A recent systematic review of methodology and reporting standards for quality of life and behavioural outcome measures in epilepsy found that from 46 RCTs of AEDs, a total of 52 different measures were used. There was a failure to apply such measures in a consistent manner and due to the wide variety of measures used, it is not possible to make valid comparisons between studies and consequently to draw any meaningful conclusions about the effect of AEDs on behaviour or quality of life.²⁸⁹ Evidence level 3,4

The Commission on the Outcome Measurement in Epilepsy (COME report) highlighted that behavioural outcome measures have been used selectively for assessment of therapeutic approaches, assessment of impairments, disabilities and handicaps and assessment of ongoing treatment.^{290, 291} Evidence level 3,4

8.1 Seizure frequency

The COME report stated that “When measured as a continuous variable, seizure frequency is by far the most sensitive measure of efficacy and should be used whenever possible”. This information is most commonly gathered through patient diaries, which have been found to vary considerably in quality. Alternatives to the reporting of seizure frequency have included percentage reduction in seizures, the categorisation of seizure frequency into percentage groups or time to first seizure recurrence.²⁹⁰ Evidence level 4

Assessments should always include seizure frequency and date of last seizure.

8.2 Seizure severity scales

With regards to seizure severity, three frequently-used measurement scales were highlighted by COME: The Veterans Administration Seizure Frequency and Rating Scale, The Liverpool Seizure Severity Scale and the National Hospital Seizure Severity Scale. There was little evidence of sensitivity to change and until more data is available no seizure severity scale can be recommended as a standard outcome measure in clinical trials.²⁹⁰

8.3 Assessment of the interictal state- adverse events

A number of Adverse Events scales have been developed. COME highlighted the Veterans Administration Systemic and Neurotoxicity Scale to evaluate common problems caused by AEDs and the Veterans Administration Composite Rating Scale which provides a composite score that represents the patient’s overall status. The Liverpool Adverse Events Profile covers similar areas and has some evidence of sensitivity to change. COME recommended that further work is required in order to improve and expand the assessment of adverse events.

8.4 Neuropsychological assessment

COME recognised the importance of determining the effects of AEDs on cognition, although clinical investigations to date have been limited due to failures to adhere to basic standards of method, design, analysis and neuropsychological evaluation. A systematic review of RCTs which evaluated neuropsychological outcome of AEDs found that 98 different tests had been used and only five were used in more than one study. Of these, only two test batteries were developed and standardised specifically for epilepsy (Dodrill’s Neuropsychological Battery and the FePsy system).¹³⁸

8.5 Quality of life

There are three main reasons for difficulty in drawing consistent conclusions from existing quality of life (QoL) research. Firstly, there is a lack of consistency regarding the implementation of QoL measures. A systematic review of methodology of quality of life and behavioural outcome measures in epilepsy, reported that often no reason was provided for test selection and when generic measures were used, there was little evidence of their reliability, validity or sensitivity in this population.²⁸⁹ Secondly, many domains in epilepsy-specific measures are important to people with epilepsy but are not amenable to change during a clinical trial eg driving. Finally, most of the available evidence on QoL measures relates to instrument validation rather than concerning their practical use. Consequently it is difficult to assess the level of change that would be clinically meaningful.^{289, 290, 291}

8.6 Conclusions

The COME report concluded that the International League Against Epilepsy should establish clear recommendations for minimum standards for the use of outcome measures in clinical trials and practice. There should be uniformity in the selection of measures with evidence of reliability, validity and sensitivity to change. Reporting of the outcomes should also be standardised. The aims of the measurements, the rationale for selecting a particular measure, the psychometric properties of the scale and its previous application to epilepsy should be considered. Caution should also be taken with regard to the use of outcome measures in routine clinical practice. Indeed, COME states that “the assessment of seizure response in routine clinical practice, for the most part, still relies on clinical judgement, a nebulous but pragmatic reality of treating an individual patient”.²⁹⁰

In conclusion, there is no single outcome scale that can be recommended for use in clinical practice.

The assessment of seizure severity, adverse effects and the impact of epilepsy on the quality of life should be considered in assessing individuals in clinical practice. Care should be taken to use appropriate outcome measures.