Epilepsy & Behavior 22 (2011) 388–390
Contents lists available at ScienceDirect
Epilepsy & Behavior j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / ye b e h
Brief Communication
Interrater reliability of the international consensus definition of drug-resistant epilepsy: A pilot study Xiao-ting Hao a, b, Irina S.M. Wong a, Patrick Kwan a, c,⁎ a b c
Division of Neurology, Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong, China Department of Neurology, West China Hospital, Sichuan University, Chengdu, China Departments of Medicine and Neurology, The University of Melbourne, Royal Melbourne Hospital, Melbourne, Australia
a r t i c l e
i n f o
Article history: Received 21 May 2011 Revised 27 June 2011 Accepted 28 June 2011 Available online 3 August 2011 Keywords: Epilepsy Seizure Drug-resistant Reliability International League Against Epilepsy definition
a b s t r a c t We evaluated the interrater reliability of the consensus definition of drug-resistant epilepsy proposed by the International League Against Epilepsy. According to the definition framework, outcome of each antiepileptic drug (AED) trial was categorized as “seizure freedom” or “treatment failure.” This level 1 assessment was used to determine the level 2 classification, which defined drug-resistant epilepsy as the failure of adequate trials of two or more AED schedules to achieve sustained seizure freedom. Two raters classified treatment outcomes of 150 patients independently. The patients had received a total of 428 trials of AEDs. Categorization of level 1 outcome to individual AED trials by the raters was consistent in 413 (96.5%). For the level 2 classification of drug-resistant or drug-responsive epilepsy, there was absolute agreement between the raters in 141 patients (94%), with a κ index of 0.91 (P b 0.001). The definition appeared to have a high degree of interrater reliability in this setting. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Approximately 50 million people in the world have epilepsy, up to one-third of whom continue to have seizures despite drug treatment [1,2]. Patients with drug-resistant epilepsy are at increased risk of premature death [3], injuries [4], psychosocial dysfunction, and reduced quality of life [5]. Recognition of drug resistance using a universally accepted definition can facilitate early evaluation of patients for nonpharmacological therapies such as surgery [6], potentially alleviating the medicosocial and economic burden of refractory epilepsy. However, diverse criteria have been used by different researchers to define drug resistance, making it difficult or even impossible to compare the results across different studies and to make practice recommendations. To improve patient care and facilitate clinical research, the International League Against Epilepsy (ILAE) recently proposed the first consensus definition of drug-resistant epilepsy (referred to hereafter as “the ILAE definition”) [7]. The definition will, arguably, have the greatest impact at the nonspecialist levels given that most patients are initially managed by general practitioners, general physicians, or general neurologists. However, the validity of the definition needs to be tested so that it can be refined as new evidence emerges [8]. In particular, for the definition to be clinically useful, it
⁎ Corresponding author at: Department of Medicine and Therapeutics, Prince of Wales Hospital, Hong Kong. Fax: + 852 2632 1546. E-mail address:
[email protected] (P. Kwan). 1525-5050/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2011.06.035
should have a high degree of agreement among different users. Here, we report early results in evaluating its interrater reliability. 2. Methods 2.1. Patients The ILAE definition of drug-resistant epilepsy [7] was applied by two raters independently to 150 patients with epilepsy followed up at the neurology clinics of the Prince of Wales Hospital, Hong Kong. Patients with a doubtful diagnosis of epilepsy, prior epilepsy surgery, significant medical or psychiatric illness, history of psychogenic nonepileptic seizures, seizures caused by drug or alcohol abuse, or documented poor compliance with medication were excluded. 2.2. Definition parameters The overall framework of the ILAE definition comprises two “hierarchical” levels [7]. Accordingly, treatment outcome with an individual AED (level 1 outcome) (Table 1) was categorized as “seizure freedom” if the patient had had no seizure for at least 12 months or three times the maximum pretreatment interseizure interval, whichever was longer. If seizure freedom was not achieved with an “adequate” trial of the AED that was appropriate for the patient's seizure type(s), the outcome was categorized as “treatment failure.” The ILAE definition does not specify the dose or duration of treatment that constitutes an “adequate” trial of each AED. For standardization, in
X. Hao et al. / Epilepsy & Behavior 22 (2011) 388–390 Table 1 Scheme for categorizing treatment outcome of an antiepileptic drug (level 1 outcome of the International League Against Epilepsy definition of drug-resistant epilepsy [7]). Outcome category 1
Seizure freedoma
A B C
Adverse effects: no Adverse effects: yes Adverse effects: undetermined
2
Treatment failureb
A B C
Adverse effects: no Adverse effects: yes Adverse effects: undetermined
3
Undeterminedc
A B C
Adverse effects: no Adverse effects: yes Adverse effects: undetermined
Table 2 Raters’ categorization of outcomes of individual trials of antiepileptic drugs (level 1 outcome of the International League Against Epilepsy definition of drug-resistant epilepsy [6]). Rater 2
1A 1B 1C
2A 2B 2C
3A 3B 3C
a
Seizure freedom is defined as freedom from seizures for a minimum of 12 months or three times the longest pretreatment interseizure interval, whichever is longer. b Treatment failure is defined as recurrent seizure(s) after the treatment has been adequately applied. c Undetermined defines cases in which treatment has not been applied adequately for a valid assessment of the outcome or information is lacking to make the assessment.
the present study an adequate trial was operationally defined as continuous use of the AED for at least 3 months at a dose of at least 50% of the World Health Organization (WHO)'s defined daily dose (DDD) [9]. The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults. If the patient was not seizure free but the drug trial was inadequate, or if there was insufficient information, outcome was categorized as “undetermined.” Whether the patient experienced adverse effects during treatment was also noted. For each patient, level 1 outcome with each AED taken either previously or currently was categorized individually. This level 1 assessment formed the basis for the level 2 determination, which classified the epilepsy as “drug responsive” if the patient was seizure free (as defined in level 1) at the last follow-up. If the patient was not seizure free after having had adequate trials of two or more tolerated, appropriately chosen, and appropriately used AED schedules (whether as monotherapy or in combination), the epilepsy was classified as “drug resistant” [7]. In patients whose epilepsy did not fulfill either situation (e.g., treatment failure of only one AED, treatment outcome undetermined because the AED had not been tried adequately, or insufficient information), the drug responsiveness of the epilepsy was classified as “undefined.” 2.3. Data analysis The patients’ medical records were independently reviewed by the two raters, a research assistant who had just completed undergraduate medical training (X.T.H.) and a research nurse who has been a study coordinator in clinical epilepsy research for 2 years (I.S.M.W.). Data were collected and processed using a software tool that incorporated the parameters of the ILAE definition (http://logic.itsc.cuhk.edu.hk/ ~b124875/DRE/). The software was designed by P.K., who trained the raters how to use it. Level 2 outcomes classified by the raters were compared using κ statistics [10]. 3. Results Of the 150 patients evaluated, 79 (52.7%) were men. Their median age at the last follow-up was 45 years (range: 14–92), and their median duration of treatment was 147 months (range: 1–521). A total of 428 trials of AEDs were prescribed, with a median number of 2 (range: 1–12) for each patient. In these drug trials, level 1 outcomes categorized by the two raters were in agreement in 413 (96.5%)
389
1A 1B 1C 2A 2B 2C 3A 3B 3C Total number of drug trials
Rater 1
Total
1A
1B
1C
2A
2B
2C
3A
3B
3C
45 0 0 0 0 0 3 0 0 48
1 5 0 0 0 0 0 1 0 7
0 0 0 0 0 0 0 0 0 0
0 0 0 192 4 0 0 0 0 196
0 0 0 0 61 0 0 1 0 62
0 0 0 0 0 0 0 0 0 0
1 0 0 2 0 0 46 1 0 50
0 1 0 0 0 0 0 46 0 47
0 0 0 0 0 0 0 0 18 18
47 6 0 194 65 0 49 49 18 428
(Table 2). There were discrepancies in whether the patients became seizure free in 9 cases and whether the patients experienced adverse effects in 6. Table 3 summarizes the level 2 outcomes classified by the two raters. There was agreement between the two raters on 141 patients (absolute observed agreement= 94%). κ statistics (agreement after correcting for agreement expected by chance alone) was 0.91, indicating almost perfect agreement [9]. Classifications differed between the two raters in 9 (6%) patients, which resulted from misunderstanding of information recorded (3 cases), misinterpretation of the occurrence of seizure relapse (3 cases), maximum pretreatment interseizure interval (2 cases), and doses of AEDs used (1 case). 4. Discussion The absence to date of a universally accepted definition of drugresistant epilepsy has resulted in the use of diverse criteria by different clinicians and research groups [11]. A consensus definition was recently proposed by the ILAE to address this situation [7]. As part of the process to validate the definition, we evaluated, for the first time, its interrater reliability in a cohort of patients followed up in routine clinical practice in the setting of a hospital neurology clinic. The results provided early evidence that the definition has a very high degree of interrater reliability. Notably, there was no disagreement in the classification of the epilepsy as drug resistant or drug responsive, but only in determining whether the outcome was undefined, which occurs, for instance, when a treatment has not been tried adequately or when there is insufficient information. Most of the discrepancies arose from misunderstanding or misinterpretation of the information collected. Another potential source of discrepancies in ratings between different users could be the criteria for defining an “adequate” trial of an AED. The ILAE definition does not specify the dose or duration of treatment that constitutes an “adequate” trial because it is influenced by a range of intrinsic and extrinsic factors, such as the pharmacological properties of the drug, the age of the patients, any interaction with concomitant medications, and the patient's hepatic and renal Table 3 Comparison of raters’ classification of drug responsiveness (level 2 outcome of the International League Against Epilepsy definition of drug-resistant epilepsy [7]). Rater 2
Drug responsive Drug resistant Undefined outcome Total number of patients
Rater 1
Total
Drug responsive
Drug resistant
Undefined outcome
47 0 5 52
0 56 1 57
2 1 38 41
Note. Absolute observed agreement = 94%. κ statistic = 0.91 (P b 0.001).
49 57 44 150
390
X. Hao et al. / Epilepsy & Behavior 22 (2011) 388–390
function. An individualized approach is needed in clinical practice. For the purpose of standardization in the research setting, we suggest that making reference to the WHO's DDD may be a reasonable approach because it provides a standard unit of measurement of dosage, with the caveat that the system is intended for monotherapy use and might not be applicable in patients taking multiple AEDs that are prone to drug–drug interactions [9]. In addition, the system is applicable primarily to adults because doses used in children are heavily influenced by body weight. This issue will need to be further addressed if the project is rolled out to a wider range of settings as different centers may adopt different standards. Research is needed to determine the most appropriate approach to define an “adequate” trial, perhaps by taking into account multiple factors simultaneously and evaluating the impact of different durations of treatment and dosages on final treatment outcome. We acknowledge the limitation of this pilot study in involving patients attending specialist clinics of a single tertiary hospital only. To further establish the definition's clinical validity, prospective evaluation in a broader range of clinical settings and health care levels, such as newly treated patients and primary care practice, and by a greater number of raters, is needed. For its effective and efficient use, the definition should be integrated into the routine medical records. An electronic platform that captures the essential information in defining drug response would facilitate this process and should be promoted as part of an electronic health record system. Training of
general practitioners and general neurologists would improve their familiarity with the definition, which will help them refer patients to specialist centers in an appropriate and timely fashion.
References [1] Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342: 314–9. [2] Kwan P, Schachter SC, Brodie MJ. Drug-resistant epilepsy. N Engl J Med in press. [3] Mohanraj R, Norrie J, Stephen LJ, Kelly K, Hitiris N, Brodie MJ. Mortality in adults with newly diagnosed and chronic epilepsy: a retrospective comparative study. Lancet Neurol 2006;5:481–7. [4] Lawn ND, Bamlet WR, Radhakrishnan K, O'Brien PC, So EL. Injuries due to seizures in persons with epilepsy: a population-based study. Neurology 2004;63:1565–70. [5] McCagh J, Fisk JE, Baker GA. Epilepsy, psychosocial and cognitive functioning. Epilepsy Res 2009;86:1–14. [6] Engel Jr J, Wiebe S, French J, et al. Practice parameter: temporal lobe and localized neocortical resections for epilepsy. Epilepsia 2003;44:741–51. [7] Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51:1069–77. [8] Kwan P, Brodie MJ. Definition of refractory epilepsy: defining the indefinable? Lancet Neurol 2010;9:27–9. [9] World Health Organization Collaborating Centre for Drug Statistics Methodology. ATC/DDD Index 2011. Available at: http://www.whocc.no/atc_ddd_index/. Accessed: 18 February 2011. [10] Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74. [11] Berg AT, Kelly MM. Defining intractability: comparisons among published definitions. Epilepsia 2006;47:431–6.