Epidemiology of traumatic brain injury-associated epilepsy and early use of anti-epilepsy drugs: An analysis of insurance claims data, 2004-2014.

Patient Population:

Eligibility/Inclusion:

• De-identified inpatient, outpatient, pharmacy claims, and insurance coverage;
• Patients aged ≥2 years who were enrolled for a minimum baseline period of 2 years without diagnostic codes indicating epilepsy or seizures and without prescriptions for AEDs; Included patients aged 65 years and older with supplemental Medicare coverage, and individuals with Medicaid coverage.
• Children aged < 2 years whose records included no diagnostic codes for epilepsy, seizures, and AED prescriptions since birth.
• Operational definition of TBI / epilepsy was based on ICD-9-CM codes.
  • Patients were eligible for inclusion if these codes were assigned in hospital, emergency department, or outpatient care site.
  • Dates of TBI were assigned as the initial date of service of the medical encounter including these codes.
• Cases of PTE were identified among the TBI cases who met the case definition of epilepsy beginning 7 days or more after TBI.
• AEDs: Categorized 28 AEDs included in study into 13 groups. Counted prescribed AED anytime from the date of TBI to the diagnosis of PTE or end of follow-up. Included patients on more than one AED.

Control Group:

• Non-TBI controls were selected, matched one-to-one by enrollment year, gender, age, region and insurance plan type with TBI cases.  

Exclusion:

• Individuals with missing age, sex, region (if CCMC), and health plan type.  

Evaluated databases examining the incidence of early seizures (within seven days after TBI) and cumulative incidence of PTE, the hazard ratios (HR) of PTE by age, gender, TBI severity, early seizure and AED use (carbamazepine, clonazepam, divalproex sodium, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, phenytoin, pregabalin, topiramate, acetazolamide).

Used backward selection to build the final Cox proportional hazard model and conducted multivariable survival analysis to obtain estimates of crude and adjusted HR (cHRs, aHRs) of PTE and 95% confidence intervals (CI).

Primary endpoints to address the incidence of the:

• Probability of developing PTE within 9 years after TBI,
• Risk factors associated with PTE,
• Prevalence of anti-epileptic drug (AEDs) use.

Secondary endpoint to evaluate the:

• Effectiveness of using AEDs prophylactically after TBI to prevent the development of PTE.

TBI and control groups

Cumulative incidence of epilepsy among TBI patients is significantly higher than that among control groups (p < 0.0001).

TBI group only

Among all 2,059,870 TBI patients (with no prior indicators of epilepsy), a total of 11,150 (0.5%) experienced early seizures, and a total of 31, 521 (1.5%) developed PTE. Among these PTE patients, 3678 (11.7%) had early seizures.

In contrast, among all 2,028,349 TBI patients who did not develop PTE, 7472 (0.4%) had early seizures.

The risk of developing PTE among older age groups compared with the youngest age group yielded adjusted Hazard Ratios that were higher, and with advancing age.

The risk of developing PTE for individuals with early seizures was 37 times higher than that for individuals without early seizure.

The risk of PTE for individuals with severe TBI was 13.7 times higher than that for individuals with mild TBI.

AED use:

Prophylactic use of AEDs, in general, did not prevent or reduce the incidence of PTE; however, use of acetazolamide was associated with a decreased risk of PTE.

Overall, 93% did not use AED. The percentage of AED use increased with age and TBI severity. Findings varied by age:

• 99.3% of 0–4-year-olds did not use AEDs, decreasing to 86.4% among 65–74-year-olds.
• Among patients without early seizures, 93.2% did not use AED, compared to 47.2% among patients with early seizures.
• 93.7% of those with mild TBI did not use AED compared to 80.7% among patients with severe TBI.

A single AED was prescribed for 5.5% TBI patients.

• TBI patients who at some point developed PTE, most commonly used levetiracetam (6.5%).
• TBI patients who did not develop PTE most commonly used gabapentin (1.9%).
• TBI patients with early seizures most commonly used levetiracetam (16.8%) while those without early seizures most commonly used gabapentin (1.9%).
• Among the patients with mild TBI, clonazepam and gabapentin were most commonly used, while among the patients with moderate and severe TBI, levetiracetam and gabapentin were most commonly used.
• Among those patients who used AEDs, levetiracetam (28.9%) and phenytoin (23.8%) were most commonly used in the first 7 days. Clonazepam and gabapentin were most commonly used after 7 days.

• Human, retrospective study.
• Dataset used is claim-based and susceptible to miscoding and missing information.
  • ICD-9-CM-based coding can result in misclassifications of the occurrence of both TBI and epilepsy. Codes do not distinguish new-onset from long-established cases.
  • AED use may be misclassified in some patients.
• AED received during hospitalization are not recorded in the database.
• Claimed-based data represent prescriptions filled only; it cannot be determined whether patients adhered to medication used the AEDs as prescribed.
• Misclassification of some preexisting epilepsy cases as new PTE cases may occur also.
• The minimum 2-year baseline is not always long enough to detect prevalent epilepsy cases.
• Claims data analyzed did not represent persons lacking health insurance (a status more frequent among younger adults) and may not fully represent the insured population of the United States.

The authors identified advancing age, early seizures, and increased severity TBI as risk factors of PTE.

Most of the individuals did not receive AED after TBI; however, there was no evidence suggesting AEDs helped to prevent PTE with the possible exception of acetazolamide.

Further studies to test the efficacy of acetazolamide in preventing PTE as well as evaluate the effectiveness of AEDs in the first seven days or more after TBI are warranted.