look in description

Neurorehabilitation with FORAMENRehab for attention impairment in children with epilepsy Marianne Saard a,b , ⁎, Mari-Liis Kaldoja c,d,MadisBachmann e, Lisanna Pertens e, Anneli Kolk a ,b aDepartment of Pediatrics, Faculty of Medicine, University of Tartu, Tartu, EstoniabDepartment of Neurology and Neurorehabilitation, Children's Clinic, Tartu University Hospital, Tartu, EstoniacSchool of Educational Sciences, Tallinn University, Tallinn, EstoniadDepartment of Neurology and Rehabilitation, Tallinn Children's Hospital, Tallinn, EstoniaeInstitute of Psychology, University of Tartu, Tartu, Estonia abstract article info Article history:

Received 22 March 2016 Revised 15 December 2016 Accepted 17 December 2016 Available online xxxx Epilepsy is a frequent neurological disorder in children and often accompanied with attention impairment. Still, few systematically controlled rehabilitation techniques for children exist. The aim of this study was to design and measure the impact of the FORAMENRehab computer-based intervention method for attention impairment rehabilitation in children with epilepsy. We chose the FORAMENRehab program because it allows separate train- ing for different attention components based on individual needs.

Forty-eight children participated in the study. At baseline, all patients underwent neuropsychological examina- tion of attention with the NEPSY test battery. The study group consisted of 17 8- to 12-year-old children with par- tial epilepsy and attention impairment who received neurorehabilitation over 5 weeks (10 sessions) with FORAMENRehab Attention module accompanied by a therapist. Two control groups were included: the ﬁrst con- trol group of 12 children with partial epilepsy and attention impairment ( waiting-list) participated in assessments with baseline tasks before and after the ﬁve-week period and received no active training. Additionally, all patients participated in the follow-up assessment 1.31 years later. The second control group consisted of 19 typically devel- oping children who only participated in the ﬁrst assessment.

After the intervention, study group patients showed signi ﬁcant improvement in complex attention and tracking (P b0.025). To achieve the effect of intervention in children with partial epilepsy, 10 sessions tailored to individ- ual levels of ability were the minimum. Three attention components –sustained, complex, and tracking –need selective and longer training for more effective remediation. Follow-up assessment revealed a long-term positive effect of intervention. After 1.31 years, the study group had signi ﬁcantly improved in three out of the four attention components (P b0.025), whereas the waiting-list group showed improvement in only two aspects of one complex attention component. In conclusion, attention impairment rehabilitation with FORAMENRehab is effective for children with epilepsy. Rehabilitation should focus on training speci ﬁc components of attention and follow an individual-based rehabilitation process.

Epilepsy Attention impairment Cognitive rehabilitation Computer-based neurorehabilitation in children FORAMENRehab program 1. Introduction 1.1. Attention As one of the key components of cognitive functioning, attention has been described as the processes that enable a person to concentrate on speci ﬁc cognitive tasks and ignore others [1]. Attention is a multidimen- sional cognitive process that affects other dimensions of cognition –learning, memory, communication, problem solving, and perception [2] . Sohlberg and Mateer [3,4]have differentiated at least four categories of models for attention: clinical models, factor analytic models, cognitive processing models, and neuroanatomic models. The current study is based on a clinical model of attention by Sohlberg and Mateer [2,3,5] , in which attention consists of ﬁve interrelated components:

focused, sustained, selective, alternating, and divided attention. Focus- ing and sustaining attention are considered lower-level processes, whereas selective, alternating, and divided attention are more complex processes that at least partially rely on the proper working of lower- level processes [2,3,5]. Tracking is also seen as part of complex attention and is especially crucial when attention is needed while doing some other mental task (e.g., digit span backwards in a task concept) [6]. Epilepsy & Behavior 67 (2017) 111 –121 ⁎ Corresponding author at: Department of Paediatrics, Faculty of Medicine, University of Tartu, N. Lunini 6, EE-51014 Tartu, Estonia.

E-mail address: [email protected] (M. Saard).

http://dx.doi.org/10.1016/j.yebeh.2016.12.030 1525-5050/© 2016 Elsevier Inc. All rights reserved. Contents lists available atScienceDirect Epilepsy & Behavior journal homepage:www.elsevier.com/locate/yebeh 1.2. Attention impairment in pediatric epilepsyImpairments of attention accompany various disorders including epilepsy [7], which is one of the most common neurological disorders in children and adolescents. Children with different types of epilepsy and epilepsy syndromes have clearly expressed dysfunctions in differ- ent attention components. Impairments in overall attention [8–11] , sustained attention [12–14] , selective attention [9], and alertness [15] have been described. Both modalities, visual and auditory attention, have been found to be affected [16–18] . Furthermore, a recent study by Rathouz et al. [10]claims that cognitive de ﬁcits in children with epilepsy that are present at baseline assessment are maintained for up to at least 5 –6 years. This constitutes a prevalent problem in the educational quality of these children. More so, attention impair- ment is closely linked to impairments in other cognitive functions (e.g., working memory and executive functions) [19]. As attention is crucial for learning, impairments in attention have been found to contribute as a major negative in ﬂuence on academic and social compe- tences [20,21] .

1.3. Attention rehabilitation in children One suitable intervention method to facilitate cognitive perfor- mance is cognitive rehabilitation (CR), which focuses on improving a person's functioning in their everyday life by increasing their ability to do what they need and like, but ﬁnd dif ﬁcult or impossible due to their cognitive disability [22,23]. Cicerone et al. [24]concludes that in order to contribute to the generalization effect, the training should be addressed to speci ﬁc attention functions and teaching strategies.

Cognitive rehabilitation in children with attention impairments is nowadays mainly developed into computer-based interventions. Chil- dren surviving cerebral malaria [25,26],childrenwithHIV [27],andsur- vivors of childhood cancer [28]have been trained with Captain's Log cognitive training software with immediate bene ﬁt to attention skills.

Lee et al. [29]and Galbiati et al. [30], who trained children with traumat- ic brain injury, found that attention-speci ﬁc neuropsychological train- ing signi ﬁcantly improved attention performance.

As for epilepsy, very few studies have investigated the effects of cog- nitive rehabilitation programs for attention impairments. The amount of research is still modest compared to other acquired brain injury diag- noses and has mostly been conducted on adult patients with epilepsy.

Engelberts et al. [31]used two training methods for attention rehabilita- tion on adults with focal seizures –the Retraining Method and the Compensation Method –and found both to be effective in improving the patients' cognitive outcomes. Gupta and Naorem [32]st ated that after speci ﬁc cognitive training (including for attention de ﬁcits), using both paper and pencil tasks and real-life activities for patients with epilepsy, overall improvement in cognitive performance occurred.

However, for children with epilepsy, very few modern neurocognitive rehabilitation techniques exist, and most of the available methods used are designed for adults. A positive outcome in remediation of attention impairment in children with epilepsy was revealed with our pilot study using the FORAMENRehab computer-program. After a six- week, therapist-guided, and personalized training program, sustained attention and complex attention of children showed signi ﬁcant improvements. No signi ﬁcant changes were detected in focused attention and tracking. That said, parents and teachers of the participating patients also stated positive changes in behavior and an increase in academic achievement, suggesting a generalized effect of the intervention [33].

There is therefore an unmet need for more accurate and systematically controlled research in pediatric cognitive rehabilitation [34 –36] . More so, the ﬁeld of neuropsychological rehabilitation needs guidelines and underlying principles in order to inform the work of clinicians [37,38]. While the results of the FORAMENRehab pilot study were promising, the need for modi ﬁcations to improve the overall outcome of attention rehabilitation became apparent, as was more thoroughly described in the previous study [33]. In the current study a larger sample size, a waiting-list control group, and follow-up assessments were included. The main aim of the current research was to test the effectiveness of a computer-based rehabilitation method with the Attention module of the FORAMENRehab program in attention impairment rehabilitation for children with epilepsy aged 8 –12 years.

The speci ﬁcaimswere:

1. to examine the rehabilitation effect on speci ﬁc attention components in 8- to 12-year-old children using the FORAMENRehab program; 2. to discover the optimal dif ﬁculty levels and duration of training for attention impairment; 3. to measure the long-term rehabilitation effect in follow-up assessments; 4. to evaluate the generalized effect and provide clinical implications for computer-based attention rehabilitation in children with partial epilepsy.

2. Methods 2.1. Participants 2.1.1. Study group The current study was carried out from May 2012 until March 2015 in the Department of Neurology and Neurorehabilitation at the Children's Clinic of Tartu University Hospital. The study was approved by the Research Ethics Committee of the University of Tartu.

In the intervention, 17 children aged 8 –12 years with attention impairment and diagnosis of partial epilepsy participated. There were 12 boys and 5 girls in the intervention group (see Table 2for further details). The age group was chosen based on methodological considerations: 1) the children were required to have suf ﬁcient reading and basic mathematical skills; 2) to keep the age range and develop- mental level of the children comparable. Participants were chosen according to the following inclusion criteria:

1. Previously diagnosed partial epilepsy (ICD-10 G40.0, G40.1, G40.2); diagnosis con ﬁrmed by a child neurologist.

2. Mild to moderate attention impairment stated by parents and teachers and con ﬁrmed by a certi ﬁed clinical neuropsychologist on the basis of neuropsychological assessment. The assessment includ- e d attention subtests from the NEPSY test battery (Tower Test, Visual Attention, Auditory Attention and Response Set, Statue, Design Fluency, Knock and Tap). Minimum and maximum scores of patients for each subtest are included in Table 1.

3. Fluency in Estonian ( ﬁrst spoken language); 4. Age between 8 and 12 years; 5. Parental written consent and child's verbal agreement to participate in the intervention. Table 1 Standardized score ranges for NEPSY subtests; normal range 7 –13.

NEPSY subtest 29 patients Min score Max score Tower 114 Visual attention 219 Auditory attention 315 Statue 112 Design ﬂuency 2 18 Knock and tap 213 112 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111–121 Exclusion criteria included other documented diseases involving the central nervous system (e.g., stroke, tumors, encephalitis, cerebral palsy), psychiatric co-morbidity (e.g., ADHD, anxiety disorder, mental retardation (ICD-10 F70-F79)), and treatment with any psychotropic medication other than antiepileptic drugs during the rehabilitation period. In summary, the sample of children with epilepsy included 4 newly-diagnosed patients, the maximum duration of epilepsy for a single child was 5.08 years, and 1 patient was on polytherapy, while 16 were on monotherapy. All 17 children attended all 10 training sessions –therefore, compliance with the intervention was 100%.

2.1.2. Control groups Two control groups were included in the study –waiting-list control group and healthy children's control group.

1. The waiting-list control group was composed of 12 children with epilepsy aged 8 –12 years with attention impairment who could not participate in the intervention. There were 9 boys and 3 girls in the group (see Table 2for further details). The inclusion and exclusion criteria were the same as for the intervention group. The study group and waiting-list group did not differ signi ﬁcantly regarding the age at epilepsy onset. Children with epilepsy who participated either in the study or waiting-list group were diagnosed with partial epilepsy and the seizures were well-controlled by medication. Also, the children all attended regular schools and did not receive special education. Additionally, all patients were presented at ﬁrst with an opportunity to participate in the rehabilitation. Frequently, children from outside the Tartu city boundary could not regularly attend the intervention due to dif ﬁculties with transportation and were there- fore assigned to the waiting-list group.

2. To assess the baseline levels for attention tasks and to obtain results representing the healthy population, the healthy children's control group was composed of 19 healthy age-matched children aged 8–12 years. There were 11 boys and 8 girls in the control group.

The children were recruited from an ordinary school in Tartu and attended 2nd to 5th grades. Parental written consent and the child's verbal consent to participate in the study were received. Children with any known neurologic or psychiatric diagnosis were excluded from the control group. The study group and two control groups did not differ signi ﬁcantly in terms of age and sex.

2.2. Rehabilitation software The FORAMENRehab Cognitive Rehabilitation Software® (FORAMENRehab) was used in the study. FORAMENRehab is a tool for cognitive rehabilitation that was developed in 2000 by Koskinen and Sarajuuri [39]in Finland. It was adapted for Estonian children in our pilot study [33]. Due to the variability of the tasks, the software can be used for children with acquired or developmental disorders. In the present study, the Attention module was implemented. Different components of attention function were assessed and trained with the module. The tasks were divided into four categories or components –fo- cused attention, sustained attention, complex attention, and tracking. The exercises are playful and last from 1 to 4 min (with the excep- tion of a sustained attention task which can take up to 20 min). The menu structure, toolbar, and icons of the software are illustrative; each task has clear written instructions as well as a model animation.

The parameters of each task are adjustable. The results are given both in numerical tables and graphs. Several outcomes are recorded for every application: solving and/or reaction time, number of correct re- sponses, and subcategories of mistakes (omission errors, premature re- sponses, commission errors, and total number of errors –sum of omission and commission errors).

2.3. Rehabilitation procedure The rehabilitation of patients took place during 6 weeks with a 5-week training period. Thirteen meetings were conducted with each patient: the ﬁrst baseline assessment, altogether 10 active training ses- sions with meetings twice a week, and the second assessment with baseline tasks (primary outcome). Also, the ﬁnal follow-up assessment with baseline tasks as the secondary outcome was conducted 1.31 years (SD = 0.40) after the training period (third assessment with baseline tasks). In total, 10 children from the study group and 9 children from the waiting-list group participated in the ﬁnal follow-up.

Training sessions occurred in an outpatient setting in the Children's Clinic. The time for each individual session varied between 30 and 40 min. At theﬁrst meeting the intervention methods and goals were intro- duced to the patient. Thereafter, the ﬁrst performance of baseline tasks was conducted to assess the child's initial pro ﬁle of attention compo- nents. The training sessions started at the second meeting. Each meeting was conducted by one of the four therapists, who were all psychologists working in the hospital. The therapists received training before administering the rehabilitation and followed a stan- dardized protocol. Throughout, the therapists did not only introduce the tasks, but also motivated and guided the child individually in order to help him/her to cope better with new and complicated situa- tions and to apply the learned techniques to everyday life. Furthermore, an important part of the training process is the therapist-guided metacognitive study experience for children, which teaches different learning and problem-solving strategies. A strict protocol for our intervention procedure was created. For assessing the effectiveness of the rehabilitation, the baseline tasks were tested once more at the last meeting of the six-week period and also ﬁnally 1.31 (SD = 0.40) years later to measure sustained long- term effects. The generalized effect of the APT was evaluated by parent and child questionnaires about the perceived attention, behavior, and school performance, in addition to objective baseline assessments.

For this purpose we developed a Likert-type scale, which was used to compare periods before and after the intervention.

2.4. Rehabilitation designs for FORAMENRehab software Two different intervention designs were used in the process of conducting the intervention and evaluating the appropriateness of the FORAMENRehab computer program. First, a pilot study was conducted Table 2 Study group and waiting-list control group characteristics.

Group Nr of patients Age at intervention (yrs) Age at epilepsy onset (yrs) Duration of epilepsy (yrs) (M/F) Mean (95%CI) a Mean (95%CI) Mean (95%CI) Study n = 17 (12/5) 9.95 (9.33…10.57) 7.85 (6.92…8.78) 2.10 (1.34…2.86) Waiting-list n = 12 (9/3) 10.20 (9.08…11.32) 8.27 (7.22…9.32) 1.97 (0.60…3.33) Sig. P ⁎ 0.96470.3637 0.3512 M –male, F –female, yrs –years. aMean score (95% con ﬁdence intervals for Mean).

⁎ Cut-off P value for signi ﬁcance is 0.05. 113 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111 –121 [33], and based on the results the intervention protocol was modi ﬁed to better differentiate children's baseline impairment pro ﬁles, to more accurately measure the rehabilitation effect, and to facilitate progress at each dif ﬁculty level during training. The new procedure protocol for the intervention that was created is shown in Fig. 1. The results re- ported in this paper are based on the new design.

In the baseline assessment, all four components of attention were tested. Two to three tasks from each component were chosen. For train- ing sessions, other tasks involving the same attention components were used, divided into three difﬁ culty levels: easy (I), medium (II), and dif- ﬁ cult (III) (see Table 3for detailed description). Some tasks were divid- ed into two dif ﬁculty levels as there were either no easy or dif ﬁcult settings for children. Development at each dif ﬁculty level was individual-based and depended on the child's personal improvement. If the child was ﬂawless at the task, he/she advanced to the next level of the same attention com- ponent at the next training session. If the child's response was incorrect, he/she had to perform the same task level at least 80 –90% correctly (depending on the task) for 3 meetings consecutively until advancing to the next dif ﬁculty level. This af ﬁrmed that the child had acquired the requested abilities.

2.5. Testing of the controls The children in the waiting-list control group participated in three assessments with baseline tasks –the ﬁrst assessment, primary outcome assessment, and follow-up or secondary outcome assessment.

During the ﬁve-week period between the ﬁrst and the second assess- ment, the waiting-list group received no intervention. One-time testing of healthy control children took place in their school setting.

2.6. Data analysis Statistical data analysis was performed with the R version 3.1.2. For some of the ﬁgures, the SAS data analysis package 9.2 was used. The Kolmogorov-Smirnov test was used for the assessment of normality.

Wilcoxon-Mann-Whitney test was used to compare the study group and controls on each of the attention variables. For each task, different components of performance were evaluated if possible (correct responses, omission and commission errors, total number of mistakes, reaction time, and processing speed). We used linear mixed models to assess whether longitudinal changes in groups were overall signi ﬁcantly different. We com- pared the baseline performance to primary (immediate interven- tion effect) and secondary outcomes within the study group and waiting-list control group. For comparing proportions (qualitative variables) the McNemar's test was used. The con ﬁdence level was set to 5%. We controlled the false discovery rate (FDR) to be lower than 5% by using linear step-up procedure (Benjamini and Hochberg 1995) [40]for multiple t-tests. Benjamini-Hochberg crit- ical values were calculated as (i/m)Q, where i is the rank in an as- cending list of P values, m is the total number of tests, and Q is a Meeting no. 1 23 4 5 6 7 8 9 10 11 12 13 Follow-up Baseline assessment 1 Interventional trainings Baseline ass ssment 2 Baseline assessment 3 5 weeks 1.31yrs 12 3 4 5 6 7 8 9 10 Fig. 1. Design of the intervention. Table 3 Difﬁculty levels in tasks under four attention components and the affected attention functions.

Easy Medium Dif ﬁcult Affected attention functions Focused attention Visual reaction time Visual reaction time; auditory warning visual choice reaction time Visual multiple choice reaction timeAttention activation, alertness and selectivity:

simple visual or auditory reaction (intrinsic vigilance); tasks with warning (phasic activation of attention) or with distracting stimuli (selective attention and reaction inhibition).

Auditory reaction time Auditory reaction time; visual warning auditory choice reaction time Auditory multiple choice reaction time Sustained attention Single ﬁgure search with letter, number, symbol I, symbol II, or picture (easy level) (Medium level) (Difﬁcult level) Continuous attention,ﬁnding high ratio of relevant stimuli, processing speed, correctness Series search with letter, number, ﬁgure, or symbol series (easy level) (Medium level) (Difﬁcult level) Paced search with single target (target length –2 characters) (Target length –4 characters; target shifting interval –faster) (Target length –7 characters) – Repeated pairs search with symbols Repeated pairs search with lettersContinuous attention, high ratio of disturbing stimuli.

Complex attention Paced search with dual targets (target length –2 characters; speed –1,5 s; direction –right) (Target length –3 characters; faster stimulus interval –speed –1s) (Faster stimulus interval – speed 0.7 s) Dual tasks: dividing and shifting attention, cognitive ﬂexibility.

Addition; single number (series length –4 digits; speed –1.5 s) (Series length –6–8 digits; faster stimulus interval) Addition; dual numbers Single addition: dividing attention, working memory.

Word recognition; single target Word recognition; dual targets Simultaneous word recognition and mental arithmeticSingle word recognition: continuous attention, comparison with existing knowledge.

Tracking Tracking task Continuous attention, attention activation.

PASAT; visual presentation PASAT; visual presentation (faster stimulus interval) – Visual dividing of attention, executive function of working memory.

PASAT - Paced Auditory Serial Addition Test. 114 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111–121 false discovery rate of 0.05. The cut-off P value for signiﬁcance in a single comparison for the longitudinal changes of attention vari- ables in study group is 0.025. Only P values that are below the ad- justed FDR signi ﬁcance threshold are therefore signi ﬁcant and marked as such (*) in the tables. Effe ct sizes for non-normal distri- butions were calculated using the Cliff's Delta statistic, which is a non-parametric effect size measure that quanti ﬁes the amount of dif- ference between two groups of observations [41]. 3. Results 3.1. Differences in attention components in children with epilepsy and healthy controls Comparison of performances on th e FORAMENRehab baseline assess- ment between children with epilepsy (including both the study group and waiting-list control group) and healthy children was conducted Table 4 Comparison of performances on ﬁrst baseline assessment between patients and healthy control children.

Parameters of FORAMENRehab Attention tasks Patients Controlsd b Sig. P First baseline assessment (B1) Median (lower and upper quartiles) a Median (lower and upper quartiles) Focused attention Visual reaction time (s) 0.59 (0.49 …0.67) 0.53 (0.43 …0.65) 0.29 0.082 Auditory reaction time (s) 0.61 (0.54 …0.74) 0.62 (0.59 …0.69) 0.08 0.626 Sustained attention Correct responses in picture search (%) 97.44 (94.87 …98.72) 98.72 (96.15…100.00) 0.32 0.040 ⁎ Omission errors in picture search c(%) 2.56 (0.00…3.85) 1.28 (0.00…3.85) 0.240.178 Processing speed in picture search (s) 205.00 (160.50…244.00) 159.00 (140.00…215.00) 0.380.026⁎ Processing speed in numbers search (s) 709.00 (597.00…886.00) 603.00 (393.00…747.00) 0.360.042⁎ Complex attention Omission errors in paced search (%) 68.20 (53.94…83.85) 44.12 (25.81…61.11) 0.57b0.001 ⁎ Total errors in paced search d(nr) 32.00 (26.00…35.00) 22.00 (12.00…27.00) 0.58b0.001 ⁎ Omission errors in word recognition (%) 58.33 (33.33…75.00) 19.87 (9.09…50.00) 0.660.0001⁎ Commission errors in word recognition e* (nr) 1.00 (1.00 …2.00) 2.00 (1.00 …3.00) 0.16 0.335 Correct responses in addition (%) 30.00 (15.00 …65.00) 80.00 (70.00…90.00) 0.77 b0.0001 ⁎ Tracking Correct responses in PASAT (%) 22.50 (10.00…30.00) 65.00 (45.00…95.00) 0.82 b0.0001 ⁎ Omission errors in PASAT (%) 50.00 (27.50…65.00) 17.50 (5.00…20.00) 0.630.0003⁎ Commission errors in PASAT (%) 22.50 (15.00…47.50) 12.50 (0.00…30.00) 0.430.013⁎ Commission errors in tracking task (nr) 1.0 (0.00…2.00) 0.00 (0.00…0.00) 0.400.013⁎ aMedian (lower 25%ile and upper 75%ile).bEffect size index Cliff's delta.cOmission errors –missed responses to target stimuli.dTotal errors –sum of omission and commission errors.eCommission errors –responses to nontarget stimuli.

⁎ Cut-off P value for signi ﬁcance is 0.05. Table 5 Primary outcome of attention function training in study group.

Parameters of FORAMENRehab Attention tasks Baseline B1 Primary outcome B2 Secondary outcome B3 B1-B2-B3 B1 vs B2 B1 vs B3 Study group (n = 17) Mean (95%CI) a Mean (95%CI)Mean (95%CI)Sig P f Sig P Sig P Focused attention Visual reaction time (s) 0.64 (0.57…0.71) 0.67 (0.58…0.76) 0.51 (0.43…0.60) 0.0071 0.5715 0.0047 ⁎ Auditory reaction time (s) 0.66 (0.59…0.73) 0.67 (0.58…0.75) 0.58 (0.50…0.65) 0.1297 0.9115 0.0869 Sustained attention Correct responses in picture search (%) 96.15 (93.59…100.00) e 98.72 (96.15 …100.00) e 99.36 (97.44 …100.00) e 0.1530 0.1864 0.0930 Omission errors in picture search b(%) 3.85 (0.00…6.41) e 1.28 (0.00…3.85) e 0.64 (0.00 …2.56) e 0.1578 0.1922 0.0966 Processing speed in picture search (s) 185.71 (165.54 …205.87) 210.94 (161.77 …260.12) 182.10 (148.09 …216.11) 0.4384 0.2927 0.2160 Processing speed in numbers search (s) 738.94 (588.32 …889.56) 854.65 (675.19 …1034.10) 575.00 (459.70 …690.30) 0.0770 0.2783 0.1677 Complex attention Omission errors in paced search (%) 66.74 (57.79 …75.68) 41.99 (30.96 …53.03) 33.94 (26.56 …41.31) b0.0001 b0.0001 ⁎ b0.0001 ⁎ Total errors in paced search c(nr) 32.19 (27.37…37.00) 18.35 (13.51 …23.20) 17.67 (13.60 …21.73) b0.0001 0.0002 ⁎ b0.0001 ⁎ Omission errors in word recognition (%) 56.15 (41.70 …70.60) 38.78 (25.59 …51.97) 20.76 (5.27… 36.26) 0.0003 0.0028 ⁎ 0.0002 ⁎ Commission errors in word recognition d(nr) 2.00 (1.00… 2.00) e 1.00 (0.00…2.00) e 0.00 (0.00 …2.00) e 0.1081 0.0744 0.0741 Correct responses in addition (%) 41.76 (26.75 …56.78) 63.53 (49.57 …77.49) 80.00 (68.82 …91.18) 0.0002 0.0007 ⁎ 0.0002 ⁎ Tracking Correct responses in PASAT (%) 25.88 (17.79 …33.97) 55.29 (38.70 …71.89) 67.00 (49.71 …84.29) 0.0001 0.0002 ⁎ 0.0002 ⁎ Omission errors in PASAT (%) 36.80 (23.97 …49.62) 29.70 (15.70 …43.71) 17.00 (5.66… 28.34) 0.1877 0.3600 0.0710 Commission errors in PASAT (%) 34.12 (20.51 …47.73) 15.00 (8.08… 21.92) 16.00 (4.71… 27.29) 0.0164 0.0049 ⁎ 0.0475 Commission errors in tracking task (nr) 1.00 (0.00 …1.50) e 1.00 (1.00 …2.00) e 0.50 (0.00 …1.00) e 0.4045 0.2526 0.2624 aMean score (95% con ﬁdence intervals for Mean).bOmission errors –missed responses to target stimuli.cTotal errors –omission and commission errors.dCommission errors –responses to nontarget stimuli.eMedian score (lower 25%ile and upper 75%ile), non-normal distribution.fMixed models method.

⁎ Cut-off P value for signi ﬁcance in single comparison for longitudinal changes of attention variables is 0.023. 115 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111 –121 (seeTable 4 ). Infocused attention , the trend showed a quicker reaction time to visual stimuli in healthy children. In s ustained attention, the chil- dren with epilepsy had a signi ﬁcantly worse overall performance com- pared to their healthy peers, as they demonstrated slower processing speed as well as a lower percentage of correct responses. In c omplex attention , the patients performed signi ﬁcantly worse than healthy children in each aspect of the task: they gave signi ﬁcantly less correct responses as they had more omission errors and total errors (sum of omission and commission errors). In tracking, the patient group had sig- niﬁ cantly worse results in every aspect compared to healthy controls.

3.2. Primary outcome in two epilepsy groups before and after the intervention period 3.2.1. Immediate rehabilitation effect (primary outcome) on speci ﬁc components of attention in study group Atﬁrst there were no signi ﬁcant differences in performances at baseline levels between study and waiting-list control groups in any of the measured attention components. But immediately after the inter- vention, noticeable improvements were seen in the study group as they had signi ﬁcantly improved complex attention andtracking components compared to the ﬁrst baseline performance (for further details see Table 5 ).

After training, the study group patients performed signi ﬁcantly bet- ter in Paced Search with Dual Targets (complex attention): they gave more correct responses, and had less omission errors and less total er- rors compared to the ﬁrst assessment. In Word Recognition with Dual Targets (complex attention), the study patients had signi ﬁcantly more correct responses and less omission errors. In this task, most of the children with epilepsy had some commission errors in baseline as well as in primary outcome assessment. Thus, for more targeted exam- ination we divided the groups into two by number of commission er- rors, based on a cut-off point of 6 errors. A McNemar's test showed signi ﬁcant overall improvement: 86.7% of children belonged to the group with a greater number of mistakes at baseline assessment, but only 26.67% at primary outcome assessment (P b0.05). In Addition Fig. 2. Means with 95% con ﬁdence interval for percent of correct responses in three complex attention tasks at three assessment points (B1-baseline, B2-primary outcome, B3-secondary outcome) in study and control groups.

Fig. 3. Means with 95% con ﬁdence interval for percent of correct responses, and commission and omission errors in PASAT under tracking component at three assessment points (B1- baseline, B2-primary outcome, B3-secondary outcome) in study and control groups. 116 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111–121 task (complex attention), the intervention group gave signiﬁcantly more correct responses after rehabilitation (see Table 5).Fig. 2 shows the percentage of correct responses in the baseline, primary, and sec- ondary outcome assessments. Also, the study group patients had improved signi ﬁcantly after training in PASAT (tracking component): they gave more correct re- sponses and had less commissions. Fig. 3shows the percentage of correct responses for each asp ect of PASAT in three assessment points. No statistically signi ﬁcant dynamic changes were revealed in focused attention for reaction times ( Fig. 4) or in sustained attention for processing speed. 3.2.2. Second assessment or primary outcome with baseline tasks in the waiting-list control group After the 5-week period without intervention, the waiting-list control group did not show any signi ﬁcant differences between the ﬁ rst and second assessment (see Table 6).

Different results for study and waiting-list control groups are visually shown in Figs. 2–4.

3.2.3. Comparison of second assessment with baseline tasks between study and waiting-list control groups After the intervention, differences between performances in base- line tasks had emerged between the study and waiting-list control Fig. 4. Means with 95% con ﬁdence interval for visual and auditory reaction times in focused attention at three assessment points (B1-baseline, B2-primary outcome, B3-secondary outcome) in study and control groups.

Table 6 Primary outcome of attention function training in waiting-list group.

Parameters of FORAMENRehab Attention tasks Baseline B1 Primary outcome B2 Secondary outcome B3 B1-B2-B3 B1 vs B2 B1 vs B3 Waiting-list group (n = 12) Mean (95%CI) a Mean (95%CI)Mean (95%CI)Sig P f Sig P Sig P Focused attention Visual reaction time (s) 0.56 (0.50 …0.61) 0.60 (0.49 …0.70) 0.60 (0.49 …0.71) 0.7498 0.6795 0.5517 Auditory reaction time (s) 0.66 (0.55…0.77) 0.70 (0.60…0.79) 0.67 (0.54…0.81) 0.8988 0.6571 0.8576 Sustained attention Correct responses in picture search (%) 97.44 (96.15…98.72) e 96.15 (96.15 …98.72) e 98.72 (97.44 …98.72) e 0.7771 0.6082 0.8191 Omission errors in picture search b(%) 2.56 (1.28…3.85) e 3.85 (1.28…3.85) e 1.28 (1.28 …2.56) e 0.3968 0.7818 0.3690 Processing speed in picture search (s) 238.20 (200.69…275.72) 224.67 (177.42 …271.91) 218.33 (152.26 …284.41) 0.4888 0.4334 0.2547 Processing speed in numbers search (s) 711 (566…912) e 969 (469.5 …1144.5) e 544 (506 …1101) e 0.7667 0.6911 0.7021 Complex attention Omission errors in paced search (%) 68.88 (57.92…79.84) 65.33 (55.57 …75.10) 51.42 (32.60 …70.25) 0.0309 0.3246 0.0094 ⁎ Total errors in paced search c(nr) 32.60 (23.80…41.41) 30.58 (23.81 …37.35) 25.56 (14.86 …36.25) 0.1855 0.2024 0.0774 Omission errors in word recognition (%) 59.30 (46.75 …71.84) 52.08 (33.12 …71.04) 42.30 (23.00 …61.60) 0.1506 0.3099 0.0562 Commission errors in word recognition d(nr) 1.00 (1.00… 2.00) e 1.00 (0.00…3.00) e 1.00 (1.00 …2.00) e 0.3725 0.2075 0.2574 Correct responses in addition (%) 33.33 (18.44…48.23) 36.67 (22.50 …50.83) 58.89 (36.33 …81.45) 0.0097 0.4418 0.0031 ⁎ Tracking Correct responses in PASAT (%) 19.00 (12.04…25.96) 31.67 (17.37 …45.96) 45.00 (21.88 …68.11) 0.0894 0.0777 0.0327 Omission errors in PASAT (%) 56.00 (44.91…67.09) 40.83 (20.89 …60.78) 39.38 (19.59 …59.16) 0.1638 0.0800 0.1686 Commission errors in PASAT (%) 25.00 (15.58…34.42) 32.50 (16.97 …48.03) 15.63 (8.07… 48.03) 0.0287 0.4165 0.1297 Commission errors in tracking task (nr) 1.00 (1.00…2.00) e 0.50 (0.00 …3.00) e 1.00 (0.00 …2.00) e 0.9119 0.8750 0.6724 aMean score (95% con ﬁdence intervals for mean).bOmission errors –missed responses to target stimuli.cTotal errors –omission and commission errors.dCommission errors –responses to nontarget stimuli.eMedian score (lower 25%ile and upper 75%ile), non-normal distribution.fMixed models method.

⁎ Cut-off P value for signi ﬁcance in single comparison for longitudinal changes of attention variables is 0.023. 117 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111 –121 groups. The second assessment showed that the study group had signif- icantly better results incomplex attentionandtracking compared to the waiting-list group. In Paced Search, they had a higher percentage of correct responses (P b0.01), less omission errors (P b0.01), and less total mistakes (P b0.01). In Addition and PASAT tasks, they had a higher percentage of correct responses (P b0.05 and P b0.05, respectively).

Trends also showed fewer mistakes in the Auditory Choice Reaction Time task (P = 0.089) and less commission errors in Picture Search in the study group (P = 0.061).

3.3. Patients' individual improvement during rehabilitation process We also examined the study group children's individual improve- ment during the rehabilitation process and advancement in reaching higher dif ﬁculty levels. A slower rehabilitation effect out of the four attention components occurred in two: complex attentionandtracking - where children's average attained level at the end of training was only 1.55 (95% CI: 1.36 …1.74) out of a maximum of 4 and 1.31 out of a maximum of 3, respectively (see Table 7for further details).

In order to assess the degree of dif ﬁculty of the tasks, the average number of training sessions at the ﬁrst dif ﬁculty level needed to prog- ress to the second level was measured for each task (see Table 8).

All children had positive individual progress throughout the inter- vention as they gradually reached higher dif ﬁculty levels. Still, the speed of progress was different individually for each child. At the end of training, children with faster progress had attained approximately 1.5 –2 times higher dif ﬁculty levels compared to the children with slower progress (see Fig. 5).

3.4. Follow-up or secondary outcome In the follow-up assessment 1.31 years later –the secondary outcome –the study group showed signi ﬁcant positive long-term effects of intervention (see details in Table 5). Infocused attention ,the reaction time to visual stimuli was signi ﬁcantly faster compared to the ﬁ rst baseline performance. Also, improvements existed in all complex attention tasks: a higher percentage of correct responses, less omission errors, and lower percentage of total errors. In the tracking component, they had signi ﬁcantly more correct responses and a trend toward less commission errors. It is noteworthy that the waiting-list group performed signi ﬁcantly better compared to baseline performance in only two aspects of complex attention tasks (see details in Ta ble 6).

Figs. 2 –4 show long-term results for both groups and the compari- son with primary and secondary outcomes in baseline tasks.

3.5. Generalized effect of attention rehabilitation All children and 9 parents gave feedback about the intervention effect. Subjective feedback from parents suggested positive behavioral change in children. As a manifestation of the generalized effect, the parents stated that their children were less distracted and more likely to engage in social communication. Also, according to the parent re- ports, reading, writing, mathematics, and visuomotor skills had im- proved. Children stated improved concentration skills and better functioning in school tasks.

4. Discussion This research was conducted to test a modern computer-based in- tervention method for children with attention impairment and examine the rehabilitation effect on different attention components.

4.1. Effective computer-based rehabilitation method for children with an individual-based intervention protocol In order to speci ﬁcally illustrate the attention impairment pro ﬁles of children with epilepsy, in addition to neuropsychological assessment, the baseline tasks of the FORAMENRehab Attention module were used. Our results demonstrated that the assessment successfully differ- entiated children with epilepsy from their healthy peers as their perfor- mance levels were signi ﬁcantly lower in three out of the four measured attention components. Therefore, the baseline tasks demonstrated a signi ﬁcant need for further intervention for the children with epilepsy.

The current study demonstrated that the Attention module of the computer-based intervention program FORAMENRehab is a suitable method for children with epilepsy. The rehabilitation design success- fully supported continuous progress during the intervention process Table 7 Average attained dif ﬁculty levels at the end of intervention in four attention components.

Attention component Nr of task Mean level 95% con ﬁdence interval Focused attention 1 3.88 3.69 4.06 24 – Sustained attention 1 3.312.94 3.69 2 3 2.81 3.19 3 3.56 3.23 3.9 Complex attention 1 1.19 a 0.971.4 2 1.62 a 1.3 1.95 3 1.88 a 1.61 2.14 Tracking 11.31 a 0.99 1.63 aTasks with slower progress. Table 8 Average number of trainings attended before moving from ﬁrst dif ﬁculty level to second level.

Attention component Nr of task Mean sessions 95% con ﬁdence interval Focused attention 1 1.75 1.22 2.28 2 1.56 1.01 2.11 Sustained attention 1 2.12 1.51 2.74 2 1.38 0.99 1.76 3 1.62 1.2 2.05 Complex attention 1 7.69 a 6.56 8.82 2 5.81 a 4.46 7.16 3 4.62 a 3.37 5.88 Tracking 15.62 a 4.69 6.56 aTasks with slower progress. Fig. 5.Children's individual progress trajectories for 10 training sessions (summary score per visit).

118 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111–121 in different attention components. One of the strengths of this method is that it is tailored to follow each individual's abilities and attention impairment proﬁle. Amonn et al. [42]stated that in order to prove clin- ical value, cognitive training programs should “focus more strongly on individually existing neuropsychological de ﬁcits ”. The training protocol used in this study allowed the observation of progress in each individual child as well as the children's overall progress within attention compo- nents, and enabled comparison of outcomes between children and groups.

The rehabilitation of children is different from interventions for adults because children do not enter the process through their own initiative. Therefore, they need continuous guidance and motivation throughout the rehabilitation period. According to Cicerone et al. [43], active therapist involvement enhances the overall effectiveness of reha- bilitation. Our design involves a therapist to help make individual plans by following each child's progress, and who follows the training protocol. The therapist motivates the child and supports the use of acquired skills in everyday life situations. Charvátová et al. [44]pointed out that children do not make a strict distinction between games, work, leisure, and educational activities, but the crucial characteristics are mo- tivation, competitiveness, emotions, and natural curiosity. This con- vinces us of the importance of a therapist in the rehabilitation process.

Children with epilepsy have also previously shown improved behavior- al performance when presented with rewards which could provide sig- niﬁ cant bene ﬁt in cognitive remediation programs [45]. Our study implies that guided intervention is especially valuable in children with attention impairment, as they continually need extra help in directing attention to tasks. It leads the child to become more aware of different options and to learn to compensate for cognitive weaknesses. Therefore, it could help them to become more independent in the learning process.

Our experience showed that if the child understands the solution pro- cess by using a speci ﬁc strategy, he/she has a successful experience and gains motivation and self-con ﬁdence. The children were more at- tentive in classes and in solving homework –they improved in pinpointing mistakes, revising work, ﬁnding solutions to problems, etc. Therefore, the support from the therapist might have also contribut- ed to the effect of the training. It is a valuable additional bene ﬁttothe intervention; children with epilepsy have been reported to perceive stigma associated with the “need for information and support ”[46] , and present with emotional problems [47].

The Neuropsychology Task Force of the International League Against Epilepsy (ILAE) stated the importance of providing the patients' families with implications of assessment results and clinical recommendations regarding what can be done for cognitive improvement [48].Inour study, the parents received personal feedback about their child's atten- tion pro ﬁle, progress throughout the training, and suggestions for future training possibilities at home (e.g., via the Internet). They were also pro- vided with additional advice for supporting their child's general learn- ing abilities and for considering their child's individual differences.

4.2. Effects of rehabilitation in the study group Effects of rehabilitation were studied by comparing the perfor- mances in baseline tasks (used only for examination) before and after the intervention period. After active training for ﬁve weeks, the study group children's performance improved signi ﬁcantly in two attention components: complex attention andtracking tasks showed signiﬁcant improvement in various aspects. Van't Hooft et al. [49]previously also described positive change in children after rehabilitation in complex at- tention tasks, in contrast to the simpler reaction time tests. A better out- come in complex attention could show improved abilities to concentrate on tasks, and to divide and shift attention between stimuli.

Less commission errors may also suggest improvement in impulse control and behavior regulation, as commissions in tasks have been described as indicating impulsivity by rapid, but incorrect responses [50] . Also, the tracking component of attention or tracking the processes of a task improved during training, and therefore the ability to simultaneously process information received either a short time ear- lier or in the moment was positively affected.

Infocused attention , no signiﬁcant change was discovered when measuring visual and auditory reaction times; similar results have been reported by Van't Hooft et al. [49]and Cicerone [51].Justlike Van't Hooft et al. [49], we also believe that although training might not improve reaction times, it still provides the children with valuable solution techniques and strategies. Furthermore, by measuring children's individual progress at each dif ﬁculty level, distinctive differences in the more complicated attention components were revealed. At the end of rehabilitation, the study group remained at lower dif ﬁculty levels in complex attention and tracking components compared to focused and sustained attention. Although the comparison with baseline tasks showed signi ﬁcant improvements in complex attention and tracking, these components would need lon- ger training for more effective remediation. The positive rehabilitation effect in the intervention group was further con ﬁrmed in comparison with the waiting-list group. After the 5-week period without intervention, the waiting-list group demon- strated no signi ﬁcant change between the ﬁrst and second assessment with baseline tasks. Furthermore, the follow-up assessment after 1.31 years showed the sustained, positive, long-term effect of rehabilita- tion in the study group, in contrast to the children in the waiting-list group who demonstrated remarkably less positive dynamics over time (aside from the developmental curve). This emphasizes the effective- ness of intervention and diminishes the chance of a positive outcome solely due to normal developmental processes. Thereafter, the parents conclusively con ﬁrmed the sustained positive effect and reported noticeable positive changes in their children's everyday life situations.

The generalized effect of rehabilitation manifested in the children's behavior and overall performance in school. It was also valuable that the family became aware of the children's cognitive strengths and weaknesses as well as the importance of cogni- tive rehabilitation. In the end, full compliance and positive feedback from children showed that the computer-based neurorehabilitation is pleasurable for children.

4.3. Clinical implications for rehabilitation Developing an intervention design with a speci ﬁc protocol and well- de ﬁned instructions for therapists is recommended. In a review by Sohlberg et al. [52]the importance of giving systematic instructions in cognitive rehabilitation is emphasized. These should consist of “simple, consistent instructional wording and scripts to reduce confusion and focus the learner on relevant content ”.

In addition, based on qualitative feedback, we noted that an important task for the therapist is to ﬁnd speciﬁ c motifs for each individual child. (e.g., for the best cooperation some children pre- ferr ed joking, some needed little breaks after a while, and others hoped for frequent appraisal and endorsement). Yet some children also required more speci ﬁc boundaries to keep them on track. In some cases, children preferred the therapist to be of the same gender as themselves. This might have contributed to making a better connec- tion and feeling of companionship between the child and the therapist.

Ultimately, an individual approach to motivation was essential to enhance the positive intervention effect and see the best possible improvement levels. Furthermore, it was subjectively noticeable that the children's moti- vation was also in ﬂuenced by their parents' attitude toward training sessions. If parents were less keen to participate in the regular meetings and verbally explained their lack of motivation, their children also showed less acceptance with training. Educating parents about the importance and outcome of rehabilitation facilitates cooperation and compliance. 119 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111 –121 4.4. Limitations and future directionsThe study also has limitations. For one, the study group was com- posed of children who lived in Tartu city or near Tartu, which means that the group was not completely randomly chosen. Secondly, the information about the children belonging either to the study group or waiting-list group was known to the therapist and this was not a blind study. Therefore, the results might have been affected by this to some extent. Thirdly, the children who were below the normal intelligence level were not included in the study, but the comparison of results for intelligence tests between study and control groups was not shown in detail. Also, typical for intervention studies, the sample size was relatively small. Therefore, continued research is recommended with a larger sample size. Still, it should be noted that the intervention is a time-consuming and long-lasting process and therefore, resources may be limited for conducting research with substantial sample sizes.

5. Conclusions Our multifaceted neurorehabilitation design with FORAMENRehab is effective for children with epilepsy. Training speci ﬁc attention com- ponents combined with individually tailored motivation and teaching metacognitive strategies improved attention function in children with epilepsy. Nonetheless, three attention components –sustained, com- plex and tracking –need selective and longer training for more effective remediation. Signi ﬁcant long-term improvements in complex attention and tracking components persisted after 1.3 years. The current comput- er program is a modern and suitable method for children with epilepsy.

The intervention design combines principles of modern computer- assisted neurocognitive rehabilitation and an individualized approach.

It holds practical future bene ﬁts as an effective intervention, which is a prerequisite for outpatient training sessions in clinical settings.

Intervention effectiveness is best described with positive results from assessments with baseline tasks and development on dif ﬁculty levels of tasks. Hence, these outcome assessment methods should be used simultaneously to examine the rehabilitation effect. Individual improvement is important to follow as complicated tasks are associated with slower progress and thus need a longer training period. As men- tioned, the effect of individually tailored motivation and metacognitive strategies must be taken into account for the effectiveness of the reha- bilitation program. Our work implies that if the child understands the solution process in cognitive tasks by using speci ﬁc strategies, he/she experiences success and gains motivation and self-con ﬁdence. The im- portance of a therapist in the guided rehabilitation process is especially valuable for children with attention impairment because these patients need continual extra help in directing attention to tasks. We recom- mend the computer-based FORAMENRehab in attention impairment in- tervention to support the basic learning skills for children with partial epilepsy and cognitive impairments.

Author contributions Marianne Saard had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Marianne Saard, Anneli Kolk.

Acquisition of the data: Marianne Saard, Madis Bachmann, Lisanna Pertens. Analysis or interpretation of data: Marianne Saard, Anneli Kolk.

Statistical analysis: Marianne Saard, Pille Kool.

Drafting of the manuscript: Marianne Saard, Anneli Kolk, Mari-Liis Kaldoja. Critical revision of the manuscript for important intellectual content:

Anneli Kolk. Obtained funding: Anneli Kolk.

Supervision: Anneli Kolk. Declaration of con ﬂicting interests We have no con ﬂicts of interest to declare.

Funding This study was supported by the Estonian Science Foundation (GARLA 9016) and by the Estonian Ministry of Education and Research TARLA 2695.

Ethical approval The study was approved by The Research Ethics Committee of the University of Tartu 190T-21.

Acknowledgements The work was done at the Department of Neurology and Neurorehabilitation, Children's Clinic, Tartu University Hospital, Tartu, Estonia. The authors thank all the patients, control group children, and their parents for their valuable contribution to this study.

References [1] Loring DW, Meador KJ. In: Loring DW, editor. INS dictionary of neuropsychology.

USA: Oxford University Press; 1999. p. 24.

[2] Sohlberg MM, Mateer CA. Introduction to cognitive rehabilitation: theory and practice. New York: Guilford Press; 1989.

[3] Sohlberg MM, Mateer CA. Improving attention and managing attentional problems.

Ann N Y Acad Sci 2001;931:359 –75.

[4] Sohlberg MM. Delivery and evaluation of direct attention training for people with acquired cognitive impairments. Contemporary forums issues in brain injury, Washington DC; 2013. https://education.uoregon.edu/users/sohlberg .

[5] Sohlberg MM, Mateer CA. Effectiveness of an attention-training program. J Clin Exp Neuropsychol 1987;9:117 –30.

[6] Lezak MD. Neuropsychological assessment. USA: Oxford University Press; 2004.

[7] Guzeva VI, Belash VO, Guzeva VV, Guzeva OV, Anastazi OI. Characteristics of cogni- tive functions in children with epilepsy. Neurosci Behav Physiol 2009;39:885 –9.

[8] Engle JA, Smith ML. Attention and material-speci ﬁc memory in children with lateralized epilepsy. Neuropsychologia 2010;48:38 –42.

[9] Kolk A, Beilmann A, Tomberg T, Napa A, Talvik T. Neurocognitive development of children with congenital unilateral brain lesion and epilepsy. Brain Dev 2001;23:

88–96.

[10] Rathouz PJ, Zhao Q, Jones JE, Jackson DC, Hsu DA, Stafstrom CE, et al. Cognitive devel- opment in children with new onset epilepsy. Dev Med Child Neurol 2014;56:

635–41.

[11] Gülgönen S, Demirbilek V, Korkmaz B, Dervent A, Townes BD. Neuropsychological functions in idiopathic occipital lobe epilepsy. Epilepsia 2000;41:405 –11.

[12] Baglietto MG, Battaglia FM, Nobili L, Tortorelli S, De Negri E, Calevo MG, et al. Neuro- psychological disorders related to interictal epileptic discharges during sleep in benign epilepsy of childhood with centrotemporal or Rolandic spikes. Dev Med Child Neurol 2001;43:407 –12.

[13] Semrud-Clikeman M, Wical B. Components of attention in children with complex partial seizures with and without ADHD. Epilepsia 1999;40:211 –5.

[14] Piccirilli M, D'Alessandro P, Sciarma T, Cantoni C, Dioguardi MS, Giuglietti M, et al.

Attention problems in epilepsy: possible signi ﬁcance of the epileptogenic focus.

Epilepsia 1994;35:1091 –6.

[15] Bennett-Levy J, Stores G. The nature of cognitive dysfunction in school-children with epilepsy. Acta Neurol Scand 1984;69:79 –82.

[16] Massa R, de Saint-Martin A, Carcangiu R, Rudolf G, Seegmuller C, Kleitz C, et al. EEG criteria predictive of complicated evolution in idiopathic rolandic epilepsy. Neurology 2001;57:1071 –9.

[17] Aldenkamp AP, Van Bronswijk K, Braken M, Diepman LA, Verwey LE, Van Den Wittenboer G. A clinical comparative study evaluating the effect of epilepsy versus ADHD on timed cognitive tasks in children. Child Neuropsychol 2000;6:209 –17.

[18] Metz-Lutz MN, Kleitz C, de Saint MA, Massa R, Hirsch E, Marescaux C. Cognitive development in benign focal epilepsies of childhood. Dev Neurosci Basel 1999;21:

182–90.

[19] Chun MM, Turk-Browne NB. Interactions between attention and memory. Curr Opin Neurobiol 2007;17:177 –84.

[20] Genizi J, Shamay-Tsoory SG, Shahar E, Yaniv S, Aharon-Perez J. Impaired social behavior in children with benign childhood epilepsy with centrotemporal spikes.

J Child Neurol 2012;27:156 –61 .

[21] Fastenau PS, Shen J, Dunn DW, Austin JK. Academic underachievement among children with epilepsy proportion exceeding psychometric criteria for learning disability and associated risk factors. J Learn Disabil 2008;41:195 –207.

[22] Sarajuuri JM, Koskinen SK. Holistic neuropsychological rehabilitation in Finland: the INSURE program —a transcultural outgrowth of perspectives from Israel to Europe via the USA. Int J Psychol 2006;41:362 –70.

120 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111–121 [23]Ylvisaker M, Szekeres SF, Haarbauer-Krupa J. Cognitive rehabilitation: organization, memory, and language. In: Ylvisaker M, editor. Traumatic brain injury rehabilitation:

children and adolescents. 2nd ed. Woburn: Butterworth-Heinemann; 1998.

p. 181 –220.

[24] Cicerone KD, Langenbahn DM, Braden C, Malec JF, Kalmar K, Fraas M, et al. Evidence- based cognitive rehabilitation: updated review of the literature from 2003 through 2008. Arch Phys Med Rehabil 2011;92:519 –30.

[25] Bangirana P, Allebeck P, Boivin MJ, John CC, Page C, Ehnvall A, et al. Cognition, behav- iour and academic skills after cognitive rehabilitation in Ugandan children surviving severe malaria: a randomised trial. BMC Neurol 2011;11:96.

[26] Bangirana P, Giordani B, John CC, Page C, Opoka RO, Boivin MJ. Immediate neuropsy- chological and behavioral bene ﬁts of computerized cognitive rehabilitation in Ugandan pediatric cerebral malaria survivors. J Dev Behav Pediatr 2009;30:310.

[27] Boivin MJ, Busman RA, Parikh SM, Bangirana P, Page CF, Opoka RO, et al. A pilot study of the neuropsychological bene ﬁts of computerized cognitive rehabilitation in Ugandan children with HIV. Neuropsychology 2010;24:667.

[28] Hardy KK, Willard VW, Bonner MJ. Computerized cognitive training in survivors of childhood cancer a pilot study. J Pediatr Oncol Nurs 2011;28:27 –33.

[29] Lee J, Harn B, Sohlberg MM, Wade SL. An overview of the attention improvement management (AIM) program with outcomes for three pilot participants. SIG 2.

Perspect Neurophysiol Neurogen Speech Lang Disord 2012;22:90 –105.

[30] Galbiati S, Recla M, Pastore V, Liscio M, Bardoni A, Castelli E, et al. Attention remedia- tion following traumatic brain injury in childhood and adolescence. Neuropsychology 2009;23:40.

[31] Engelberts NH, Klein M, Adèr HJ, Heimans JJ, Trenité DG, Van der Ploeg HM. The effectiveness of cognitive rehabilitation for attention de ﬁcits in focal seizures: a randomized controlled study. Epilepsia 2002;43:587 –95.

[32] Gupta A, Naorem T. Cognitive retraining in epilepsy. Brain Inj 2003;17:161 –74.

[33] Kaldoja ML, Saard M, Lange K, Raud T, Teeveer OK, Kolk A. Neuropsychological ben- eﬁ ts of computer-assisted cognitive rehabilitation (using FORAMENRehab program) in children with mild traumatic brain injury or partial epilepsy: a pilot study.

J Pediatr Rehabil Med 2015;8:271 –83.

[34] Slomine B, Locascio G. Cognitive rehabilitation for children with acquired brain injury. Dev Disabil Res Rev 2009;15:133 –43.

[35] Limond J, Leeke R. Practitioner review: cognitive rehabilitation for children with acquired brain injury. J Child Psychol Psychiatry 2005;46:339 –52.

[36] Van't Hooft I, Andersson K, Sejersen T, Bartfai A, Von Wendt L. Attention and memory training in children with acquired brain injuries. Acta Paediatr 2003;92:

935–40.

[37] Prigatano GP. A brief overview of four principles of neuropsychological rehabilita- tion. In: Christensen AL, editor. International handbook of neuropsychological rehabilitation. New York: Springer US; 2000. p. 115 –25. [38] Backeljauw B, Kurowski BG. Interventions for attention problems after pediatric traumatic brain injury: what is the evidence? PM R 2014;6:814 –24.

[39] Sarajuuri JM, Koskinen SK. Holistic neuropsychological rehabilitation in Finland: The INSURE program —a transcultural outgrowth of perspectives from Israel to Europe via the USA. Int J Psychol 2006;41:362 –70.

[40] Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and po werful approach to multiple testing. J Roy Stat Soc B 1995;57:289 –300.

[41] Cliff N. Answering ordinal questions with ordinal data using ordinal statistics.

Multivar Behav Res 1996;31:331 –50.

[42] Amonn F, Frölich J, Breuer D, Banaschewski T, Doepfner M. Evaluation of a computer-based neuropsychological training in children with attention-de ﬁcit hyperactivity disorder (ADHD). NeuroRehabilitation 2013;32:555 –62.

[43] Cicerone KD, Dahlberg C, Kalmar K, Langenbahn DM, Malec JF, Bergquist TF, et al.

Evidence-based cognitive rehabilitation: recommendations for clinical practice.

Arch Phys Med Rehabil 2000;81:1596 –615.

[44] Charvátová G. The game world and its in ﬂuence on our lives [online]. MU Brno cit 2007 2012-04-26. Available from: http://is.muni.cz/th/135429/pedf_b/bak.prace.

pdf .

[45] Triplett RL, Velanova K, Luna B, Padmanabhan A, Gaillard WD, Asato MR. Investigat- ing inhibitory control in children with epilepsy: an fMRI study. Epilepsia 2014;55:

1667–76.

[46] Austin JK, Perkins SM, Dunn DW. A model for internalized stigma in children and adolescents with epilepsy. Epilepsy Behav 2014;36:74 –9.

[47] Borgatti R, Piccinelli P, Montirosso R, Donati G, Rampani A, Molteni L, et al. Study of attentional processes in children with idiopathic epilepsy by Conners' continuous performance test. J Child Neurol 2004;19:509 –15.

[48] Wilson SJ, Baxendale S, Barr W, Hamed S, Lang ﬁtt J, Samson S, et al. Indications and expectations for neuropsychological assessment in routine epilepsy care: report of the ILAE Neuropsychology Task Force, Diagnostic Methods Commission, 2013 –2017. Epilepsia 2015;56:674 –81.

[49] van't Hooft I, Andersson K, Bergman B, Sejersen T, von Wendt L, Bartfai A. Sustained favorable effects of cognitive training in children with acquired brain injuries.

NeuroRehabilitation 2007;22:109 –16.

[50] Halperin JM, Wolf L, Greenblatt ER, Young G. Subtype analysis of commission errors on the continuous performance test in children. Dev Neuropsychol 1991;7:207 –17.

[51] Cicerone KD. Remediation of'working attention'in mild traumatic brain injury. Brain Inj 2002;16:185 –95.

[52] Sohlberg MM, Ehlhardt L, Kennedy M. Instructional techniques in cognitive rehabil- itation: a preliminary report. Semin Speech Lang Nov 2005;26(4):268 –79. 121 M. Saard et al. / Epilepsy & Behavior 67 (2017) 111 –121