Prevalence of Cervical Artery Dissection Among Hospitalized Patients With Stroke by Age in a Nationally Representative Sample From the United States

Abstract

Objective

To test the hypothesis that the prevalence of cervical artery dissection remains constant across age groups, we evaluated the relationship between age and cervical artery dissection in patients with stroke using a nationally representative sample from the United States.

Methods

We used inpatient claims data included in the 2012–2015 releases of the National Inpatient Sample (NIS). We used validated ICD-9-CM codes to identify adults hospitalized with ischemic stroke and a concomitant diagnosis of carotid or vertebral artery dissection. Survey weights provided by the NIS and population estimates from the US census were used to calculate nationally representative estimates. The χ² test for trend was used to compare the prevalence of concomitant dissection among stroke hospitalizations across patient subgroups defined by age. Poisson regression and the Wald test for trend were used to evaluate whether the prevalence of hospitalizations for stroke and concomitant dissection per million person-years varied by age groups.

Results

There were 17,320 (95% confidence interval [CI], 15,614–19,026) hospitalizations involving ischemic stroke and a concomitant dissection. The prevalence of dissection among stroke hospitalizations decreased across 10-year age groups from 7.2% (95% CI, 6.2%–8.1%) among persons younger than 30 years to 0.2% (95% CI, 0.1%–0.2%) among persons older than 80 years (p value for trend <0.001). However, the prevalence of hospitalizations for stroke and concomitant dissection increased from 5.4 (95% CI, 4.6–6.2) hospitalizations per million person-years among adults younger than 30 to 24.4 (95% CI, 21.0–27.9) hospitalizations per million person-years among adults older than age 80 (p value for trend <0.01).

Conclusion

In a nationally representative sample, the prevalence of hospitalizations for dissection-related stroke increased with age.

Cervical artery dissection is a well-described cause of stroke in the young; it comprises 9%–25% of all strokes in persons younger than 55 years and these patients typically have few established stroke risk factors.^1–5 The role of dissection as a cause of stroke among older adults is less clear. As trauma is a well-established cause of dissection, older adults may be at heightened risk of dissection given their risk of falls. On the contrary, older adults may be less susceptible to dissection due to changes in arterial compliance over time, gene–environment interactions, and differences in activities performed in older vs younger adults.^6–9 In addition, as advanced neuroimaging becomes more widely available, dissection may be more frequently identified, particularly among older people, who may be more likely to be imaged for neurologic symptoms.^10–12 We therefore sought to evaluate the prevalence of hospitalizations for cervical artery dissection across age groups. We hypothesized that the prevalence of hospitalizations for dissection-related ischemic stroke would remain constant across age groups.

Methods

Study Design

This is a cross-sectional study involving the National Inpatient Sample (NIS), which is a nationally representative, deidentified dataset maintained by the Agency for Healthcare Research and Quality.¹³

Standard Protocol Approvals, Registrations, and Patient Consents

This study was approved by the institutional review boards at Weill Cornell Medical College and Montefiore Medical Center, and the need for obtaining informed consent was waived at both institutions.

Population and Measurements

We evaluated all adult hospitalizations between 2012 and 2015 with a discharge diagnosis of stroke and a concomitant diagnosis of carotid or vertebral artery dissection. Ischemic stroke was defined by ICD-9-CM discharge diagnosis codes 433.x1, 434.x1, or 436 in any diagnosis code position in the absence of a primary discharge code for rehabilitation (V57) or any codes for subarachnoid hemorrhage (430), intracerebral hemorrhage (431), or brain trauma (800–804 and 850–854). This combination of codes has a sensitivity of 86% and specificity of 95% for the diagnosis of acute ischemic stroke.¹⁴ Cervical artery dissection was identified using ICD-9-CM codes 443.21 and 443.24 in any discharge diagnosis position. These ICD-9-CM codes have previously been validated to have a positive predictive value of 82.1% for dissection¹⁵; however, no data on their sensitivity and specificity have been reported. Therefore, to assess the reliability of these codes to identify patients with cervical artery dissection, we used data from the Cornell Acute Stroke Academic Registry (CAESAR) to identify 25 patients with acute ischemic stroke who were adjudicated to have a cervical artery dissection as the mechanism of stroke. In addition, we identified 25 patients with acute ischemic stroke who were adjudicated to have a mechanism of stroke other than dissection. The methods for stroke adjudication and etiologic subtype classification in CAESAR have been previously published.¹⁶ For each patient, we then evaluated whether ICD-9-CM codes 443.21 and 443.24 were present in any discharge diagnosis code position during the hospitalization for acute ischemic stroke. We performed a second validation study using data from 2 Montefiore Medical Center hospitals involving review of 40 hospitalizations with discharge diagnosis codes 443.21 and 443.24 and a random sample of 40 hospitalizations with other cerebrovascular diagnoses.

Statistical Analysis

Survey weights were utilized to ensure nationally representative estimates. The χ² test for trend was used to compare the prevalence of concomitant dissection among stroke hospitalizations across patient subgroups defined by age. We then used population estimates from the US census to calculate the prevalence of hospitalizations with stroke and concomitant dissection per million person-years across subgroups of age.¹⁷ Poisson regression and the Wald test for trend were used to evaluate whether the prevalence of hospitalizations for stroke and concomitant dissection per million person-years varied by age groups. To put these results in context, we compared the prevalence of hospitalizations for stroke and concomitant dissection per age group to the prevalence of hospitalizations for stroke without concomitant dissection per age group and to the prevalence of hospitalizations for dissection without stroke per age group.

We performed a subgroup analysis by sex and race. In a secondary analysis, we examined carotid vs vertebral artery dissections separately. In order to evaluate whether the presence of trauma could affect the association between dissection and age, we performed a sensitivity analysis in which we excluded patients who had concomitant head or neck trauma, defined as per prior studies by ICD-9-CM codes 800–806, 839.1, 839.2, 847.0, 850–854.1, 900.x, 920, 952.2, or 959.0.⁷ Statistical analysis was performed using Stata (version 14.0, College Station, TX).

Data Availability

The statistical analyses that support the findings of this study are available from the corresponding author upon reasonable request. The data that support the findings of this study may be requested from the Agency for Healthcare Research and Quality.

Results

In the validation assessments described above, the sensitivity of ICD-9-CM codes 443.21/443.24 was 80% (95% confidence interval [CI], 59.3%–93.2%) and the specificity was 100% (1-sided 97.5% CI, 86.3%–100%) at Weill Cornell Medical Center, and the sensitivity was 85.0% (95% CI, 70.2%–94.3%) and the specificity was 100% (1-sided 97.5% CI, 91.2%–100%) at Montefiore Medical Center. After pooling the data from both centers, the combined sensitivity was 83.1% (95% CI, 71.7%–91.2%) and the combined specificity was 100% (1-sided 97.5% CI, 94.5%–100%).

Between 2012 and 2015, we identified 17,320 (95% CI, 15,614–19,026) hospitalizations with ischemic stroke and a concomitant cervical artery dissection. The mean age of patients at the time of hospitalization was 52.7 (±15.9) years, and 6,696 (39%) of these hospitalizations involved women (table 1).

Table 1.

Characteristics of Patients With Cervical Artery Dissection and Stroke by Age Group

The prevalence of dissection among stroke hospitalizations decreased across age groups from 7.2% (95% CI, 6.2%–8.1%) among persons younger than 30 years to 0.2% (95% CI, 0.1%–0.2%) among persons older than 80 years (p value for trend <0.001) (figure 1A). Conversely, the prevalence of hospitalizations for stroke and concomitant dissection increased across age groups (p value for trend <0.01). There were 5.4 (95% CI, 4.6–6.2) hospitalizations for stroke and concomitant dissection per million person-years among adults younger than 30 and 24.4 (95% CI, 21.0–27.9) hospitalizations per million person-years among adults older than 80 (figure 1B).

Figure 1. Prevalence of Hospitalizations for Stroke-Related Dissection by Age Group.

(A) Prevalence of cervical artery dissection in patients with stroke by age group. (B) Prevalence of hospitalizations for stroke-related dissection by age group per million population.

Our results were unchanged in a subgroup analysis stratified by sex and race (figure 2, A–H; table 2) and in a sensitivity analysis in which we excluded patients who had concomitant head or neck trauma (table 2).

Figure 2. Prevalence of Hospitalizations for Stroke-Related Dissection by Age Group Among Subgroups.

(A) Prevalence of cervical artery dissection among men with stroke by age group. (B) Prevalence of cervical artery dissection among women with stroke by age group. (C) Prevalence of hospitalizations for men with stroke and cervical artery dissection by age group per million person-years. (D) Prevalence of hospitalizations for women with stroke and cervical artery dissection by age group per million person-years. (E) Prevalence of cervical artery dissection among White adults with stroke by age group. (F) Prevalence of cervical artery dissection among Black adults with stroke by age group. (G) Prevalence of hospitalizations for White adults with stroke and cervical artery dissection by age group per million person-years. (H) Prevalence of hospitalizations for Black adults with stroke and cervical artery dissection by age group per million person-years. (I) Prevalence of carotid artery dissection in adults with stroke by age group. (J) Prevalence of vertebral artery dissection in adults with stroke by age group. (K) Prevalence of hospitalizations for stroke and carotid artery dissection by age group per million person-years. (L) Prevalence of hospitalizations for stroke and vertebral artery dissection by age group per million person-years.

Table 2.

Trends in the Prevalence of Hospitalizations for Stroke and Concomitant Dissection per Million Person-Years

Carotid dissection made up a larger proportion of dissection-related stroke among older adults, whereas vertebral dissection made up a larger proportion of dissection-related stroke among younger adults (p value < 0.001) (table 1). The prevalence of both carotid and vertebral artery dissection among stroke hospitalizations decreased across age groups (p value for trends <0.001) (figure 2, I and J). The prevalence of hospitalizations for stroke and concomitant carotid dissection increased across age groups (p value for trend = 0.03); however, the prevalence of hospitalizations for stroke and concomitant vertebral artery dissection did not vary across age groups (p value for trend = 0.42) (tables 1 and 2; figures 2, K and L).

Both the prevalence of hospitalizations for dissection without stroke and the prevalence of hospitalizations for stroke without dissection increased across age groups (p values for trends <0.01). There were 15.3 (95% CI, 13.6–17.0) hospitalizations for dissection without stroke per million person-years among adults younger than 30 and 38.9 (95% CI, 34.4–43.4) hospitalizations for dissection without stroke per million person-years among adults older than 80 (table 3). There were 69.9 (95% CI, 66.3–73.5) hospitalizations for stroke without concomitant dissection per million person-years among adults younger than 30 and 15,281 (95% CI, 14,922–15,639) hospitalizations for stroke without concomitant dissection per million person-years among adults older than age 80 (table 3).

Table 3.

Trends in the Prevalence of Hospitalizations for Dissection and Stroke per Million Person-Years

Discussion

In a nationally representative sample, we found that although cervical artery dissection accounts for a smaller proportion of ischemic stroke hospitalizations among older adults, the prevalence of hospitalizations for dissection-related stroke increases with age.

Epidemiologic data regarding the incidence of cervical artery dissection are scarce and largely come from single-center registries.¹⁸ Although it is established that dissection is a relatively common cause of stroke in the young, it is less often considered as an etiology of stroke among older adults. Recent data, however, have highlighted the occurrence of dissection in older adults.^10–12 Our study adds novel data suggesting that although dissection accounts for a relatively smaller proportion of all stroke in older adults, the overall prevalence of dissection as a cause of stroke increases by age. The cause of dissection in older adults is uncertain and may range from similar genetic or environmental risk factors as in the young, or conditions common in the elderly, including infections, trauma, and atherosclerosis.^12,19,20 Finally, as prior data suggest that older patients may not have typical signs and symptoms of dissection, such as neck pain, the higher prevalence of dissection-related stroke by age may in part be due to a delayed diagnosis of dissection. However, this would not explain the similar trends we found in the prevalence of hospitalizations for dissection-related stroke and dissection without stroke by age.¹⁰

We found that the peak prevalence of dissection-related stroke was between 30 and 49 years of age in women as compared to 50 to 89 years of age in men. These results are consistent with prior reports suggesting an age-associated difference between men and women with dissection-related stroke.¹¹ Puerperium may explain some of the differences in the sex-related prevalence of dissection-related stroke as we recently found that pregnancy was associated with a heightened risk for cervical artery dissection.²¹ In addition, as men typically have a younger age at development of cardiovascular disease, this may explain the higher prevalence of dissection-related stroke in men vs women between 30–39 and 40–49 years of age, especially if atherosclerosis is a contributing cause of dissection.²² We found similar trends in the prevalence of hospitalizations for dissection-related stroke in both White and Black adults, suggesting that the prevalence of hospitalizations for dissection-related stroke increases with age, regardless of race. However, given the heterogenous ancestry of those who self-identify as Black in the United States, future studies are necessary to determine the role of ancestry, race, and ethnicity as it pertains to dissection. We found that carotid dissection made up a larger proportion of dissection-related stroke among older persons whereas vertebral dissection made up a larger proportion of dissection-related stroke among younger persons; these findings corroborate the results of previously published studies.^12,19 Further mechanistic studies are warranted to help explain the difference in the incidence of carotid vs vertebral artery dissection across age groups. Finally, our results were unchanged when we excluded patients who had concomitant head or neck trauma, suggesting that trauma may not fully account for the increased incidence of dissection-related stroke among older persons.

This study has several important limitations. First, it relied on claims data that inherently lack important key elements including radiographic characteristics of the dissection, stroke severity, or etiology of dissection. Second, we lacked data on genetic risk factors or ancestry, which may influence the risk of dissection. Third, we calculated the sensitivity and specificity of ICD-9-CM codes 443.21 and 443.24 at primary and comprehensive stroke centers and the reliability of these codes may vary across hospitals included in the NIS that may not be stroke centers. Fourth, as our study cohort is derived from the NIS, we may have undercounted the prevalence of dissection due to lack of availability of advanced neuroimaging at many hospitals. Although this may have led to a lower prevalence of dissection-related-stroke, it would have likely caused a nondifferential misclassification. Fifth, because the NIS identifies hospitalizations, it is possible that some patients were included more than once due to recurrence or worsening symptoms. Sixth, patients with dissection who were asymptomatic or who had had minor symptoms may have been less likely to have sought medical care; the decision to seek medical care may have differed between older and younger individuals and thus led to ascertainment bias. Seventh, some of the dissections may have been chronic and not directly related to the acute ischemic stroke at the time of hospitalization. Eighth, we were unable to differentiate between intracranial and extracranial dissections.

In a nationally representative sample, we found that although cervical artery dissection accounts for a greater proportion of ischemic stroke hospitalizations in younger adults, the prevalence of hospitalizations for dissection-related stroke increases with age.

Glossary

CAESAR

Cornell Acute Stroke Academic Registry

CI

confidence interval

ICD-9-CM

International Classification of Diseases, 9th Revision, Clinical Modification

NIS

National Inpatient Sample

Appendix. Authors

Study Funding

Dr. Murthy is supported by NIH grant K23NS105948. Dr. Liberman is supported by NIH grant K23NS107643. Dr. Merkler supported by AHA grant 18CDA34110419 and the Leon Levy Foundation.

Disclosure

Dr. Atalay, Dr. Piran, and A. Chatterjee report no disclosures. Dr. Murthy is supported by NIH grant K23NS105948. Dr. Navi serves as a DSMB member for the PCORI-funded TRAVERSE trial and has received personal fees for medicolegal consulting on stroke. Dr. Liberman is supported by NIH grant K23NS107643. J. Dardick and C. Zhang report no disclosures. Dr. Kamel serves as co-PI for the NIH-funded ARCADIA trial, which receives in-kind study drug from the BMS-Pfizer Alliance and in-kind study assays from Roche Diagnostics; serves as Deputy Editor for JAMA Neurology; serves as a steering committee member of the Medtronics Stroke AF trial (uncompensated); serves on an endpoint adjudication committee for a trial of empagliflozin for Boehringer-Ingelheim; and has served on an advisory board for Roivant Sciences related to Factor XI inhibition. Dr. Merkler is supported by AHA grant 18CDA34110419 and the Leon Levy Foundation and has received personal fees for medicolegal consulting on neurologic disorders. Go to Neurology.org/Nhttps://n.neurology.org/lookup/doi/10.1212/WNL.0000000000011420 for full disclosures.