Antiseizure medications (ASM) are the first-line treatment for epilepsy (the most common severe chronic neurological disorder), and many patients become completely seizure-free when prescribed an appropriate drug. An increasing number of ASMs have been introduced over the years, with approximately 20 now in common use. Few ASMs are effective for all seizure types and some are suboptimal due to specific patient characteristics such as age, sex, epilepsy syndrome, comorbidities, adverse effect profile, and drug interaction potential.
Having alternative options improves the opportunity to tailor treatment to the individual and select other medications in case of poor tolerability, but can also lead to inappropriate or suboptimal selection, particularly when epilepsy is managed by non-specialist health professionals. This risk is compounded by the shortage of neurologists and epilepsy specialists in many places.
To assist healthcare professionals in the management of epilepsy, a pragmatic web-based algorithm was developed aimed at facilitating appropriate selection of ASM for monotherapy. The algorithm takes into account several patient-specific variables and provides a ranking of ASM in order of likely suitability for an individual based on the best available scientific evidence supplemented by expert judgment. In addition to listing a ranking of individualized treatment options, the web-based application provides a summary of prescribing information for each of the suggested medications.
| Methods |
Using the available evidence and a consensus process based on a Delphi panel, a group of epilepsy experts developed an algorithm for the selection of ASM, according to the type of seizure and the presence of relevant clinical variables (age, sex, comorbidities and co -medications). The algorithm was implemented in a web-based application that was tested and improved in an iterative process.
| Results |
The algorithm classifies ASMs that are considered appropriate for each seizure type or combination of seizure types into three groups, with Group 1 ASMs considered preferred, Group 2 considered second-line, and Group 3 considered third-line.
Depending on the presence of relevant clinical variables, the classification of individual ASMs is adjusted in the prioritization scheme to adapt recommendations to the characteristics of the individual. The algorithm is available in a web application.
| Discussion |
The proposed algorithm, which incorporates key demographic details and 17 clinical variables, is intended to assist healthcare professionals who are not experts in epileptology in the selection of ASM monotherapies for patients with seizure onset at 10 years of age or older. . The instrument, implemented in an electronic application freely available on the Internet, provides several ASM options with different levels of prioritization (group 1, 2 or 3), and gives health care the final responsibility of the professional to decide which ASM is the most appropriate for the individual patient.
Because the algorithm is designed to facilitate ASM selection, it includes only those clinical variables considered most important for treatment decisions, an approach that differs from phenomenological and syndromic classifications intended to differentiate seizure types independently of the therapeutic implications.
Although a suite of algorithm-based applications have been developed to help healthcare professionals diagnose seizures, none so far have addressed the need for patient-tailored medication selection. Correctly identifying the epilepsy syndrome and seizure types is a necessary first step in clinical practice, but does not necessarily lead to the choice of an appropriate treatment.
Many studies conducted in different settings have documented evidence of incorrect or suboptimal drug selection in patients with epilepsy. The instrument of this work has the potential to reduce inappropriate prescribing, as it provides advice not only on medication choice, but also on dosing and titration schedules. The algorithm is easy to use and requires less than 2 minutes to respond (less than 1 minute once familiar with it), and provides the user with a range of ASMs ranked in order of suitability.
This should facilitate wide use among professionals interested in accessing simple advice in the management of their patients. In addition to medication selection and links to additional resources, the explanation of how the algorithm adjusted the classification based on individual variables has a strong educational component, helping physicians without epilepsy training understand important aspects related to the choice of the appropriate medication.
Some limitations must be recognized. The algorithm is applicable only to patients whose seizures begin at age 10 or older and is primarily designed for use in a monotherapy setting. Developing a similar tool for a younger population, or for a polytherapy setting, would be much more complex. Although the algorithm was constructed taking into consideration the available evidence on individual ASM properties and clinical guideline recommendations, adjusting ASM selection to individual characteristics cannot be completely evidence-based and must take into account personal judgment, which justified the use of a Delphi approach.
Medication choice in the algorithm is based on seizure type rather than epilepsy syndrome, but combinations of seizure types associated with syndromes with onset at age ≥10 years are fully taken into account. The clinical variables (modifiers) considered are limited and some of them are only broadly defined, due to the need to compromise between efficiency and practicability.
Consequently, we are aware that not all information that should ideally be considered in ASM selection is included in the algorithm, and this may in some cases lead to less than optimal advice. In particular, aspects subject to geographic variability such as availability, label indications, cost and reimbursement of medications are not considered; Furthermore, the algorithm does not help determine whether treatment should be initiated or deferred in an individual person.
Due to these limitations, the instrument is offered as a tool to facilitate therapeutic decisions and cannot replace the user’s clinical judgment.
A final limitation that needs to be discussed relates to the lack of formal external validation. To date, the instrument has primarily undergone internal testing. Although feedback received from several epilepsy specialists who have tested it has been generally favorable, we recognize the need for larger, formal validation studies. A study currently underway is evaluating how ASM selections made by expert epileptologists around the world compare to the algorithm’s performance in a variety of case scenarios.
This research is expected to confirm our preliminary findings on reliability and external validity within the limitations of using predefined cases. The next step will be to extend these studies to real-world field tests, and obtain broader feedback on the performance and usefulness of the instrument from expert users. Finally, it is contemplated to evaluate the acceptability and perceived usefulness within the community for which the instrument is primarily designed, that is, health professionals who are not experts in the management of epilepsy, including (but not limited to) those who live in resource-limited environments.
We view the algorithm as a tool that will evolve as experience with its use progresses, new scientific evidence emerges, and new drugs are developed. The ultimate goal is to improve the quality of care for people with epilepsy and contribute to reducing the "treatment gap" worldwide.
| Conclusion |
The proposed algorithm is easy to use, requires less than 2 minutes to complete, and provides the user with a range of suitable treatment options to choose from. This should facilitate its widespread use and contribute to improving the management of epilepsy for healthcare providers who wish to receive advice, particularly those who lack special expertise in the field.
