Indian subcontinent and asymptomatic calcified granulomas are seen frequently on neuroimaging. This was followed by CT findings of cerebral atrophy and focal gliosis or chronic infarct, which were seen in 3.4% and 2.8% of the scans, respectively. These numbers were higher when compared with work of Simpson et al.,⁶ who found 1.5% of head CTs with cerebral atrophy and 1% of patients with infarction.

Not much work has been published with regards to the presence of different CT diagnosed pathologies in different age groups. Our study showed that in children and young adult age groups (0–29 years) the two most common CT finding after normal head CT was calcified granulomas and CT findings suggestive of infections seen in 74 patients (2.96%).

In a study published by Prpić et al. on neuroimaging done in children with chronic headache,¹¹ out of 215 children CT/MRI was done in 164 patients and majority of them (n = 117, 71.3%) had normal neuroimaging. Our study in comparison had 79.8% of normal scans. In 28.1% (n = 46) of children there were incidental findings,¹¹ which in our study was 6.68% among children aged between 0 and 19 years (n = 299). This difference might be because of the difference in incidental findings in both the studies. We took into account only intra-axial findings, whereas the findings of Prpić et al. were a mixture of both intra- and extra-axial findings like asymmetry of ventricles, enlarged cisterna magna, enlarged adenoids, fluid in the mastoids, etc.¹¹ They reported only one tumor in 164 scans a percentage of 0.6% whereas we found three tumors in 299 patients, a percentage of 1.0%. Again this difference could be partially because of the difference in sample size and also because of referral bias in the present study.

In the older age group (60–90 years), cerebral atrophy followed by focal gliosis were the most frequent CT findings after normal study seen in 74 (2.96%) of the patients. Neoplasms were seen most frequently in the age group between 50 and 59 years with 4.19% of the patients in that age group showing primary or metastatic tumors on head CTs.

Our study has some limitations. It was retrospective in nature. A referral bias was inherently present in our study as it was conducted at a tertiary care center; therefore, the proportion of patients with significant findings might be higher.

In conclusion, the present study and previous studies highlight the fact that CT in cases of isolated chronic headache has a low yield and close to 80% of the patients have a normal CT scan. Also, if the headache is associated with a neurological sign or any other symptom, like syncope, focal weakness, etc., the chance of positive CT scan almost doubles. Though the positive CT findings in these patients are less, the rate is not as low as reported by a few earlier studies. Also, certain CT findings are more common in certain age groups, e.g. infections are more common in children and neoplasms and cerebral atrophy in the elderly, knowledge of which might help in making diagnosis in difficult or indeterminate scans.

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.