You are here

Article Text

Article menu

Download PDFPDF

Neurosurgery
Research paper
Risk of epilepsy after traumatic brain injury: a retrospective population-based cohort study
  1. Chun-Chieh Yeh1,2,
  2. Ta-Liang Chen3–5,
  3. Chaur-Jong Hu6,
  4. Wen-Ta Chiu7,
  5. Chien-Chang Liao3–5
  1. 1Department of Surgery, China Medical University Hospital, Taichung, Taiwan
  2. 2Graduate Institute of Clinical Medical Science, China Medical University, Taichung, Taiwan
  3. 3Department of Anesthesiology, Taipei Medical University Hospital, Taipei, Taiwan
  4. 4Health Policy Research Center, Taipei Medical University Hospital, Taipei, Taiwan
  5. 5School of Medicine, Taipei Medical University, Taipei, Taiwan
  6. 6Department of Neurology, Taipei Medical University, Taipei, Taiwan
  7. 7Graduate Institute of Injury Prevention and Control, Taipei Medical University, Taipei, Taiwan
  1. Correspondence to Assistant Professor Chien-Chang Liao, Department of Anesthesiology, Taipei Medical University Hospital, 252 Wuxing Street, Taipei 110, Taiwan; jacky48863027{at}yahoo.com.tw

Abstract

Objective To investigate the associated risk of epilepsy after traumatic brain injury (TBI) in a population-based retrospective cohort study.

Methods Using Taiwan's National Health Insurance Research Database of reimbursement claims, we conducted a retrospective cohort study of 19 336 TBI patients and 540 322 non-TBI participants aged ≥15 years as reference group. Data on newly developed epilepsy after TBI with 5–8 years’ follow-up during 2000 to 2008 were collected. HRs and 95% CIs for the risk of epilepsy associated with TBI were analysed with multivariate Cox proportional hazards regressions.

Results Compared with the non-TBI cohort, the adjusted HRs of developing epilepsy among TBI patients with skull fracture, severe or mild brain injury were 10.6 (95% CI 7.14 to 15.8), 5.05 (95% CI 4.40 to 5.79) and 3.02 (95% CI 2.42 to 3.77), respectively. During follow-up, men exhibited higher risks of post-TBI epilepsy. Patients who had mixed types of cerebral haemorrhage were at the highest risk of epilepsy compared with the non-TBI cohort (HR 7.83, 95% CI 4.69 to 13.0). The risk of post-TBI epilepsy was highest within the first year after TBI (HR 38.2, 95% CI 21.7 to 67.0).

Conclusions The risk of epilepsy after TBI varied by patient gender, age, latent interval and complexity of TBI. Integrated care for early identification and treatment of post-trauma epilepsy were crucial for TBI patients.

  • Head Injury
  • Epilepsy
  • Epidemiology

https://doi.org/10.1136/jnnp-2012-302547

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Traumatic brain injury (TBI) is one of the most serious forms of trauma worldwide, causing 1.1 million emergency visits, 235 000 hospitalisations and 50 000 deaths every year in the USA.1–3 Epilepsy causes significant functional disability among trauma survivors, and epilepsy after TBI is the most common cause of acquired seizure disorders in adults.4 While several studies have investigated associations between epilepsy and TBI on a limited scale,4–7 large-scale population-based studies concerning post-traumatic epilepsy are scarce.8 ,9

    Aneger et al conducted a population-based cohort study showing strong correlation between TBI and subsequent epilepsy,8 but this did not relate different ages, subtypes of severe brain injury or various skull bone fractures to the risk of post-trauma epilepsy. Another population-based study by Christensen et al reported that the risk of epilepsy among patients who had mild brain injury, skull bone fracture or severe brain injury was increased for more than 10 years after trauma.9 This study was confined to children and young adults, and mainly focused on patients with facial bone to skull bone fractures. Facial bone injuries such as mandible or nasal bone fracture were reported to be infrequently associated with increased risk of post-trauma epilepsy. Moreover, neither of these population-based studies considered medical risk factors that might affect relative risk of epilepsy, such as migraine, mental disorders, liver cirrhosis or dialysis.10–14

    We conducted a retrospective, population-based cohort study to examine the long-term risk of epilepsy in adult populations after TBI categorised by different subtypes of head injury and severity. The chronology of onset for epilepsy among different subtypes of severe brain injury and skull bone fracture was examined using time-dependent analysis. From these data, we could further assess and predict the relative risk for post-TBI epilepsy among different patterns of TBI and other factors such as gender. These findings should be helpful in developing an integrated healthcare strategy for this common post-injury complication.

    Methods

    Data source

    A retrospective, population-based cohort study was conducted using medical claims from Taiwan's National Health Institute Research Database (NHIRD), which has been described previously.15 Our database randomly selected 1 million insured subjects from the NHIRD database comprising healthcare data from the medical records of all beneficiaries, more than 99% of the 23.37 million people in Taiwan. The International Classification of Diseases, Clinical Modification, ninth edition (ICD-9-CM) code was applied to define procedures and diseases. Insurance reimbursement claims data used in this study were available from NHIRD for public access, and patient identification has been encoded to ensure confidentiality. The study was approved by the National Health Research Institutes, Taiwan. Some details have been described previously.16 ,17

    Study subjects

    We identified all patients aged 15 years and older; these were divided into a TBI group and a group without TBI from the database as the cases of the study cohort. The TBI group enrolled patients aged over 15 years and diagnosed with a de novo TBI between 2000 and 2003, with exclusion of relevant diagnosis of TBI before 2000. Each subject was either followed-up from the index date until 31 December 2008 or was censored. The follow-up time, in person-years, was calculated for each subject until the diagnosis of epilepsy or until being censored because of death, withdrawal from the insurance system or loss to follow-up. The non-TBI group included the remaining people who did not have a history of TBI throughout the following period. To avoid confounding influence from other known risk factors for epilepsy, we excluded patients who had a previous history of epilepsy (ICD-9-CM 345), brain tumour (ICD-9-CM 191, 225.0, 225.1, 225.2), stroke (ICD-9-CM 430–438) or brain surgery before the end of 2008 in both the TBI and non-TBI groups.18–20 In addition, to focus on long-term risk of epilepsy after TBI, patients who died within 30 days after TBI were also excluded.

    The enrolled patients with TBI were categorised into mild brain injury (ICD-9-CM 850), severe brain injury (ICD-9-CM 851–854) and skull bone fracture (ICD-9-CM 800–804). Mild brain injury is defined as brain concussion without brain structural damage. By contrast, severe brain injury means structural brain injury, including brain contusion, subdural haemorrhage (SDH), epidural haemorrhage (EDH), subarachnoid haemorrhage (SAH) and intracranial haemorrhage. Skull bone fractures included pure skull base fracture, pure skull vault fracture and multiple skull bone fracture, but pure facial bone fracture was excluded. If more than one type of head injury appears, categorisation of the patients follows this hierarchy of severity: severe brain injury, skull bone fracture, mild brain injury.

    Covariates

    Other medical conditions—mental disorder (ICD-9-CM 290–319), migraine (ICD-9-CM 346), liver cirrhosis (ICD-9-CM 571) and end-stage renal disease under haemodialysis treatment—were considered as confounding factors due to the documented increased risk of epilepsy in patients with these coexisting diseases.10–13 These were defined as covariates only before the occurrence of epilepsy in the study group or the whole follow-up period in the control group without occurrence of epilepsy. Low income was defined as patients being certified by the Taiwan Health Insurance Program for waived medical copayment.14 The calculation of population density (persons/km2) was described in a previous study and categorises quartiles as low, moderate, high and very high urbanisation areas.14 ,16 ,21

    Statistical analyses

    Our study used χ2 tests and t-tests to compare the sociodemographic characteristics and coexisting diseases between people with and without TBI. We calculated the HRs with 95% CIs for risk of epilepsy after TBI, adjusting for age, sex, low income, urbanisation, mental disorders, migraine, dialysis and liver cirrhosis in multivariate Cox proportional hazard regression models. The risk and incidence of epilepsy among subtypes of TBI were also determined. Cox proportional hazard models were used to analyse the risk of post-trauma epilepsy among various subtypes of TBI (brain contusion, SAH, epidural haematoma, subdural haematoma, intracerebral haemorrhage (ICH) or a combination of at least two types of these brain injuries), in different types of skull bone fracture (skull vault vs skull base facture) and by gender. Finally, the latent interval of epilepsy after TBI (0–1, 1–2, 2–3, 3–4 and ≥4 years), patient's age at TBI (15–29, 30–39, 40–49, 50–59, 60–69, ≥70 years) and risk of post-trauma epilepsy were also analysed in Cox regression models. HRs of post-trauma epilepsy in each type of brain injury were calculated in three different models after adjusting age, sex, level of urbanisation, income and various coexisting diseases. A p value <0.05 indicated statistical significance, and all tests were two-tailed. Analyses were performed with SAS V.9.1.

    Results

    Of 559 658 people aged ≥15 years enrolled between 2000 and 2003, 19 336 patients visited outpatient services or were admitted with a diagnosis of TBI; the 540 322 people without TBI were controls (table 1). Risk of epilepsy in the TBI group was significantly higher than that of the non-TBI group (1.9% vs 0.3%, p<0.0001). Men comprised 55.4% of patients with TBI. The mean age of patients with brain injury was 39.1 years, which was comparable to controls. Patients with TBI had higher percentages of low income and residence in less urbanised areas, and higher rates of coexisting diseases including mental disorders, migraine, chronic renal insufficiency receiving haemodialysis and liver cirrhosis.

    View this table:
    Table 1

    Characteristics of patients with brain injury and controls*

    Table 2 shows risk of epilepsy in various types of TBI when compared with the non-TBI group. Compared with patients without brain injury, patients with skull bone fracture had a 17 times higher risk of epilepsy (crude HR 17.2, 95% CI 11.6 to 25.5). This risk was nearly eight times higher for patients with severe brain injury (crude HR 7.78, 95% CI 6.8 to 8.9) and was four times higher for patients with mild brain injury (crude HR 3.9, 95% CI 3.1 to 4.8). After adjustment for age, gender, low income and urbanisation in model 2, the HRs of epilepsy in patients with skull bone fracture, severe brain injury and mild brain injury were 15.9 (95% CI 10.7 to 23.5), 6.7 (5.9 to 7.7) and 3.6 (2.9 to 4.5), respectively, compared with patients without brain injury. With further adjustment for coexisting in model 3, the risk of epilepsy was highest in patients with skull bone fracture (HR 10.6, 95% CI 7.1 to 15.8), followed by severe brain injury (HR 5.1, 95% CI 4.4 to 5.8) and mild brain injury (HR 3.0, 95% CI 2.4 to 3.8) compared with patients without brain injury.

    View this table:
    Table 2

    Risk of epilepsy in association with various severities of brain injury in Cox proportional hazard regression models

    Table 3 shows relative risks of developing epilepsy by gender and subtypes of severe brain injury and skull bone fracture. Compared with brain contusion, ICH was associated with the highest risk of epilepsy (HR 4.7, 95% CI 2.8 to 8.6), followed by SDH (HR 4.1, 95% CI 2.2 to 7.4), EDH (HR 3.3, 95% CI 1.3 to 8.8) and SAH (HR 2.8, 95% CI 1.5 to 5.0) after adjusting for age, sex, low income, urbanisation and coexisting disease. Combination of at least two subtypes of brain injury previously described showed an even higher risk of epilepsy (HR 7.8, 95% CI 4.7 to 13.0). The risk of epilepsy after skull base fracture is not significantly different from skull vault fracture. Men were at significantly higher risk of epilepsy than women after brain injury (HR 1.7, 95% CI 1.3 to 2.1).

    View this table:
    Table 3

    Risk of epilepsy in association with different types of brain injury in Cox proportional hazard regression models

    Table 4 shows the relative risk of epilepsy according to patient's age at brain injury and by latent interval after different types of brain injury. TBI increased risk of epilepsy in all age groups, but risk of epilepsy increased before adjustment with increasing patient age at brain injury. After adjustment, the risk of epilepsy after skull bone fracture (HR 38.2, 95% CI 21.7 to 67.0), severe brain injury (HR 14.8, 95% CI 11.7 to 18.8) and mild brain injury (HR 6.7, 95% CI 4.4 to 10.0) was highest in the first year after TBI. The risk of epilepsy increased after skull bone fracture and severe brain injury, and lasted for at least 4 years after TBI.

    View this table:
    Table 4

    Onset time and relative risk of developing epilepsy after first medical visit for various severities of traumatic brain injury

    Discussion

    This cohort study investigated the long-term risk of developing epilepsy after TBI among various subtypes of TBI patients. Skull bone fracture, severe brain injury and mild brain injury posed a significantly higher long-term risk of developing epilepsy compared with the control group. Focusing on patients with severe brain injury, the risk of epilepsy in patients with combined injuries was the highest. The risk of developing epilepsy increased among all age groups of TBI patients and continued to be higher until the fourth year after injury.

    Our study identified low income, younger age, male gender, living in less-urbanised areas, mental disorders, migraine, liver cirrhosis and dialysis status as potential factors associated with traumatic injuries.22 Liver cirrhosis, as a risk factor for TBI, might be linked to alcohol consumption and cirrhosis-associated encephalopathy.23 It is well known that mental disorders, migraine, liver cirrhosis and end-stage renal disease all correlate with increased risk of epilepsy.10–13 After adjusting for potential confounding factors, we validated the associated risks of developing epilepsy in patients with TBI; these risks differed from those noted in previous population-based cohort studies.8 ,9 Limited information is available about gender's effect on post-traumatic epilepsy, although there is a complex interaction between gender and occurrence of epilepsy.24–26 In this study, men had a higher risk of developing epilepsy after TBI compared with women. A higher risk of post-traumatic epilepsy in male adults with TBI might partially result from a higher incidence of traumatic head injury in the men in this study. Gender-specific hormones might also influence occurrence of epilepsy, but this would require further study to confirm.26 After adjustment, our study showed that patients with skull fracture had the highest risk of developing epilepsy, followed by those with severe brain injury and those with mild brain injury. As skull fractures cause high-impact damage to the skull and deeper trauma to brain matter, especially in patients with skull base fracture, it is understandable that patients with skull fractures had the highest epilepsy risk. Consistent with the findings of Christensen et al, risks of epilepsy after TBI were significantly higher than in the control population, even in the mild head injury group.9

    Unlike a previous study which showed that the risk of post-TBI epilepsy lasted for more than 10 years after head injury in children and young adults,9 our investigation indicated that increased risk of post-TBI epilepsy persisted only 4 years in adult populations. The discrepancy might result from the different age composition of the studied groups, as it is well known that older populations have higher mortality and poorer outcomes after head injury.27 Higher mortality rates in studied populations might be the reason that the long-term risk of post-trauma epilepsy in adults was not increased beyond 4 years after head injury. Severe head injuries such as subdural haematoma or cerebral contusion result in higher risks for post-trauma epilepsy than do mild or moderate head injuries.8 However, detailed analysis of various subtypes of severe head injuries regarding the risk of developing epilepsy is still lacking. In this study, ICH had the highest risk of post-trauma epilepsy, followed by SDH,  EDH, SAH and brain contusion. The severity ranking for risk of epilepsy in various subtypes of severe head injuries in this study could be partially explained by the differences in neuronal injuries after trauma.28 Brain parenchyma injuries could lead to bleeding and iron deposition being released. The deposited iron in brain parenchyma would cause neurone death and glial scarring, and subsequently increased epileptic activity.28 ICH and SDH exhibited a higher risk of devastating brain parenchyma injury than other severe brain injury; therefore, both ICH and SDH had the highest risk of post-trauma epilepsy.

    This study has several limitations. First, we used insurance claims data which lack information on clinical risk scores (such as Glasgow Coma Scale), lesion characteristics (location, size or types), and biochemical measures that have been reported as predictors of post-TBI epilepsy.6 Second, the database used for the longitudinal follow-up was mainly focused on survivors after TBI, as patients with fatalities after head injuries were excluded. Thus the prevalence of epilepsy and the population with severe trauma were underestimated. However, the study aimed to investigate the chronological risk of epilepsy after TBI, rather than to analyse the relative risk of epilepsy among various types of brain injuries. Even being underestimated, severe head injury was proven to be associated with increased risk of epilepsy until 4 years after trauma. Similarly, either ICH or SDH might be associated with increased risk of short-term mortality in comparison to EDH, and the risk of epilepsy in either ICH or SDH might also be underestimated.29 ,30 Third, this study only included TBI patients with emergency care or inpatient care. Thus the number of TBI patients might be underestimated because TBI patients with outpatient care who may have very minor TBI were not included and some patients with very minor TBI may not seek medical treatment. Fourth, though the accuracy of major diagnosis codes in the NHIRD was validated in the previous study,31 validity of comorbidity and complication codes might still be one of the study limitations.

    In summary, this population-based cohort study demonstrated a risk of epilepsy after TBI in adults and a chronological risk in the first 4 years after trauma. Among the pathologies of injuries, patients with mixed types of brain haemorrhage had the highest risk of developing post-trauma epilepsy. These findings could provide clinicians with a spectrum to understand potential risk factors relating to post-trauma epilepsy. The combined care policy can be developed by neurosurgeons, neurologists and associated medical team members for these specific patient populations.

    Acknowledgments

    This study is based in part on data obtained from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health and managed by the National Health Research Institutes. The interpretation and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health or National Health Research Institutes.

    References

    1. ↵
      1. DeKosky ST,
      2. Ikonomovic MD,
      3. Gandy S
      . Traumatic brain injury: football, warfare, and long-term effects. N Engl J Med 2010;363:1293–6.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      Centers for Disease Control and Prevention. Rates of hospitalization related to traumatic brain injury: nine states, 2003. MMWR Morb Mortal Wkly Rep 2007;56:167–70.
      OpenUrlPubMed
    3. ↵
      1. Yates PJ,
      2. Williams WH,
      3. Harris A,
      4. et al
      . An epidemiological study of head injuries in a UK population attending an emergency department. J Neurol Neurosurg Psychiatry 2006;77:699–701.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Lowenstein DH
      . Epilepsy after head injury: an overview. Epilepsia 2009;50:4–9.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Frey LC
      . Epidemiology of posttraumatic epilepsy: a critical review. Epilepsia 2003;44:11–17.
      OpenUrl
    6. ↵
      1. Raymont V,
      2. Salazar AM,
      3. Lipsky R,
      4. et al
      . Correlates of posttraumatic epilepsy 35 years following combat brain injury. Neurology 2010;75:224–9.
      OpenUrlCrossRef
    7. ↵
      1. Angeleri F,
      2. Majkowski J,
      3. Cacchio G,
      4. et al
      . Posttraumatic epilepsy risk factors: one-year prospective study after head injury. Epilepsia 1999;40:1222–30.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Annegers JF,
      2. Hauser WA,
      3. Coan SP,
      4. et al
      . A population-based study of seizures after traumatic brain injuries. N Engl J Med 1998;338:20–4.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Christensen J,
      2. Pedersen MG,
      3. Pedersen CB,
      4. et al
      . Long-term risk of epilepsy after traumatic brain injury in children and young adults: a population-based cohort study. Lancet 2009;373:1105–10.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Eleftheriadis N,
      2. Fourla E,
      3. Eleftheriadis D,
      4. et al
      . Status epilepticus as a manifestation of hepatic encephalopathy. Acta Neurol Scand 2003;107:142–4.
      OpenUrlPubMed
    11. ↵
      1. Alldredge B
      . Seizure risk associated with psychotropic drugs: clinical and pharmacokinetic considerations. Neurology 1999;53:S68–75.
      OpenUrl
    12. ↵
      1. Glenn CM,
      2. Astley SJ,
      3. Watkins SL
      . Dialysis-associated seizures in children and adolescents. Pediatr Nephrol 1992;6:182–6.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Haut SR,
      2. Bigal ME,
      3. Lipton RB
      . Chronic disorders with episodic manifestations: focus on epilepsy and migraine. Lancet Neurol 2006;5:148–57.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Chang CC,
      2. Hu CJ,
      3. Chang H,
      4. et al
      . Postoperative adverse outcomes in surgical patients with epilepsy: a population-based study. Epilepsia 2012;53:987–94.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Wang IK,
      2. Chang SN,
      3. Liao CC,
      4. et al
      . Hypertensive disorders in pregnancy and preterm delivery and subsequent stroke in Asian women: a retrospective cohort study. Stroke 2011;42:716–21.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Liao CC,
      2. Su TC,
      3. Sung FC,
      4. et al
      . Does hepatitis C virus infection increase the risk for stroke? a population-based cohort study. PLoS One 2012;7:e31527.
      OpenUrlCrossRefPubMed
    17. ↵
      1. Chou YC,
      2. Liao CC,
      3. Su LT,
      4. et al
      . Stroke rehabilitation is associated with a reduction in dementia risk: a population-based retrospective cohort study. J Rehabil Med 2012;44:319–24.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Morris HH,
      2. Estes ML,
      3. Gilmore R,
      4. et al
      . Chronic intractable epilepsy as the only symptom of primary brain tumor. Epilepsia 1993;34:1038–43.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Camilo O,
      2. Goldstein LB
      . Seizures and epilepsy after ischemic stroke. Stroke 2004;35:1769.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Herman ST
      . Epilepsy after brain insult: targeting epileptogenesis. Neurology 2002;59:S21–6.
      OpenUrl
    21. ↵
      1. Shih CC,
      2. Su YC,
      3. Liao CC,
      4. et al
      . Patterns of medical pluralism among adults: results from the 2001 National Health Interview Survey in Taiwan. BMC Health Serv Res 2010;10:191.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Poole GV,
      2. Lewis JL,
      3. Devidas M,
      4. et al
      . Psychopathologic risk factors for intentional and nonintentional injury. J Trauma 1997;42:711.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Neiman J
      . Alcohol as a risk factor for brain damage: neurologic aspects. Alcohol Clin Exp Res 1998;22:346s–51s.
      OpenUrlCrossRefPubMed
    24. ↵
      1. Englander J,
      2. Bushnik T,
      3. Duong TT,
      4. et al
      . Analyzing risk factors for late posttraumatic seizures: a prospective, multicenter investigation. Arch Phys Med Rehabil 2003;84:365–73.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Diaz-Arrastia R,
      2. Gong Y,
      3. Fair S,
      4. et al
      . Increased risk of late posttraumatic seizures associated with inheritance of APOE epsilon4 allele. Arch Neurol 2003;60:818–22.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Vagnerova K,
      2. Koerner IP,
      3. Hurn PD
      . Gender and the injured brain. Anesth Analg 2008;107:201–14.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Tokutomi T,
      2. Miyagi T,
      3. Ogawa T,
      4. et al
      . Age-associated increases in poor outcomes after traumatic brain injury: a report from the Japan Neurotrauma Data Bank. J Neurotrauma 2008;25:1407–14.
      OpenUrlCrossRefPubMed
    28. ↵
      1. Willmore LJ
      . Post-traumatic epilepsy: cellular mechanisms and implications for treatment. Epilepsia 1990;31:S67–73.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Bullock MR,
      2. Chesnut R,
      3. Ghajar J,
      4. et al
      . Surgical management of acute epidural hematomas. Neurosurgery 2006;58:S7–15.
      OpenUrlPubMed
    30. ↵
      1. Haselsberger K,
      2. Pucher R,
      3. Auer LM
      . Prognosis after acute subdural or epidural haemorrhage. Acta Neurochir 1988;90:111–16.
      OpenUrlCrossRefPubMed
    31. ↵
      1. Cheng CL,
      2. Kao YH,
      3. Lin SJ,
      4. et al
      . Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 2011;20:236–42.
      OpenUrlCrossRefPubMed
    View Abstract

    Footnotes

    • CCY and TLC contributed equally

    • Correction notice This paper has been amended since it was published Online First. In table 1, “Characteristics of patients with brain injury and controls”, in the third column, the total number of patients is wrong. It has now been changed to the correct number of 559,658.

    • Contributors CCY and TLC: conception and design, analysis and interpretation of the data, drafting the article and final approval of the version to be published. CJH and WTC: interpretation of the data, critical revision of the manuscript for important intellectual content and final approval of the version to be published. CCL: conception and design, analysis and interpretation of the data, critical revision of the manuscript for important intellectual content and final approval of the version to be published.

    • Funding This research was supported by a Foundation for Anaesthesia Education and Research fellowship grant, Taipei Medical University, Taipei, Taiwan.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Ethics approval National Health Research Institutes, Taiwan.