Memory impairments are common cognitive problems associated with TBI. As such, myriad cognitive rehabilitation therapy (CRT) interventions aim to restore or compensate for memory deficits. This chapter presents descriptions for studies by method of memory strategy (e.g., internal, external, or combined). Within these sections, the controlled studies (e.g., RCTs and nonrandomized, parallel group) are divided by treatment comparator arm (e.g., no treatment, non-CRT treatment, other CRT treatment); following controlled studies, the noncontrolled studies (e.g., pre-post or single-subject, multiple baseline experiments) are described. The chapter closes with the committee’s conclusions for all memory studies reviewed, drawing out notable findings for mild or moderate-severe traumatic brain injury (TBI), as possible.
The committee reviewed 13 randomized controlled trials (RCTs) of treatments intended to improve or compensate for memory deficits. These trials varied in their intent to restore memory, show improvements in learning, or train individuals to use external or internal aids to compensate for poor memory. These trials enrolled a total of 315 study participants, with the size of the treatment group ranging from 8 to 39. The average age of participants ranged from early 20s to late 50s. Of the 13 trials, 12 enrolled participants in the chronic phase of recovery, averaging 4 to 7 years postinjury. One RCT enrolled participants who were in the subacute recovery phase, at 6 to 9 months postinjury (Watanabe et al. 1998).
The committee reviewed two nonrandomized, parallel group controlled studies of treatments intended to compensate for poor memory by training
the use of internal strategies. Goldstein et al. (1996) enrolled 20 participants and O’Neil-Pirozzi et al. (2010) enrolled 94 participants. In both studies participants were considered chronic, averaging 1 to more than 11 years postinjury; the average participant age ranged from the 20s to the 40s. The committee reviewed six pre-post single group design studies and six single-subject, multiple baseline (SS/MB) designs. Table 10-1 (at the end of the chapter) presents a summary of all included studies in this review.
Internal memory strategies may include the use of visual imagery or other repetitive, drilled practices. The committee reviewed seven RCTs and two nonrandomized, parallel group studies that used internal memory strategies; comparator arms included no treatment (n = 3), non-CRT treatment (n = 1), and other CRT treatment (n = 5). The committee also reviewed one pre-post single group design and five single-subject multiple, baseline experiments. Table 10-2 presents all internal memory strategy studies by design, strategy and treatment comparator.
Comparator Arm: No Treatment
Tam and Man (2004) conducted a small RCT in which 26 participants were randomly assigned to four computerized learning conditions: self-paced practice, stimuli/multi-sensory feedback, personalized training contents, and visually enhanced presentation. Treatment dosage ranged between 3 and 5 hours. Performance on drilled content improved significantly for all treatment groups compared to no treatment, with the feedback group showing the most gain. On a self-efficacy scale however, the feedback group demonstrated significant change after treatment, whereas others’ self-efficacy did not change. None of the groups improved significantly on the Rivermead Behavioural Test. The group that received stimuli/multi-sensory feedback appeared to improve memory for drilled content, which also may be related to their changes in self-efficacy for memory ability. It is unclear if improvement was related to the treatment, spontaneous neurological recovery, or other treatment participants were receiving at the time. With six and seven participants per group, interpretation and generalizability are limited. Also, specific time since injury was not reported, though individuals fewer than 3 months from injury were excluded.
Thickpenny-Davis and Barker-Collo (2007) conducted a small RCT that included moderately and severely injured participants who were more than 1 year postinjury. The 14 participants were randomly assigned either
|Study||Design||Multiple||Visual Imagery||No Treatment||Non- CRT||Other CRT|
|Bourgeois et al. 2007||RCT||×||×|
|Dirette et al. 1999||RCT||×||×|
|Dou et al. 2006||RCT||×||×|
|Ruff et al. 1994||RCT||×||×|
|Ryan and Ruff 1988||RCT||×||×|
|Tam and Man 2004||RCT||×||×|
|Thickpenny-Davis and Barker-Collo 2007||RCT||×||×|
|O’Neil-Pirozzi et al. 2010||Parallel||×||×|
|Goldstein et al. 1996||Parallel||×||×|
|Milders et al. 1998||Pre-Post||×|
|Benedict and Wechsler 1992||SS/MB||×|
|Ehlhardt et al. 2005||SS/MB||×|
|Hux et al. 2000||SS/MB||×|
|Manasse et al. 2005||SS/MB||×|
to receive a structured memory program or to join a waitlist. The memory intervention consisted of educating participants about memory (four parts of memory: attention, encoding, storage, and retrieval), assisting participants in understanding their own memory impairment and its effects, introducing and practicing strategies to aid memory and learning, and assisting participants in identifying the most appropriate and useful strategies for them. Strategies included didactic teaching, small group activities, discussions, problem solving and practice implementing memory strategies, errorless learning, and repetition. Postintervention, the experimental group as compared to the control group improved in many neuropsychological measures of memory (California Verbal Learning Test [CVLT]) long delayed free recall, Wechsler Memory Scale (WMS) logical memory delayed recall, and response time on the attention test (Continuous Performance Test [CPT]). The experimental group also showed increased knowledge of memory/memory strategies, increased use of memory aids/strategies, and decreased behaviors indicative of memory impairment. Results were maintained at follow-up with the exception of response time on the attention test
and immediate recall of narratives on the WMS. In addition to the initially small sample sizes, four of the seven participants in the waitlist control drop dropped out before providing posttreatment and follow-up measures.
O’Neil-Pirozzi et al. (2010), a large nonrandomized, parallel group study, examined the effects of memory training on individuals with mild, moderate, and severe injuries. Of the 94 enrolled participants, 54 received memory intervention and 40 received no specific intervention. Memory intervention, called I-MEMS focused on memory education and teaching individuals to use internal memory strategies, particularly “semantic association (i.e., categorization and clustering); semantic elaboration/chaining and imagery were emphasized secondarily” (O’Neil-Pirozzi et al. 2010). The memory intervention included 12 group sessions, 90 minutes each, held twice each week for 6 weeks, totaling 18 hours. Primary outcome measures were memory performance on the Hopkins Verbal Learning Test–Revised and the Rivermead Behavioural Memory Test II. Additional standardized tests of memory and executive functions were included. The treatment group demonstrated significant improvement on T-tests after treatment. Over time, these improvements went beyond changes in the control group. Regressions were used to determine if performance could be predicted after treatment (or second testing of control group). Consistent with the hypothesis, treatment predicted performance on both primary outcome measures at the second testing. Participants who received memory intervention improved more than those who did not. Furthermore, mild and moderately injured participants improved beyond those severely injured, even though the severely injured participants still improved beyond severely injured participants who received no treatment. At 1 month posttreatment, no significant changes were seen in memory performance. Aside from the limitation of not being completely randomized, the pre-post study design provides some evidence that the instruction of internal memory strategies has positive treatments effects when compared to no treatment, even for individuals who are at least 1 year postinjury.
Comparator Arm: Non-CRT Treatment
Ryan and Ruff (1988), a small RCT, enrolled 20 mildly to moderately injured participants who averaged 5 to 6 years postinjury. Participants were randomly assigned to the memory strategies arm or to the control arm. The memory strategies arm included training to use internal memory strategies such as associational tasks, chaining, rehearsal, visual imagery, and ritualized recall. The control group received psychosocial support and played cognitive games. Each group received 48 hours of treatment over 6 weeks. On neuropsychological measures of memory, both groups improved after treatment; however. those who were mildly injured and received strategy
training improved significantly more than moderately injured participants in both groups, as well as mildly injured participants in the psycho-social support group. Participants were not available for follow-up and no patient-centered measures were included. This study’s limitations include its small number of participants and data analysis by severity post hoc, even though it makes sense scientifically to examine treatment effects by injury severity. It should be noted however, that this was one of the earliest studies in memory intervention to find a severity effect.
Comparator Arm: Other CRT Treatment
Bourgeois et al. (2007), another modest-sized RCT, involved adults (average age 42) with persisting memory problems several years after a documented closed head injury. Participants also needed a family member willing to participate. Participant-caregiver pairs were assigned to either spaced retrieval training or a didactic control therapy that consisted of strategy education. Assignments were made using stratified pairing based on race and sex (quasi-experimental). Both interventions were delivered via telephone by clinician trainers. After initial face-to-face assessments of cognitive difficulties and social participation (Community Integration Questionnaire), the trainer discussed treatment goals with the patient and care-giver, and the group selected three specific goals. The trainer then provided memory logs and asked patients and caregivers to record the frequency with which each problem occurred over the next week. The trainer called the participant the following day to make sure instructions and data collection methods were understood. The trainer then called participants four to five times weekly for 30-minute sessions. Participants in the spaced retrieval group received an instructional technique focused on selected goals. During sessions, the therapist modeled correct responses to questions related to the goals and instructed the participants not to struggle to retrieve responses, but to respond immediately. Participants in the control arm received the same total amount of therapy time in sessions that included discussion about memory strategies such as association, verbal rehearsal, imagery, and written reminders. Outcomes included goals mastered, generalization, the frequency of reported memory problems, cognitive difficulties scale, and community integration. Immediately and at 1 month posttraining, the space retrieval group (and their caregivers) reported more treatment goal mastery and use than the didactic instruction group (and their caregivers). Both groups reported some generalization to other nontargeted behaviors, but the difference between these improvements among groups was not statistically significant. There were no reported important or statistically significant improvements in quality of life between or within groups on these measures. One limitation was that data about “objective, observable
behaviors” related to selected goals was obtained from memory logs, and these data were sometimes incomplete or not turned in. Of the 51 pairs that agreed to participate, only 38 completed the study: 22 spaced retrieval training pairs and 16 didactic control pairs.
Dirette et al. (1999), a small RCT, included 30 participants, the vast majority of whom had mild, moderate, or severe TBI. Injury severity was distributed equally across two treatment arms: one in which internal compensatory strategies (verbalization, chunking, pacing) were taught and one in which remedial computer work involving visual processing was provided. Both treatments were delivered via a computer for a total of 3 hours, in four 45-minute sessions, once per week for 4 weeks. The compensatory strategies came from a program called “IQ Builder,” which included “memory for numbers” and “memory for letters.” Outcomes included weekly measurement of working memory using the PASAT and two pre-post measures of computer-based visual processing for data entry and reading. Following treatment, both groups improved significantly on weekly and posttreatment measures, although performance did not differ by group, i.e., there was no treatment effect for learning internal compensatory strategies. Demographic variables, including injury severity and time since injury, did not account for participants’ performance either. Post hoc analyses of self-report and observations of strategy use indicated that about 80 percent of all participants, regardless of which treatment they participated in, used compensatory strategies. Unfortunately, treatment dosage was very low; there was no description of the instruction of the strategies. Furthermore, only F statistics and p-values were presented, which limits the applicability of these results to inform future research and interpretation.
Ruff et al. (1994) conducted a small RCT that involved 15 participants with severe TBI. Participants were randomized into two groups, in which the order of receiving restorative attention therapy and compensatory memory therapy was counterbalanced; i.e., both groups received both kinds of therapy in a crossover design. Participants received 20 hours of therapy via a computer program called “THINKable.” Outcomes were computer scores, neuropsychological tests of attention and memory, and behavioral assessments. After intervention, the computer scores showed significant improvement in attention but no significant improvement in memory. Results of the neuropsychological measures were mixed: immediate memory improved while delayed memory did not; only one attention measure improved. Self and other behavioral assessments of memory-based behavior did change after intervention, but only observer rating of attention-related behavior showed significant change after intervention. Thus, this study provides nonspecific, limited evidence on the efficacy of internal compensatory memory training (versus attention training) in that although subjective
ratings showed improved memory, improvement on computerized memory scores and neuropsychological test scores was inconsistent.
Dou et al. (2006), a small RCT, involved 30 participants with TBI who were several months post neurosurgery. Exclusion criteria include a history of psychiatric problems or computer phobia. Participants were randomly assigned to three groups: computer assisted memory training, therapist assisted memory training, and a control group that did not receive any specific memory training. In the computer assisted training, participants were asked to identify or define the information to be learned with computerized assistance. This decontextualized training consisted of instruction in internal, compensatory memory strategies aimed at memory and management of typical daily activities. The computer then provided the necessary information for the participants to generate correct decisions through an errorless approach. Participants were not encouraged to engage in guesswork, to avoid mistakes, and were told to consider alternatives to and consequences of an intended action. The therapist assisted training covered the same content but converted the instruction into a picture album; therapists gave directions face to face. The 15 hours of training were delivered in 20 sessions occurring 6 days a week, with each session lasting about 45 minutes. Immediately after treatment, both groups improved on multiple standardized measures of memory (Neurobehavioural Cognitive Status Examination, Rivermead Behavioural Memory Test) compared to the no-treatment group, although not on every measure. The treatment groups performed similarly in comparison to each other. Performance was the same at 1 month posttreatment. Thus, there appears to be some benefit to those at a chronic recovery stage to learning to use to internal, compensatory memory strategies; the delivery (therapist versus computer) does not appear to matter. Estimates and effect sizes were not provided, so the results cannot be used to inform the design of future studies.
Goldstein et al. (1996), a small nonrandomized, parallel group study, enrolled 20 participants with TBI and persistent amnesia who were provided with computerized instructions on how to create stories from word lists (“The Ridiculously Imaged Story” technique). Of the 20 participants, 10 received the computerized presentation on how to make associations between names and faces, as well as additional initial coaching and instruction about the cues the computer would provide for the list-story task. The other participants were instructed to make these associations using the original therapist delivery mode (Goldstein et al. 1988). Both groups were trained in these imagery techniques using roughly equivalent procedures. Data from 10 participants in a previous study that used therapist delivery were included as a comparison group. The number of words recalled from lists appeared to improve during generalization trials, though
no individual trials were significantly different between computerized and the noncomputerized comparison group (from original data in Goldstein et al. 1988). After treatment, both groups recalled significantly more from examiner-provided lists when compared to pretraining, and the computerized group appeared to improve slightly more. On participant-provided lists, pretreatment to posttreatment recall improved significantly, though the computerized group lost its advantage. On the name-face learning task, the computerized group had a clear advantage over the original method group, both in learning trials and pre- and posttreatment comparisons; in fact, the therapist delivery group did not recall significantly more names after treatment. Authors stated that the decontextualized methods did not provide evidence of long-term use of learned strategies to improve memory, though there was no long-term follow-up.
Other Study Designs
Benedict and Wechsler (1992), a single-subject, multiple baseline study, examines the effects of teaching the method of loci (MOL, for word list learning) and Preview, Question, Repeat, State, and Test (PQRST, for paragraph learning). Two individuals participated in the study—one with moderate TBI and moderate memory impairment and the other with severe TBI and severe memory impairment. They received 27 and 34 weeks of training, respectively, in which the order of MOL and PQRST were counterbalanced. Results revealed that the moderately impaired participant’s memory for word lists benefitted from the MOL training, but the participants’ paragraph learning did not benefit from PQRST training. The severely impaired participant’s performance was highly variable throughout, resulting in little change in recall from word lists or paragraphs.
Ehlhardt et al. (2005) investigated the efficacy of instructing adults with severe TBI to use recall and e-mail in a multiple-baseline-across-subjects-designed study. All five participants were many years postinjury and all demonstrated severely impaired memory and executive functions on standard neuropsychological measures. Treatment included the TEACH-M approach, which entails seven steps and learning principles of errorless learning; distributed practice and metacognitive instruction were emphasized. Training was delivered four to five times weekly, ranging from 7 to 15 weeks (as many as required to reach criteria). Four of the five participants completed the training and three of these four participants maintained these steps at 1 month after treatment ended, and all four participants maintained implementation of of the e-mail steps when “altered interface and/or a computer game with no shared features” was added (Ehlhardt et al. 2005). Interviews revealed that all four participants who completed the training endorsed the training. Inter-rater reliability and procedural fidelity
were reportedly strong: baselines were adequate prior to the start of treatment; therefore within-subject experimental control was clearly established.
Hux et al. (2000) examined the efficacy of internal memory strategies (mnemonics and visual imagery) to improve face-name recall in seven individuals with TBI who ranged from 2 to 26 years postinjury. Participants’ memory impairment ranged from nonexistent to severe. Intervention was delivered via training sessions that occurred five times per day in one phase, one time per day in another phase, and two times per week in yet another phase using within-participant comparisons. Face-name recall improved more after the intervention was provided one time per day or two times per week as opposed to five times per day, however results were highly variable across individual participants. Authors also reported frequent participant behavior problems.
Manasse et al. (2005) examined the efficacy and effectiveness of a sequential treatment approach that consisted of visual imagery for face-names, followed by real-word training that involved three cuing strategies: name restating, phonemic cuing, and visual imagery. There were five participants with chronic, severe TBI, ranging from more than 1 to 29.5 years postinjury. Treatment was provided in 9 sessions of visual imagery and 30 sessions of real-world intervention. All participants improved in name-face recall after intervention regardless of the kind of cuing, and four of five participants demonstrated more spontaneous use (effectiveness) of therapists’ names.
Milders et al. (1998), a pre-post single group study, involved 13 adults with memory problems following closed head injuries and 13 healthy controls matched on age and level of education. Most patients had been discharged from a nearby rehabilitation center. The mean time from injury was about 4 to 5 years, and the mean length of posttraumatic amnesia (PTA) they had suffered was reported as 36 days. The healthy controls were friends or relatives of the patients. Patients were taught strategies to improve the learning of new names and the retrieval of familiar people’s names. Strategies were taught in eight, 1-hour sessions delivered one on one over a 4-month period. The importance of applying the strategies in everyday life was repeatedly stressed and homework exercises were encouraged. Pre-post assessments in both groups included the following: three target evaluation tasks that had items not presented in the training (i.e., Name Learning Test, Name-Occupation-Town Learning Test, Famous Faces Naming Test); and two memory tests assumed insensitive or unrelated to the strategies practiced during training (i.e., Digit Span Forwards and Auditory Verbal Learning Task). Performance on two of the three target tasks improved with training compared to controls, but performance on the Name Learning Test did not change in either group. Both groups had similar improvement in the two control memory tests. Limitations included
the small selected sample, an unclear history of the severity and sequelae of TBI in some patients, and narrowly focused outcome measures.
External memory strategies may include the use of notebook or other tool to enhance memory abilities. The committee reviewed four RCTs and no nonrandomized, parallel group studies that used external memory strategies; comparator arms included no treatment (n = 1), non-CRT treatment (n = 1), and other CRT treatment (n = 2). The committee also reviewed three pre-post single group designs and one single-subject, multiple baseline experiment. Table 10-3 presents all external memory strategy studies by design, strategy, and treatment comparator.
Bergquist et al. (2010) and Bergquist et al. (2009), a small randomized crossover study, enrolled 20 volunteers who had moderate-severe TBI and were more than 1 year postinjury. Participants with a history of ongoing
|Study||Design||Notebook, Diary, Calendar, Other||External Cuing, PROMpting Device(s)||No Treatment||Non- CRT||Other CRT|
|Bergquist et al. 2009, 2010||RCT||×||×|
|Ownsworth and McFarland 1999||RCT||×||×|
|Schmitter-Edgecombe et al. 1995||RCT||×||×|
|Watanabe et al. 1998||RCT||×||×|
|Gentry et al. 2008||Pre-Post||×|
|Hart et al. 2002||Pre-Post||×|
|Zenicus et al. 1991||SS/MB||×|
psychiatric symptoms were included as long as symptoms were not severe (e.g., psychotic symptoms) and did not interfere with study participation. Participants also had to have reliable access to the Internet, as the trial compared two Internet-based interventions: an active calendar treatment intervention and a control diary condition. The calendar intervention, which involved an online therapist, focused on developing calendar skills to address difficulties with memory in everyday life and strategies to improve memory functioning. Participants in the diary control condition spent an equivalent amount of time interacting with a therapist online but simply used their calendar to record day-to-day events and not as a compensatory tool. Only 14 of the 20 participants completed the study; 6 of 8 assigned to the calendar intervention, and 2 of 8 assigned to the diary. Outcome measures included self-reported measures that assessed use of compensation strategies (Compensation Techniques Questionnaire) and satisfaction (four questions—satisfaction with therapist, satisfaction with therapy received, emotional distress during therapy, and willingness to receive such therapy again), as well as measures completed by family members (Neurobehavioral Functioning Inventory [NFI] and Compensation Integration Questionnaires [CIQ]). Analytic methods were not well described, particularly regarding missing data for patients who did not complete the trial. Most participants in both groups were satisfied with the Internet-based interventions. No statistically significant differences between groups were found for the four satisfaction questions. Also, no statistically significant differences in functional change between groups were reported after 30 sessions (NFI, CIQ outcomes).
Ownsworth and McFarland (1999) conducted a small RCT in which 20 participants with TBI who were many years postinjury were provided with a diary. Severity of brain injury was not described. Participants were randomized to either use a procedural worksheet during diary use (Diary and Self-Instructional Training) or to use the diary without this self-instruction (diary only), which required the use of higher cognitive skills of self-awareness and self-regulation. The diary-only participants were taught a behavioral sequence to use the diary. During the Diary and Self-Instructional Training session subjects learned how to compensate for everyday memory problems using a small notebook, as an internal strategy to mediate diary use. Some instructions for daily memory checklists were given verbally over the phone (in one session), but the 4-week intervention period mainly involved self-use of diaries. At the end of the intervention period, groups did not differ in mean number of diary entries; however, the diary-plus-self-instruction group maintained their use of the diary strategy to a greater extent than the diary-only group. Using daily checklists, the diary-plus-self-instruction group self-reported these strategies as more helpful and reported less confusion on a questionnaire. Thus, support is provided for
the use of self-instruction when using a memory diary if the purposes are to enhance self-efficacy of strategy use and reduce confusion and moments of disorientation.
Schmitter-Edgecombe et al. (1995) conducted a small RCT in which eight participants with severe TBI who averaged 13 to 16 years postinjury were randomly assigned to a treatment arm or a control condition for a 9-week intervention. The treatment arm consisted of training to use memory notebooks to compensate for memory, whereas the control condition consisted of group meetings to provide psychosocial support. In total, 16 hours of treatment or group support were provided (in 1-hour sessions, twice each week). Memory notebook training was provided in stages of skill-based learning consisting of anticipation, acquisition, application, and adaptation. Didactic instruction and homework, along with weekly goals, were incorporated at each stage in learning activities packets. Participants were taught to use the notebook, identify information, and take notes (Schmitter-Edgecombe et al. 1995). Modifications in notebooks were made based on participants’ needs. The control group met in group sessions to discuss social or psychological challenges in everyday living due to their memory impairment (Schmitter-Edgecombe et al. 1995). The primary outcome measures were laboratory-based measures (recall, everyday memory failures [EMFs]), retrospective report of EMFs, symptom distress indicators, and observational reports of EMFs. The study also measured neuro-psychological outcomes, but anticipated these would remain unchanged at posttreatment due to the focus on functional everyday memory activities. Pretreatment EMFs established a baseline to reduce error due to individual differences in subjects. On outcome measures for laboratory-based recall, laboratory-based everyday memory, and retrospective report of EMFs, there was no significant different between groups. However, a significant difference on observed EMFs was noted at immediate posttreatment; at 6 month follow-up, these findings retained direction but were no longer statistically significant. These findings provide preliminary evidence for the usefulness of notebook training to decrease EMFs for individuals with severe TBI. The limitation of the trial primarily was due to small size of the sample.
Watanabe et al. (1998), a small RCT, compared the effect on orientation of the presence/absence of a wall calendar in participants’ hospital room. All participants were receiving other inpatient rehabilitation, presumably CRT. The study compared temporal orientation (memory for the date) of 30 inpatients on an acute rehabilitation unit who were randomly assigned to groups that either have a wall calendar posted in their room or to not have a calendar. The average age in both groups was in the 50s. Neither time since injury nor severity of injury was reported; however, because participants were reportedly still in PTA, they were likely at least moderately injured and more than 6 months postinjury. The primary out-
come measure was the Temporal Orientation Test (TOT). Results indicated that the presence of a wall calendar had no effect on orientation; indeed, only the emergence out of PTA corresponded to orientation. This relatively weak study found no relationship between the presence of a wall calendar and orientation. The limited information provided on the participants, and the vague description of the intervention, make it difficult to interpret the results of this study for an inpatient population participating in rehabilitation. It is unclear how therapists provided orientation therapy that involved the wall calendar. The older ages of the participants implies that many had strokes, which can result in different kinds of orientation problems (e.g., neglect), which confounds these results. Also, because both groups were actively engaged in inpatient rehabilitation, there were likely numerous commonly shared features of rehabilitation between the two groups.
Other Study Designs
Bergman (2000) conducted a pre-post study involving 41 individuals with chronic cognitive deficits after severe TBI. All were described as having “difficulties with conventional strategies” for aiding memory such as notebooks, calendars, and Post-it reminders. The tested intervention was a “cognitive orthotic,” a computer software program designed as a compensatory strategy for aiding weak or ineffective cognitive functions. The underlying foundation for the program was described as “error-free learning, rapid system and skill acquisition, and facilitated generalization.” The computer program used six activity modules intended to minimize potential for error, reduce memory burden, maximize ease of memory storage and retrieval, limit preservative tendencies, promote transfer of training, and facilitate task completion through guided sequences. Modules addressed topics such as telephone logs, savings and checking, and appointments. Examiners (neuropsychologists or speech-language therapists) oriented individuals to the program and assessed participants’ mastery of the modules. Mastery was defined as the unassisted reliable completion of a targeted task. Reported outcomes were that 36 of the 41 participants achieved mastery of four or more activity modules, and 36 demonstrated rapid achievement of success on initial assigned tasks. Limitations included the absence of a control group, narrowly focused or restricted outcome measures, and an unclear history of the severity and sequelae of TBI in some patients.
Gentry et al. (2008), a pre-post single group study, involved 23 community-dwelling individuals with severe TBI at least 1 year postinjury. All had memory problems that affected ability to perform everyday tasks, such as remembering appointments, managing time and tasks, and managing money and medications. The intervention involved training individuals to use a freely provided personal digital assistant (PDA) as a compensatory
cognitive aid. Training sessions were provided by an occupational therapist in three to six 90-minute home visits conducted within a 1-month period. After training, participants were asked to use their PDAs for an 8-week period. All participants completed the study. Reported outcomes were improvements (pre-post) in assessments of self-rated occupational performance, satisfaction with occupational performance, and in participation in everyday life tasks. The outcomes were measured with standardized tests (Canadian Occupational Performance Measure and Craig Handicap Assessment and Rating Techniques-Revised Measure) and, while self-reported, were agreed upon by a family member or caregiver. Limitations were the absence of a comparison group and perhaps lack of outcome measures assessed by an objective (outside) observer. Generalizability may be limited because all participants were motivated volunteers recruited through fliers who had a working home personal computer and who were able to use a stylus without difficulty.
Hart et al. (2002) investigated the usefulness of a voice organizer in a pre-post design study. The 10 participants, who had moderate-severe TBI and were 3 to 18 years postinjury, were enrolled in a comprehensive TBI rehabilitation program. Case managers or clinicians developed a list of six therapy goals for each client. The goals chosen were considered likely to be discussed in upcoming therapeutic sessions, known to have been forgotten or not followed through by the client in the past, and agreed upon as important by the client and family. Case managers read the individualized goals to clients. Half of the goals that were read and reviewed were randomly assigned to be recorded on a voice organizer for clients while half were not recorded. Clients were given and trained to use devices with the voice recordings. They were prompted by an alarm to listen to the recorded goals three times daily. Seven days after the original session in which goals were recorded, each client’s recall for all six goals was tested by a staff member who was blind both to the therapy goals relevant to that client and to the specific goals that had been recorded. Recorded goals were recalled more often than the goals that were not recorded. Clinicians involved in the study thought that participants were more conscious of their recorded goals and more likely to follow through with them. Limitations include the small selected sample and narrow outcome measures that did not assess behavior changes.
Raskin and Sohlberg (1996), a single-subject, multiple baseline experiment, studied the efficacy of prospective memory training with two adults with severe TBI who were, respectively, 11 and 12 years postinjury. Two types of intervention were provided: prospective memory training and repetitive memory drill. Prospective memory was measured using the Patient Reported Outcome Measures (PROM), which measures memory at 1, 2, 10, and 20 minutes, and at 24 hours. Memory for future actions improved
more after prospective memory training than after repetitive drill, although generalization to real-world remembering was variable across participants and type of training. Both participants indicated their preference for prospective memory training during interviews.
Zencius et al. (1991), a single-subject, multiple baseline report, examined the usefulness of memory notebook training for completing homework assignments with four adults with TBI who were also receiving interdisciplinary rehabilitation services. Little descriptive information was provided about the participants other than age and variable test results. After notebook training, three of the four participants improved in completing the number of components to the homework assignments. Without participant or training information, coupled with the ongoing rehabilitation services participants were receiving, these results are difficult to interpret.
Combined memory strategies may include a blend of both internal and external approaches. The committee reviewed two RCTs and no nonrandomized, parallel group studies that used combined memory strategies; comparator arms included no treatment (n = 1) and other CRT treatment (n = 1). The committee also reviewed one pre-post single group design. Table 10-4 presents all combined memory strategy studies by design, strategy and treatment comparator
Berg et al. (1991) (with Milders et al. 1995) enrolled 39 severely injured participants in a small RCT in which they compared the efficacy of a memory strategy program that consisted of instructing two control groups
|Study||Design||Multiple||Visual Imagery||No Treatment||Non- CRT||Other CRT|
|Berg et al. 1991||RCT||Multiple||Multiple||×||×|
|Milders et al. 1995||strategies||strategies||×|
|Kaschel et al. 2002||RCT||Visual
|Freeman et al. 1992||Pre-Post||Multiple||Multiple|
on compensatory internal strategies and external aids. Thus, there were three arms in this trial; two that received treatment, the memory strategy rehabilitation group and a “pseudo rehabilitation” group, and one group that did not receive treatment. One of the “pseudo rehabilitation” control groups drilled and practiced (restorative), and the other received no treatment. The memory strategy program emphasized both internal strategies and the use of external memory aids, whereas the “pseudo rehabilitation” control treatment consisted of repetitive drill and practice, and the control group patients were tested according to the time schedule of the trained groups, but received no training. All participants were severely injured and averaged 5 to 6 years postinjury (i.e., in the chronic phase of recovery). Outcomes included self- and other subjective memory questionnaires (including measurements of anxiety related to memory and coping with daily memory problems), and standardized scores (mean sum score, acquisition score, and delayed memory score) from the Rey Auditory Verbal Learning Test, face-name learning, and memory for a shopping list. Immediately after treatment, the subjective ratings of memory problems improved significantly for both the strategy and the drill/practice groups. The strategy group improved on two of three neuropsychological memory measures (sum and delayed memory scores) immediately after treatment, and at follow-up improved significantly in the other neuropsychological memory measure (acquisition). There were no significant improvements found for the drill/practice and the no treatment group. Unfortunately, the authors did not report the reasons for dropouts, nor make adjustments for this in the data analysis; this information may have helped to explain why scores on memory tests appeared to improve over time after the immediate post-treatment results.
Kaschel et al. (2002) conducted a small RCT of 24 patients, including 12 patients with severe TBI who averaged 5 to 6 years postinjury. Participants were randomly assigned to receive visual imagery to improve memory or to receive a typical memory rehabilitation program, which emphasized a combination of compensatory internal strategies and external compensatory strategies. There were 30 treatment sessions in total. Primary outcomes were measures from the Rivermead Behavioural Memory Test (RBMT), the logical memory (stories) subtest from the Wechsler Memory Scale (WMS), and the Appointments test. Secondary outcomes were measures on the Concentration Endurance Test d2, Memory Assessment Clinics ratings scales (MAC-S, MAC-F). Immediate outcomes after intervention revealed that the visual imagery group performed better on the immediate recall of stories (both RBMT and WMS), delayed recall on the RBMT, and delayed (but not immediate) recall on the Appointments test. There were inconsistent treatment effects on the self-reported and other-reported ratings. No treatment
effects were found on the secondary measures. At 3 months after treatment, all treatment effects were maintained.
Other Study Design
Freeman et al. (1992) conducted a pre-post study that enrolled 12 adults in a private rehabilitation program center. All had cognitive deficits and a history of a closed head injury. Of the 12, 6 had been referred for cognitive rehabilitation; they were enrolled in a 6-month rehabilitation program that included a memory module as one of seven modules. The memory module was completed in 2.5 weeks. It was delivered in a 2-hour group setting, three times weekly. During the treatment, trainees and staff repeated various paragraphs and taught skills and techniques to enhance paragraph retention. Skills and techniques included such things as note taking in a memory book, self-monitoring skills, prompts to stop and think, restatement of presented material, and use of imagery. The other six people in the study had been referred for neuropsychological testing only. They received none of the rehabilitation modules but did paragraph memory tests (described below) as part of their neurological assessment at an initial visit and then again 2.5 weeks later. Of note, the mean time since injury for the memory module group was 33 months whereas the mean time since injury for the control group was 12 months. The outcome measure was a memory score based on comprehension and retention of main and secondary ideas presented in a paragraph. The reported outcome was a statistically significant difference between treatment and control posttest memory scores that favored the treatment group. Limitations included the small sample size, differences in characteristics of the intervention and control groups that were not accounted for in analyses, an intervention that was not described sufficiently to be replicable, and a single, limited outcome measure. Whether staff that administered and scored the outcome were the same staff that administered the intervention was not clear.
Restorative memory strategies aim to reestablish memory functioning following brain injury. The committee reviewed two RCTs that included repetitive drill as a treatment arm; comparator groups were both no treatment and have been previously described in this chapter (see Berg et al. 1991; Tam and Man 2004). The committee also reviewed one pre-post design and one single subject, multiple baseline experiment. Table 10-5 presents all restorative memory strategy studies by design, strategy, and treatment comparator.
|Study||Design||Restorative||No Treatment||Non- CRT||Other CRT|
|Berg et al. 1991;||RCT||Multiple strategies||X||X|
|Milders et al. 1995|
|Tam and Man 2004||RCT||Multiple strategies||X|
|Raskin and Sohlberg 2009||Pre-Post||Cuing, PROMpting|
|Raskin and Sohlberg 1996||SS/MB||Cuing, PROMpting|
Raskin and Sohlberg (1996), a single-subject, multiple baseline experiment, studied the efficacy of prospective memory training with two adults with severe TBI who were 11 and 12 years from injury. Two types of intervention were provided: prospective memory training and retrospective memory drill. Prospective memory was measured using the PROM of the Assessment of Intentional Memory (AIM) scale, which measures memory at 1, 2, 10, and 20 minutes, and at 24 hours. Memory for future actions improved more after prospective training than after the memory drill, although generalization to real-world memory was variable across the two participants and type of training. Both participants validated their preference for prospective memory training during interviews.
In a follow-up pre-post crossover design, Raskin and Sohlberg (2009) provided both prospective memory training and retrospective memory drills to adults with brain injury and healthy adults. Eight adults with brain injury received 1-hour training sessions, twice each week for 6 months. Again, prospective memory was measured using the PROM tasks of the AIM scale, at 2 and 10 minutes. Additional neuropschological tests, memory questionnaires, and a journal/log served as generalization measures. Adults with brain injury improved on prospective memory time and tasks after 2 minutes; however, this group did not show improvement at the longer delay of 10 minutes. On neuropsychological measures immediately post treatment, adults with brain injury improved in attention and executive functions. Generalization to everyday memory performance as measured by a memory questionnaire and memory diaries also improved. Maintenance of prospective memory improvements was demonstrated at 1 year posttreatment. None of the subjects showed improvement for retrospective memory drills. Half of the brain injury group initially enrolled in the study dropped out for various reasons leading to the potential for selection bias.
The majority of the evidence on the efficacy of memory intervention is with moderate-severely injured individuals who are at a chronic stage of recovery. In the chronic recovery phase, those with impaired ability to learn (store and retrieve) new information, routines, and skills are likely targets for interventions targeting the individual’s precise memory impairment. For example, encoding strategies are taught to individuals who have lost the ability to transfer new information into long-term knowledge. Individuals at a subacute phase of recovery also experience memory impairments; however, related attention, information processing, and organization impairments usually impede successful isolation and treatment of memory impairments.
The committee found no evidence that demonstrates the benefit of using internal memory strategies for everyday memory given the absence of patient-centered outcomes.
The committee found limited evidence that the ability to recall new information improves in patients with chronic, mild TBI when they learn to use internal memory strategies such as visual imagery and other encoding strategies. This benefit was short term or immediate as measured by standard memory tests (O’Neil-Pirozzi et al. 2010; Ryan and Ruff 1988).
The committee found limited evidence that in patients with chronic, mild TBI, learning to use internal memory strategies benefits memory long term (O’Neil-Pirozzi et al. 2010).
The committee found no studies that investigated the benefit of using external memory aids for patients with mild TBI.
None of the studies investigated the efficacy of memory intervention for individuals with mild TBI at the subacute recovery stage. Within a short time after injury, most individuals with mild TBI recover and remain asymptomatic. There was limited evidence that individuals with mild TBI in the chronic stage of recovery benefit from learning to use internal strategies
such as visual imagery and other encoding strategies (O’Neil-Pirozzi et al. 2010; Ryan and Ruff 1998). In these studies, dosage was provided for 13 to 18 hours, compared to psychosocial support or no treatment. Gains on formal tests of memory immediately after treatment were positive, although only one study provided evidence that these benefits were maintained at 1 month. There is no evidence demonstrating benefit to everyday memory, given the absence of patient-centered outcomes. Future research will be necessary to determine whether or not these strategies improve an individual’s ability to learn new information with clear benefit to daily activities (e.g., learning procedure manual instructions, retaining information for an exam). The absence of evidence describing the efficacy of external memory or compensatory strategies for those who have lingering memory impairment after mild TBI should not be equated with negative findings; that is, no current evidence does not mean that individuals with mild TBI do not benefit from using external aids.
The literature suggests that there is limited evidence of a differential benefit of internal memory strategies to patients with mild TBI over those with moderate or severe TBI. Two studies, one RCT and one nonrandomized, parallel group design (O’Neil-Pirozzi et al. 2010; Ryan and Ruff 1998) found that those with mild TBI benefited more than those with moderate or severe TBI. Single-subject, multiple baseline studies found that while individuals with moderate injuries made some improvement in memory, those with severe injuries did not benefit as much (Benedict and Weschler 1992) or did not demonstrate transfer of these skills (Manasse et al. 2005). Even RCTs with good experimental control showed that the generalization of the use of these strategies is insufficiently documented for those with moderate-severe TBI.
The committee found evidence that was not informative that memory intervention restores memory functioning in patients with moderate-severe TBI (Berg et al. 1991; Tam and Man 2004).
The identified evidence did not show a benefit of attempting to restore memory in individuals with moderate-severe injuries. Berg et al. (1991) (with Milders et al. 1995) suggests that restoring memory in patients with severe TBI is not efficacious, even though subjectively patients in the repetitive drill and practice arm reported changes in their memory. This RCT found that a comprehensive memory program including internal and external memory strategies improved both memory test scores and patientcentered
centered measures of improved everyday memory, at least maintained at follow-up. On standard measures of memory, only the strategy group improved. Tam and Man (2004) compared various kinds of computerized intervention, which was provided for 3 to 5 hours. All groups improved memory for the learned content after treatment, although not as much as the feedback group improved. The drill and practice group’s self-efficacy ratings of memory did not change. The low dosage of intervention makes these results difficult to interpret.
The committee found limited evidence that using internal memory strategies resulted in practical, improvement in everyday activities that involve memory and/or learning. Benefits in patient-centered outcomes were demonstrated by changes in participants’ self-efficacy about their memory (Tam and Man 2004), increased knowledge about memory strategies, validated reports by others in the use of strategies, and fewer behavior-based memory problems (Thickpenny-Davis and Barker-Collo 2007).
The committee found limited evidence that showed the majority of treatment effects were maintained at 1-month posttreatment follow-up (Bourgeois et al. 2007; Ehlhardt et al. 2005; O’Neil-Pirozzi et al. 2010; Thickpenny-Davis and Barker-Collo 2007).
The committee found modest evidence that most studies that were compared to no treatment or non-CRT treatment showed immediate benefit of improved memory using internal strategies as measured on standard memory tests (O’Neil-Pirozzi et al. 2010; Thickpenny-Davis and Barker-Collo et al. 2007; Ryan and Ruff 1988). Beneficial treatment effects were difficult to determine in studies comparing memory intervention to other CRT, possibly due to overlapping cognitive processes (Bourgeois et al. 2007; Dirette et al. 1999; Dou et al. 2006; Kaschel et al. 2002; Ruff et al. 1994).
The efficacy of using internal memory strategies to immediately improve memory performance in individuals with moderate-severe TBI on standard memory tests has been shown in several RCTs and a nonrandomized, parallel group design when compared to no treatment or non-CRT treatment (Dou et al. 2006; O’Neil-Pirozzi et al. 2010; Ryan and Ruff 1988; Tam and Man 2004; Thickpenny-Davis and Barker-Collo 2007). Dosage ranged from 13 to 30 sessions. The findings from RCTs that compared internal memory strategies given by instruction to other CRT treatments were less
consistent in finding a benefit to memory above and beyond the other CRT group on standard memory tests (Bourgeois et al. 2007; Dirette et al. 1999; Dou et al. 2006; Kaschel et al. 2002; Ruff et al. 1994). Considering the overlap in cognitive functions, it is challenging to isolate the active ingredient that enhances memory in those in the comparison treatments receiving another form of CRT.
A few RCTs had mixed results when they compared the interface or delivery of instruction of treatment strategies to moderate-severely injured individuals. Delivery methods included computer versus therapist, spaced retrieval instruction versus strategy discussion, and four computerized versions of memory intervention (Bourgeois et al. 2007; Dou et al. 2006; Tam and Man 2004). Although the treatment conditions resulted in improved memory over no treatment or baseline, there were not clear advantages of one instructional practice over another. Pre-post designs and single-subject designed studies add to the evidence base with similar results as the RCTs (Milders et al. 1998). The benefits of improved memory were in general maintained, though not all studies reported maintenance effects.
There is modest evidence that the use of internal memory strategies results in practical improvement in everyday activities that involve memory and/or learning. Two studies reported improved patient-centered outcomes that included changes in self-efficacy about their memory (Tam and Man 2004), increased knowledge about memory strategies, validated reports by others in the use of strategies, and fewer behavior-based memory problems (Thickpenny-Davis and Barker-Collo 2007). Three studies reported that they followed participants after treatment ended and the majority of the treatment effects were maintained (Bourgeois et al. 2007; O’Neil-Pirozzi et al. 2010; Thickpenny-Davis and Barker-Collo 2007).
Comparator: No Treatment or Non-CRT Treatment
Three RCTs (Dou et al. 2006; Tam and Man 2004; Thickpenny-Davis and Barker-Collo 2007) and one nonrandomized, parallel group study (O’Neil-Pirozzi et al. 2010) demonstrated improvement in learning and memory for those who received internal memory strategy training when compared to a no treatment control group. Outcomes included standardized tests of memory. Two of the four studies reported improved patient-centered outcomes that included changes in self-efficacy about their memory (Tam and Man 2004), increased knowledge about memory strategies, validated reports by others in the use of strategies, and fewer behavior-based memory problems (Thickpenny-Davis and Barker-Collo 2007). Two of the three studies that reported treatment effects were maintained at 1 month had no treatment as the control group (O’Neil-Pirozzi et al. 2010; Thickpenny-Davis and Barker-Collo 2007). One RCT provided evidence of
memory intervention when compared to control intervention that was not CRT (e.g., “sham” treatment). Ryan and Ruff (1988) found that the benefit of internal memory strategies was confined to those with mild injuries, not those with moderate-severe injuries.
Comparator: Other CRT Treatment
Five RCTs (Bourgeois et al. 2007; Dirette et al. 1999; Dou et al. 2006; Kaschel et al. 2002; Ruff et al. 1994) and one nonrandomized, parallel study (Goldstein et al. 1996) provided generally positive evidence that internal memory strategies improve aspects of memory above and beyond the control CRT. In Ruff, participants demonstrated changes on memory tests after a memory training module and after an attention module. Kaschel et al. (2002) attempted to investigate the active ingredient of visual imagery from matched participants who were receiving memory rehabilitation involving both external memory compensatory aids and other internal memory strategies. Participants who were trained in visual imagery performed better on several laboratory measures of memory, but not all. Dou et al. (2006) found that both the computer and therapist delivered internal memory programs resulted in similar improvement in memory over those who received no treatment; these results were maintained at 1 month. Dirette et al. (1999) compared to 3 hours of a computer-delivered internal memory strategy program to a “remedial computer program of visual processing” and found no group differences. The low dosage in this study is noticeable compared to the other trials, which ranged from 15 to 30 hours. Bourgeois et al. (2007) investigated the efficacy of spaced retrieval with individuals with severely impaired memory, compared to strategy instruction/discussion over the telephone with the intent to improve the recall and mastery of participants’ individualized goals. The spaced-retrieval group was better at reporting their goals and their use than the strategy discussion group, although no differences occurred between groups with generalized strategy use or reported memory problems. Bourgeois et al. (2007) also reported most of the treatment effects were maintained at 1 month. In a small non-randomized, parallel group study, Goldstein et al. (1996) had mixed results when comparing a computer- to therapist-delivered intervention on how to make associations.
External Strategies or Aids
The committee found modest evidence of the effectiveness of external memory aids (e.g., notebooks, alerting devices) to reduce everyday memory failures for patients with moderate-severe injuries in three RCTs (Bergquist et al. 2009, 2010; Ownsworth and McFarland 1999;
Schmitter-Edgcombe et al. 1995) and other studies (Bergman 2000; Gentry et al. 2008; Hart et al. 2002). Patient-centered outcomes included reduced numbers of memory failures and patient satisfaction.
The committee found modest evidence from RCTs (Bergquist et al. 2009, 2010; Ownsworth and McFarland 1999; Schmitter-Edgcombe et al. 1995) and other studies (Bergman 2000; Gentry et al. 2008; Hart et al. 2002) that showed immediate benefit of using external strategies or aids to compensate for poor memory.
There is modest evidence from three RCTs of the effectiveness of external memory aids to reduce everyday memory failures for patients with moderate-severe injuries in three small to modest-sized RCTs (Bergquist et al. 2009, 2010; Ownsworth and McFarland 1999; Schmitter-Edgecombe et al. 1995). Patient-centered outcomes included use of a compensatory aid, reduced numbers of memory failures, and patient satisfaction. Schmitter-Edgecombe et al. (1995), in a small but well-designed trial, found evidence that therapy to use memory notebooks resulted in compensation for everyday memory failures over those who received psychosocial support. Beyond using the compensatory aides, results suggest that guided self-instruction is associated with participants’ reporting the compensatory aid is more helpful and more effective in reducing daily disorientation than being given the aid without instruction (Ownsworth and McFarland 1999). In a telehealth study, Bergquist et al. (2009, 2010) compared dynamic instruction in using a calendar to a control condition (other CRT) in which participants used a diary. Both groups reported satisfaction with the Internet therapy; groups did not differ in self-reported satisfaction or in changes in general overall function on patient-centered outcomes of community integration.
In addition to these RCTs, several studies of other designs found complementary findings, including using cognitive or those strategies to guide the completion of complex, goal-directed activities (Bergman et al. 2000; Gentry et al. 2008; Hart et al. 2002). Therefore, while it would not be expected that external memory aids would actually improve memory; there is evidence that their use is effective in assisting patients to complete everyday, complex activities as indicated in functional, patient-centered outcomes. There is some evidence that patients continue to use compensatory aids several months after treatment ends.
|Study||N||TBI Severity Level||Brief Narrative||Comparator||Outcome Measures||Findings|
|Bergquist et al. 2010
Bergquist et al. 2009
|14||Moderate-Severe||This study examined patient satisfaction with a web-based cognitive rehabilitation program previously shown to effect functional improvements.||Y
Other CRT Content: Diary condition
• Compensation Techniques Questionnaire
• Study Satisfaction Questionnaire
|Between the time of assessment and the treatment condition, no significant differences were found in satisfaction. Greater satisfaction upon study completion was additionally found to be correlated with a higher level of calendar use prior to the study’s initiation.|
|Berg et al. 1991
Milders et al. 1995
“Memory strategy training” with several memory tasks chosen by participant with following rules applied:
• try to accept that a deficient memory cannot be cured
• make a more efficient use of your remaining capacities
• use external aids when possible
• pay more attention
• spend more time
• make associations
• link input
• retrieval situations
No Content and Other CRT Content: Two control groups-one that drilled and practiced (restorative), and one that received no treatment
• Control tasks
■ Distraction-Reaction Time Task
■ Reaction Time Task
• Effect tasks
■ 15 Words Test
■ Face-Name Learning
■ Shopping Lists
|At 4-month follow-up both groups of trainees saw significant, subjective effects of therapy on their everyday memory functioning, rating them as highly positive; only the strategy training group saw significant, objective effects on memory performance scores.|
|Bourgeois et al. 2007||38||NR||This study’s objective was to evaluate the effectiveness of an errorless training approach called Spaced Retrieval (SR), delivered via telephone, in improving everyday memory problems.||Y
Other CRT Content: Didactic memory strategy instruction
• Cognitive Difficulties Questionnaire (CDS)
• Community Integration Questionnaire (CIQ)
|Both SR and SI groups experienced a non-statistically significant improvement in their frequency of memory problems; furthermore, the differences between the two from pre-assessment remained, but were still not significant. At 1-month follow-up both groups reported some generalized strategy use to other nontargeted behaviors, but there were no significant results.|
|Dirette et al. 1999||30||Mild, Moderate, Severe||This study examined the effects of a compensatory intervention compared to a remedial intervention for visual processing deficits.||Y
Other CRT Content: Remedial computer work
• Functional computer-based visual processing tasks: PASAT and the “Matching Accuracy Test” segments of The brain Game program
■ Two data entry tasks
■ One reading task
|Both groups experienced statistically significant improvement on posttests as well as weekly measures; further, 80 percent of both groups used compensatory strategies regardless of the intervention method. Those who used strategies performed better than those who did not.|
|Dou et al. 2006||37||NR||This study evaluated the effects of a computerized, errorless learning-based memory rehabilitation program in China.||Y
Other CRT Content: Nonspecific memory training
• Neurobehavioural Cognitive Status Examination (NCSE)
• Rivermead Behavioural Memory Test-Cantonese version (RBMT)
• Hong Kong List Learning Test (HKLLT)
|Though no significant differences were found between patients in computerassisted memory group (CAMG) and the therapistadministered memory group (TAMG), including a comparison between post training outcome measures and follow-up results, patients in both groups performed better than the Control Group (CG) in the NCSE and RBMT. Compared to the TAMG and CG, the CAMG showed significant improvement in their HKLLT assessment.|
|Kaschel et al. 2002||24||NR||This study aimed to compare two training programs, with the experimental condition based on imagery training and the other based on pragmatic memory training for the control group.||Y
Other CRT Content: Standard of care, with frequency, intensity, individualization of training comparable to treatment group
• l“d2” form of the Concentration Endurance Test
• lEveryday memory test “Appointments”
• lMemory Assessment Clinics Rating Scales (MAC-S, MAC-F)
• lWechsler Memory Scale
|Imagery training significantly improved delayed recall of everyday relevant verbal materials, such as stories and appointments.|
|Ownsworth and McFarland 1999||20||NR||The aim of this study was to investigate the remediation and assessment of everyday memory problems, using the RBMT and WMS-R to measure global memory performance in 20 subjects.||Y
No Content: Control group instructed to use the diary, without treatment
|Compared to the Diary Only (DO) group, the Diary and Self-Instructional Training (DSIT) group made diary entries more consistently, reported fewer memory problems, and made more positive ratings associated with treatment efficacy.|
|Ruff et al. 1994||15||Severe||Two groups received two counterbalanced treatments, attention training compared or memory training, delivered via a computer-based multi-media program called THINKable.||Y
Other CRT Content: Attention and memory training, delivered to both groups in a crossover design
• 2+7 Selective Attn Test
• Behavioral assessments
• Continuous Performance Test
• Corsi Block Learning Test
• Rey Auditory Verbal Learning Test
|On the computer scores after intervention, there was significant improvement in attention, whereas no significant improvement for memory. On neuropsycho-logical measures, the results were mixed: immediate memory improved while delayed memory did not; only one attention measure improved. Self- and other behavioral assessment of memory-based behavior did change after intervention, whereas only observer rating of attention-related behavior reached significance after intervention.|
|Ryan and Ruff 1988||20||Mild- Moderate||Memory retraining with visuospatial and verbal representations, using multiple strategies: associational tasks, chaining, rehearsal, visual imagery, multiple associations, ritualized recall.||Y
Non-CRT Content: Psychosocial activities with no feedback from therapist
• Benton Visual Retention Test (BVRT)
• Rey-Osterrieth Complex Figure Test (CFT)
• Ruff-Light Trail Learning Test (TLT)
• Selective Reminding Test
• Taylor Complex Figure
• Wechsler Memory Scale (WMS), Logical Memory Subtest
|On neuropsychological measures of memory, both groups improved after treatment; however, those who were mildly injured improved significantly more than moderately injured participants who had received strategy training, moderately injured participants in the psychosocial support group, or mildly injured participants in the psychosocial support group.|
|Schmitter- Edgecombe et al. 1995||8||Severe||The purpose of this study was to evaluate the effectiveness of a 9-week memory notebook treatment for patients with known memory deficits.||Y
Non-CRT Content: Supportive group therapy
• Everyday Memory Questionnaire (EMQ)
• Global Severity Index from the Symptom Checklist 90-Revised
• Logical Memory 1 and 2 scales
• Notebook training
• Visual Reproduction 1 and 2 scales
• Weekly learning activities packets (LAPs)
|The notebook training group reported significantly fewer everyday memory failures (EMFs) on a daily checklist measure than the supportive therapy group; at follow-up, this finding was no longer significant. For the labbased memory measures, no significant treatment effects were found at posttreatment or follow-up.|
|Tam and Man 2004||26||NR||The goal of this study was to gain understanding of computer-assisted training effects, using four theoretically different memory retraining strategies, for the treatment of posttraumatic amnesia.||Y
No Content: Unspecified; the study indicated treatment was not provided
• Built-up computer performance records
• Self-efficacy rating scale
|Though all four memory training methods showed nonstatistically significant positive results compared with the no-treatment control, clinical improvement was found in all four methods, and the Feedback group showed significant improvement in self-efficacy compared to the other groups.|
|Thickpenny- Davis and Barker- Collo 2007||14||Moderate- Severe||This study aimed to evaluate the effect of an eight-session structured group format memory rehabilitation program on memory function deficits.||Y
No Content: Waitlist control
• Behavioral indicators of memory impairment checklist
• California Verbal Learning Test (CVLT)
• Integrated Visual and Auditory Continuous Performance Test (IVA-CPT)
• Memory in Everyday Life and Use of Aids and Strategies Questionnaire
• “Memory quiz” (Extent of participants’ knowledge about memory and memory strategies)
• Visual paired associates (VPA)
• Wechsler Memory Scale (WMS) Logical Memory Subtest, revised
|Compared to the control, the experimental group improved in many neuropsychological measures of memory (CVLT long delayed free recall, WMS logical memory delayed recall, and response time on the attention test [CPT]), also showing increased knowledge of memory/memory strategies, increased self/observer (significant other) use of memory aids/strategies and decreased self/observer behaviors indicative of memory impairment. Results were maintained at follow-up, with the exception of response time on the attention test and immediate recall of narratives on the WMS.|
|Watanabe et al. 1998||30||NR||This study compared the effect of the presence or absence of a wall calendar in participants’ hospital room on temporal orientation.||Y
Other CRT Content: Control received nonspecific cognitive rehabilitation, with no calendar in room
|Temporal Orientation Test (TOT) measuring day, date, month, year||Results indicated that the presence of a wall calendar and corresponding orientation therapy had no effect on orientation; only the emergence out of PTA corresponded to orientation.|
|Nonrandomized, Parallel Controlled Group|
|O’Neil- Pirozzi et al. 2010||94||Mild, Moderate, Severe||This study examined the effects of memory training focused on internal strategy use for memory impairment.||Y
No Content: Control received no specific intervention
• Hopkins Verbal Learning Test-Revised (HVLT-R)
• Rivermead Behavioral Memory Test II (RBMT II)
|Memory group intervention participants showed improved memory performance immediately postintervention as well as one month postintervention. Mild and moderately injured participants improved beyond those who were severely injured, even though the severely injured participants still improved beyond severely injured participants who received no treatment.|
|Goldstein et al. 1996||30||NR||This study replicated earlier work that demonstrated the efficacy of two methods of memory training using imagery mnemonics.||Y
Other CRT Content: Parallel groups, with one receiving (partly) computerized intervention, one receiving non-computerized intervention
• Free recall of patient-provided list
• Number of words recalled on a selective reminding task
• Scoring on RIS and FNM tasks
|After treatment, both groups recalled significantly more from examiner-provided lists when compared to pretraining, and the computerized group appeared to improve slightly more. On participant-provided lists, pretreatment to posttreatment recall improved significantly, though the computerized group lost its advantage. On name-face learning task, the computerized group had a clear advantage over the original method group both in learning trials and pre- and posttreatment comparisons; the therapist delivery group did not recall significantly more names after treatment.|
|Pre-Post Single Group|
|Bergman 2000||41||NR||This study examined the effect of orientation to a cognitive orthosis (CO); subjects were progressed on specific tasks until they could perform without assistance (mastery).||N||
• Appointment Scheduling
• Check Writing
• Savings Deposit/Withdrawal
• Telephone Log
|Thirty-six of the 41 participants achieved mastery of four or more activity modules and 36 demonstrated rapid achievement of success on initial assigned tasks.|
|Freeman et al. 1992||12||NR||This study compared the efficacy of memory remediation treatment, which consisted of compensatory and executive training skills delivered over a 2.5-week period, with no treatment.||N||Identification of key ideas in a paragraph read to the subject||There was a statistically significant difference between treatment and control posttest memory scores that favored the treatment group.|
|Gentry et al. 2008||23||Severe||This study examined the efficacy of personal digital assistants (PDAs) as a compensatory cognitive aid, used by subjects for 8 weeks after three to six in-home training sessions conducted by an occupational therapist.||N||
• Canadian Occupational Performance Measure (COPM)
• Craig Handicap Assessment and Rating Technique-Revised (CHART-R)
|Self-ratings of occupational performance and satisfaction with occupational performance (COPM), as well as self-rating of participation (CHART-R), showed statistically significant improvement.|
|Hart et al. 2002||10||Moderate- Severe||This study investigated the usefulness of a portable voice organizer in helping people recall therapy goals and plans previously discussed with their case managers.||N||Number of goals (recorded and nonrecorded) that were remembered||Subjects recalled recorded goals more often than goals that were not recorded; furthermore, recorded goals seemed to be correlated with better awareness or follow-through with therapy objectives.|
|Milders et al. 1998||26||Severe||In this training study, patients were taught strategies to improve the learning of new names and the retrieval of familiar people’s names.||N||
• Digit Span Forwards
• Dutch version of Rey’s auditory verbal learning task
• Famous Faces Naming Test
• Name Learning Test
• Name-Occupation-Town Test
|Performance on two of the three target tasks improved with training compared to controls but performance on the Name Learning Test did not change in either group. Both groups had similar improvements in the two control memory tests.|
|Raskin and Sohlberg 2009||8||Mild, Moderate, Severe||Prospective memory training delivered to eight subjects in a within-subjects crossover design. Subjects were expected to perform a specific task, tested at 2 and 10 minutes, with and without external cues. In addition, half of the subjects were tested on retrospective memory.||N||
• Boston Diagnostic Aphasia Examination, animal naming
• Consonant Trigrams Test
• Controlled Oral Word Association Test
• Randt Memory Test, story recall and picture recognition
• Revised Attention Process Test
• Trail Making Test
|Subjects showed improvement on prospective memory time and tasks after 2 minutes, but not at the longer delay of 10 minutes. On neuropsychological measures immediately post treatment, subjects improved in attention and executive functions. Generalization to everyday memory performance also improved, as measured by a memory questionnaire and diaries. Maintenance of prospective memory improvements was demonstrated at one year post-treatment. None of the subjects showed improvement for retrospective memory drills.|
|Benedict and Wechsler 1992||2||Mild and Moderate||The study examined the efficacy of long-term memory retraining in two adults using two training strategies: method of loci (MOL) and the PQRST verbal strategy.||N||Scores on word-list recall and paragraph recall||The moderately impaired participant’s memory for word lists benefitted from the MOL training, but paragraph learning did not benefit from PQRST training. The severely impaired participant’s performance was highly variable throughout, resulting in little change in recall form word lists or paragraphs.|
|Ehlhardt et al. 2005||4||Severe||The TEACH-M program, which facilitates learning and retention of multistep procedures, was administered to assess participants’ ability to learn multi-step procedures utilizing specific cognitive rehabilitation principles.||N||
• Number of correct e-mail steps in sequence (out of 7)
• Number of correct steps regardless of sequence
• Number of training sessions needed for mastery
|All four participants replicated treatment effects immediately posttreatment and 30 days thereafter; generalization and social validity data also supported the effectiveness of TEACH-M.|
|Hux et al. 2000||7||NR||This study examined the efficacy of internal memory strategies, specifically mnemonics and visual imagery, to improve face-name recall in seven subjects ranging from 2 to 26 years postinjury.||N||
• Number of faces correctly identified
• Number of training sessions needed to reliably identify subset of faces
|Sessions held daily and twice a week were found to be more effective than those held five times a day. Mnemonics and visual imagery strategies were effective for four of the participants, irrespective of session frequency.|
|Manasse et al. 2005||5||Severe||The study examined three cueing strategies (name restating, phonemic cueing, and visual imagery) in a real-world context to increase mastery and use of face-name associations in five chronic participants.||N||Number of names mastered (name use, name knowledge seen in each cueing condition)||All participants improved in name-face recall after intervention regardless of the kind of cuing and four of five participants demonstrated more spontaneous use (effectiveness) of therapists’ names when given the same number of opportunities.|
|Raskin and Sohlberg 1996||2||NR||This study was an investigation of prospective memory training using two different types of intervention: prospective and repetitive memory drill.||N||
• Performance on criterion standard neuropsychological measures
• Performance on prospective memory screening test (Perspective Memory Screening, or PROMS)
• Probes of everyday prospective memory
|Memory for future actions improved more after prospective memory training than after repetitive drill, although generalization to real world remembering was variable across participants and type of training. Both participants validated their preference for prospective memory training during interviews.|
|Zencius et al. 1991||4||NR||In this study, the usefulness of memory notebook training for completing homework assignments was examined for four adults, who were also receiving interdisciplinary rehabilitation services.||N||
Number of homework assignment components completed correctly:
• Meeting a specified person at a certain place and time
• Turning in a completed written homework assignment
|After notebook training, three of the four participants improved in completing the number of components to the homework assignments.|
Benedict, R. H., and F. S. Wechsler. 1992. Evaluation of memory retraining in patients with traumatic brain injury: Two single-case experimental designs. Journal of Head Trauma Rehabilitation 7(4):83–92.
Berg, I. J., M. Koning-Haanstra, and B. G. Deelman. 1991. Long-term effects of memory rehabilitation: A controlled study. Neuropsychological Rehabilitation 1(2):97–111.
Bergman, M. M. 2000. Successful mastery with a cognitive orthotic in people with traumatic brain injury. Applied Neuropsychology 7(2):76–82.
Bergquist, T., C. Gehl, J. Mandrekar, S. Lepore, S. Hanna, A. Osten, and W. Beaulieu. 2009. The effect of Internet-based cognitive rehabilitation in persons with memory impairments after severe traumatic brain injury. Brain Injury 23(10):790–799.
Bergquist, T., K. Thompson, C. Gehl, and J. M. Pineda. 2010. Satisfaction ratings after receiving Internet-based cognitive rehabilitation in persons with memory impairments after severe acquired brain injury. Telemedicine and eHealth 16(4):417–423.
Bourgeois, M. S., K. Lenius, L. Turkstra, and C. Camp. 2007. The effects of cognitive teletherapy on reported everyday memory behaviours of persons with chronic traumatic brain injury. Brain Injury 21(12):1245–1257.
Dirette, D. K., J. Hinojosa, and G. J. Carnevale. 1999. Comparison of remedial and compensatory interventions for adults with acquired brain injuries. Journal of Head Trauma Rehabilitation 14(6):595–601.
Dou, Z. L. D. W. K. Man, H. N. Ou, J. L. Zheng, and S. F. Tam. 2006. Computerized errorless learning-based memory rehabilitation for Chinese patients with brain injury: A preliminary quasi-exerimental clincial design study. Brain Injury 20(3):219–225.
Ehlhardt, L. A., M. M. Sohlberg, A. Glang, and R. Albin. 2005. TEACH-M: A pilot study evaluating an instructional sequence for persons with impaired memory and executive functions. Brain Injury 19(8):569–583.
Freeman, M. R., W. Mittenberg, M. Dicowden, and M. Bat-Ami. 1992. Executive and compensatory memory retraining in traumatic brain injury. Brain Injury 6(1):65–70.
Gentry, T., J. Wallace, C. Kvarfordt, and K. B. Lynch. 2008. Personal digital assistants as cognitive aids for individuals with severe traumatic brain injury: A community-based trial. Brain Injury 22 (1):19–24.
Goldstein, G., M. McCue, S. M. Turner, C. Spanier, J. A. Malec, and C. Shelly. 1988. An efficacy study of memory training for patients with closed head injury. Clinical Neuropsychologist (2):251–56.
Goldstein, G., S. R. Beers, S. Longmore, and M. McCue. 1996. Efficacy of memory training: A technological extension and replication. Clinical Neuropsychologist 10(1):66–72.
Hart, T., K. Hawkey, and J. Whyte. 2002. Use of a portable voice organizer to remember therapy goals in traumatic brain injury rehabilitation: A within-subjects trial. Journal of Head Trauma Rehabilitation 17(6):556–570.
Hux, K., N. Manasse, S. Wright, and J. Snell. 2000. Effect of training frequency on face-name recall by adults with traumatic brain injury. Brain Injury 14 (10):907–920.
Kaschel, R., S. Della Sala, A. Cantagallo, A. Fahlbock, R. Laaksonen, and M. Kazen. 2002. Imagery mnemonics for the rehabilitation of memory: A randomised group controlled trial. Neuropsychological Rehabilitation 12(2):127–153.
Manasse, N. J., K. Hux, and J. Snell. 2005. Teaching face-name associations to survivors of traumatic brain injury: A sequential treatment approach. Brain Injury 19(8):633–641.
Milders, M. V., I. J. Berg, and B. G. Deelman. 1995. Four-year follow-up of a controlled memory training study in closed head injured patients. Neuropsychological Rehabilitation 5(3):223–238.
Milders, M., B. Deelman, and I. Berg. 1998. Rehabilitation of memory for people’s names. Memory 6(1):21–36.
O’Neil-Pirozzi, T. M., G. E. Strangman, R. Goldstein, D. I. Katz, C. R. Savage, K. Kelkar, C. Supelana, D. Burke, S. L. Rauch, and M. B. Glenn. 2010. A controlled treatment study of internal memory strategies (I-MEMS) following traumatic brain injury. Journal of Head Trauma Rehabilitation 25(1):43–51.
Ownsworth, T. L., and K. McFarland. 1999. Memory remediation in long-term acquired brain injury: Two approaches in diary training. Brain Injury 13(8):605–626.
Raskin, S. A., and M. M. Sohlberg. 1996. The efficacy of prospective memory training in two adults with brain injury. Journal of Head Trauma Rehabilitation 11(3):32–51.
———. 2009. Prospective memory intervention: A review and evaluation of a pilot restorative intervention. Brain Impairment 11(1):76–86.
Ruff, R. M., R. Mahaffey, J. Engel, C. Farrow, D. Cox, and P. Karzmark. 1994. Efficacy study of THINKable in the attention and memory retraining of traumatically head-injured patients. Brain Injury 8(1):3–14.
Ryan, T. V., and R. M. Ruff 1988. The efficacy of structured memory retraining in a group comparison of head trauma patients. Archives of Clinical Neuropsychology 3(2):165–179.
Schmitter-Edgecombe, M., J. F. Fahy, J. P. Whelan, and C. J. Long. 1995. Memory remediation after severe closed head injury: Notebook training versus supportive therapy. Journal of Consulting & Clinical Psychology 63(3):484–489.
Tam, S., and W. Man. 2004. Evaluating computer-assisted memory retraining programmes for people with post-head injury amnesia. Brain Injury 18(5):461–470.
Thickpenny-Davis, K. L., and S. L. Barker-Collo. 2007. Evaluation of a structured group format memory rehabilitation program for adults following brain injury. Journal of Head Trauma Rehabilitation 22(5):303–313.
Watanabe, T. K., K. L. Black, R. D. Zafonte, S. R. Millis, and N. R. Mann. 1998. Do calendars enhance posttraumatic temporal orientation? A pilot study. Brain Injury 12(1):81–85.
Zencius, A., M. D. Wesolowski, T. Krankowski, and W. H. Burke. 1991. Memory notebook training with traumatically brain-injured clients. Brain Injury 5(3):321–325.