Remediation of Learning and Memory Deficits

Memory impairment is one of the most common symptoms following brain injury and it is estimated that time and cost of care would be reduced if effective treatments were found to improve memory (Walker et al.,1991). When evaluating intervention strategies to improve memory performance following brain injury, the literature indicates that there are two main approaches to rehabilitation: restoration or retraining of the function and compensation. Compensation includes “training strategies or techniques that aim to circumvent any difficulty that arises as a result of the memory impairment.” Compensatory techniques include internal aids, which are “mnemonic strategies that restructure information that is to be learned.” Various interventions have focused on the remediation of memory deficits in individuals with TBI, including external compensatory aids (computers, pagers, and notebooks), individualized remediation programs, family/social support and environmental adaptations, didactic lessons and homework, training in compensatory strategies including rehearsal, organizational strategies, visual imagery, verbal labeling, and use of mnemonics, as well as implicit memory tasks.

Cicerone et al. (2000) reviewed 42 studies examining the effectiveness of various interventions to improve memory impairment following stroke and TBI. In 2005 and again in 2011, Cicerone and colleagues updated their original review. It should be noted that studies were not included in our review if the population did not comprise of more than 50% brain-injured patients, or if the sample size (n) was less than 3. As well only those studies dealing with moderate-to-severe brain-injured individuals were included in this review. 

Cappa and colleagues (2005) reviewed various strategies used to improve memory deficits without the use of electronic, external aids were judged to be “possibly effective.” Specific learning strategies (e.g. errorless learning) were found to be “probably effective” depending upon the task used, the type of memory involved and the severity of impairment. 

Several studies were identified examining interventions to improve learning and memory following acquired brain injury. Studies were categorized into the following groupings: assistive technology (external aids, computer assisted training and virtual reality and cognitive functioning), internal strategies used during learning to enhance recall, memory interventions and cranial electrotherapy stimulation and memory.

External Aids

External aids assist memory by use of external methods of recording and accessing information.  In an updated review by Cappa et al. (2005), the use of external, electronic assistive devices were assessed as “probably effective.”

Table: Remediation of Memory Deficits

Individual Studies

Table: The Use of External Aids to Enhance Memory

Table: Summary of the Use of External Aides to Enhance Memory


During the past year several studies were conducted looking at the effectiveness of various active reminders used to those with memory impairment.

In a study recently published RCT, Powell et al. (2012) compared trial and error learning to systematic instruction. Twenty-nine individuals were randomized to either the systematic instruction group (n=15) or the trial and error group (n=14). Twelve 45 minute training sessions were given to each individual. Each session targeted selected skills on a PDA. The systematic instruction condition emphasized mastery. Following treatment five areas were evaluated: post-test accuracy; maintenance accuracy; fluency; generalization; social validity.  At the initial follow-up there were no significant differences between the groups, however, at the 30 day follow-up significantly differences began to immerge.  Those in the systematic instruction group performed better at immediate post test generalizing trained PDA skills when interacting with people, especially those other than the instructor. As well this group was found to be more fluent on task performance than those in the conventional instruction group (Powell et al., 2012).

McDonald et al. (2011) conducted an RCT cross over trial in which participants were randomly assigned to a Google calendar group, or a standard diary group. Prior to randomization, participants were asked to identify routines they would like to complete within the next 15 weeks. Following this individuals were randomized to either the Google calendar group (group A) or the standard diary group (group B). At the end of the 5 weeks, group A began using the standard diary and group B began using the Google calendar. Results indicate that memory aids helped to improve prospective memory performance of all participants. Google calendar was found to be more effective in improving prospective memory then the standard diary. Participants were able to achieve 82% of their targets using Google calendar but only 55% of targets were achieved using the standard diary.

Lemoncello and colleagues (2011) randomly assigned 23 individuals into one of two groups. Those assigned to group A, the Television Assisted Prompting (TAP) group, had the TAP system installed on their television where they received reminders of events to be completed. Those in group B, the Assistive Technology for Cognition (ATC) group, received reminders through more traditional methods (paper planner, cell phones or computers). Following the 10 week intervention Group A received reminders in the more traditional way and group B began receiving reminders through the TAP system. The TAP system was found to improve task completion. This finding adds to the growing body of literature supporting the use of ATC to improve prospective memory post ABI.

In another RCT, 45 individuals were randomly assigned into one of 4 treatment groups (Shum et al., 2011). The treatment groups consisted of 4 different intervention programs: self-awareness plus compensatory prospective memory training; self-awareness training plus active control; active control plus compensatory prospective memory training and active control only. Pre intervention scores on the CAMPROMPT did not reveal any significant differences between any of the groups. Those assigned to the compensatory prospective memory training groups showed greater changes in strategies used to improve memory. Compensatory prospective memory training included use of a diary or organizational devices, and group members were encouraged to use written reminders, appointments and note taking. Although at total of 45 participants started the study, only 36 completed it.

Dowds et al. (2011) recruited 36 adults to participate in a RCT using two different PDAs or a paper-based schedule book to assist them in remembering to complete various pre-selected tasks. Tasks completion rates were higher under the MOS and POS conditions. Participants using the POS PDA had a significantly higher task completion rate then those using the MOS PDA.

Wright et al. (2001a) examined the effect of two pocket computer systems containing three memory aides: appointment diary, notebook, and a to-do-list with a group of 12 ABI participants (9 TBI, 2 ABI). The type of pocket computer was counterbalanced and participants used each one for 8 weeks. No significant difference in use was found between type of pocket computer (they differed in terms of text entry – physical keyboard or touch- screen keyboard), and the majority (83%) used the three aids. Those participants who had previously used a memory aid made significantly more diary entries compared to those who had not previously used a memory aid. Severity of injury as well as level of cognitive function was not reported in this study. In another study by Wright et al. (2001b), findings were similar (i.e. no differences between computer systems in terms of use of memory aids).

Wilson et al. (1997) evaluated the efficacy of NeuroPage, a portable paging system, in reducing everyday memory problems in 15 ABI participants (10 TBI, 5 ABI). Using an A-B-A design, results indicated that all subjects significantly benefited from using the NeuroPage system and that following 12 weeks of use, performance remained at improved levels compared to baseline for another 3 weeks. Wilson et al. (2001) conducted a randomized controlled cross-over trial with 143 memory impaired patients, many – how many having sustained a TBI. The objective for this study was to evaluate a paging system designed to improve independence in people with memory problems as well as to reduce deficits in executive function. Results demonstrated that the pager system significantly increased patients’ ability to carry out daily tasks, and successful task achievement was more efficient after the pager intervention was introduced.

Hart et al. (2002) used hand-held recorders to remind moderate-to-severely impaired patients of their therapy goals (within subject design). Six individual goals were determined and half were recorded onto a hand-held organizer with an alarm preprogrammed to review the goals 3 times a day throughout the week. The other half of the goals were not recorded but were summarized at the weekly clinical management meetings. Goals were correctly recalled when using the hand held recorder compared to when goals were reviewed. It should be noted that the study examined only if the goals could be elicited during recall (either free recall or cued) and did not examine whether the subjects actually followed through with their goals.

Burke et al. (2001) used a complex computerized tracking system (patient locater and reminder system – PLAM) to remind and direct 5 patients on an acute rehabilitation unit to their next therapy appointment. The electronic tracking system prompted patients 10 minutes in advance of their appointments and continued to do so until the patient started moving toward the therapy room. If patients were going in the wrong direction, the system would prompt them on how to get to the appointment and would offer positive reinforcement as the patient made their way to the therapy room. Using a case series design, baseline data was gathered for a week and included the number of staff prompts needed to get the person to scheduled therapy and the time the person arrived at the therapy. Once the patients were introduced to the PLAM system, data was collected for a 3-day period. Results indicated that the subjects arrived earlier to their appointments and required fewer prompts (i.e. the number of sessions that did not require prompting increased from 7% to 44%).

In a prospective controlled study completed by Bourgeois et al. (2007), 38 subjects, along with one significant other for each subject, were assigned to either the spaced retrieval (SR) group or the Didactic strategy instruction (SI) group. Subjects were asked to maintain a daily log where they would note all areas they were having difficulties in. Treatment goals were developed based on areas of difficulty. Those in the SR group were given prompt questions and responses for each goal selected. Answers to the prompt questions were expected to be given in exactly the same way each time. Those in the SI group received time with a therapist to discuss memory strategies. All sessions were conducted by phone for both groups. Results indicate that the frequency of memory problems decreased in both groups over time. Significant improvement in goal mastery (p<0.05) was noted in the SR group but not in the SI group. Scores on the Cognitive Difficulties Questionnaire (CDS) indicate both groups were experiencing fewer difficulties following treatment. Post treatment, scores on the community integration questionnaire (CIQ), showed no significant differences between the two groups.

Boman et al. (2007) invited 8 participants live in one of two apartments equipped with electronic aids to daily living (EADLs). Participants, one in the apartment, were given 4 or 5 sessions weekly, each lasting 1 to 2 hours on how use the EADLs. Init would do; however, post study results indicate improvement in the self-perceived ability to perform important activities and in the satisfaction with performing tasks (p<0.05). For 6 participants improvement was also seen on the Sickness Impact Profile (SIP) 136. Study results indicate occupational health and quality of life had improved. Overall the authors found that EADLs may play a role in facilitating everyday functions.

In a study conducted by Egan et al. (2005) individuals who had sustained a TBI were instructed on how to use the internet. Each participant was given one-on- one instruction in their own home. Following training, individuals showed significant improvement on their level of independence in using the internet (p<0.028). Less improvement was noted when looking at the tasks that required greater abstract understanding and required more steps to complete. Participants’ were able to complete concrete tasks using fewer steps with greater ease.

Using a memory notebook as the external memory aid, Schmitter-Edgecombe et al. (1995) assigned 8 individuals with severe closed-head-injury and memory deficits into either a notebook-training group or an interpersonal support group (control). Groups were matched on a number of demographic variables. Outcome measures included both performance on memory tests as well as observation and responses to a questionnaire on everyday memory failures. Both groups received 2, 1-hour sessions per week for 8 weeks (16 sessions). Results indicated that, on cognitive measures of memory functioning, there was no difference between groups. However, on observed everyday memory failures (questionnaire), performance improved (i.e., less failures) following treatment, although performance was not maintained at 6-month follow-up.

In a randomized controlled trial, Watanabe et al. (1998) examined whether use of a calendar would enhance orientation following an acquired brain injury. Results indicated that the presence of a calendar did not enhance performance on a temporal orientation test (date and time). It is difficult to judge the outcome of this study as no scores were reported for either the control or treatment group, and it is not clear whether post-traumatic amnesia, and/or severity of injury had an impact on performance.

Ownsworth and McFarland (1999) evaluated two different training approaches in the use a diary to compensate for memory problems. They randomly assigned 20 ABI volunteers (15 TBI; 5 ABI) to either a Self-Instructional Training (SIT) approach or to a task-specific learning approach. The Diary-SIT approach trains compensation using higher cognitive skills of self-regulation and self-awareness. That is, participants where taught to question themselves with the following script (WSCT): What are you going to do?  Select strategies; Try it out; Check how it’s working. By using this training approach, the researchers speculated that it provides direct, internal feedback, which can generalize to other situations involving memory. In contrast the Diary-Only approach taught subjects how to use the diary. Results indicated that those in the Diary-SIT group made consistently more diary entries, reported a reduction in everyday memory problems and made more positive ratings on treatment efficacy compared to the Diary-Only group.

In a study conducted by Fish et al. (2007), all participants (n=20) were given cell phones and asked to make calls a specific times of the day. To aid them in remembering, participants were sent text messages. Results indicate that the texting a reminder to participants resulted in significantly more calls (p<0.001) being made then when they were not sent a message.

Van den Broek et al. (2000) evaluated the effectiveness of the compensatory external aid, the Voice Organizer for five individuals following brain injury.  All five participants benefited from the use of the Voice Organizer as measured by the Message-Passing Test. For four of five patients, there was no significant improvement or deterioration in positive or negative affect during the course of the study.

In a case series, Manasse et al. (2005) exposed subjects to 2 treatment measures to aid them in memory recall. The traditional treatment was designed to assist subjects with memory recall, by pairing pictures of staff with an imagery statement, while the real-world treatment consisted of name restating, phonemic cueing and visual imagery to assist subjects in remembering names. Results from the traditional treatment indicated that 2 of the 5 subjects mastered 6 names during treatment, 1 of the 5 mastered 3 names and 4 of the 5 mastered one of the names. During the cueing condition of the real world treatment sessions only 2 names were consistently used by each subject. The visual memory program was the only program where subjects consistently used both target names. When questioned directly 4 of the 5 subjects could consistently identify one or both of the target names. Because subjects did not use the names was not indicative of whether or not they knew them.

Cicerone et al. (2000) recommended that the use of memory notebooks or other external aids “may be considered for persons with moderate to severe memory impairments after TBI [and] should directly apply to functional activities, rather than as an attempt to improve memory function per se.”


There is Level 1a evidence supporting the use of active or high tech external aids (assistive technology) as a compensatory strategy for memory impairments.

There is Level 2 evidence supporting the use of passive or no tech/low tech aids in improving memory impairments post ABI.

External memory aids have been shown to be an effective compensatory strategy for memory impairments.

Computer-assisted training

A specific intervention for improving general cognitive functioning is computer- assisted training. The use of computer-assisted cognitive retraining has multiple potential benefits within the rehabilitation setting following brain injury. Computer retraining allows for flexibility in retraining procedures, increased individuality of therapy programs and also decreases the amount of direct time a therapist is with the patient. It also has the potential of continuing cognitive retraining within the community setting. Furthermore, as presented at the NIH Consensus Development Panel (1999) computer-assisted strategies are used to improve neuropsychological processes, including attention, memory and executive skills.

In recent years, clinicians have recommended the use of computers as an efficacious tool in cognitive rehabilitation. One study investigated the efficacy of computer-assisted rehabilitation in comparison to non-computerized methods (Tam & Man, 2004). Eight studies were identified that used computer-assisted measures for cognitive rehabilitation following brain injury.

Individual Studies

Table: The Use of Computer Assisted Training to Enhance Executive Function


In a current RCT conducted by Dou et al. (2006), participants were randomized to one of two groups: the computer assisted memory training group (CAMG-treatment - group 1) or the therapist administered memory training group (TAMG-treatment - group 2) with each receiving one month memory training. Memory training was similar between the groups but they were delivered differently. The treatment groups received 20 training sessions with each running for 6 days per week and lasting approximately 45 minutes. The control group received no training. Sessions consisted of: training basic component memory skills in (1) the management of typical daily tasks utilizing/integrating the component memory skills, (2) customized programs and (3) skill consolidation as well as in the generalization of those skills in practice. Scores on the neurobehavioural cognitive status examination (NCSE) showed significant improvement in the TAMG and CAMG groups (p<0.015, p<0.02 respectively) compared to the control group. Results from the Rivermead Behavioural Memory Test (Cantonese version) showed the CAMG improved significantly compared to the control group (p<0.0001).Those in the TAMG showed no significant improvement.

Ruff et al. (1994) evaluated the effect of computer assisted rehabilitation using the THINKable computer program which is a multi-media system that focuses on memory and attention retraining. Although this study was designed as a randomized controlled cross over design, due to the small number recruited (15), the groups were analyzed together in a pre-post intervention fashion.   Psychometric testing revealed modest but significant gains made for some memory and attention measures in each of the groups.

Self practice, presentation of attractive stimuli, multi-sensory feedbacks and personalized training contents were the four different forms of computer-assisted cognitive re-training programmes that Tam and Man (2004) used to evaluate people with post-head injury amnesia. Participants were randomly assigned to one of four treatment groups (matched diagnostically and demographically): (1) self-paced group, which allowed individuals to move at their own pace in a non-threatening environment; (2) feedback group, which involved immediate provision of feedback in a non-judgmental fashion; (3) personalized group, whereby the computer presented training contents showing the participant’s actual living environment and routines; and (4) the visual presentation group, which was a provision of attractive and bright presentation designed to help individuals engage in the activity. Each group went through one of the four computer-assisted memory re-training strategies. Results revealed that the patients in the experimental group showed positive improvements on all of the four memory training methods as compared to the control group. However, there were no statistically significant differences among the four training methods. Nonetheless, this study showed that computer-assisted memory retraining yield positive results for patients with memory post-head injury amnesia. Similarly, in a non-controlled study by Kim et al. (2000), all 12 patients that took part in a trial investigating the efficacy of a palmtop computer for use in daily activities, recommended that this treatment continue to be used in outpatient brain injury rehabilitation.

Middleton et al. (1991) also examined computer assisted cognitive rehabilitation by comparing two forms of retraining. Patients were assigned to either a group targeting reasoning and logical thinking or a group targeting attention and memory. Gains were made by both groups using within-group comparisons. No differences were noted in outcomes based on treatment group. 

Chen et al. (1997) studied the effect of computer assisted cognitive rehabilitation versus traditional therapy methods. Within-group comparisons of pre- and post-intervention measures demonstrated significant gains on multiple psychometric tests taking into account multiple statistical comparisons. However, multivariate analysis comparing the experimental and control groups across the domains of attention, visual-spatial, memory and problem solving did not demonstrate significant differences between the groups.

Gray et al. (1992) investigated the effects of attentional retraining using a microcomputer-based intervention. Patients were stratified into closed-head-injury (CHI) or other diagnosis (17 patients diagnosed as CHI) and mild/moderate or severe injury (15 diagnosed as severe) and randomly assigned to receive either attentional retraining or recreational computing (control). Time since injury varied widely from 7 weeks to 10 years. Immediately following training, the treatment group showed marked improvement on two measures of attention in comparison to the control group, however once premorbid intelligence score and time since injury were included as covariates, this treatment effect was no longer significant.  The experimental group showed continued improvement at 6-month follow-up on tests involving working memory.

In a recent study conducted by Bergquist et al. (2009), individuals were asked to participate in either an active calendar acquisition intervention or a control diary intervention program. Participants were assigned to one of the two interventions and once completed they began the second intervention. Sessions were completed on line using an instant messaging system. Improvement was noted in calendar use and using a cue card (p=0.02, p=0.01 respectively). Family members noted an improvement in mood and memory problems post intervention.

Gentry et al. (2008)conducted pre and post test assessment on a group of 23 individuals with a TBI living in the community. Due to problems with memory, all participants were found to have difficulties with every day tasks. To assist them in improving their memory a PDA was given to each individual and training was provided by an occupational therapist. Following training, participants reported improved in satisfaction with performing everyday tasks. Improvements were noted when looking at post training performance and post training satisfaction (p<0.001) and on the scores on the CHART-R self-assessment rating scale. Overall significant improvement was seen on the scores of the occupation, cognitive independence, and mobility subsections of the test (p<0.001).

Electronic mail (email) may prove useful for reducing the experience of social isolation for patients sustaining acquired cognitive-linguistic impairments (Sohlberg et al., 2003). The authors were interested in the usability and patient preference of a simplified email interface on eight brain injured patients. Patients were asked to read and reply to four prompt conditions: no prompt, idea prompt, fill-in-the-blank prompt and multiple-choice prompt. Difficulties encountered included computer usability and message composition. Results identified three categories of usability problems: lack of knowledge concerning functionality of keys for word-processing operations, poor conceptual understanding for the mouse operation and poor use of interface prompts. Results also found that there was considerable variation among patient preferences and the types of errors observed in composing emails, and that all patients legitimized the use of email interfaces as a means of reducing social isolation.


There is conflicting evidence supporting the use of computer assisted cognitive retraining as an adjunct to the rehabilitation program, especially regarding attentional retraining following brain injury.  Although some improvement in memory was found in a few of the studies it was not found in all. General cognitive functioning did appear to benefit from computer assisted cognitive retraining; however, further study confirming these findings need to be conducted.

Computer-assisted training has been shown to have a positive effect on general cognitive functioning, but has not yet been shown to be an effective treatment for the remediation of memory and attentional deficits.

Virtual Reality and Cognitive Functioning

One study was identified that used an innovative approach to improving cognitive function following brain injury, using a non-immersive, virtual reality component to promote exercise and cognitive functioning.

Individual Studies

Table: Virtual Reality (VR) Exercises and Their Impact on Cognitive Functioning


Grealy et al. (1999) addressed the effects of exercise and virtual reality post brain injury. This study evaluated the impact of an exercise program, which used a stationary bicycle in conjunction with non-immersive virtual reality administered over a minimum of 4 weeks. The results demonstrated significant benefits in the experimental group pre and post intervention for learning and memory tasks. Similarly, when compared to historical controls, the experimental group fared significantly better on digit symbol as well as visual and verbal learning tasks.

In a study by Zhang et al. (2001) 60 individuals (30 of them had suffered a TBI) participated in a study which tested their skills in a virtual reality kitchen. Participants were given a task to complete (twice in ten days) and were given cues to assist in completing the task. Individuals who had sustained a TBI did not perform as well as those without a TBI. Individuals with a TBI were slower, experienced difficulty in processing information presented to them and were unable to work through the task in a logical way.


There is Level 2 evidence of a positive impact on visual and verbal learning post exercise intervention for brain injury survivors.

There is Level 3 evidence from one study indicating that VR programs do not enhance cognitive functioning post TBI in individuals who have sustained a TBI.

Virtual reality programs may enhance the recovery of visual and verbal learning following brain injury; however more study needs to be completed as currently there is limited evidence supporting the use of VR programs.

Internal Aids

The following studies examined how internal aids could be used to enhance memory following an ABI.

Individual Studies

Table: The Use of Internal Aids to Enhance Memory Post ABI


In a recent study Potvin et al. (2011) assigned 30 moderate to severe TBI patients to either an experimental group (n=10) or a control group (n=20). Both groups were matched based on age and education. All participants were initially assessed using the TEMP. Those in the experimental group participated in ten prospective memory training sessions. Each session lasted 90 minutes. The PM program was divided into 5 phases: understanding PM functioning; training to visualize simple images; learning visual imagery techniques; applying visual imagery in PM; and applying visual imagery in everyday situation.  The scores on the TEMP, following treatment, improved for those in the experimental group. Study authors also noted that those in the experimental group reported fewer symptoms of depression than the control group.

Twum and Parente (1994) randomly assigned 60 TBI patients into one of 4 groups (one control and three mnemonic strategy groups) counterbalanced. The researches demonstrated improved performance for subjects who were taught a strategy (either verbal labeling or visual imagery) while learning paired-associations. Treatment groups showed greater efficiency in learning and greater delayed recall information.

Ryan and Ruff (1988) used mnemonic strategies, including visual imagery in a memory group and found that these strategies enhanced performance for mildly impaired subjects only (severely impaired group showed non-significant findings between control and treatment groups). Thoene and Glisky (1995) using a case series design also showed enhanced performance following the use of a mnemonic strategy (verbal elaboration and visual imagery) compared to vanishing cues and/or video presentation during paired associations.

Goldstein et al. (1996) and Malec et al. (1991)evaluated a visual-imagery technique (“Ridicuously Imaged Story” technique (RIS)) in training severely brain injured individuals to learn and recall lengthy word lists. Participants were asked to read a story where 20 words are presented in bold-face and subjects were instructed to remember the bold-face words for later recall. If subjects could not recall all the words they were provided with (1) the part of the story in which the word appeared and if that didn’t aid recall, they were then provided with (2) a category cue for the word.  It should be noted that in both studies reviewed, a number of their subject pool (N=10) came from a previous study (Goldstein et al., 1988). Goldstein et al. (1996) evaluated whether there were differences between a computerized and non-computerized version of RIS and another visual imagery technique (Pictorial Imagery). Results indicated that although the computerized versions resulted in a slightly better performance on learning trials, the difference was non-significant. Malec et al. (1991) used the RIS technique to examine the predictors of memory training success and found that the “better subjects did at tasks similar to those which they were trained, the better their learning and capacity to generalize.”

By using the various visual imagery techniques to aid learning and recall, researchers have demonstrated that increasing the saliency of features encoded, results in an increase in the amount recalled. Milders et al. (1998) examined performance on a name learning task by increasing the meaningfulness of people’s names with various strategies (e.g. when learning a new name-face association try to think of an occupation or object with the same name or a famous person with a similar name etc). When subjects (13 severely TBI vs. 13 matched controls) were tested on 3 different memory tasks, results indicated a significant difference following training, more so for the control group than the TBI group.  Also, learning procedures were more effective on one task (where subjects were required to learn the name-occupation-and town) compared to the other two tasks (famous-faces or name learning), which supports Malec et al. (1991) findings of generalization when tasks are similar. Goldstein et al. (1990) found that semantic processing aids recognition of to-be-recalled words compared to processing words at a more perceptual level in both closed head injury patients and control subjects (of course the degree of facilitation is reduced in the TBI group compared to controls).

Zencius et al. (1990a) examined the differential effects of various strategies on recall of information. Six TBI patients were asked to find two jobs from the help wanted column of a newspaper extracting 3 pieces of information for each job.  They were asked either to learn the information for later recall using one of the following strategies: verbal rehearsal, written rehearsal, acronym formation or notebook logging. All strategies resulted in improved performance (number of information correctly recalled) with the exception of written rehearsal (performance similar to baseline). Notebook logging resulted in the best performance.

Berg et al. (1991) demonstrated that severely brain injured patients demonstrated improved effects on objective measures of memory at 4 months following training in a strategy-use group compared to a pseudo-treatment and a no treatment control group.  In the strategy group, individuals were taught general cognitive principles of memory functioning and aids (i.e., internal and external strategies were taught and practiced). In contrast, the pseudo-treatment group practiced memory games and tasks with no explanation. In a 4 year follow up study Milder et al. (1995) results demonstrated that the effects at 4 months were no longer evident at 4 years (all groups were equivalent).

How individuals learn (i.e., encode) information will determine to a large extent what is later recalled. Twum and Parente (1994) demonstrated that if an active strategy (either verbal labeling for visual information or visual imagery for verbal information) is taught to individuals while learning the paired associations, learning and recall is enhanced (i.e., fewer trials needed to reach criterion during learning and improved recall following a delay).Tailby and Haslam (2003)also examined how learning can improve or limit later recall of information. They had 24 ABI subjects matched on basis of age, gender, premorbid and current intellectual status divided into 3 groups based on performance of verbal memory (mild, moderate & severe). Each group (n=8) was randomly assigned to one of 3 learning conditions: errorless learning, self-generated; errorless learning, experimenter generated; and errorful learning. Results showed that regardless of severity level, subject recalled more information in the errorless learning conditions (with self-generated superior to experimenter generated) than in the errorful learning condition.

Constantinidou and Neils (1995) examined the effects of stimulus modality on verbal learning of patients with moderate-to-severe closed head injury and a matched control group. Results indicated that when information is presented visually (with and/or without auditory presentation of names) more information is learned than when information is presented within the auditory modality alone.  As expected, patients learn new information at a significantly slower rate compared to controls.

It is generally thought that while patients are experiencing post-traumatic amnesia (PTA), they are not able to learn and retain new information, and as a result, cognitive rehabilitation is usually postponed until PTA has resolved. This tends to be true if using tasks of explicit or declarative learning and recall. Two studies were reviewed that reported that PTA patients were capable of learning and retaining new information when task demands were dependent on implicit/procedural learning. Glisky and Delaney (1996) evaluated implicit memory (priming using a stem completion task) and the use of vanishing cues when learning semantic information in a small number of TBI patients (n=8 & 4) who were still experiencing PTA and a matched control group. Findings revealed that learning and recall of information (once PTA has resolved) had occurred, albeit at reduced levels compared to controls. Ewert et al. (1989)also demonstrated procedural learning and retention in a group of 16 severely closed head injured participants and matched controls.


There is Level 2 evidence (from several studies) that internal strategies appear to be an effective aid in improving recall performance.


Internal strategies appear to be an effective aid in improving recall performance.


Memory Programs

Following an ABI or TBI one of the most persistent problems are memory deficits. Although the literature examining the efficacy of memory programs is limited, there is some support for training that stresses external memory strategies. Again the support for these programs is limited as many individuals post injury neglect their devices or simply stop using them (O'Neil-Pirozzi et al., 2010). Internal memory strategies have also met with limited success.

Individual Studies

Table: The Effect of Amantadine on Executive Functioning Following Brain Injury


In a recent study by O’Neil-Pirozzi et al. (2010), individuals with a TBI participated in 12 ninety minute sessions which were held twice a week. The intervention included memory education and to improve memory function the study emphasized internal strategy acquisition. Primary emphasis was placed on semantic association followed by semantic elaboration/chaining and imagery.  Results from the Hopkins Verbal Learning test indicated significant differences between the groups and those with a severe TBI performed more poorly than those with a moderate injury. Those with severe TBIs, although they performed more poorly than those with mild or moderate injuries, did perform better than those in the control group who were individuals who had sustained a severe TBI. In all memory performance was seen to improve for all in the intervention group compared to the control group.

Thick-Penny and Barker-Collo (2007) randomly assigned 14 individuals to either the treatment or control groups. Those in the treatment group participated in a memory rehabilitation program. The memory groups consisted of 8 learning modules each 60 minutes long. They ran twice a week for 4 weeks. Memory improvement and difficulties were evaluated. Overall a reduction in memory impairment was noted at the end of the 4 weeks of intervention and again at the one month follow-up time period.

Ryan and Ruff (1988) randomly assigned 20 severely brain injured individuals, matched for age, gender, education, and time since injury, to either a memory retraining group or a psychosocial group (control). Treatment lasted for 6 weeks (4 days per week, 5.5 hours per day for each group. Initially no differences were observed between groups on neuropsychological measures of memory. When groups were subdivided based on neurocognitive severity (mild vs. severe), results indicated that the mildly impaired group benefited more than the severely impaired group from memory retraining.

Sumowski et al. (2010) investigated the effects of retrieval practice (a technique that has been shown to be effective with non-TBI individuals) on those who had sustained a TBI. It has been shown with healthy individuals that retrieval practice allows for the retrieval of information shortly after it has been presented which leads to better delayed recall. Here 14 TBI and non-TBI individuals were presented with a series of paired words and divided across 3 learning conditions: massed restudy, spaced restudy and retrieval practice. Results indicate that retrieval practice was effective in improving memory in persons with a TBI.

Schefft et al. (2008) conducted two studies looking at the effect of a self generation memory encoding strategy on memory. In both studies flash cards were presented to participants with a pair of words on each. In Study 1 subjects were assigned to either the generate condition or the read condition or vise versa. In the read condition cards were presented to the subjects with a pair of words on each card, which they were asked to read aloud. In the generate condition participants were asked to look at the word presented and voice the second word that had been on the card. In Study 2, pairs of words were presented to the participants followed by a cued recall trail or a free recall trial. Results of Study 1 indicate that memory test performance improved as a result of self-generation encoding procedures. Again the results of Study 2 indicated that self generation strategy improved cued recall. It did not however improve free recall.

In a case series conduced by Hillary et al. (2003), 20 individuals with a TBI were presented with a series of words (4 groups of 20 words each). Each group was presented once (single condition), twice (massed condition) or twice with 11 words between each presentation (spaced condition). Spaced presentation led to significantly (p=0.018) greater recall and recognition of words than massed and once presented words. This result was also seen during delayed recall, and delayed recognition.

Freeman et al. (1992) conducted a matched-controlled treatment outcome study to evaluate executive and compensatory memory retraining in traumatic brain injured patients. Twelve patients were included in this study; six who received remediation treatment, which involved repeated presentation of various paragraphs, and six who received no treatment. A significant difference was found between the treatment group and the control group’s post-training measures with the experimental group improving considerably more than the control group. Results suggest that memory remediation is effective for brain-injured patients with memory impairments.

Evans and Wilson (1992) examined the effects of a memory group that met weekly for 11 months (2 hours a week for approximately 48 weeks).  Family and individuals reported an increase in using memory aides and strategies at 7 months and at 11 months compared to baseline (no objective measures were given and it is unclear if beneficial). Scores on neuropsychological measures of memory did not change over time. A main drawback of this study is the researchers failure to describe the nature and content of the memory program.

Quemada et al. (2003) examined memory rehabilitation following severe TBI in 12 individuals (no controls). The program ran for 6 months (50 minute sessions 5 days a week for 5 months and then 3 days a week for one month) and followed a specified format utilizing behaviouralcompensation techniques, mnemonic strategies, environmental adaptations, external and internal aides. Results indicated little improvement in standard measures of memory functioning, although patients and family members report meaningful functional gains (self-report and observed behaviour in everyday functioning).

In a case study(n=3) conductedby Fleming et al. (2005)subjects were enrolled in an eight week memory rehabilitation program. Each attended 1 to 2 hours per week.  In each session elements of self-awareness and compensatory strategy training (using post-it notes, labeling and making lists) was incorporated. Subject responded well to the introduction of various strategies (personnel notebooks, diaries and organizers) to assist them in organizing activities in their daily lives and over time the total number of entries into their notebooks decreased.

Hux et al. (2000) examined the effect of training frequency on face-name recall.  Seven TBI patients with demonstrated memory impairment in a modified multiple-baseline design utilizing 3 training phases (daily sessions, twice a week session and 5 times a day) participated in the study. The phases were counterbalanced, thereby eliminating any order effect. Daily sessions as well as twice a week sessions were found to be more effective than sessions that occurred 5 times a day.  Mnemonics and visual imagery strategies were effective for 4 of the 7 participants regardless of frequency of intervention sessions.


There is Level 2 evidence indicating that memory-retraining programs appear effective, particularly for functional recovery although performance on specific tests of memory may or may not change.

There is Level 3 evidence supporting spaced retrieval practice as an effective method of improving memory post ABI.

There is Level 3 evidence suggesting that the spacing of repetitions improves memory post ABI.

Memory-retraining programs appear effective, particularly for functional recovery although performance on specific tests of memory may or may not change.

Although several mnemonic strategies have been used to help improve memory post ABI, retrieval practice seems to be the most effective.

Recall and recognition of words can be enhanced by using a spaced learning condition.

Cranial Electrotherapy Stimulation and Memory

Cranial electrotherapy stimulation (CES) is the application of less than 1 mA of electric current to the cranium. This application has been used to treat a variety of disorders, including treatment of withdrawal of patients with substance abuse (Michals et al., 1993). The effect of CES for the improvement of memory following brain injury was investigated.

Individual Studies

Table: The Effect of Cranial Electrotherapy Simulation on Memory


Michals et al. (1993) studied cranial electrotherapy stimulation and its effect on post-traumatic memory impairment in clinical care patients with closed head injury.  Patients received CES or sham CES treatments for 40 minutes daily over a period of four weeks. The group receiving CES treatment did not improve in their memory performance, nor did their immediate or delayed recall improve.  Further, with retesting, both the CES and the sham CES group showed a similarly significant trend with no group performing any better than the other. These results suggest that CES stimulation in brain-injured patients does not improve memory functioning.


There is Level 1b evidence, from one RCT, that cranial electrotherapy stimulation did not help to improve memory and recall following brain injury.


Cranial electrotherapy stimulation was not shown to be an effective treatment to enhance memory and recall abilities following brain injury.


Summary of Learning and Memory Post ABI

Not all patients respond equally to all intervention strategies and no study in the current review indicated whether severity of memory impairment (or memory profile) interacts with a particular external memory aid. Technology has increased the availability of external aids, although some seem more feasible to use than others (e.g., cell phones or hand-held recorders). Unfortunately, the studies reviewed did not specify the length of time subjects required to master compensatory strategies nor the nature of the long-term effects, if any. 

Most studies examined only tasks of word list recall and paired-associate learning suggesting that the mnemonic strategies reviewed may not generalize to other types of information (particularly real-world or functional information outside the laboratory). Errorless learning appears to be one procedure that can be used to enhance learning conditions.  One study highlighted the difference between severity of impairment and ability to benefit from internal strategies.

Frequency of intervention has an impact on learning and retention, although the exact parameters of this are unclear at the present time. The optimal duration of a program is also open for speculation.  No studies reviewed examined the number of sessions required for memory groups to be effective and only one study evaluated a difference in effectiveness between mild and severely impaired individuals after sessions.

Pharmacologic intervention does not appear to be effective in improving learning and memory deficits.