Abstracts

Rationale: Currently, a large variety of anti-epileptic drugs (AEDs) is available. In daily clinical practice, many different treatment strategies are used, often by “trial and error”. For general (not-specialised) neurologists it may be challenging to keep an overview of which medications are indicated or contra-indicated in particular patient profiles. We developed an electronic decision tool to support appropriate treatment choice in daily clinical practice. Methods: Using the RAND Appropriateness Method, a panel of 13 Belgian experts in epilepsy assessed the appropriateness of 9 AEDs for mono-therapy and 15 AEDs for combination therapy. Assessments were related to several clinical conditions including type of seizures, response to previous treatment, co-morbidities, concomitant therapies, age, female reproductive system, and interactions between AEDs. The panel first rated individually the appropriateness of all options (n=1128), using a 9-point scale. During a plenary meeting, results were discussed and adaptations were made to the rating structure. Then a second rating round was conducted in which the revised options (n=1428) were re-rated. Based on the median score and the extent of agreement, a panel recommendation was calculated for each option: “inappropriate”, “appropriate”, or “use with caution”. These recommendations were embedded in an electronic tool supporting 3 types of clinical decisions: 1. initial mono-therapy, 2. mono-therapy after failure of the initial treatment, and 3. combination therapy. After selecting a patient profile, the tool displays the panel judgment on the appropriateness of the various AEDs in red (inappropriate), green (appropriate), and yellow (use with caution). For options being “inappropriate” or “use with caution”, pop-up screens provide information on the reasons behind. Validity of the recommendations and feasibility of the tool in clinical practice was discussed with two groups of Belgian general neurologists. Results: Disagreement amongst the panellists was low and dropped from 2.6% in the first round to 0.7% in the second round (all options together). Options in relation to type of seizures and co-morbidities were more clear-cut (either appropriate or inappropriate) than those related to concomitant treatments, age, female reproductive factors and interactions between AEDs: for this second group of variables, the percentage of “use with caution” varied between 41 and 69, compared to 19-25% for the first group. The two groups of general neurologists expected the program to be very useful in daily practice, particularly because of its conciseness and user-friendliness. Conclusions: The RAND Appropriateness Method proved to be useful in determining the appropriateness of AEDs in relation to several patient-specific clinical conditions. The electronic decision tool, showing the panel recommendations, was perceived as being very feasible for daily practice. Periodic review of the tool is needed to adapt the recommendations to new scientific insights. Funding: supported by Pfizer Belgium