J Head Trauma Rehabil 2014 29 E49

J Head Trauma Rehabil
Vol. 29, No. 5, pp. E49–E64

c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
Copyright
Context-Sensitive Goal Management

Training for Everyday Executive
Dysfunction in Children After Severe
Traumatic Brain Injury
Agata Krasny-Pacini, MD; Jenny Limond, PhD; Jonathan Evans, PhD; Jean Hiebel, MD;
Karim Bendjelida, MD; Mathilde Chevignard, MD, PhD
Objectives: To assess the effectiveness of a metacognitive training intervention, based on an adapted Goal Manage-
ment Training and Ylvisaker’s principles, on 3 activity domains of executive functions: (1) prospective memory (PM)
performance in ecological setting, (2) complex cooking task management, and (3) daily executive functioning (EF)
at home and at school. Participants: Five children aged 8 to 14 years, who were 3 to 11 years post–severe traumatic
brain injury, experiencing severe EF difficulties in daily life. Design: Single-case experimental design and assessment
of EF twice prior to intervention, postintervention, and 3 and 6 months postintervention. Progress was monitored
by a weekly ecological PM score. The effect on EF was assessed using the Children’s Cooking Task. Transfer to the
child’s natural context was assessed by parental and teacher questionnaires and Goal Attainment Scaling. Results:
All children improved both on the measure of PM and on questionnaires of daily EF. Two children improved
on the Children’s Cooking Task but returned to their preintervention level in a novel cooking task at follow-up.
Participation of school personnel and parents in the program was low. Conclusions: It is feasible but challenging
to use Goal Management Training in children with traumatic brain injury. Further research is needed in relation
to how to promote generalization and how to increase the involvement of the child’s “everyday people” in the
intervention. Key words: child, daily life activities, ecological, executive functions, Goal Management Training, intervention,
prospective memory, traumatic/acquired brain injury
Author Affiliations: Institut Universitaire de Réadaptation

Clemenceau-Strasbourg, Strasbourg, France, and Paediatric Surgery
Department of Hautepierre University Hospital, Strasbourg, France
E XECUTIVE FUNCTIONS are a collection of re-
lated but distinct abilities that allow individuals
both to engage efficiently in intentional, goal-directed,
(Drs Krasny-Pacini and Hiebel); Rehabilitation Department for Children problem-solving actions1,2 through conscious and ef-
with Acquired Brain Injury, Hôpitaux de Saint Maurice, Saint Maurice,
France (Dr Chevignard); Mental Health and Wellbeing, Institute of Health fortful processing3 and to adapt to new situations in
& Wellbeing, Gartnavel Royal Hospital, University of Glasgow, Glasgow, the real world.4 Executive functioning (EF) deficits are a
Scotland, United Kingdom (Drs Limond and Evans); ER 6-UPMC, frequent consequence of traumatic brain injury (TBI).5
Pitié-Salpêtrière Hospital, University Pierre et Marie Curie–Paris 6, Paris,
France (Drs Krasny-Pacini and Chevignard); and EMOI-TC 68, Hôpital The outcome of TBI is predicted by EF level.6
du Hasenrain, Mulhouse cedex, France (Dr Bendjelida). There is a lack of validated methods for EF rehabilita-
This study was supported by scholarships from the “SOFMER” (French Society tion in children,7–9 although some general rehabilitation
of Physical and Rehabilitation Medicine), the “SFERHE” (French Research principles have proven to be useful. Ylvisaker10 empha-
Society against Children’s handicap), and the “Fondation Gueules Cassées.” sized 2 principles: (1) the key role both of parents and of
The authors thank Brian Levine, Ian Robertson, and Tom Manly, for pro- all the “everyday people” surrounding the child in the
viding Goal Management Training (GMT) materials, and Leigh Schrieff,
Kevin Thomas, and Clare Corbett, for providing their pilot pediatric version
The authors declare no conflict of interest.
of GMT. The authors also thank all rehabilitation teams that participated in
the recruitment of children for the study and especially the children’s depart- Supplemental digital content is available for this article. Direct URL citations
ment of Clemenceau University rehabilitation Center in Strasbourg; Marilyne appear in the printed text and are provided in the HTML and PDF versions
Periot, from Mulhouse Hospital Rehabilitation Department and EMOI-TC of this article on the journal’s Web site (www.headtraumarehab.com).
Unit; Doctor Anne Laurent-Vannier and the occupational therapy depart- Corresponding Author: Agata Krasny-Pacini, MD, Institut Universitaire de
ment of INR A in Saint-Maurice; Sylvie Lazourenko, for the illustrations of Réadaptation Clemenceau-Strasbourg, 45 bd Clemenceau, 67082 Strasbourg,
the theoretical modules; Arnaud Roy for providing the Fonctions Exécutives France (agata.krasny@ugecam-alsace.fr OR agatakrasny@yahoo.com).
Enfant battery and for his useful advice; and the 2 anonymous reviewers for
their valuable and constructive feedback on the paper’s previous drafts. DOI: 10.1097/HTR.0000000000000015
E49
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
E50 JOURNAL OF HEAD TRAUMA REHABILITATION/SEPTEMBER–OCTOBER 2014
“cognitive coaching” of their child (see the studies by for the parents of children with TBI.30 Prospective mem-
Braga et al11 and Wade et al12,13 for examples of family- ory is impaired in children with TBI31,32 compared with
delivered interventions in TBI); and (2) the necessity of a children with orthopedic injuries, even after cues are
“context-sensitive” approach,14 embedded in functional given33 and even under strong incentive conditions.34
routines of everyday life and using meaningful activities The primary aim of this study was to assess the fea-
rather than decontextualized exercises. sibility and effectiveness of implementing a metacogni-
In current clinical practice, 4 types of approaches tive training intervention, based on an adapted form of
are used to help children with their EF difficulties, al- GMT and Ylvisaker’s rehabilitation principles, in 3 do-
though each has its limitations: (1) Providing environ- mains: (1) PM performance, (2) complex cooking task
mental support and compensatory aids (eg, use of elec- management, and (3) daily EF at home and at school.
tronic prompting devices)—although only a restricted Secondary aims were to determine whether the effects of
number of situations lend themselves to such an ap- such a metacognitive training generalize enough to help
proach. (2) Training component EF skills (eg, repeated children to (1) achieve personalized, untrained goals and
exercises aiming at changing the brain’s working mem- (2) manage a demanding novel task that requires execu-
ory capacity)—although transfer to natural contexts and tive functions.
generalization to untrained activities effects of this ap-
proach have rarely been demonstrated.14 (3) Training METHODS
children on specific goals (eg, if preparing a schoolbag
is problematic, the child will be trained on this specific Participants
activity until the goal is achieved)—although generaliza- The study was approved by the ethics committee of
tion to similar goals in different situations (eg, preparing Pitié-Salpêtrière University Hospital in Paris, France. In-
a suitcase for holidays) is often not achieved. (4) Provid- formed parental written consent and participation assent
ing children with metacognitive strategies applicable to were obtained for all participants before initiating any
a variety of everyday situations—while this is effective procedure.
in some adults after TBI,15 there is little evidence that Inclusion criteria were as follows: (1) severe TBI
metacognitive training is effective for children with a (initial Glasgow Coma Scale score <9); (2) sustained
dysexecutive syndrome after TBI,16 although research at least 2 years previously; (3) children attending 1 of
on children with other forms of brain injury suggests it the 2 participating rehabilitation departments; (4) aged
may be a useful approach.16,17 8 to 14 years; (5) evidence of a dysexecutive syndrome
Goal Management Training (GMT)18 is one of the on neuropsychological assessment performed at least 2
most studied metacognitive training programs, of which years postinjury; and (6) parental report of EF difficulties
many variants exist.19 It includes self-instruction strate- in daily life.
gies, self-monitoring exercises, metacognitive strategies Exclusion criteria were as follows: (1) diagnosed learn-
aimed at improving planning, prospective memory ing disabilities, neurological, or psychiatric condition
(PM) and hierarchical goal management, mindfulness prior to TBI; (2) severe intellectual disability; and (3)
practice exercises, stories promoting discussion about insufficient French language level of the child or of his
executive dysfunction in daily life, and homework or her family.
assignments (see the study of Levine et al20 for a more Characterization data included classical standard-
detailed description). Goal Management Training was ized tests from the Wechsler Intelligence Scale for
developed from the theory of “goal neglect” of Duncan Children,35 the Children’s Memory Scale,36 and a
et al,21,22 which suggests that dysexecutive patients French battery of EF for children: “Fonctions Exécutives
are impaired in the construction and use of “goal Enfant” battery (A. Roy, J-L. Roulin, D. Le Gall, N.
lists,” necessary for goal-directed behavior. They do Fournier, Groupe FEE, unpublished data). The battery
remember the intended goal but tend to lose sight has a much larger normative data than any other EF
of it as they progress through a task leading to a PM test available in French. Each child participating in the
failure. Prospective memory (remembering to carry out intervention was compared with a sample of controls
intended actions) tasks require retrospective memory to matched for sex, age, and socioeconomic status from
remember the task but depend on EF23 for successful the Fonctions Exécutives Enfant database, using the
goal maintenance, retrieval, and implementation at the method of Crawford and Garthwaite,37 with 1-tailed
right moment. It depends upon frontal lobe integrity,24 probability38 taking P < .05 criteria for significance.
with a key role for rostral prefrontal cortex (BA10).25 In
typically developing children aged 6 to 12 years, perfor-
Intervention
mance on EF tasks such as planning and switching,26
working memory,27 and inhibition28 is correlated with The intervention was inspired by GMT18 but ex-
PM.29 Problems of PM are reported as a major concern tended to follow Ylvisaker’s principles of involving
Context-Sensitive GMT for Executive Dysfunction After Severe TBI E51
“everyday people” in the child’s social network in the Practical modules served to practice GMT content
cognitive coaching of the child and using a context- in meaningful activities (Ylvisaker’s content-sensitive
sensitive approach prioritizing functional ecological principle): cooking of various meals, route finding,
activities. searching for information, poster making, and photo
Prior to the intervention, parents participated in a 2- ordering. This was an explicit generalization training
to 3-hour informal interview during which the following that was aimed both to show the children that GMT
were discussed: (1) the child’s daily life difficulties at metacognitive strategies are applicable to many situa-
home and at school; (2) program content and key GMT tions in life where their EF impairment may impede
concepts; and (3) the need to apply the metacognitive success and to promote the use of these strategies
GMT techniques at home and at school. An emphasis to personal (present and future) goals and untrained
was placed on the key role parents played for the success activities.40 The activities involved planning, strategy
of the intervention. generation, following steps, and monitoring of actions.
The intervention comprised (1) GMT theoretical Difficulty increased as children progressed in the pro-
modules and between-session “missions” (promoting gram. Similarly to the theoretical modules, the meaning-
GMT use at home and at school); (2) practical modules ful activities contained naturally occurring PM tasks to
in which children practiced GMT content in meaningful allow discussion about PM difficulties in real life activi-
activities; and (3) an “everyday people cognitive coach- ties (eg, checking regularly if the first set of finger biscuits
ing guide.” The materials included PowerPoint slides, is cooked while preparing the next). The trainer guided
a workbook, posters for home and school use, mission the children when needed, using nonspecific prompts
sheets, and the cognitive coaching guide. The 15 mod- and general cues. Explicit help was given only if these
ules (theoretical modules + meaningful activities) of the were not sufficient. At the end of each activity, the child
intervention required 15 to 20 hours to be completed was invited to review his or her performance using the
and were administered individually over a 4- to 6-month “mission” sheets, identify Oops errors and any effective
period on a weekly basis, either in a rehabilitation center strategy that had helped in the task and to think about
or at home. situations in real life where the same kind of strategy may
Theoretical modules were derived from the adult be useful. The child always took the “product” of his or
GMT PowerPoint Manual,20 developed by Levine, her activity (eg, crepes) home to increase motivation
Robertson, and Manly, that has already been used (with from the praise he or she received at home. Moreover,
minor changes) in children.39 A new, shorter, color- the product was expected to remind the parents about
ful version was created for the intervention to make the child’s program.
materials child-friendly, age-appropriate, enjoyable, and To encourage transfer to the child’s natural contexts,
simpler. For example, “slips” (referring to slips of at- we tried to involve the child’s “everyday people” as cog-
tention) became “Oops errors.” “To do lists” and the nitive coaches for their child. Everyday people were
“mental blackboard” were combined in a unique “note- parents, teachers, school assistants, and any adult the
book” concept to explain to the children how a real parents identified as a potential cognitive coach (babysit-
paper notebook can help them not to overwhelm their ter, student helping the child with homework). A letter
“mental notebook.” Discussion about EF failures in presenting the program was sent to the child’s teacher
daily life was triggered through illustrated stories re- and school assistant, explaining briefly TBI executive
lating to school and leisure activities. In each mod- problems and their implications and asking the teacher
ule, we included a PM task to be performed during and the school assistant to participate by applying GMT
the session (eg, when you see a slide with X, you do at school. The intervention content was not explained
Y) to encourage discussion about PM failures at the orally. We asked for a contact e-mail and a telephone
end of each session. The content-free cue “Look into contact to discuss the child’s difficulties and set realistic
your mental notebook” was used as a prompt when goals on goal attainment scales (GAS). The letter was
children failed PM tasks. Throughout the training pe- sent a second time after 1 month as the first yielded few
riod, children had to complete “mission sheets,” in- responses, so we had responses from at least 1 school
spired from GMT between-session assignments. These staff member per child. Other potential everyday peo-
were of 3 types, introduced progressively: (1) monitoring ple identified by the parents were sent a similar letter.
Oops errors, their consequences, and factors influenc- Twice a month all everyday people who agreed to par-
ing their occurrence; (2) listing occasions on which the ticipate received 1 chapter, 2-page long, of a cognitive
child used a metacognitive strategy of his or her own or coaching guide that was created for the intervention.
from the program with success; and (3) identifying situa- The guide was colorful, using the same drawings, dia-
tions where a stepwise processing approach can be used grams, and analogies as the theoretical modules, explain-
to manage a goal (preparing the schoolbag, preparing a ing the rehabilitation content, and suggesting how to ap-
sandwich). ply metacognitive strategies at school and at home. The
www.headtraumarehab.com
intervention was organized in a way that each metacog- (4) need for novelty for a task to truly test “executive”
nitive strategy was (1) first introduced during a theoreti- functions.19
cal module, (2) then practiced on a meaningful activity,
and (3) finally introduced to everyday people. Every- Outcome measures
day people were sent the corresponding chapter of the
coaching guide that described the activity for which the Repeated measures of PM performance: Single-case
child had already practiced the strategy and suggested experimental design
other activities to which it could be applied (see Sup- We monitored the effectiveness of the program
plemental Digital Content data, available at http://links. through a weekly score on a time-based PM task. The PM
lww.com/JHTR/A92, for examples). Through this guide, task was inspired by the phone call task of Fish et al43 that
everyday people were encouraged (1) to use nonspecific has recently been used in children.44 Three times a week,
prompts for PM failures (“Look into your mental note- children had to look up the day’s Saint on a calendar (eg,
book”) rather than specific instructions (“You need to 24th June is Saint John’s day) and send it to the therapist
feed the dog”) or negative sentences (“You’ve forgotten at an agreed target time as a text message, an e-mail, or
to feed your dog again!”); (2) to promote strategy gener- a phone call. The child was awarded 3 points if correct
ation instead of giving the solution to their child, consis- information was given within 1 hour of the agreed time,
tent with Ylvisaker’s aim of “helping the child to become 2 points if given within the day, 1 point if given on
a strategic thinker”10 ; (3) to prompt and help the child a different day, and zero points if the child completely
to fill in his or her “mission sheets” regularly; and (4) to forgot about the task. The retrospective memory compo-
practice goal identification (“state your goal”) and step- nent was controlled for by checking at each session that
wise processing in daily activities (preparing schoolbag, children remembered the task and agreed times. Parents
table setting). Parents were explicitly asked to go through were given the timings in case the child wanted to check
the metacognitive strategies of the cognitive coaching the target time. To encourage use of mental strategies,
guide and to sign the child’s GMT workbook every week. children were asked not to use cues such as alarms or
The everyday people were not requested to participate preprogrammed text messages. Parents were instructed
in the rehabilitation session, but they were advised that not to give any cues or help to complete the task. The 3
the therapist was available if they had any questions. target days and times were chosen individually for each
child with the parents before the first assessment to en-
Qualitative data about the program sure (1) that the child was easily available for the task (ie,
not during school time or leisure activity); (2) that it did
Throughout the sessions, the therapist recorded (1) not disturb family routines (eg, bedtime); (3) that the
how the child reacted to the intervention content; (2) time did not correspond to a regular activity that could
whether the metacognitive strategies were easily under- act as a cue (eg, TV show); (4) that timings respected the
stood and used; (3) whether the child seemed aware following common rules for all the children: noncon-
of his or her EF difficulties during discussion and per- secutive days, 1 weekend day, and 2 school days, except
formance of meaningful activities; and (4) whether the the day of the intervention and different target time on
“mission sheets” were filled in between the sessions. each of the 3 days. As in the study by Fish et al,43 days
when the tasks had not been performed for reason other
Study design than PM failure were not used in the analysis (eg, medi-
cal appointment at target time, no Internet connection
Intervention effectiveness was assessed in 2 ways: during a weekend outing); therefore, the total score was
(1) a single-case experimental design,41,42 with repeated expressed as the percentage of total possible points that
ecological measurements of PM, was used to monitor week.
progress throughout the intervention; and (2) pre-post
measurement of EF with 2 baseline assessments 4 to
“Children’s Cooking Task”: Pre-post ecological
8 weeks apart (B1 and B2), and 3 postintervention as-
measurement of EF
sessments, immediately after the intervention (R1) and
at 3 (R2) and 6 months (R3), to assess maintenance of As the aim of this study was to improve EF in daily
effects. Baseline was expected to be stable, as children life, assessment included an ecological45 test of EF,
had sustained their TBI at least 2 years earlier, minimiz- called the Children’s Cooking Task (CCT).46,47 In the
ing chances of spontaneous recovery during the study. CCT, children have to prepare a real chocolate cake
The pre-post measurement served to capture the key is- and a fruit cocktail following a structured, photo-cued,
sues of EF rehabilitation, namely: (1) EF performance child-friendly recipe contained in a cookbook including
in ecological tasks; (2) transfer of training effects to nat- distractors. The task has been shown to be highly sen-
ural contexts; (3) generalization to untrained tasks; and sitive to executive dysfunction in TBI, as it is novel,
challenging, open-ended, and requires multiple goal Goal Attainment Scaling: Pre-post measurement
management and innovative higher-level strategies to of generalization (metacognitive strategy use
succeed. The CCT has good interrater and test-retest in untrained tasks)
reliability, high internal consistency, as well as good
Goal Attainment Scaling57,58 was used as a generaliza-
discriminant and concurrent validity.46 It can be per-
tion measure to assess whether a child who has applied
formed from the age of 8. Scoring is based on the
metacognitive strategies to meaningful activities in a re-
number of errors, including omissions, additions, com-
habilitation setting is capable of applying those strategies
mentaries, substitution, and estimation errors. Norma-
to untrained tasks that are judged to be problematic for
tive data are not yet available. Therefore, raw scores
him or her by his or her everyday people. Problems re-
were used to track individual child changes. The num-
lated to EF reported by the child, parents, and school
ber of errors in the CCT made by each child was
staff served to elaborate personalized GAS goals for each
compared with the number of errors made by age-
child. These GAS goals were not trained specifically, but
matched healthy controls, extracted from unpublished
children were repeatedly encouraged to apply metacog-
data.48
nitive strategies to daily life. We also used “general”
GAS for metacognitive strategy use and GMT appli-
Questionnaires: Pre-post measurement of EF cation (see Table 3 in the “Results” section). Themes
in natural contexts of “general” goals were similar for all children, but the
initial levels and expected outcome levels were specific
Ultimately, the aim of any cognitive rehabilitation
to each child, taking into account children’s age and
intervention is to allow a transfer of learned skills to
possibilities. The detailed procedure for goal selection,
the natural context of the child,10 and this was particu-
GAS elaboration, and GAS levels adjustment are de-
larly important to assess because metacognitive training
scribed in Supplemental Digital Content data (available
aims at providing children with strategies they can ap-
at http://links.lww.com/JHTR/A90). The GAS scores
ply to many tasks in their natural context. This was
were used to calculate a global T-score for each child,
assessed through 2 questionnaires of EF: (1) the Be-
using the Kiresuk and Sherman formulae (see the works
havior Rating Inventory of Executive Function (BRIEF)
of Kiresuk and Sherman57 and Krasny-Pacini et al58 for
questionnaire,49,50 completed by parents (transfer to
details of GAS methodology). A T-score of 50 meant
home-context) and teachers (transfer to school-context);
that the goals were overall attained as expected, and a
(2) a Cognition subscore derived from the Dysexecutive
score of more than 50 meant that goals were attained
Questionnaire for Children (DEX-C),51 which was com-
better than expected.
pleted only by parents. The BRIEF assesses 8 domains
of EF in the real world, which give together a Global
Executive Composite score. Higher scores correspond “Christmas biscuits task”: Measurement of adaptation to
to increased EF difficulties in daily life. A T-score supe- novelty, performed only once at the end of the intervention
rior to 65 is defined as the clinical range. The BRIEF has EF outcome measures need to be novel to really cap-
large normative data for children aged 5 to 18 years, high ture EF.59,60 When the same task is repeated after in-
internal consistency,52 good validity53 and good test- tervention, it is more “familiar,” which can make it less
retest reliability,52 although parent-teacher agreement is demanding on EF.61,62 Familiarity effects increase when
only moderate.52 It is the most commonly used ques- patients are tested on several occasions (as it is the case
tionnaire of executive functions and seems to be sen- in our design for the CCT). To get a “purer” EF mea-
sitive to deficits in EF in children with TBI.53–56 Its sure postintervention,62 we developed a parallel form
relationship with common EF cognitive tests is, how- of the CCT for assessment at R3, involving the same
ever, inconsistent. The DEX-C is a 20-item question- number of steps and ingredients but requiring different
naire, probing 4 broad areas of EF difficulties (emo- types of ingredients and procedures. This version has
tional/personality, motivational, behavioral, and cogni- no established psychometric properties. Children had
tive), and is part of the “Behavioural Assessment of the to bake “Christmas biscuits.” While both tasks required
Dysexecutive Syndrome for Children.”51 Higher z scores cooking, as children were not experienced cooks, a new
correspond to increased cognitive difficulties relating to recipe could not be viewed as a familiar task.
EF. The DEX-C has less evidence regarding psychome-
tric properties.51 However, the DEX-C Cognition sub-
Controlling for confounding factors
scale completed by parents has a high correlation with
the CCT score48 and therefore this subscale together At the beginning of the program, everyday people
with the CCT was expected to detect improvement in were not informed of exactly when the intervention
cognitive EF impairment, which the BRIEF might not component would commence: from the first interview
capture. onward, all children were seen weekly, whereas the
intervention started only 5 to 8 weeks later. We hoped to 1-4) with intervention (weeks 5-18). A 2 standard devia-
control in this way for parent’s enthusiasm for a novel re- tion band (2SDB) was determined for each child on the
habilitation program, which was expected to be reflected basis of standard deviation of the 4 baseline points. The
by an improvement between B1 and B2 in this study de- criterion of Gottman and Leiblum was used: the prob-
sign. The intervention effect was measured comparing ability that 2 consecutive points fall outside the 2SDB
postintervention results (R1, R2, and R3) with the second is less than .05 (see the work of Perdices and Tate41 for
baseline (B2), as this was considered a “purer” baseline details). Trends were detected by celeration lines, us-
eliminating the enthusiasm and novelty effect. Inconsis- ing the split middle trend line procedure.41 To obtain
tent answers to the BRIEF questionnaire were detected the magnitude of effect, the “nonoverlap of all pairs”
by computing the inconsistency score described in the (NAP) method was used,63 through SPSS software. For
BRIEF manual (a score of >9 being a threshold to con- the other outcome measures, an intervention effect size
sider the questionnaire unreliable because of contradic- (ES) was calculated for each child from B2 to R1, R2, and
tory answers on special items serving to assess consis- R3 as a standard difference between T-scores divided by
tency of answers). Furthermore, the intervention did not 10 for the BRIEF and as a standard difference between z
significantly change the amount of time spent in reha- scores for the DEX-C Cognition subscale. For the CCT,
bilitation: all children had already been attending the ES was obtained by dividing the score difference by the
outpatient department for half a day a week for many standard deviation of all 5 CCT scores of the child.
years (including sports, group games, and group discus- Furthermore, because the CCT inevitably has a practice
sion to promote socialization and language pragmatics effect as the recipe becomes more known, which could
for P.B., C.S., and R.K.; analytic psychotherapy for P.B.; account for improvement throughout the trials, we read-
and paper-and-pencil neuropsychological exercises aim- justed ES by subtracting the practice effect of each child
ing at improving attention for I.P.); therefore, the effect (score change between B1 and B2 being considered as
of the potentially confounding factor of time spent with the practice effect for that child). Effect sizes were inter-
therapist was considered likely to be negligible. preted subjectively with reference to Cohen’s64 guide-
lines: 0.2 = small; 0.5 = medium; 08 = large.
External investigator postintervention interview
After the intervention, an external interviewer, who RESULTS
had neither been involved in the rehabilitation nor been
Five children, aged 8 to 14 years, met the inclusion
involved with the research team called all the every-
criteria (P.B., C.S., R.K., I.P., and Y.R.). All had
day people involved in the program. A structured in-
sustained severe TBI at an early age, 3 to 11 years
terview focused on how they perceived the program,
before the study, and had a highly complex family
their views on applying cognitive coaching at home and
situation. All had specialized schooling, either attending
at school, clarity of the cognitive coaching guide, how
a special support class or having a school assistant.
children reacted to the intervention, and whether they
Characteristics of the participants are summarized
thought their child had improved in various domains
in Table 1. All children had a severe dysexecutive
(autonomy, school results, etc) even if it was a domain
syndrome on paper-and-pencil EF tests (see Table 1bis
not included in the GAS and questionnaires. The inter-
in Supplemental Digital Content, available at:
view contained embedded questions aimed at quantify-
http://links.lww.com/JHTR/A91), on the CCT (see B1
ing how much the everyday people participated in the
and B2 scores in Figure 1), and (apart from child I.P.’
cognitive coaching and at checking whether they under-
BRIEF score) on EF questionnaires (see B1 and B2 in
stood the concepts that were explained to them in the
Figures 2 and 3). In contrast, they were not impaired
cognitive coaching guide. They were asked for examples
in reasoning abilities (apart from C.S.) or retrospective
of metacognitive strategies they could recall, situations
memory (see Table 1ter in Supplemental Digital Con-
they applied them to, and were asked how often they
tent, available at: http://links.lww.com/JHTR/A91).
managed to go through the child’s workbook together
One child (Y.R.) dropped out of the study after 4 ses-
and whether GMT posters had been hung at home.
sions. Y.R. seemed to be unaware of his impairments and
Feedback from the child was obtained informally from
decided that he no longer wanted to be involved in any
the first author conducting the intervention, because
rehabilitation. His challenging behavior (see Table 1)
answering to an unknown external investigator on the
and school absconding were the main issues at the time
phone was considered age-inappropriate.
of the study. He was, however, included in this pilot
study initially because it was hoped that an intervention
Statistical analysis and effect size calculation
focusing on meaningful activities might be accepted by
The Saint’s day task (SDT) PM scores were visually an- Y.R. in contrast with all the other rehabilitation and
alyzed on time-series graphs comparing baseline (weeks school support he refused.
TABLE 1 Characteristics of the participants

P.B. C.S. R.K. I.P. Y.R.
Sex Girl Girl Boy Boy Boy
Age at inclusion, y 11 11 13 8 14
TBI mechanism MVA (passenger) Fall of metal bar MVA (pedestrian Fall of furniture Collision
on the child’s hit by car) on the child’s against
head head running child
Age at injury, y 2.5 6.5 7 5.5 2.5
Initial GCS score 6 4 3 6 <7
Brain imaging Large right Cerebellar and Subdural Brain stem Unknown
hemisphere right hematoma with hemorrhage,
hemorrhage parieto-occipital diffuse edema diffuse
and edema, lesion with and pneumo- subarachnoid
right parietal depression cephalus hemorrhage
depression fracture
fracture
Duration of coma, d Unknown 1 10 6 Unknown
Time since injury, y 9 5 6 3 11
Schooling Ordinary school Part-time Part-time private Ordinary school Special
+ part-time SA special—part- schooling with + part-time SA schooling
time ordinary SA, part-time Excluded from
schooling private lessons school half
of the year
for
behavioral
issues
Associated Lack of FSIQ 69 Attention ADHD Severe
impairments, awareness, Impaired ToM problems behavioral
reported in epilepsy and language Left arm disorder
medical records, absences pragmatics weakness Lack of
and previous neu- treated by Moderately awareness
ropsychological carbamazepine spastic equinus
assessments foot
Glasgow Outcome 2 3 3 2 3
Scale score
Family structure Monoparental Parents Large family Monoparental Two parent
separated (10 siblings) family household
geographically Father in prison
Sister followed Primary
up for a caregiver:
transplant cousin
Parental education, y Father: 14 Father: 11 Father: 22 Father: 7 Father: 4
Mother: 15 Mother: 11 Mother: 15 Mother: 11 Mother: 17
Abbreviations: ADHD, attention-deficit/hyperactivity disorder; FSIQ, Full Scale Intellectual Quotient; GCS, Glasgow Coma Scale; MVA,
motor vehicle accident; SA, school assistant; ToM, Theory of Mind.
Qualitative data about the program regularly and stating the goal while sorting cards with an
embedded PM task) but were reluctant to apply them
The intervention appeared to be feasible to imple- to more complex and ecological activities such as cook-
ment and it was reported that children enjoyed it, espe- ing, judging the strategies as an additional task per se,
cially the meaningful activities and stories used in the and demonstrating no consistent application of strate-
theoretical modules. Most GMT concepts were under- gies in the meaningful activities. Only the 13-year-old
stood by the children, although examples of personal R.K., probably the most aware and the most impaired in
cognitive failures were difficult to obtain. Interestingly, daily life, actually engaged with the techniques and used
children seemed to consider the metacognitive strate- them whenever he noticed task similarities. C.S. seemed
gies as exercises to practice rather than something that to understand only a few GMT concepts and metacogni-
could be applied to other tasks. As such, they would tive strategies. I.P. and P.B. seemed to lack awareness of
use the strategies on theoretical modules (eg, pausing impairments and reported not finding the intervention
Figure 1. Number of errors in the CCT: changes over time

(∗ age-matched controls, n = 8–13 per child). B1 and B2: Figure 3. DEX-C Cognition subscore change over time. B1
first and second baseline assessments; R1, R2, R3: assessments and B2: first and second baseline assessments; R1, R2, and R3:
performed at 0, 3, and 6 months postintervention. In the CCT, assessments performed at 0, 3, and 6 months postintervention.
errors include action errors as well as commentaries/questions. DEX-C indicates Dysexecutive Questionnaire for Children.
CCT indicates Children’s Cooking Task.
progress on the SDT only by week 8. She maintained
useful but found the program was fun and they partici- performance until school holidays (week 17) when her
pated willingly. performance dropped momentarily to zero (overall
medium effect; NAP = 0.74 [0.50-0.98]). R.K. had a very
Repeated PM measures variable performance but a strong effect of intervention
(NAP = 0.87 [0.713-1]). The use of the following
The SDT was performed by 3 children. I.P., aged 8
strategies to manage the task was reported by parents
years, did not complete the task, as he was not familiar
and/or children: stopping all activity up to 1 hour
with mobile phones, did not know how to use the
before the target time and watching the clock (R.K.),
Internet, and making a phone call to an unfamiliar
using cues such as the view of her computer (C.S.).
person was not age-appropriate. Y.R. dropped out of the
study. Weekly PM score changes over time are shown in
Figure 4. During baseline, none of the children reached Complex cooking task management: Ecological EF
a score of 50%. Using a 2SDB, all children showed test CCT and its parallel form (“Christmas biscuits”)
statistically significant progress, as all had at least 2 con- All children were very impaired on the CCT, scoring
secutive points outside their 2SDB. The best progress from 3 (C.S.) to 25 (I.P.) SDs below age-matched con-
seemed to be made by P.B. Unfortunately, when her trols. R.K. and P.B. required help from the examiner to
performance was reaching 100% on week 12, she lost the finish the task, and I.P. completed the task with nearly
charger of her mobile phone and her parents did not re- 200 errors and failed the task. C.S.’s errors were mainly
place it until the end of the study, giving only a medium commentaries/questions on every action she undertook,
effect (NAP = 0.47 [0.12-0.81]). C.S. began to make but she successfully finished the task.
Changes in the number of errors from baseline to
postintervention and follow-up for each child are pre-
sented in Figure 1. During baseline, all children showed
some practice effect between B1 and B2. After interven-
tion, P.B. increased the number of errors. Interestingly,
she was so focused on not repeating the errors from
her previous trial that she often skipped whole recipe
steps and thus forgot more ingredients/steps throughout
the trials. Furthermore, from trial to trial, P.B. seemed
more confident each time and stated how well she knew
the recipe and how easy it would be. For C.S., the de-
crease of errors from B2 to R1 had a small ES (Table 2).
Figure 2. Parental BRIEF questionnaire changes of GEC
Conversely, R.K. and I.P. significantly decreased the
Tscore over time. B1 and B2: first and second baseline as- number of errors after the intervention with large ES,
sessments; R1, R2, and R3: assessments performed at 0, 3, and R.K. showing a performance similar to controls after
6 months postintervention. The horizontal line represents the intervention, and I.P. improving from 25 to 6 SD com-
clinical cutoff score of 65. GEC indicates Global Executive pared with controls. R.K. clearly used the metacognitive
Composite. techniques taught in the intervention while performing
TABLE 2 Outcome measures ES

comparing B2 with R1, R2, and R3a
ES ES ES
at R1 at R2 at R3
Children’s Cooking Task score
P.B. − 1.15 − 2.90 − 2.90
C.S. 0.33 1.09 0.76
R.K. 1.42 1.35 1.59
I.P. 1.31 1.16 0.30
Parental BRIEF score
P.B. 0.5 0.2 0.1
C.S. 0.9 1.3 0.4
R.K. 0.1 0.1 − 0.1
I.P. 1.2 0.8 1
DEX-C Cognition subscore
P.B. 1.33 2.00 2.33
C.S. 0.67 1.00 1.33
R.K. 0.67 0.00 − 0.67
I.P.b 2.00 1.67 1.84
I.P. compared with B1b 0.67 0.33 0.50
Abbreviations: BRIEF, Behavior Rating Inventory of Executive

Function; DEX-C, Dysexecutive Questionnaire for Children; ES,
effect size; R1, R2, and R3, assessments performed at 0, 3, and
6 months postintervention.
a The Cohen rating of ES: 0.2 = small; 0.5 = medium; 0.8 = large.
b Because I.P.’s DEX-C Cognition subscore at B2 was deviant
(see Fig 3), we also report here ES comparing postintervention

outcomes to his best baseline score (B1).
Transfer to natural contexts

At baseline, all children but one (I.P.) scored in the
clinical range (T-scores >65) for Global Executive Com-
posite scores on the parental and teacher BRIEF ques-
tionnaires.
Transfer to home-context: All parental scores were con-
sistent (inconsistency score <9). Immediately af-
ter the intervention, 3 children (P.B., C.S., and
I.P.) showed a decrease on parental BRIEF scores
(Figure 2), reflecting possibly less executive dysfunc-
tion in daily life at home. Effect sizes are reported in
Figure 4. Saint Day Task: Prospective memory score changes Table 2. All but one child (I.P.) showed a decrease on
over time for P.B. (A), C.S. (B), and R.K. (C). The vertical DEX-C Cognition subscores between the 2 baselines
line corresponds to the beginning of the intervention. Arrows that was considered to be the enthusiasm effect we
correspond to split-middle celeration lines. Dashed lines cor- had expected because of intervention novelty. How-
respond to +2 and 2 SD band.
ever, the decrease was accentuated much further after
the intervention for all children and continued to de-
crease at 3- and 6-month follow-ups for P.B. and for
the CCT (checking he finished a step before moving to C.S. with large ES (2.33 for P.B. and 1.33 for C.S.;
the next, saying “Stop!” and thinking before adding a Table 2).
new ingredient . . . ). Effects were totally maintained at Transfer to school-context: Teacher BRIEF scores re-
3 and 6 months for R.K., whereas effect progressively mained stable for R.K. and C.S. and were unreliable
diminished at 3 and 6 months for I.P. However, when for P.B. and I.P. (inconsistency index >9), meaning
using a completely different and unknown recipe at R3 that it was not appropriate to draw any reliable con-
(Christmas biscuits), all children returned to their initial clusions on EF in the school context for these chil-
number of errors. dren, although P.B.’s BRIEF scores seemed to show
significant improvement. P.B. was reported to have after the intervention, and RK was for the first time al-
made excellent progress at school on school academic lowed to be in the kitchen alone by his parents. Parents
reports. reported that children’s self-esteem increased because
they could make a meal for the family.
Generalization: Goal Attainment Scaling and Participation of everyday people
postintervention interview
Overall, the level of participation of the everyday peo-
There was a high rate of missing GAS data. The ple was very low. Only R.K.’s parents asked for feedback
GAS goals could not be developed in collaboration after the intervention. In interviews, parents, teachers,
with R.K.’s school everyday people because he attended and school assistants all reported that the intervention
school very rarely in that period. GAS goals were devel- was fun for the child and that metacognitive techniques
oped with I.P.’s teacher and school assistant but postin- “were useful.” However, when asked to provide exam-
tervention GAS forms were handed to I.P. who lost them ples of metacognitive strategies contained in the “cog-
(as school closed for 2 months after R1, new forms could nitive coaching” guide they had received, 7 of 10 ev-
not be obtained). C.S.’s teacher, with whom the goals eryday people recalled less than half of the strategies.
were developed, changed after R1, explaining missing Moreover, examples of strategy use were not always
data for R1 and R2. Overall, GAS goals were obtained appropriate. Between-session assignments (ie, “mission
for at least 1 “everyday person” per child (see Table 3). sheets” including the simple task of helping the child to
We were only able to agree on personal goals with 1 child detect and write in a table when an “Oops error” had
(R.K.). The other 3 viewed the goals proposed by their occurred) were never or rarely done. Concrete interven-
everyday people as not problematic or not important. tion content was much better followed than the abstract
All children progressed toward their goals at R1 (see demand of “cognitive coaching”: one school assistant
Figure 5). However, only one of the GAS scores (I.P.’s) (P.B.’s) regularly used the paper notebook to compen-
reached the expected goal attainment level (T-score of sate for P.B.’s constant PM failures relating to school
50). During the external interview at the end of the study, goals (bring sports clothes, get a form signed), and one
P.B.’s mother reported significant daily life benefits of mother (C.S.’s) started to cook with her daughter. Sev-
the program: P.B. forgot her antiepileptic drugs much eral parents reported that using the term “Oops error”
less often (by linking her breakfast orange juice with re- helped lower family’s tension to the child’s cognitive
membering to take her medications), was less often late failures, and some began using the term with their other
at school, and made important progress at school, allow- children and themselves. All parents reported being gen-
ing her to continue schooling in an ordinary class with a erally too busy to apply the cognitive coaching at home.
school assistant rather than going to special education as Teachers reported the children did not use the strategies
had originally been planned. These last 2 improvements at school, but they had not prompted the children to
were not captured by the child’s GAS scores because do so. Both teachers and school assistants tended to em-
these were unanticipated positive outcomes. Some pos- phasize the behavioral, attentional, and “lack of effort”
itive outcomes were reported for I.P. by his main carer, problems at school as the key problem for the child and
which was consistent with a GAS score that reached 50. did not consider metacognitive strategies use as a pri-
Parents reported some general progress in well-being at ority for the child. A lack of knowledge about TBI was
R1 for R.K. and C.S. Most children carried on cooking identified with children’s difficulties not being seen as
cognitive (“he does not try to pay attention,” “he has no
friends”).
DISCUSSION
The “context-sensitive GMT” intervention comprised
(1) an adapted GMT, (2) metacognitive strategies prac-
tice through meaningful activities, and (3) a “cognitive
coaching guide” for the child’s everyday people. The
program was feasible to implement and apparently en-
joyable for children. However, participation of everyday
people was limited. Children significantly improved on
the SDT (time-based PM). Executive functioning perfor-
Figure 5. Goal Attainment Scaling T-score evolution over mance in the ecological CCT improved in 2 children.
time. R1, R2, and R3: assessments performed at 0, 3, and Three children showed a decrease on parental BRIEF
6 months postintervention. scores, reflecting possibly less executive dysfunction in
TABLE 3 Personal and general GAS goals

Examples of personal GAS goals corresponding to
EF-related problems reported by everyday Examples of general GAS goals corresponding to
people (only the goals and not the full GAS are metacognitive strategy use and GMT application
reported) (only the goals and not the full GAS are reported)
P.B.—mother: to be aware of one’s “Oops” errors (attentional
to forget taking antiepileptic drug less often slips)
to lose fewer objects to detect “Oops” errors as they occur
to be flexible enough to change strategy if the first to stop and think before beginning a new task
strategy does not work to formulate a task’s main goal before beginning a
to brainstorm for possible solutions before rushing task (eg, school exercise, home activity)
to start a task or school exercise to write down the things one might forget to do
P.B.—school assistant: to remember to look in the notebook to perform the
to hand-in homework on time/not so late intended action
to remember to give routine weekly documents to to follow a series of steps that are given to perform
her mother a task, finishing each step before moving to
to remember nonroutine one-off items (eg, bring another
money for excursion, ask parents to sign the to split complex tasks into steps and substeps
excursion form) to check a task/exercise before moving on to
C.S.—mother: another
to estimate whether a school exercise will be hard
or easy before beginning
to check school work for errors before handing it in
to check she has understood what she is supposed
to do before beginning a task
to ask questions if she is not sure she understood
what she is supposed to do
C.S.—teacher:
to accept the need to check her work when she is
prompted to do so
to estimate the difficulty of exercises/tasks
R.K.—both parents and the child:
to be able to tidy up his room (without the need for
someone to tell him in which order to do it)
to be less stressed about his prospective memory
problems
to be able to perform an instruction made of 3
consecutive tasks (eg, “drink your milk, empty the
dish washer, and get ready to go out”).
I.P.—teacher and school assistant:
to be able to prepare schoolbag alone
to write down information/instructions from the
teacher without being prompted by the school
assistant
to remember to check agenda to see what needs to
be done
Abbreviations: EF, executive functioning; GAS, Goal Attainment Scale; GMT, Goal Management Training.
daily life at home. All children decreased their DEX-C generalization. Intervention effects persisted at 3-month
Cognition subscore, suggesting that parents perceived follow-up and were partially maintained at 6-month
improvement in cognitive EF impairment. There were follow-up.
some indications of generalization to untrained tasks in One reason why participation was low for everyday
all children but not sufficient to achieve the expected people was that families had difficult situations to deal
level of achievement in EF-related GAS goals. When with (see Table 1), leaving little time for the cognitive
presented with a truly novel task (the parallel version coaching of their child. The chronic phase of TBI may
of CCT—“Christmas biscuits”), all children dropped to not be the optimum time for new cognitive coaching
their initial level of performance, indicating a lack of practices to be taught to parents and others, as many
habits have already settled. The cognitive coaching improvements in behavior had occurred, these were bal-
guide, although simply explained, was abstract and ev- anced out with greater awareness of difficulties on the
eryday people were not involved in direct training ses- part of the carer. These issues could perhaps explain
sions, as opposed to other programs.11,65,66 The cog- why the other outcome measures (CCT, GAS) were not
nitive coaching guide was rarely used by the teachers consistent with BRIEF scores (eg, the BRIEF scores of
and school assistants. It is not entirely clear why this C.S. and P.B. decreased but they did not improve on the
was, but one possibility is that the relatively limited CCT, while RK made best progress on CCT, there was
contact with the investigators (a phone call before and no corresponding decrease in BRIEF scores). However,
after the study and the rest through a written guide) was it is possible that metacognitive strategies are effective
not sufficient to engage them in the intervention con- in a time-limited task such as the CCT but impractical
trary to other school-delivered interventions.67 This is in the context of daily life’s constant attentional de-
clearly an important issue both for future studies and for mands, as it is an effortful, top-down process. This may
clinical interventions that depend heavily on a child’s explain why R.K. made good progress on the CCT but
everyday people for success. It is probably easier to en- did not apply the metacognitive strategies in daily life
gage school staff when the interventions are aimed at so that parents did not notice a real change in everyday
responding to their needs (especially managing behav- life postintervention. Alternatively, it may be the case
ior problems such as those reported in the studies of that parents and other everyday people may not have
Feeney and Ylvisaker67,68 ). In everyday clinical practice, had sufficient training to enable them to support the
frequent contact with the child’s everyday people (es- children to implement the strategies in everyday situa-
pecially school staff) is often not feasible and so exam- tions consistently. Furthermore, differences between ob-
ining whether written information (such as our cogni- jective measures and improvements reported by parents
tive coaching guide) can facilitate intervention support and patients have been frequently noted.23,69 Correla-
from these everyday people is an important research tions between parental BRIEF scores and EF classical
question. tests53,54 and with the CCT46 are typically small so the
Prospective memory performance might have im- BRIEF might not have captured the children’s progress.
proved because the task became familiar and routine. However, all children, including I.P. and R.K., signifi-
However, previous studies with adults using a similar cantly improved on the DEX-C Cognition subscores at
design did not show an improvement in performance R1, which is consistent with earlier finding that the num-
with time.43 Furthermore, children performed so poorly ber of errors in CCT and DEX-C Cognition subscales
and with so much variation from week to week (very are highly correlated46 in both adults70,71 and children46
rarely giving the Saint day within 1 hour of the target and might be a better measure of the children’s executive
time) that no possible routine could have been estab- progress than the BRIEF.
lished. We could not control for the performance of For P.B., it is difficult to explain the contrast between
the ongoing task (activity the child was doing at the consistency of improvement on DEX-C Cognition sub-
target time). It has been emphasized that PM perfor- score, GAS, qualitatively reported generalization, and
mance needs to take into account performance in the parental BRIEF and the increased number of errors on
ongoing task as well as the PM performance because PM the CCT. This should, however, be interpreted with cau-
paradigms can be considered as a dual-task paradigms.23 tion because all improvements were based on subjective
As such, it is possible that PM performance increased at informant reports, mostly of her mother. As P.B. seemed
the expense of ongoing activities. For one child (R.K.), to have very poor awareness according to the therapist,
his parents actually reported that he stopped all activity the intervention might have improved her awareness
up to 1 hour before the target time, watching the clock rather than EF, which would explain perceived improve-
to perform the task (but often actually forgot the task ment in her natural contexts but not on objective mea-
anyway). Pausing activity to avoid missing an important sures of EF (CCT). Nevertheless, she is the only child
phone call appointment may be considered an effective who seems to have benefited most from the interven-
strategy in real life, albeit not for an hour beforehand. tion, with lasting effects at 6 months. This was unex-
Although ratings on the parental BRIEF questionnaire pected, as lack of insight is known to impede patients
improved, better scores may not have been due to im- from actively engaging in rehabilitation72 and is a factor
provement in EF. Rather, it is possible that there was of poorer outcome. Indeed, P.B. never found the ses-
some bias in questionnaire responses. For example, par- sions “useful” but only “fun.” In the absence of aware-
ents were involved in the training and their responses ness, her motivation did not seem to be to overcome
may have reflected a desire to be perceived as good her difficulties (perceived as nonproblematic) but rather
cognitive coaches. Furthermore, the “home-school cog- to enjoy herself during the sessions and through that
nitive coaching” guide may have increased carers’ in- enjoyment some implicit learning may have occurred.
sight into the child’s difficulties, meaning that even if In children, enjoyment may be more important to an
intervention’s success than awareness and our interven- nitive training. Elaborating 2 sets of GAS goals has al-
tion seems to have fulfilled this requirement. As Bjork- ready been proposed by Schlosser74 in the concept of
lund noted, “Trying something new may be a goal into “control goals.” This could have supported further the
itself, and the fact that it does not improve performance finding of our study that children could be trained effec-
may be relatively unimportant to children.”73(p270) This tively in a meaningful task such as making a chocolate
may be why children were happy to try the metacog- cake (CCT) by combining metacognitive training and
nitive strategies on paper-and-pencil tasks but showed repeated cake baking but could not be trained to man-
no consistent application of strategies in the meaning- age a new untrained recipe (Christmas biscuits). More
ful activities. The same finding has been reported in the broadly, in rehabilitation research, using 2 sets of GAS
study by Missiuna et al65 of cognitive strategy training goals would make of GAS methodology both a power-
in children with TBI: making the intervention fun was ful motor for achieving meaningful goals by focusing
identified as being useful, whereas the “Goal-Plan-Do- intervention on them and a pertinent measure of gen-
Check” strategy (that is similar to GMT) and promotion eralization. Future research should also focus on other
of good strategy use were not. Considering together the goal-setting procedures: we did not manage to agree on
evidence of ours and the study of Missiuna et al (both EF goals with the children in our study, whereas in the
on very small samples), it seems that strategy use does study by Missiuna et al,65 children were able to self-
not come easily to children with more severe TBI and identify goals using a more framed and age-appropriate
therefore may not be the best rehabilitation approach goal-setting system than GAS, which could be used in
for them. In any case, strategies need to be simple, con- future studies. However, children in the study by Mis-
crete, and repeatedly practiced in order to benefit those siuna et al65 had sustained mild to moderate TBI and
children. were probably less impaired. Goal setting requires some
Usually, elaborating a GAS serves to focus rehabili- basic level of awareness, which children with severe TBI
tation on that goal. Such goal-focused rehabilitation is often lack. Lack of awareness was identified for all chil-
indeed an effective approach. However, this presents dren in our study except R.K. and seemed to be the
a methodological challenge for EF research: when a main reason why goals could not be identified by the
task is trained, its familiarity may make it less demand- children. Besides, in the study by Missiuna et al, chil-
ing on EF, as it is likely to require the application of dren were allowed to choose any goal (eg, learn a new
learned knowledge and task-specific procedures (which sport) whereas we purposefully retained only EF-related
may have become automatic, therefore, not “executive”) goals. More in-depth assessment of awareness would
rather than more general problem-solving and goal man- bring valuable contribution for research on goal-setting
agement processes.59,61 On the contrary, daily life is full procedures.
of EF-demanding tasks that require conscious, novel, The intervention was able to improve one particular
and effortful processing,3,59 without lapses into auto- PM task performed in an ecological setting (SDT), it
maticity. Apparent progress after a goal-focused training allowed some children to perform better in a cooking
may not necessarily reflect changes in underlying ex- task (CCT), and it resulted in some gains in daily EF.
ecutive processes needed to face daily life. Conversely, However, this metacognitive training did not allow
our aim was to improve children’s ability to cope with enough generalization effects to reach expected levels in
new, EF-demanding situations. As an outcome measure EF untrained personalized goals or to manage a novel,
needs to be novel (therefore untrained) to make signifi- complex EF-demanding task. The aim of providing chil-
cant demands on EF,59,60 personal GAS goals were not dren with metacognitive strategies applicable to “any”
trained to keep GAS as a generalization measure of EF. situation in life is an ideal goal but is perhaps not feasi-
However, in future studies, it would be more pertinent ble: children’s ability to cope with new EF-demanding
to divide child’s goals into a trained set of goals (and situations of daily life may not be possible to improve
corresponding GAS) and an untrained set of goals (and with training in case of severe impairments. In those se-
corresponding GAS) and then to focus the interven- vere cases, a repeated, goal-focused rehabilitation using
tion on training the former while using the latter as an activities that are meaningful to the children, and not fo-
ecological generalization measure. In such an approach, cusing on explicit generalization training, may be a more
main issues would be to match GAS sets for level of dif- reasonable therapeutic option (see the study of Missiuna
ficulty, child’s interest, and level of priority as seen by et al65 for an example of effective goal-focused inter-
everyday people who participate in goal selection. Fur- vention in a small sample of children with TBI). It
thermore, for GAS aiming at measuring generalization, should also be emphasized that GMT targets more
it would be important to control how much explicit link- specifically the PM aspects of EF-demanding tasks
ing to these goals is done during the intervention. The and much less problem-solving abilities. This study
intervention would probably be more effective if it com- supports a recent review19 in adults which concluded
bined goal-focused rehabilitation and general metacog- that GMT is probably more effective when combined
with other interventions targeting other aspects of EF GMT usually target patients with moderate and mild
such as problem solving and initiation (see the works of TBI.20,75 As this study demonstrates the feasibility of
Spikman et al62 and Miotto et al75 for examples of such the program for a particularly complex group of chil-
interventions in adults). dren, it would be helpful to replicate this study with
Limitations of previously published studies included children who have sustained a moderate TBI and chil-
insufficient assessment of generalization,76 of specific dren who may have more access to everyday people for
effects on EF,11,77 lack of objective cognitive perfor- providing cognitive coaching. On the contrary, in the
mance measures (using only questionnaires as the out- clinical setting, it is precisely children with severe TBI
come measure),12 use of problem-solving tasks that lack and complex family situations that are most needy of
ecological validity,78 lack of demonstration of EF diffi- intervention and future research should focus on this
culties prior to intervention,44 or lack of multiple base- group despite its challenges.
line or follow-up in pre-post designs.65,79 Others focused For clinical use, the intervention may need further
mainly on the behavioral aspects of the dysexecutive adaptation: the program may benefit from being longer
syndrome.13,67 This study is to the best of our knowl- to allow the children to integrate each strategy before
edge the first study that explores whether metacogni- practicing a new one, and everyday people should be
tive training generalizes and helps children to adapt and supported further to participate in the sessions, in a sim-
manage a novel EF-demanding task and to achieve un- ilar way to that used in the study by Braga et al.11 Cog-
trained goals. Of course, the small sample of this study nitive coaching should be presented through concrete
does not allow drawing general conclusions about the activities to be done at home and at school rather than
program efficacy. Our results must be interpreted with general concepts and advice. The intervention should
caution, especially because we included the most chal- be more closely embedded in the family life to improve
lenging population for this pilot intervention, which family participation without adding an additional family
may have limited its effectiveness, that is, children with burden.82,83 Not all families are willing and/or capable
severe EF impairment and with 3 known major factors of engaging in a family-delivered program,83 and evaluat-
of poorer outcome55,56,80,81 : (1) severe TBI; (2) sustained ing how to predict this prior to intervention would be of
at an early age; and (3) in nonoptimally functioning fam- benefit to service providers. Direct contact with school
ilies (similarly to GMT of Corbett et al39 that targeted staff is needed. The impact of parental metacognitive
children from low socioeconomic background in Cape knowledge, skills, and beliefs on outcomes in family-
Town with little success). Those children are usually ex- delivered interventions would also be a valuable com-
cluded from protocols65 and adult interventions using ponent of future studies.
REFERENCES
1. Anderson P. Assessment and development of executive function 8. Limond J, Leeke R. Practitioner review: cognitive rehabilitation
(EF) during childhood. Child Neuropsychol. 2002;8(2):71–82. for children with acquired brain injury. J Child Psychol Psychiatry.
2. Gioia GA, Isquith PK. Ecological assessment of executive function 2005;46(4):339–352.
in traumatic brain injury. Dev Neuropsychol. 2004;25(1/2):135–158. 9. Slomine B, Locascio G. Cognitive rehabilitation for children
3. Rabbitt P. Methodologies and models in the study of executive with acquired brain injury. Dev Disabil Res Rev. 2009;15(2):
function. In: Rabbitt P, ed. Methodology of Frontal and Executive 133–143.
Functions. East Sussex, England: Psychology Press Ltd. http:// 10. Ylvisaker M, ed. Traumatic Brain Injury Rehabilitation: Children
www.scribd.com/doc/50671857/Methodology-of-Frontal-and- and Adolescents. 2nd ed. Woburn, MA: Butterworth-Heinemann;
Executive-Functions. Accessed July 30, 2012. 1998.
4. Burgess PW, Simons JS. Theories of frontal lobe executive func- 11. Braga LW, da Paz Júnior AC, Ylvisaker M. Direct clinician-
tion: clinical applications. In: Halligan PW, Wade DT, eds. The delivered versus indirect family-supported rehabilitation of chil-
Effectiveness of Rehabilitation for Cognitive Deficits. Oxford, England: dren with traumatic brain injury: a randomized controlled trial.
Oxford University Press; 2005:211–230. Brain Inj. 2005;19(10):819–831.
5. Babikian T, Asarnow R. Neurocognitive outcomes and recovery 12. Wade SL, Walz NC, Carey J, et al. A randomized trial of teen
after pediatric TBI: meta-analytic review of the literature. Neuropsy- online problem solving for improving executive function deficits
chology. 2009;23(3):283–296. following pediatric traumatic brain injury. J Head Trauma Rehabil.
6. Anderson V. Assessing executive functions in children: biological, 2010;25(6):409–415.
psychological, and developmental considerations. Dev Neuroreha- 13. Wade SL, Walz NC, Carey J, et al. Effect on behavior problems of
bil. 2001;4(3):119–136. teen online problem-solving for adolescent traumatic brain injury.
7. Ross KA, Dorris L, McMillan T. A systematic review of psycho- Pediatrics. 2011;128(4):e947–e953.
logical interventions to alleviate cognitive and psychosocial prob- 14. Ylvisaker M, Adelson PD, Braga LW, et al. Rehabilitation and
lems in children with acquired brain injury. Dev Med Child Neurol. ongoing support after pediatric TBI: twenty years of progress.
2011;53(8):692–701. J Head Trauma Rehabil. 2005;20(1):95–109.
15. Cicerone KD, Langenbahn DM, Braden C, et al. Evidence- 34. McCauley SR, Pedroza C, Chapman SB, Cook LG, Vásquez AC,
based cognitive rehabilitation: updated review of the literature Levin HS. Monetary incentive effects on event-based prospective
from 2003 through 2008. Arch Phys Med Rehabil. 2011;92(4): memory three months after traumatic brain injury in children. J
519–530. Clin Exp Neuropsychol. 2011;33(6):639–646.
16. Butler RW, Copeland DR, Fairclough DL, et al. A multicen- 35. Wechsler D. WISC-IV: échelle d’intelligence de Wechsler pour en-
ter, randomized clinical trial of a cognitive remediation program fants: manuel d’interprétation. ECPA—Les Ed. du Centre de psychologie
for childhood survivors of a pediatric malignancy. J Consult Clin appliquée. 2005.
Psychol. 2008;76(3):367–378. 36. Cohen M. Echelle de Mémoire pour Enfants. Les éditions du centre de
17. Chan DYK, Fong KNK. The effects of problem-solving skills train- psychologie appliquée. 2001.
ing based on metacognitive principles for children with acquired 37. Crawford JR, Garthwaite PH. Investigation of the single case in
brain injury attending mainstream schools: a controlled clinical neuropsychology: confidence limits on the abnormality of test
trial. Disabil Rehabil. 2011;33(21/22):2023–2032. scores and test score differences. Neuropsychologia. 2002;40(8):
18. Levine B, Robertson IH, Clare L, et al. Rehabilitation of executive 1196–1208.
functioning: an experimental-clinical validation of goal manage- 38. Crawford JR, Howell DC. Comparing an individual’s test score
ment training. J Int Neuropsychol Soc. 2000;6(3):299–312. against norms derived from small samples. Clin Neuropsychol.
19. Krasny-Pacini A, Chevignard M, Evans J. Goal Management 1998;12(4):482–486.
Training for rehabilitation of executive functions: a systematic 39. Corbett C, Schrieff L, Thomas K. Rehabilitation of Executive Func-
review of effectiveness in patients with acquired brain injury. http: tioning Following Pediatric Traumatic Brain Injury: A Goal Manage-
//informahealthcare.com/eprint/ijeZAnewNg5hUQWaYuny/ ment Training Intervention [thesis]. South Africa: University of Cape
full. Published 2013. Accessed April 18, 2013. Town; 2009.
20. Levine B, Schweizer TA, O’Connor C, et al. Rehabilitation of ex- 40. Toglia J, Johnston MV, Goverover Y, Dain B. A multicontext
ecutive functioning in patients with frontal lobe brain damage with approach to promoting transfer of strategy use and self-regulation
goal management training. Front Hum Neurosci. 2011;5:9. http:// after brain injury: an exploratory study. Brain Inj. 2010;24(4):664–
www.ncbi.nlm.nih.gov/pmc/articles/PMC3043269. Accessed July 677.
27, 2012. 41. Perdices M, Tate RL. Single-subject designs as a tool for evidence-
21. Duncan J, Emslie H, Williams P, Johnson R, Freer C. Intelligence based clinical practice: are they unrecognised and undervalued?
and the frontal lobe: the organization of goal-directed behavior. Neuropsychol Rehabil. 2009;19(6):904–927.
Cogn Psychol. 1996;30(3):257–303. 42. Graham JE, Karmarkar AM, Ottenbacher KJ. Small sample re-
22. Duncan J, Johnson R, Swales M, Freer C. Frontal lobe deficits after search designs for evidence-based rehabilitation: issues and meth-
head injury: unity and diversity of function. Cogn Neuropsychol. ods. Arch Phys Med Rehabil. 2012;93(8):S111–S116.
1997;14(5):713–741. 43. Fish J, Evans JJ, Nimmo M, et al. Rehabilitation of executive
23. Fish J, Wilson BA, Manly T. The assessment and rehabilitation of dysfunction following brain injury: “content-free” cueing im-
prospective memory problems in people with neurological disor- proves everyday prospective memory performance. Neuropsycholo-
ders: a review. Neuropsychol Rehabil. 2010;20(2):161–179. gia. 2007;45(6):1318–1330.
24. Okuda J, Fujii T, Yamadori A, et al. Participation of the prefrontal 44. Rous R, Adams M, Fish J, Adlam A. Prospective memory inter-
cortices in prospective memory: evidence from a PET study in vention for adolescents with acquired brain injury: a preliminary
humans. Neurosci Lett. 1998;253(2):127–130. study. Paper presented at: 9th International Brain Injury Congress;
25. Burgess PW, Gonen-Yaacovi G, Volle E. Functional neuroimag- March 2012; Edinburgh.
ing studies of prospective memory: what have we learnt so far? 45. Chevignard MP, Soo C, Galvin J, Catroppa C, Eren S. Ecological
Neuropsychologia. 2011;49(8):2246–2257. assessment of cognitive functions in children with acquired brain
26. Mackinlay RJ, Kliegel M, Mäntylä T. Predictors of time-based injury: a systematic review. Brain Inj. 2012;26(9):1033–1057.
prospective memory in children. J Exp Child Psychol. 2009; 46. Chevignard MP, Catroppa C, Galvin J, Anderson V. Development
102(3):251–264. and evaluation of an ecological task to assess executive functioning
27. Mahy CEV, Moses LJ. Executive functioning and prospective post childhood TBI: the Children’s Cooking Task. Brain Impair.
memory in young children. Cogn Dev. 2011;26(3):269–281. 2010;11(02):125–143.
28. Ford RM, Driscoll T, Shum D, Macaulay CE. Executive and 47. Chevignard MP, Servant V, Mariller A, Abada G, Pradat-Diehl P,
theory-of-mind contributions to event-based prospective memory Laurent-Vannier A. Assessment of executive functioning in chil-
in children: exploring the self-projection hypothesis. J Exp Child dren after TBI with a naturalistic open-ended task: a pilot study.
Psychol. 2012;111(3):468–489. Dev Neurorehabil. 2009;12(2):76–91.
29. Kerns KA. The CyberCruiser: an investigation of development 48. Servant V, Chevignard M. Evaluation ecologique des fonctions cog-
of prospective memory in children. J Int Neuropsychol Soc. 2000; nitives chez les enfants et adolescents traumatisés crâniens à travers une
6(1):62–70. tâche de la vie quotidienne: réalisation d’une activité de cuisine [thesis].
30. Ward H, Shum D, Dick B, McKinlay L, Baker-Tweney S. Interview Grenoble, France: Université Joseph Fourrier; 2009.
study of the effects of paediatric traumatic brain injury on memory. 49. Gioia GA, Isquith PK, Guy SC, Kenworthy L. Behavior Rating
Brain Inj. 2004;18(5):471–495. Inventory of Executive FunctionR (BRIEFR ). 2000.
31. McCauley SR, McDaniel MA, Pedroza C, Chapman SB, Levin 50. Gioia GA, Isquith PK, Guy SC, Kenworthy L. In: Roy A, Fournet
HS. Incentive effects on event-based prospective memory perfor- N, Legall D, Roulin J-L, eds. Behavior Rating Inventory of Executive
mance in children and adolescents with traumatic brain injury. FunctionR (BRIEFR ). French Adaptation. Hogrefe; 2000.
Neuropsychology. 2009;23(2):201–209. 51. Emslie H. Behavioural Assessment of the Dysexecutive Syndrome for
32. Ward H, Shum D, McKinlay L, Baker S, Wallace G. Prospective Children: (BADS-C). Thames Valley Test Company; 2003.
memory and pediatric traumatic brain injury: effects of cognitive 52. Gioia GA, Isquith PK, Guy SC, Kenworthy L. TEST REVIEW Be-
demand. Child Neuropsychol. 2007;13(3):219–239. havior Rating Inventory of Executive Function. Child Neuropsychol.
33. McCauley SR, Levin HS. Prospective memory in pediatric 2000;6(3):235–238.
traumatic brain injury: a preliminary study. Dev Neuropsychol. 53. Anderson VA, Anderson P, Northam E, Jacobs R, Mikiewicz
2004;25(1/2):5–20. O. Relationships between cognitive and behavioral measures of
executive function in children with brain disease. Child Neuropsy- brain injury and behavioral difficulties. J Head Trauma Rehabil.
chol. 2002;8(4):231–240. 2010;25(6):416–425.
54. Vriezen ER, Pigott SE. Relationship between parental report on 69. Lewis MW, Babbage DR, Leathem JM. Assessing executive per-
the BRIEF and performance-based measures of executive function formance during cognitive rehabilitation. Neuropsychol Rehabil.
in children with moderate to severe traumatic brain injury. Child 2011;21(2):145–163.
Neuropsychol. 2002;8(4):296–303. 70. Chevignard M, Pillon B, Pradat-Diehl P, et al. An ecological ap-
55. Nadebaum C, Anderson V, Catroppa C. Executive function out- proach to planning dysfunction: script execution. Cortex. 2000;
comes following traumatic brain injury in young children: a five 36(5):649–669.
year follow-up. Dev Neuropsychol. 2007;32(2):703–728. 71. Chevignard MP, Taillefer C, Picq C, Poncet F, Noulhiane M,
56. Mangeot S, Armstrong K, Colvin AN, Yeates KO, Taylor HG. Pradat-Diehl P. Ecological assessment of the dysexecutive syn-
Long-term executive function deficits in children with traumatic drome using execution of a cooking task. Neuropsychol Rehabil.
brain injuries: assessment using the Behavior Rating Inventory of 2008;18(4):461–485.
Executive Function (BRIEF). Child Neuropsychol. 2002;8(4):271– 72. Ownsworth T, Clare L. The association between awareness deficits
284. and rehabilitation outcome following acquired brain injury. Clin
57. Kiresuk TJ, Sherman R. Goal Attainment Scaling: a general Psychol Rev. 2006;26(6):783–795.
method for evaluating comprehensive community mental health 73. Bjorklund DF. Learning to think on their own: executive func-
programs. Commun Mental Health J. 1968;4(6):443–453. tion, strategies and problem-solving. In: Children’s Thinking: Cog-
58. Krasny-Pacini A, Hiebel J, Pauly F, Godon S, Chevignard M. Goal nitive Development and Individual Differences. 5th International ed.
Attainment Scaling in rehabilitation: a literature-based update. Cengage Learning; 2011:270.
Ann Phys Rehabil Med. 2013;56(3):212–230. 74. Schlosser RW. Goal attainment scaling as a clinical measurement
59. Denckla MB. Measurement of executive function. In: Frames of technique in communication disorders: a critical review. J Commun
Reference for the Assessment of Learning Disabilities: New Views on Disord. 2004;37(3):217–239.
Measurement Issues. Baltimore, MD: Paul H Brookes; 1994:117– 75. Miotto EC, Evans JJ, Souza de Lucia MC. Rehabilitation of exec-
142. utive dysfunction: a controlled trial of an attention and problem
60. Shallice T, Burgess P. Higher order cognitive impairments and solving treatment group. Neuropsychol Rehabil. 2009;19(4):517–
frontal lobe lesions in man. In: Levin HS, Benton AL, eds. Frontal 540.
Lobe Function and Dysfunction. New York, NY: Oxford University 76. Selznick L, Savage RC. Using self-monitoring procedures to in-
Press; 1991:125–138. crease on-task behavior with three adolescent boys with brain in-
61. Philips L. Do “frontal tests” measure executive function? Issues jury. Behav Interv. 2000;15(3):243–260.
of assessment and evidence from fluency tests. In: Rabbitt P, ed. 77. Walker AJ, Onus M, Doyle M, Clare J, McCarthy K. Cogni-
Methodology of Frontal and Executive Functions. East Sussex, England: tive rehabilitation after severe traumatic brain injury: a pilot pro-
Psychology Press Ltd; 1997:200–203. gramme of goal planning and outdoor adventure course participa-
62. Spikman JM, Boelen DHE, Lamberts KF, Brouwer WH, Fasotti tion. Brain Inj. 2005;19(14):1237–1241.
L. Effects of a multifaceted treatment program for executive dys- 78. Suzman KB, Morris RD, Morris MK, Milan MA. Cognitive-
function after acquired brain injury on indications of executive behavioral remediation of problem solving deficits in children
functioning in daily life. J Int Neuropsychol Soc. 2010;16(1):118– with acquired brain injury. J Behav Ther Exp Psychiatry. 1997;
129. 28(3):203–212.
63. Parker RI, Vannest K. An improved effect size for single-case 79. Catroppa C, Anderson V, Muscara F. Rehabilitation of ex-
research: nonoverlap of all pairs. Behav Ther. 2009;40(4):357–367. ecutive skills post-childhood traumatic brain injury (TBI):
64. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. a pilot intervention study. Dev Neurorehabil. 2009;12(5):
Routledge; 1988. 361–369.
65. Missiuna C, DeMatteo C, Hanna S, Mandich A, Law M, 80. Donders J, Warschausky S. Neurobehavioral outcomes after early
Mahoney W, Scott L. Exploring the use of cognitive intervention versus late childhood traumatic brain injury. J Head Trauma Reha-
for children with acquired brain injury. Phys Occup Ther Pediatr. bil. 2007;22(5):296–302.
2010;30(3):205–219. 81. Ewing-Cobbs L, Prasad MR, Landry SH, Kramer L, DeLeon R.
66. Tamm L, Nakonezny PA, Hughes CW. An open trial of a metacog- Executive functions following traumatic brain injury in young
nitive executive function training for young children with ADHD. children: a preliminary analysis. Dev Neuropsychol. 2004;26(1):487–
J Attention Disord. 2012. http://www.ncbi.nlm.nih.gov/pubmed/ 512.
22647287. Accessed August 23, 2012. 82. Cole WR, Paulos SK, Cole CAS, Tankard C. A review of family
67. Feeney TJ, Ylvisaker M. Context-sensitive behavioral supports for intervention guidelines for pediatric acquired brain injuries. Dev
young children with TBI: short-term effects and long-term out- Disabil Res Rev. 2009;15(2):159–166.
come. J Head Trauma Rehabil. 2003;18(1):33–51. 83. Jansen LM, Ketelaar M, Vermeer A. Parental experience of partic-
68. Feeney TJ. Structured flexibility: the use of context-sensitive ipation in physical therapy for children with physical disabilities.
self-regulatory scripts to support young persons with acquired Dev Med Child Neurol. 2003;45(1):58–69.

J Head Trauma Rehabil 2014 29 E49

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

J Head Trauma Rehabil 2014 29 E49

Uploaded by

Copyright:

Available Formats

J Head Trauma Rehabil

Vol. 29, No. 5, pp. E49–E64

Context-Sensitive Goal Management

Author Affiliations: Institut Universitaire de Réadaptation

TABLE 1 Characteristics of the participants

Figure 1. Number of errors in the CCT: changes over time

TABLE 2 Outcome measures ES

Abbreviations: BRIEF, Behavior Rating Inventory of Executive

(see Fig 3), we also report here ES comparing postintervention

Transfer to natural contexts

TABLE 3 Personal and general GAS goals

You might also like