A summary on the effectiveness of the Amsterdam memory and attention training for children (Amat-c) in children with brain injury

ABSTRACT

Objective: To summarise the current research on the effectiveness of the Amsterdam Memory and Attention Training for Children (Amat-c).

Methods: A literature search was conducted to find articles published about the Amat-c, using PubMed, psychINFO, and PsychBITE databases. Relevant search terms included Amat-c, attention and memory, and childhood ABI.

Results: Our literature search identified 7 articles that described 5 separate studies including 61 children in total (mostly TBI). Only one study had a control group. All results indicated positive effects on memory and attention, although in three of the studies, these results were not statistically tested. Positive results were generally maintained six months follow up.

Conclusions: This review showed that the Amat-c is effective for treating attention and memory disturbances in children with ABI. However, evidence is limited and training material is outdated. We suggest that the Amat-c should be digitised and implemented in a school setting and further evaluated.

KEYWORDS:

• ABI
• Amat-c
• Amsterdam Memory and Attention Training for Children
• paediatric rehabilitation
• TBI

Introduction

Children with an acquired brain injury (ABI) often struggle with difficulties in areas such as attention, memory, learning, and social behaviour (1-5). Approximately 1 million children experience a brain injury every year in the US alone(6). Many of these injuries are classified as mild, and do not require hospitalisation. However, even mild brain injuries can have lasting negative effects that are not always immediately obvious, and are therefore often left untreated (7,8).

Despite the high prevalence of paediatric ABI and its numerous consequences, there are very few evidence-based interventions developed specifically to treat these children (9,10). However, one intervention that has been evaluated more than once and in different countries and settings is the Amsterdam Memory and Attention Training for children (Amat-c)(11). The Amat-c was initially developed for children with attention and memory deficits after receiving treatment for cancer(12). The training involves three phases, which focus on improving sustained attention, selective attention, and mental tracking. During the training, the child practices various tasks, games, and exercises every day with a coach. The exercises gradually increase in difficulty, and focus on training both visual and auditory modalities. The Amat-c was first shown to be effective for children with attention and memory deficits after they had received cancer treatment(12). Given these initial successful findings, researchers since then have studied if the Amat-c is also effective for children with ABI(13–19).

Since the Amat-c is one of the few interventions developed for children, it is currently widely used in clinical practice as we know from an inventory on paediatric cognitive rehabilitation in the Netherlands. However, it uses the same format that was initially developed two decades ago. The training now seems to be rather out of date, as many of the exercises use a pencil and paper. Therefore, it may be time to digitise and modernise the Amat-c. Yet, before spending resources to implement any updates, evidence must be collected and summarised to show that the training is indeed effective and worth updating.

Here, we provide a review of the current evidence for the Amat-c. Given that the Amat-c has been studied from multiple points of view and used in several countries, the purpose of this review was to consolidate the findings from multiple groups of researchers. The summary presented below is helpful in determining if the Amat-c is effective, and if it is worth updating. Furthermore, this summary is helpful in indicating the strengths and weaknesses of the treatment, and what changes would be the most beneficial.

Methods

A literature search was conducted to find articles published about the Amat-c, using PubMed, psychINFO, and PsychBITE databases. Relevant search terms included Amat-C, attention and memory, and childhood ABI. We also personally contacted several researchers that are known to study the Amat-c, to send us their own work as well as help us find other authors. Our literature search identified seven peer-reviewed articles that described five separate studies. Two of these articles reported data collected from the same group of participants, six months apart (15,16). Catroppa et al. (2015) included data from a previously published pilot study (13,17). Throughout the data extraction process, we consulted with the original authors to confirm that we correctly interpreted their research.

Results

Study characteristics

The studies were conducted in Sweden, Denmark, and Australia, and included a total of 61 children (see Table 1). Four of the studies were preliminary; only one study had a control group. All studies examined some aspect of the efficacy or feasibility of the Amat-c. Every study conducted assessments prior to and immediately after intervention and several also measured long-term effects after six months. The research utilised a wide variety of neuropsychological tests and questionnaires.

Table 1. Study characteristics.

Author/s Year Country	N (E/C)	Design	Aim	When Assessed	Measures	Positive Effects on memory/ Attention	General Conclusions
Hooft et al 2003 Sweden	3	Small open study, preliminary report	Test feasibility of Amat-c for Swedish children with ABI	Prior to intervention, and within 2 weeks of completion	Binary Choice, Bourdon-Vos, Coding, Digit Span, D-SBR (parent/teacher), RAVLT, RBMT, Seashore Rhythm, Stroop, TMT, V&ART	Yes (observed)	Preliminary results showed that Amat-c is feasible for children with ABI
Hooft et al. 2005 Sweden	18/20	Randomised control study	“”	Prior to intervention, and within 2 weeks of completion	Binary Choice, Coding, Digit Span, GDS, RAVLT, RBMT, ROCF, Stroop, TMT, V&ART	Yes (statistical)	Amat-c improved memory and attention in children with ABI
Hooft et al. 2007 Sweden	“”	“”	Test if positive effects of the Amat-c are maintained	6 months after completion	“”	Yes (statistical)	Results indicate positive long-term effects of the Amat-c
Hooft and Norberg 2010 Sweden	3	Pilot study	Test feasibility of condensed Amat-c, and coach parents to help alleviate parent stress	Prior to intervention and immediately after completion	BRIEF (parent/teacher), Coding, Digit Span, GDS, JTJA, RAVLT, ROCF, SP-PC, SMBQ, Stroop,	Yes (observed)	Condensed training was reported as somewhat stressful; the parent coaching reduced parent stress; performance on NP tests improved
Madsen Sjö 2010 Denmark	7 completed the training out of 8 that began	Pilot study	Test efficacy of Amat-c when fully integrated into school	3 weeks prior to intervention and 3 weeks after completion	Block Span, BRIEF (parent/teacher), Coding, Digit Span, FAST, RAVLT, ROCF, TEA-Ch, TMT	Yes (statistical)	The program was successfully integrated into the school setting
Catroppa et al. 2014 Australia	3	Single subject pre-post case study	Test if positive effects of Amat-c generalise to daily life by using ecologically valid measures	Prior to intervention, immediately after completion, and 6 months after completion	ABAS (parent), BASC (parent), BRIEF (parent), CBCL (parent), CMS, OMQ-PF (parent), TEA-Ch	Yes (observed)	Preliminary results suggest that the Amat-c improvements may generalise to daily life, and that improvements are sustained long term
Catroppa et al. 2015 Australia	10 (includes 3 from 2014 study)	Preliminary study	“”	Prior to intervention, immediately after completion, and 6 months after completion	ABAS (parent), BASC (parent), BRIEF (parent), CMS, OMQ-PF (parent), TEA-Ch	Yes (statistical)	Preliminary results were positive; emphasised interplay of variability in participants and the intervention

Abbreviations: C, control group; E, experimental group

Abbreviations of Measures: ABAS, Adaptive Behaviour Assessment System; BASC, Behaviour Assessment System for Children; BRIEF, Behaviour Rating Inventory of Executive Function; CBCL, Child Behaviour Checklist; CMS, Children’s Memory Scale; D-SBR, Deasey-Spinetta Behaviour Rating Scales; FAST, F-A-S Test; GDS, Gordon Diagnostic System; JTJA, Jag tycker jag är; OMQ-PF, Observer Memory Questionnaire-Parent Form; RAVLT, Rey Auditory-Verbal Learning Test; RBMT, Rivermead Behavioural Memory Test; ROCF, Rey-Osterrieth Complex Figure test; SMBQ, Shirom-Melamed Burnout Questionnaire; SP-PC, Self-Perception Profile for Children; TEA-Ch, Test of Everyday Attention for Children; TMT, Trail Making Test; V&ART, Visual and Auditory Reaction Time Tests

Participant characteristics

Patient characteristics are shown in Table 2. In total, the summarised research includes 36 males and 25 females between the ages of 8 and 16. Thirty-four of these children had traumatic brain injury (TBI), and 27 experienced deficits due to other causes (spontaneous intra-cerebral bleeding, encephalitis, anoxia, stroke, medulloblastoma, hypoxia, and birth delivery complications). The time since treatment, accident, or onset ranged from 0.8 to 12.7 years. Inclusion criteria varied slightly, but all the children had attention and/or memory deficits.

Table 2. Participant Characteristics.

Author/s	Genders (M/F)	Ages (years)	Cause of difficulties (TBI/other)	Time since onset/accident/treatment	Inclusion criteria
Hooft et al. (2003)	2/1	9–16	ABI (2/1)	No.1: 3 years No.2: 5 years 9 months No.3: 3 years 2 months	Children with attention/memory disturbances
Hooft et al. (2005/2007)	E: 12/6 C: 10/10	9–16	ABI (E: 12/6 C: 9/11)	E: 2.2 years C: 2.6 years	IQ>70, and 20% of attention and memory test-performance 1SD below average for age
Hooft and Norberg (2010)	2/1	8–15	Treated for medullo- blastoma	No.1: 3 years No.2: 5 years 10 months No.3: 1 year	Children with attention/memory disturbances
Madsen Sjö (2010)	3/4	11–15	ABI (3/4)	0.8–8 years	Attention and⁄or memory sequelae after ABI (time since injury at least 1 year) and age between 8–16 years.
Catroppa et al. (2014)	2/1	8–13	ABI (1/2)	No.1: 10 years No.2: 12 years No.3: 2 years	IQ score > 70; at least one score of 1.5 SD below mean on Stories or Family Pictures of the CMS and/or Sky Search or Score DT of the TEA-Ch and/or a score of at least 1 SD below mean on BASC 2 Composite and/or a score < 94 on the OMQ-PF
Catroppa et al. (2015)	7/3	8–13	ABI (8/2)	1.6–12.7 years	ABI requiring hospital admission, 8–16 years at time of intervention, minimum of 1 year post-ABI, IQ>70, at least one score of 1.5 SD below the mean on Stories or Family Pictures of the CMS and/or Sky Search or Score DT of TEA-Ch and/or a score of at least 1 SD below the mean on the BASC-2 and/or a score <94 on the OMQ-PF

Abbreviations: ABI, Acquired Brain Injury; C, Control Group; E, Experimental Group; F, Female; M, Male; TBI, Traumatic Brain Injury

Intervention characteristics

The treatment was given in Swedish, Danish, and English (see Table 3). Treatment was conducted at school or at home, and the regular coach was either a parent or a teacher. The children also had weekly meetings with either a therapist or supervisor, at a clinic, hospital or school. The duration of the intervention ranged from 10 weeks to 20 weeks, and intensity ranged from 30 minutes per day to an hour per day.

Table 3. Intervention characteristics.

Author/s	Language	Setting	Coach	Duration	Intensity
Hooft et al. (2003)	Swedish	In school or at home, hospital once a week for feedback	Parent or teacher	20 weeks	30 min per day
Hooft et al. (2005/2007)	Swedish	At school or home, hospital once a week for feedback	Parent or teacher	17 weeks	30 min per day
Hooft and Norberg (2010)	Swedish (name translated to SMART)	At school or home, seen by supervisor once a week for feedback	Parent or teacher	10 weeks	Up to an hour a day Monday-Friday
Madsen Sjö et al. (2010)	Danish	School	Teacher every day, with a supervisor once per week	4 trained for 18 weeks, 3 trained for 20 weeks	30–45 min per day
Catroppa et al. (2014)	English	Home, met with therapist once per week	Parent	18 weeks	30 min per day
Catroppa et al. (2015)	English	Home, met with therapist once per week	Parent	18 weeks	30 min per day

Outcomes

All results indicated positive effects on memory and attention, although in three of the studies these results could only be observed and not statistically analysed due to small sample size. In Table 4, the outcomes on the different neuropsychological domains are summarised.

Table 4. Evidence for separate domains.

Domain	Hooft 2003	Hooft 2005	Hooft 2007 (6 m)	Hooft 2010	Madsen Sjo 2010	Catroppa 2014	Catroppa 2014 (6 m)	Catroppa 2015	Catroppa 2015 (6 m)
Reaction time	NS (V&ART)	NS (V&ART)	NS (V&ART)
Tempo					+ (ROCF)
Sustained Attention	+ (Bourdon- Vos)	+ (GDS)	+ (GDS)	+ (GDS)	NS (TEA-Ch)	+ (TEA-Ch)	+ (TEA-Ch)	Mixed (TEA-Ch)	Mixed (TEA-Ch)
Selective Attention	+ (Binary Choice, Coding, Stroop,TMT)	+ (Binary Choice, Coding, Stroop, TMT)	+ (Binary Choice, Coding, Stroop)	+ (Coding, Stroop)	NS (TEA-Ch)	+ (TEA-Ch)	+ (TEA-Ch)	Mixed (TEA-Ch)	Mixed (TEA-Ch)
Attentional Control/ Switching					+ (TEA-Ch)
Mental Tracking
Visual Spatial Memory		+ (ROCF)	+ (ROCF)	+ (ROCF)	+ (ROCF)	+ (CMS)	+ (CMS)	+ (CMS)	+ (CMS)
Verbal Memory	+ (Digit Span, RAVLT)	+ (Digit Span, RAVLT)	+ (RAVLT)	+ (Digit Span, RAVLT)	+ (RAVLT)	+ (CMS)	+ (CMS)
Everyday Memory	+ (RBMT)	+ (RBMT)	+ (RBMT)			+ (OMQ)
Executive Functions				+ /- (BRIEF)	+/- (BRIEF) and NS (FAST)	NS (BRIEF)	+ (BRIEF)	Mixed (BRIEF)	Mixed (BRIEF)
behaviour	+ (D-SBR)					+ (BASC)	+ (BASC)	Mixed (BASC)	Mixed (BASC)
Adaptive Behaviour						+ (ABAS)	+ (ABAS)	Mixed (BASC)	Mixed (BASC)
Adaptive and Maladaptive Functions						+ (CBCL)	+ (CBCL)
Self Image/Social Relations				+ (JTJA, SP-PC)				+ (ABAS)	+ (ABAS)
Parent Stress				+ (SMBQ)
Visuo-Constructive Abilities					+ (ROCF)

Abbreviations: +, indicates improvement; –, indicates worsening; 1 yr, results 1 year after completion of the treatment; 6 m, results 6 months after completion of treatment; NS, not significant; Mixed (in Catroppa 2015 columns), indicates a mix of not significant and significant improvements on an individual level

Abbreviations of Measures: ABAS, Adaptive Behaviour Assessment System; BASC, Behaviour Assessment System for Children; BRIEF, Behaviour Rating Inventory of Executive Function; CBCL, Child Behaviour Checklist; CMS, Children’s Memory Scale; D-SBR, Deasey-Spinetta Behaviour Rating Scales; FAST, F-A-S Test; GDS, Gordon Diagnostic System; JTJA, Jag tycker jag är; OMQ-PF, Observer Memory Questionnaire-Parent Form; RBMT, Rivermead Behavioural Memory Test; ROCF, Rey-Osterrieth Complex Figure test; SMBQ, Shirom-Melamed Burnout Questionnaire; SP-PC, Self-Perception Profile for Children; TEA-Ch, Test of Everyday Attention for Children; TMT, Trail Making Test; V&ART, Visual and Auditory Reaction Time Tests.

Short term

The majority of the papers found positive results immediately after intervention in the domains of sustained attention and selective attention. Visual spatial memory, everyday memory, and verbal memory also showed consistent improvement immediately after intervention. The immediate outcomes per test are presented in Appendix 1.

Long term

Positive results were generally maintained six months after intervention, in the domains of sustained and selective attention, as well as spatial memory, everyday memory, and verbal memory. See Appendix 2 for long-term outcome.

Discussion

Positive effects

In general, the collected evidence indicates that the Amat-c is effective and feasible for treating attention and memory disturbances in children with ABI. The evidence is however limited because only one of the seven studies incorporated a control group and the total number of children studied was 61. Across all studies, the overall feedback about the training from parents, children, and teachers was positive. Several comments mentioned noticeable improvements in self-confidence(17), ease in studying for exams(18), and behaviour in general. The long-term outcomes were also positive; studies found sustained improvements at six months post intervention, and continued improvement after six months for selective attention and verbal working memory(16).

Intervention characteristics

This review was primarily conducted to investigate how the Amat-c should be updated. The following section discusses the weaknesses of the current intervention, in order to illuminate the changes that could be made. Across all studies, the characteristics that appeared to have the largest effect on children’s motivation were the length and the location of the training.

The length of the intervention varied from 10 weeks–20 weeks. Initial feedback suggested that 20 weeks was far too long, and the children had low motivation to complete the training(18). A subsequent study found that a length of 17 weeks was more reasonable (15,16). Another study showed that a condensed version of 10 weeks was feasible, although the supervisors found it stressful. Therefore, they recommended giving the supervisors more time to work with the children(14). Altogether, there is the most evidence for the training to last about 18 weeks, although 10 weeks is technically feasible.

The location of the training was either at home or at school. In general, it appears that school seems to be the better location for integrating the training into daily life. Parents that coached their children at home commented that it was difficult to incorporate the training into a regular routine (17,18). However, when the training took place in school, it didn’t disrupt the child’s normal routine(19). One study also noted that having the meeting with the supervisor at school rather than at a hospital seemed to be beneficial(19). They suggest that weekly meetings in the hospital may have reminded kids of their injuries, which may explain the low motivation in previous studies.

Limitations

There are several limitations of the current summary. Firstly, only one study was a randomised controlled trial (15,16). Thus, the majority of these papers did not include a control group. Secondly, three of the studies could not statistically analyse the results, and only indicated observed positive or negative changes (14,17,18). In general, the sample sizes for these studies were very small, and several were preliminary case studies that only included a few subjects. Additionally, the majority of the chosen measures do not generalise to daily life. One study used neuropsychological tests with high ecological validity, but with a small sample size with no control group (13,17).

Clinical implications and future directions

Given that the Amat-c was developed two decades ago, and uses mainly paper and pencil tasks, it may be tempting to label this training as outdated. However, it is still valuable, because it is the only treatment that has been tested for children with ABI. We conducted this review to investigate if updating the Amat-c is a good idea, and we have found that although this training is effective, the children and parents sometimes perceive it as lengthy and boring.

Ideally, rehabilitative treatments should be easy to incorporate into daily life. Currently, the Amat-c does not always appear to fit nicely into a regular routine. We therefore suggest that the Amat-c needs to be easier to integrate. One possibility is to digitise the training. Children are now accustomed to working with technology in the classroom, at home, and in other settings. Therefore, using a computer to do the training may help increase the appeal. Once the Amat-c is digitised, it needs to be further studied with randomised controlled trials, and compared to other computer treatments such as Cogmed. This computerised working memory training as recently studied in children with TBI(20).

We also suggest implementing the training in a school setting. Since the children are already in school every day, this might not interrupt their schedule as much. It would also reduce stress on the parents, since they wouldn’t have to worry about finding time to do the training as part of the (evening) routine at home. After the Amat-c has been digitised and consistently implemented in a school setting, the optimal length will need to be defined. However, there may no longer be complaints about the length causing a decrease in motivation. That is, if the rehabilitation is more easily incorporated into the child’s life, then the child is more likely to be motivated to complete the training.

In conclusion, the limited evidence presented here indicates that the Amat-c is effective for treating attention and memory deficits in children with ABI. However, the feedback from children and parents suggests that the training is sometimes difficult to incorporate smoothly into daily life. Therefore, we suggest that digitising the Amat-c and implementing it in a school setting would be beneficial for integrating rehabilitation into the children’s normal routine. Proper evaluation of this new Amat-c using a high quality design is recommended.

Declaration of interest

The authors report no declarations of interest.

Acknowledgements

The authors would like to thank C. Catroppa, C. Hendriks, N. Sjö, and I. van’t Hooft.

Appendix 1. Immediate Outcome Measures.

Test	Hooft 2003 (Cases, Raw scores not given)	Hooft et al 2005	Hooft and Norberg 2010 (Cases, ± indicate better/worse, not significance)	Madsen Sjö 2009	Catroppa et al. 2014 (Cases, improvement >2/3 SD is considered significant)	Catroppa et al. 2015
ABAS					No.1: NS No.2: NS No.3: practical + (pre 70; post 87) social + (pre 68; post 87) composite + (pre 68; post 85)	NS (mixed on an individual level)
BASC					No.1: NS No.2: Internalising Composite + (pre 68; post 56) No.3: NS	NS (mixed on an individual level)
Binary Choice	+	RT NS, Number Correct + (E: pre 53; post 57; effect 3.5, C: pre 55; post 56; effect 0, p = 0.002)
Block Span				NS
Bourdon-Vos	+
BRIEF			No.1: Teachers rated more improvement than parents No.2: Teachers rated more improvement than parents No.3: parent ratings worse than teacher ratings	Parents Initiate -(pre 47.4; post 54.1, p = 0.023) Parents Plan/organise –(pre 54.9; post 60.1, p = 0.039) Trainers Emotional Control + (pre 76.9; post 65.3, p = 0.023)	NS	NS (mixed on an individual level)
CBCL					No.1: NS No.2: Internalising Problems + (pre 68; post 61) No.3: NS
CMS					No.1: Story Recall Immediate + (pre 12; post 19) Story Recall Delayed + (pre 12; post 16) Story Recall Recognition + (pre 6; post 16) Family Pictures Immediate + (pre 6; post 11) Family Pictures delayed + (pre 7; post 8) No.2: Story Recall Immediate + (pre 6; post 9) Story Recall Delayed + (pre 7; post 9) Family Pictures Delayed + (pre 13; post 16) No.3: Story Recall Immediate + (pre 13; post 17) Story Recall Delayed + (pre 14; post 19) Story Recall Recognition + (pre 9; post 14) Family Pictures Immediate + (pre 14; post 17) Family Pictures Delayed + (pre 15; post 17)	Family Pictures Immediate + (pre 10.8; post 13.5) Family Pictures Delayed + (pre 11; post 14.2)
Coding	+	+ (E: pre 38; post 41; ES 6, C: pre 43; post 39; ES −0.5, p = 0.002)	No.1: + (pre 30; post 35) No.2: + (pre 34; post 39) No.3: + (pre 41; post 44)	NS
D-SBR	+
Digit Span	+	+ (E: pre 11; post 14; ES 3, C: pre 10; post 11; ES 0, p = 0.0004)	No.1: + (pre 10; post 13) No.2: 0 (pre 7; post 7) No.3: + (pre 12; post 16)	NS
FAST				NS
GDS		Correct Answers + (E: pre 39; post 43; ES 3, C: pre 42; post 43, ES 1, p = 0.01) Commission Errors NS	No.1: Correct Answers + (pre 35; post 44) Commission Errors + (pre 14; post 8) No.2: Correct Answers – (pre 27; post 26) Commission Errors + (pre 10; post 2) No.3: Correct answers + (pre 44; post 45) Commission Errors + (pre 1; post 0)
JTJA			No.1: general improvement No.2: slight improvement No.3: slight improvement
OMQ-PF					NS
RAVLT	+	Immediate Recall NS, Delayed Recall + (E: pre 9; post 11; effect 3, C: pre 9; post 10; effect 0.5, p = 0.02)	No.1: + (pre 47; post 58) Recall + (pre 9; post 14) No.2: + (pre 36; post 39) Recall 0 (pre 7; post 7) No.3: + (pre 41; post 52) Recall + (pre 11; post 13)	Trial no.5 + (pre 3.3; post 4.3, p = 0.033) Total Words + (pre 2.9; post 4.4, p = 0.02)
RBMT	+	+ (E: pre 52; post 64; ES 11, C: pre 60; post 58; ES 0.5, p = 0.00004)
ROCF		Delayed Recall + (E: pre 11; post 18; ES 7, C: pre 19; post 16; ES 1, p = 0.003)	No.1: Copy + (pre 14; post 27.5) Recall + (pre 3; post 7.5) No.2: Copy – (pre 8; post 7.5) Recall + (pre 5.5; post 7.5) No.3: Copy + (pre 35; post 36) Recall + (pre 26; post 31)	Copy: Time + (pre 209; post 170, p = 0.032) Copy + (pre 3.0; post 5.6, p = 0.013) 3min Recall + (pre 3.4; post 5.1, p = 0.029) Delayed Recall + (pre 3.9; post 5.3, p = 0.021) Organisation + (pre 3.1; post 4.4, p = 0.017)
Seashore Rhythm	NS
SMBQ			No.1: mother from high to low stress No.2: mother from high to low stress No.3: mother from high to low stress
SP-PC			No.1: general improvement No.2: slight improvement No.3: slight improvement
Stroop	+	Stroop1 + (E: pre 53; post 68; ES 11, C: pre 71; post 73; ES 2, p = 0.02) Stroop2 NS, Stroop 3 NS	No.1: Stroop1 + (pre 79; post 69) Stroop2 + (pre 150; post 121) Stroop3 + (pre 257; post 216) No.2: Stroop1 - (pre 67;post 73) Stroop2 - (pre 92; post 95) Stroop3 + (pre 195; post 176) No.3: Stroop1 + (pre 65; post 61) Stroop2 + (pre 120; post 109) Stroop3 + (pre 227; post 133)
TEA-Ch				Creature Counting Accuracy + (pre 10.1; post 12.7, p = 0.022)	No.1: Sky Search Time Taken + (pre 8; post 11) Sky Search Attention Score + (pre 8; post 13) No.2: Sky Search Time Taken + (pre 7; post 9) Score DT + (pre 8; post 11) No.3: Sky Search Number of Targets + (pre 11; post 14) Sky Search Time Taken + (pre 5; post 10) Sky Search Attention Score + (pre 5;post 10) Score DT + (pre 9; post 15)	NS (mixed on an individual level)
TMT	NS	Part A + (E: pre 40; post 29; ES −8, C: pre 31; post 32; ES 1, p = 0.002) Part B + (E: pre 11; post 72; ES −32, C: pre 89; post 88; ES −3, p = 0.006)
V&ART	Not Stated	NS

Abbreviations: E, experimental group; C, control group; NS, not significant

Abbreviations of Measures: ABAS, Adaptive Behaviour Assessment System; BASC, Behaviour Assessment System for Children; BRIEF, Behaviour Rating Inventory of Executive Function; CBCL, Child Behaviour Checklist; CMS, Children’s Memory Scale; D-SBR, Deasey-Spinetta Behaviour Rating Scales; FAST, F-A-S Test; GDS, Gordon Diagnostic System; JTJA, Jag tycker jag är; OMQ-PF, Observer Memory Questionnaire-Parent Form; RAVLT, Rey Auditory-Verbal Learning Test; RBMT, Rivermead Behavioural Memory Test; ROCF, Rey-Osterrieth Complex Figure test; SMBQ, Shirom-Melamed Burnout Questionnaire; SP-PC, Self-Perception Profile for Children; TEA-Ch, Test of Everyday Attention for Children; TMT, Trail Making Test; V&ART, Visual and Auditory Reaction Time Tests

Appendix 2. Long-Term Outcome Measures.

Test	Hooft et al. 2007 (effects after 6 months)	Catroppa et al. 2014 (effects after 6 months) (Cases, improvement >2/3 SD is considered significant)	Catroppa et al. 2015 (effects after 6 months)
15 Word	Immediate Recall + (E: pre 36; post 48, C: pre 43; post 48, p = 0.005) Delayed Recall + (E: pre 9; post 13, C: pre 9; post 10, p = 0.005)
ABAS		No.1: NS No.2: NS No.3: social + (pre 68; post 78)	NS (mixed on an individual level)
BASC		No.1: Internalising Composite + (pre 76; post 57) Behavioural Symptoms Index + (pre 71; post 61) Hyperactivity + (pre 64; post 53) No.2: NS No.3: NS	NS (mixed on an individual level)
Binary Choice	RT NS, Correct Answers + (E: pre 53; post 57, C: pre 55; post 57, p = 0.002)
Block Span
Bourdon-Vos
BRIEF		No.1: Behavioural Regulation Index + (pre 79; post 65) General Executive Composite + (pre 77; post 69) No.2: NS No.3: NS	NS (mixed on an individual level)
CBCL		No.1: Externalising Problems + (pre 62; post 51) Total Problems + (pre 73; post 64) No.2: Internalising Problems + (pre 68; post 61) No.3: NS
CMS		No.1: Story Recall Immediate + (pre 12; post 18) Story recall delayed + (pre 12; post 16) Story Recall Recognition + (pre 6; post 10) Family Pictures Immediate + (pre 6; post 13) Family Pictures Delayed + (pre 7; post 13) No.2: Story Recall Immediate + (pre 6; post 9) Story Recall Delayed + (pre 7; post 11) Family Pictures Immediate + (pre 10; post 16) Family Pictures Delayed + (pre 13; post 15) No.3: Family Pictures immediate + (pre 14; post 18) Family Pictures Delayed + (pre 15; post 18)	Family Pictures Immediate + (pre 10.8; post 14.5, p = 0.001, ES = 0.53) Family Pictures Delayed + (pre 11; post 13.7, p = 0.008, ES = 0.41)
Coding	+ (E: pre 38; post 44, C: pre 43; post 44, p = 0.009)
D-SBR	NS
Digit Span
FAST
GDS	Correct Answers + (E: pre 39; post 45, C: pre 42; post 43, p = 0.0003) Commission errors + (E: pre 5; post 1, C: pre 3; post 2, p = 0.04)
JTJA
OMQ-PF		No.1: Raw Score + (pre 58; post 69) No.2: NS No.3: NS
RAVLT
RBMT	+ (E: pre 52; post 65, C: pre 60; post 53, p = 0.00002)
ROCF	Delayed Recall + (E: pre 11; post 22, C: pre 19; post 17, p = 0.00002)
Seashore Rhythm
SMBQ
SP-PC
Stroop	Stroop1 NS, Stroop 2 NS, Stroop3 + (E: pre 24; post 32, C: pre 27; post 31, p = 0.004)
TEA-Ch		No.1: NS No.2: Sky Search Time Taken + (pre 7; post 11) Score DT + (pre 8; post 13) No.3: Sky Search Time Taken + (pre 5; post 9) Sky Search Attention Score + (pre 5; post 10) Score DT + (pre 9; post 9)	NS (mixed on an individual level)
TMT	Part A NS, Part B + (E: pre 110; post 70, C: pre 89; post 90, p = 0.03)
V&ART	NS

Abbreviations: E, experimental group; C, control group; NS, not significant

Abbreviations of Measures: ABAS, Adaptive Behaviour Assessment System; BASC, Behaviour Assessment System for Children; BRIEF, Behaviour Rating Inventory of Executive Function; CBCL, Child Behaviour Checklist; CMS, Children’s Memory Scale; D-SBR, Deasey-Spinetta Behaviour Rating Scales; FAST, F-A-S Test; GDS, Gordon Diagnostic System; JTJA, Jag tycker jag är; OMQ-PF, Observer Memory Questionnaire-Parent Form; RAVLT, Rey Auditory-Verbal Learning Test; RBMT, Rivermead Behavioural Memory Test; ROCF, Rey-Osterrieth Complex Figure test; SMBQ, Shirom-Melamed Burnout Questionnaire; SP-PC, Self-Perception Profile for Children; TEA-Ch, Test of Everyday Attention for Children; TMT, Trail Making Test; V&ART, Visual and Auditory Re