Newer antiepileptic drug use and other factors decreasing hospital encounters

Racial and Ethnic Differences in Antiseizure Medications Among People With Epilepsy on Medicaid: A Case of Potential Inequities

Background and Objectives

Being on a newer, second-, and third-generation antiseizure medication (ASM) may represent an important marker of quality of care for people with epilepsy. We sought to examine whether there were racial/ethnic differences in their use.

Methods

Using Medicaid claims data, we identified the type and number of ASMs, as well as the adherence, for people with epilepsy over a 5-year period (2010-2014). We used multilevel logistic regression models to examine the association between newer-generation ASMs and adherence. We then examined whether there were racial/ethnic differences in ASM use in models adjusted for demographics, utilization, year, and comorbidities.

Results

Among 78,534 adults with epilepsy, 17,729 were Black, and 9,376 were Hispanic. Overall, 25.6% were on older ASMs, and being solely on second-generation ASMs during the study period was associated with better adherence (adjusted odds ratio: 1.17, 95% confidence interval [CI]: 1.11-1.23). Those who saw a neurologist (3.26, 95% CI: 3.13-3.41) or who were newly diagnosed (1.29, 95% CI: 1.16-1.42) had higher odds of being on newer ASMs. Importantly, Black (0.71, 95% CI: 0.68-0.75), Hispanic (0.93, 95% CI: 0.88-0.99), and Native Hawaiian and Other Pacific Island individuals (0.77, 95% CI: 0.67-0.88) had lower odds of being on newer ASMs when compared with White individuals.

Discussion

Generally, racial and ethnic minoritized people with epilepsy have lower odds of being on newer-generation ASMs. Greater adherence by people who were only on newer ASMs, their greater use among people seeing a neurologist, and the opportunity of a new diagnosis point to actionable leverage points for reducing inequities in epilepsy care.

Healthcare utilization and cost outcomes for a multicenter first seizure and new onset epilepsy clinic

RATIONALE

A First Seizure/New Onset Epilepsy (FS/NOE) protocol was implemented to ensure proper evaluation by an epileptologist and improve overall care for patients. We compared healthcare utilization and cost incurred by patients pre and post protocol implementation.

METHODS

Clinical data were retrospectively collected from the EMR and cost data from the financial database. Patients were identified by FS event and grouped into either the pre-implementation (pre-FSC) or post-implementation cohort (post-FSC). Pre-FSC patients were seen between January 2014-December 2015 and post-FSC between March 2016-January 2018. Utilization outcomes include time from FS to neurology appointment, MRI, and electroencephalogram (EEG). Cost outcomes included the annualized median difference in pre versus post costs for ER, inpatient, outpatient or ambulatory, and total hospital services. Cost and utilization outcomes were collected within 90 days or 6 months post first-seizure event. Pre and post cohorts were compared using Kaplan-Meier analysis and Cox proportional hazard models for time-to-event outcomes, multivariable median regression models for cost differences and negative binomial regression models for utilization analyses. Models were adjusted for age, sex, health insurance, and comorbidities.

RESULTS

One-hundred and fifty six patients were included with 84 (53.8%) pre- and 72 (46.2%) post-FSC patients. Kaplan-Meier and Cox regression results indicated post-FSC patients had significantly faster time-to-first neurology appointment (5.0 vs. 20.9 days, p < .001; Adjusted Hazard Ratio (HR) = 5.98, p < .001), time-to-MRI (9.0 vs. 27.0 days; p = 0.005; HR = 1.88, p = .021) and EEG (3.6 vs. 48.6 days, p < .001; HR = 9.01, p < .001). A total of 138 patients had at least one cost in the financial database. For 6-month follow-up period, post-FSC patients had higher adjusted all-cause total median costs (+$830, p = 0.009) and outpatient costs (+$1203, p < .001) but lower ED costs (-245, p = 0.073), not significant. Results were similar for seizure-related costs. Similarly, Post-FSC patients had a significantly higher likelihood of all-cause (Adjusted Rate Ratio (ARR) = 1.41, p = .029) and outpatient utilization (ARR = 1.72, p = .008) but lower ED utilization (ARR = 0.54, p < .001).

CONCLUSIONS

Implementation of the FSC decreased time to evaluation by a neurologist and time to diagnostic workup. Ultimately, total healthcare costs and ambulatory costs increased but ED costs and utilization were reduced. It is our hypothesis that faster access to initial care and diagnosis would result in better control of seizures and reduce long-term costs and utilization. Further research over a longer duration of time across a broader population is needed to evaluate the full implications of an epilepsy specialist-populated FSC.

Sociodemographic factors associated with the first administration of anti-seizure medication in patients with focal epilepsy in Western China

Background

Epilepsy is a severe chronic neurologic disease with a prevalence of 0.7% worldwide; anti-seizure medications (ASMs) are the mainstay of epilepsy treatment. The effects of sociodemographic factors on the characteristics of initial treatment in patients with newly diagnosed focal epilepsy in Western China are unknown. This study was conducted to explore sociodemographic factors associated with initial treatment characteristics.

Methods

Patients with focal epilepsy on continuous ASM treatment who visited to our epilepsy center at Sichuan Provincial People’s Hospital between January 2018 and December 2019 were recruited. Data on initial treatment status and sociodemographic variables were obtained from the patients with a questionnaire designed by our researchers. We examined whether sociodemographic factors were associated with epileptic patients’ access to neurologists and prescriptions of individual ASMs.

Results

A total of 569 patients completed this study. We found that patients with a higher education level, aged < 16 years, and with a higher household disposable income were more likely to receive treatment from a neurologist than their counterparts. Patients with a lower personal income level and who were treated at a junior hospital were more likely to receive prescriptions for carbamazepine, and those who were younger than 16 years were less likely to receive prescriptions for carbamazepine and oxcarbazepine. Patients with a higher education level, with a higher household disposable income level, who were younger than 16 years, and who were treated at a senior hospital were more likely to receive prescriptions for levetiracetam than their counterparts. Adult, female patients with focal epilepsy treated at a senior hospital were more likely to receive prescriptions for lamotrigine.

Conclusions

This observation suggests that sociodemographic characteristics are associated with access to neurologists and prescriptions of individual antiepileptic drugs. These data may help public health officials establish guidelines for doctors and distribute resources according to the needs of different patient groups.

Real-world analysis of hospitalizations in patients with epilepsy and treated with perampanel

OBJECTIVES

(1) To evaluate risk of hospitalization following initiation of perampanel (pre- and post-analysis) and (2) to compare hospitalization rates following initiation of perampanel vs lacosamide.

METHODS

Patients were identified from Symphony Health's Patient Integrated Database if they had a prescription for perampanel (July 1, 2014-June 30, 2016). Patients 4-11 years of age with any partial-onset seizure (POS) or ≥12 years of age with any POS or primary generalized tonic-clonic seizure (GTCS) (pre-post); or ≥12 years of age (perampanel vs lacosamide). The first fill of perampanel ("index date") marked the start of the analysis period. Patients had ≥1 additional fill for perampanel and ≥2 diagnoses for epilepsy or nonfebrile convulsion diagnosis during pre-index (based on ICD-9/ICD-10 codes). Patients were matched using a 1:1 propensity scoring method for the perampanel vs lacosamide analysis. Primary outcome was hospitalization during the one year following medication initiation.

RESULTS

Pre- and post-perampanel: N = 1771 (mean age 34 years, 55% female). One-year all-cause hospitalization risk ratio was 0.76 (P < .05) and 36.2% with hospitalization during the pre-period vs 29.5% in the follow-up. One-year epilepsy-related inpatient hospitalization risk ratio was 0.72 (P < .05) and 30.8% with hospitalization during the pre-period vs 23.9% during follow-up. In the perampanel and lacosamide cohorts, N = 1717 per cohort after matching, most baseline demographics were balanced. A higher percentage of subjects were prescribed ≥3 anti-seizure medications for perampanel vs lacosamide (60.5% vs 57.7%, P < .001). The perampanel cohort had a 9.6% reduction in all-cause hospitalizations vs 5.8% for the lacosamide cohort (P < .05). Epilepsy-related hospitalizations decreased from the pre-index rate by 9.9% for perampanel and 8.3% for lacosamide (P < .05). Among those with baseline hospitalizations, perampanel was associated with a 59.9% reduction in all-cause hospitalizations vs 48.6% for lacosamide (P < .05), and for epilepsy-related hospitalizations, a reduction of 65.0% vs 58.9%, respectively (P < .05).

SIGNIFICANCE

Perampanel was associated with a significant reduction in one-year hospitalization risk.

Comparative Economic Outcomes in Patients with Focal Seizure Initiating First-Line Eslicarbazepine Acetate Monotherapy versus Generic Antiseizure Drugs

Objective

To examine the association between initiating first-line (1L) monotherapy with eslicarbazepine acetate (ESL) vs a generic antiseizure drug (ASD) and healthcare resource utilization (HCRU) and charges in adults with treated focal seizures (FS).

Methods

This was a retrospective analysis of Symphony Health’s Integrated Dataverse^® open-source claims data. Two cohorts were identified as having initiated 1L monotherapy with ESL or literature-defined generic ASDs. Linear regression models with person fixed effects and inverse probability treatment weights assessed the relative additional changes in HCRU and charges among patients who received ESL compared to generic ASD.

Results

A total of 250 and 43,220 patients initiated ESL (48.3 years; 57.2% female) or a generic ASD (54.5 years; 58.1% female), respectively. Compared to patients initiating a generic ASD, patients treated with ESL had additional reductions of 11.8 percentage points in the likelihood of any all-cause outpatient visits (P<0.001), 7.4 percentage points in the likelihood of any emergency department (ED) visits (P=0.013), and 22.7 percentage points in the likelihood of any FS-related outpatient visits (P<0.001). Patients initiating ESL had greater reductions in mean charges for all-cause medical ($2620; P=0.002), outpatient ($1995; P=0.005), and non-FS-related medical ($2708; P<0.001) services. Patients initiating ESL had greater relative increases in mean total prescription ($1368; P<0.001) and ASD-related prescription ($1636; P<0.001) charges, but greater relative reductions in non-ASD prescription ($269; P=0.032) charges. The increases in prescription charges were of a lower magnitude than the decreases in medical charges.

Conclusion

Initiation of ESL as 1L monotherapy was associated with statistically significantly greater reductions in any use of several all-cause and FS-related services, number of visits, and charges compared to initiation of a generic ASD as 1L monotherapy in patients with FS. Initiation of a generic ASD as 1L monotherapy was associated with significantly smaller increases in total prescription charges and ASD-related prescription charges.

Association of Levels of Specialized Care With Risk of Premature Mortality in Patients With Epilepsy

Importance

Patients with epilepsy are at an elevated risk of premature mortality. Interventions to reduce this risk are crucial.

Objective

To determine if the level of care (non-neurologist, neurologist, or comprehensive epilepsy program) is negatively associated with the risk of premature mortality.

Design, Setting, and Participants

In this retrospective open cohort study, all adult patients 18 years or older who met the administrative case definition for incident epilepsy in linked databases (Alberta Health Services administrative health data and the Comprehensive Calgary Epilepsy Programme Registry [CEP]) inclusive of the years 2002 to 2016 were followed up until death or loss to follow-up. The final analyses were performed on May 1, 2019.

Exposures

Evaluation by a non-neurologist, neurologist, or epileptologist.

Main Outcomes and Measures

The outcome was all-cause mortality. We used extended Cox models treating exposure to a neurologist or the CEP as time-varying covariates. Age, sex, socioeconomic deprivation, disease severity, and comorbid burden at index date were modeled as fixed-time coefficients.

Results

A total 23 653 incident cases were identified (annual incidence of 89 per 100 000); the mean age (SD) at index date was 50.8 (19.1) years and 12 158 (50.3%) were women. A total of 14 099 (60%) were not exposed to specialist neurological care, 9554 (40%) received care by a neurologist, and 2054 (9%) received care in the CEP. In total, 4098 deaths (71%) occurred in the nonspecialist setting, 1481 (26%) for those seen by a neurologist, and 176 (3%) for those receiving CEP care. The standardized mortality rate was 7.2% for the entire cohort, 9.4% for those receiving nonspecialist care, 5.6% for those seen by a neurologist, and 2.8% for those seen in the CEP. The hazard ratio (HR) of mortality was lower in those receiving neurologist (HR, 0.85; 95% CI, 0.77-0.93) and CEP (HR, 0.49; 95% CI, 0.38-0.62) care. In multivariable modeling, specialist care, the age at index, and disease severity were retained in the final model of the association between specialist care and mortality.

Conclusions and Relevance

Exposure to specialist care is associated with incremental reductions in the hazard of premature mortality. Those referred to a comprehensive epilepsy program received the greatest benefit.

Innovative approaches reaching underserved and rural communities to improve epilepsy care: A review of the methodology of the Connectors Project

OBJECTIVE

The Connectors Project, a collaboration between the Epilepsy Foundation and UCB Pharma, was a multiyear project designed to improve epilepsy care in underserved communities. A core objective of the Connectors Project was to pilot new and innovative approaches to epilepsy awareness and education in rural and underserved areas, including standardized curricula for healthcare providers and patients.

METHODS

A series of consensus conferences explored opportunities and barriers to epilepsy care throughout the United States including access to local Epilepsy Foundations, neurologists, and epilepsy centers. Data from QuintilesIMS™ were examined for access to newer antiepileptic drugs (AEDs)-a proxy for quality of epilepsy care-in different regions. State factors (e.g., local epilepsy foundation office, access to newer vs. older AEDs, and geographic density and diversity) were used in selecting four states as examples of rural and underserved areas to pilot the awareness and educational programs. For each state, a work team assessed challenges and opportunities, tailored educational curricula, and developed strategies for effective delivery of the educational programs. Interventions were held between June 2016 and June 2017. Interventions consisted of outreach and awareness programs, in-person health education to healthcare providers and patients/families, and digital health education.

RESULTS

Michigan, Nevada, Oklahoma, and West Virginia were identified as pilot states representing geographically diverse areas, ranging from a state with a large high-density population center with several epilepsy centers and a local Epilepsy Foundation office (Michigan) to a state with predominately rural areas and a few small urban cores, two epilepsy centers, and no in-state Epilepsy Foundation office (West Virginia). State work teams tailored interventions and examined options for type, intent, ease of use, and impact. All states implemented outreach and awareness initiatives and in-person health education for patients and healthcare providers; use of digital health education was variable. Measurement of the interventions was agreed to be performed by the use of patient and physician surveys and reevaluation of data from QuintilesIMS for access to newer AEDs.

CONCLUSION

Local Epilepsy Foundation offices successfully connected healthcare providers in rural and underserved areas to epilepsy education designed to enhance quality care in epilepsy. Educational opportunities for people with epilepsy and their families addressed critical gaps in accessing quality epilepsy care and self-management. Tailored and innovative educational approaches can be used to increase awareness levels and to overcome geographic challenges in reaching underserved populations. Relationship building and repeated, consistent engagement with healthcare providers and patients can assist in improving communication and self-management skills among patients with epilepsy.

Evaluation of long-term antiepileptic drug use in patients with temporal lobe epilepsy: Assessment of risk factors for drug resistance and polypharmacy

PURPOSE

To evaluate risk factors for drug resistance and polypharmacy in patients with temporal lobe epilepsy.

METHODS

Patients with temporal lobe epilepsy, treated for more than 5 years, completed questionnaires on antiepileptic drug use and effect. Logistic regression models were used for analysis of risk factors.

RESULTS

Of 135 patients included in the study, 65% were classified as drug resistant and 41% identified as using polypharmacy. Poor effects associated with first-choice antiepileptic drug were reported by 59% of the patients, and 70% reported poor effects of second-line treatment. The most frequently used first-generation antiepileptic drugs had a similar mean effect to those of second-generation. Univariate regression analyses showed a significant association between drug resistance and mesial temporal sclerosis, seizure onset below 18 years, and lack of family history of epilepsy. However, multivariate regression analysis showed no association with any demographic or clinical features. Unsuccessful treatment with the first antiepileptic drug increased the risk of drug resistance by 18 times, and the risk of poor effect from the second antiepileptic drug by 9 times. Disease duration was associated with annual risk for drug resistance of 7% and for polypharmacy of 5%.

CONCLUSIONS

A poor effect from initial pharmacotherapy is the only early risk factor for drug resistance found in this study. Long disease duration increases the risk of drug resistance and polypharmacy. Second-generation antiepileptic drugs provide no additional effect for poor responders to first-generation drugs.

Epilepsy in Asia: Disease burden, management barriers, and challenges

This article reviews the burden of epilepsy in Asia, the challenges faced by people with epilepsy, and the management of epilepsy. Comparison is made with other parts of the world. For this narrative review, data were collected using specified search criteria. Articles investigating the epidemiology of epilepsy, diagnosis, comorbidities and associated mortality, stigmatization, and treatment were included. Epilepsy is a global health care issue affecting up to 70 million people worldwide. Nearly 80% of people with epilepsy live in low- and middle-income countries with limited resources. People with epilepsy are prone to physical and psychological comorbidities, including anxiety and depression, which can negatively impact their quality of life. Furthermore, people with epilepsy are at higher risk of premature death than people without epilepsy. Discrimination or stigmatization of people with epilepsy is common in Asia and can affect their education, work, and marriage opportunities. Access to epilepsy treatment varies throughout Asia. Although highly advanced treatment is available in some countries, up to 90% of people with epilepsy are not adequately treated or are not treated with conventional antiepileptic therapy in resource-limited countries. People in remote areas often do not receive any epilepsy care. First-generation antiepileptic drugs (AEDs) are available, but usually only in urban areas, and second-generation AEDs are not available in all countries. Newer AEDs tend to have more favorable safety profiles than first-generation AEDs and provide options to tailor therapy for individual patients, especially those with comorbidities. Active epilepsy surgery centers are present in some countries, although epilepsy surgery is often underutilized given the number of patients who could benefit. Further epidemiologic research is needed to provide accurate epilepsy data across the Asian region. Coordinated action is warranted to improve access to treatment and care.

Changing the neurology policy landscape in the United States: Misconceptions and facts about epilepsy

Epilepsy has a relatively high prevalence, and diagnosis and treatment are often challenging. Seizure freedom without significant side effects is the ultimate goal for both physicians and patients, but not always achievable. In those cases, the treatment goals of patients and providers may differ. In the United States, many clinicians continue to prescribe older AEDs, even though newer AEDs have a more desirable safety and tolerability profile, fewer drug-drug interactions, and are associated with lower epilepsy-related hospital visits. Newer AEDs are more commonly prescribed by neurologists and epilepsy center physicians, highlighting the importance of access to specialty care. We report that antiepileptic drugs are not the dominant cost driver for patients with epilepsy and costs are considerably higher in patients with uncontrolled epilepsy. Poor drug adherence is considered a main cause of unsuccessful epilepsy treatment and is associated with increases in inpatient and emergency department admissions and related costs. Interventions and educational programs are needed to address the reasons for nonadherence. Coverage policies placing a higher cost burden on patients with epilepsy lead to lower treatment adherence, which can result in higher future health care spending. Epilepsy is lagging behind other neurological conditions in terms of funding and treatment innovation. Increased investment in epilepsy research may be particularly beneficial given current funding levels and the high prevalence of epilepsy.

Attention-deficit/hyperactivity disorder medication and seizures

OBJECTIVE

Individuals with attention-deficit/hyperactivity disorder (ADHD) are at increased risk of seizures, but there is uncertainty about whether ADHD medication treatment increases risk among patients with and without preexisting seizures.

METHODS

We followed a sample of 801,838 patients with ADHD who had prescribed drug claims from the Truven Health MarketScan Commercial Claims and Encounters databases to examine whether ADHD medication increases the likelihood of seizures among ADHD patients with and without a history of seizures. First, we assessed overall risk of seizures among patients with ADHD. Second, within-individual concurrent analyses assessed odds of seizure events during months when a patient with ADHD received ADHD medication compared with when the same individual did not, while adjusting for antiepileptic medications. Third, within-individual long-term analyses examined odds of seizure events in relation to the duration of months over the previous 2 years patients received medication.

RESULTS

Patients with ADHD were at higher odds for any seizure compared with non-ADHD controls (odds ratio [OR] = 2.33, 95% confidence interval [CI] = 2.24-2.42 males; OR = 2.31, 95% CI = 2.22-2.42 females). In adjusted within-individual comparisons, ADHD medication was associated with lower odds of seizures among patients with (OR = 0.71, 95% CI = 0.60-0.85) and without (OR = 0.71, 95% CI = 0.62-0.82) prior seizures. Long-term within-individual comparisons suggested no evidence of an association between medication use and seizures among individuals with (OR = 0.87, 95% CI = 0.59-1.30) and without (OR = 1.01, 95% CI = 0.80-1.28) a seizure history.

CONCLUSIONS

Results reaffirm that patients with ADHD are at higher risk of seizures. However, ADHD medication was associated with lower risk of seizures within individuals while they were dispensed medication, which is not consistent with the hypothesis that ADHD medication increases risk of seizures.

Quality Indicator for Epilepsy Treatment 15 (QUIET-15): Intervening after recurrent seizures in the elderly

In this study, we examined the provision of care to older adults with epilepsy and compliance with the "Quality Indicator for Epilepsy Treatment 15" (QUIET-15) measure. We analyzed 2008-2010, 5% random sample of Medicare beneficiaries augmented with data from all beneficiaries who identified as a minority with claims related to seizures (780.3x) or epilepsy (345.xx). Of 36,912 identified epilepsy cases, 12.6% had ≥1 emergency room (ER) visit for seizure(s). For those who presented to ER, among those taking anti-epileptic drugs (AEDs), AED was changed in 15.4%, dose adjusted in 19.7%, and stopped in 14.9%; among those not taking AED, therapy was initiated in 68.5%. In adjusted logistic regressions, African-Americans were more likely to have recurrent seizures than Whites (OR 1.41, 95%CI 1.27-1.56), while Asians were less likely to have recurrent seizures (OR 0.71, 95%CI 0.57-0.89). There were no significant racial/ethnic differences in the likelihood of a post-seizure intervention. The chance of seizure recurrence leading to ER visit decreased with age and increased with the number of comorbidities. Patients with seizure recurrence were more likely to be taking an enzyme-inducing AED (OR 1.69, 95%CI 1.57-1.82) and receiving Part D Low Income Subsidy (OR 1.36, 95%CI 1.22-1.51). The probability of AED change after a seizure was higher for patients with ≥4 comorbidities (OR 1.69, 95%CI 1.25-2.27), patients who saw a neurologist (OR 1.49, 95%CI 1.30-1.70), and patients who were taking an enzyme-inducing AED (OR 1.47, 95%CI 1.27-1.71). Overall, a minority of Medicare beneficiaries experienced seizure recurrence that resulted in an ER visit. However, only half of them received treatment concordant with QUIET-15. Though racial differences were observed in occurrence of seizures, none were noted in the provision of care.

Neurologist adherence to clinical practice guidelines and costs in patients with newly diagnosed and chronic epilepsy in Germany

PURPOSE

The aim of this study was to evaluate physician adherence to the German Neurological Society guidelines of 2008 regarding initial monotherapy and to determine the cost-of-illness in epilepsy.

METHODS

This was an observational cohort study using health data routinely collected at 55 outpatient neurology practices throughout Germany (NeuroTransData network). Data on socioeconomic status, course of epilepsy, anticonvulsive treatment, and direct and indirect costs were recorded using practice software-based questionnaires.

RESULTS

One thousand five hundred eighty-four patients with epilepsy (785 male (49.6%); mean age: 51.3±18.1years) were enrolled, of whom 507 were newly diagnosed. Initial monotherapy was started according to authorization status in 85.9%, with nonenzyme-inducing drugs in 94.3% of all AEDs. Drugs of first choice by guideline recommendations were used in 66.5%. Total annual direct costs in the first year amounted to €2194 (SD: €4273; range: €55-43,896) per patient, with hospitalization (59% of total direct costs) and anticonvulsants (30%) as the main cost factors. Annual total direct costs decreased by 29% to €1572 in the second year, mainly because of a 59% decrease in hospitalization costs. The use of first choice AEDs did not influence costs. Chronic epilepsy was present in 1077 patients, and total annual direct costs amounted to €1847 per patient, with anticonvulsants (51.0%) and hospitalization (41.0%) as the main cost factors. Potential cost-driving factors in these patients were active epilepsy and focal epilepsy syndrome.

CONCLUSION

This study shows excellent physician adherence to guidelines regarding initial monotherapy in adults with epilepsy. Newly diagnosed patients show higher total direct and hospital costs in the first year upon diagnosis, but these are not influenced by adherence to treatment guidelines.

Changing the approach to treatment choice in epilepsy using big data

PURPOSE

A UCB-IBM collaboration explored the application of machine learning to large claims databases to construct an algorithm for antiepileptic drug (AED) choice for individual patients.

METHODS

Claims data were collected between January 2006 and September 2011 for patients with epilepsy > 16 years of age. A subset of patient claims with a valid index date of AED treatment change (new, add, or switch) were used to train the AED prediction model by retrospectively evaluating an index date treatment for subsequent treatment change. Based on the trained model, a model-predicted AED regimen with the lowest likelihood of treatment change was assigned to each patient in the group of test claims, and outcomes were evaluated to test model validity.

RESULTS

The model had 72% area under receiver operator characteristic curve, indicating good predictive power. Patients who were given the model-predicted AED regimen had significantly longer survival rates (time until a treatment change event) and lower expected health resource utilization on average than those who received another treatment. The actual prescribed AED regimen at the index date matched the model-predicted AED regimen in only 13% of cases; there were large discrepancies in the frequency of use of certain AEDs/combinations between model-predicted AED regimens and those actually prescribed.

CONCLUSIONS

Chances of treatment success were improved if patients received the model-predicted treatment. Using the model's prediction system may enable personalized, evidence-based epilepsy care, accelerating the match between patients and their ideal therapy, thereby delivering significantly better health outcomes for patients and providing health-care savings by applying resources more efficiently. Our goal will be to strengthen the predictive power of the model by integrating diverse data sets and potentially moving to prospective data collection.