Evaluating the efficacy of remediation or rehabilitation of attention deficits following a brain injury is complicated by a number of factors. First, there is no consensus regarding a definition of attention. Is it a general construct or does it reflect more specific sub-components or systems of functioning (e.g., sustained, divided, focused, selective, vigilance, speed of information processing, etc)? Second, different researchers and clinicians will report using the same or similar tests to measure different aspects of attention. Third, a study may use the same outcome measures repeatedly, thereby confounding practice and treatment effects (e.g., PASAT exposure to the test). Finally, studies may not consider and account for the rate of spontaneous recovery following brain injury (i.e. would participants naturally show recovery of function in the absence of treatment?).
Comparing the efficacy of various remediation efforts is also complicated by cross-study variability in treatment duration (e.g. from 30 minutes once a day for 5 days to 5 hours, every day for 6 weeks). Severity of injury and time since injury may also fluctuate from study to study.Over the past several years Cicerone et al. (2000; 2005; 2011) reviewed a series of studies investigating the effectiveness of attentional retraining interventions during rehabilitation following traumatic brain injury and stroke.
Cicerone et al. (2005) recommended strategy training for persons with TBI for improving deficits of attention. It should be noted, however, that there was insufficient evidence to distinguish the effectiveness of specific attention training during acute stage rehabilitation from improvements made from spontaneous recovery or from more general cognitive interventions (Cicerone et al., 2005).
Drill & Practice
The following studies examined the influence of “drill & practice” exercises (either computerized and/or paper-and-pencil) on attentional functioning.
Novack et al. (1996) performed an RCT of severe TBI participants in acute rehabilitation and found no difference between two treatment groups “a focused group consisting of sequential, hierarchical interventions directed at specific attention mechanisms and an unstructured intervention consisting of nonsequential, nonhierarchical activities requiring memory or reasoning abilities”. No differences were found between groups in attentional, functional and/or cognitive skills assessed, although post-intervention improvement of all subjects was demonstrated as compared to pre-intervention. It should be noted that this could reflect spontaneous recovery, as a “no-treatment”, control group was not included. Similarly, Park et al. (1999) examined whether “attention processing training (APT)” had a beneficial effect on attention measures (PASAT, Consonant Trigrams) in a severe TBI group (tested pre and post training approximately 7 months apart). They compared their results to a “convenience” sample of controls, given the same measures one-week apart without training. Results suggested that APT did not have a significantly beneficial effect as performance improved on all measures across both groups (indicating practice effects and possibly spontaneous recovery).
There is Level 2 evidence to suggest that specific structured training programs designed to improve attention are ineffective or at best equivocal in their effects on attention.
Specific structured training programs designed to improve attention are ineffective.
The following studies examined the effect of “Dual-task” training on speed of processing.
In a recent RCT, Couillet et al. (2010) randomly divided 12 participants into either a non- specific cognitive group that did not tap on divided attention or working memory or an experimental rehabilitation training program with specific dual task training. Six individuals were assigned to either: the control/experimental group (AB) or the experimental/control (BA) group. Prior to completing the dual task, all were asked to complete the single tasks until they were able to do so without difficulty. To measure changes in divided attention, the divided attention subtest of the TAP was used. At the 6th week assessment period the BA group showed significant improvement (p<0.01) in reaction times and omissions compared to the AB group.
Fasotti et al. (2000) randomly assigned 22 severe TBI patients undergoing rehabilitation to either Time Pressure Management (TPM) training (treatment group, N=12) or to a concentration group (control, N=10). Patients were pre-selected for inclusion in this study if they demonstrated slowed processing speed (as measured by 3 tests). TPM consists of a series of cognitive strategies to compensate for reduced processing speed. There are 3 main stages: increased self-awareness of errors and deficits, acceptance and acquisition of TPM cognitive strategies (4 steps), and strategy application and maintenance in increasingly more demanding/distracting situations. The concentration-training group consisted of 4 generic suggestions (e.g., focus, don’t get distracted, etc). Groups were compared on pre-training, post-training and follow-up on task performance (information from a video recording) and results indicated that there were no significant differences between groups (both improved task performance), although the TPM made more gains and appeared to generalize to positive effects other measures.
In a study conducted by Hasegawa and Hoshiyama (2009) patients were initially presented with visual stimuli of Japanese characters. Participants were shown one character for 3 seconds, and then 4 appeared on screen for 3 seconds. Patients were asked to identify on a piece of paper which characters (4) and the position they had on screen. Participants were expected to retain the memory for 7 seconds before they recalled it. During the experimental sessions participants were presented orally with 4 numbers which they were asked to repeat before writing down the symbols they saw on the screen. During the visual-spatial task, flicker stimulation (1 or 2 dots) randomly appeared on the screen. Participants were asked to follow the flickering lights. Following this, each participant wrote down the characters presented on the screen. The first and last sessions were control sessions and the two middle sessions were the experimental sessions. In the control and visuo-spatial task sessions, the correct answer rate was higher for the character being presented than for locations (p<0.05), in the phonological task session, the location was higher than that for character (p<0.05). When looking at the repetition effects, during the control session (session 1 and 4), the correct answer rate was higher. Results of the memory tests and the results on the various scales used indicated the RBMT and the social activity scores were correlated (r=.53, p<0.02). The dual task cost for both factors in the phonological and visuo-spatial session were negatively correlated with the social and personal activity score (r = -.43, p<0.05; r = -.42, p<.03) respectively. Overall the TBI groups showed memory disturbances in the simple memory test and RBMT; correct answer rates in the control and dual task session were all lower in the patient group compared to the control group; the correct answer rates for the TBI patients in the dual task experiment were lower than in the control group. ADLs were all highly correlated with the dual task and memory tests, but dual tasks and memory tests were not correlated.
Dockree et al. (2006) in a case control study compared the performance of individuals who had sustained a brain trauma to those that had not. All participants participated in two tasks of sustained attention: a single task of sustained attention (SART) and a dual task of sustained attention (DART). Results indicate that the TBI patients made significantly more errors (p<0.0001) than their non-TBI counterparts on the dual task sustained attention compared to the single task sustained attention.
Stablum et al. (2000) compared two patient groups (those suffering from a closed head injury (CHI) and those who experienced an aneurysm of the anterior communicating artery (ACoA) and matched controls on performance on a dual-task paradigm and neuropsychological tests. Results suggested that CHI and ACoA patients had significant difficulty compared to matched controls on dual-task reaction time measure and specific measures of executive functioning (e.g., WCST and PASAT) and compared to their own performance on a single-task reaction time measures (non-significant differences between groups on this latter measure). With training, however, performance improved to levels similar to matched control subjects and was maintained at follow-up 3 months later. However, what remains unclear is whether training generalized to functional gains or whether it remained specific to this specific dual-task.
In a case series by Foley et al. (2010) they found that level of injury severity as measured by the GCS or PTA did not play a role in who performed poorly on the dual task assignment given to participants. They found that approximately 26% of the study population performed below the cut-off for normal performance. The authors concluded that fewer than expected had deficits in attentional control.
There is Level 2 evidence that dual task training has a positive effect on divided attention.
There is Level 2 evidence that dual-task training is effective on the speed of processing.
There is Level 3 evidence that individuals with a TBI perform poorly on dual task activities due to their inability to maintain a measure of sustained attention.
Dual-task training assists individuals in dealing with dual task situations rapidly and accurately.
Dual-task training on speed of processing is effective.
Individuals who sustained a TBI tend to perform more poorly on dual task skills.
It has been noted that those who have sustained a severe traumatic brain injury tend to have a slower reaction time than those who have not (Stuss et al., 1989; Azouvi et al., 2004).
Two studies were found looking at the reaction times of individuals with a TBI and comparing these times with individuals who had not sustained a TBI (Stuss et al., 1989; Azouvi et al., 2004). Both case-control studies demonstrated that those with a TBI were found to have slower reaction times. Struss et al. (1989), after conducting three studies, found (in study 3), those with a brain injury were significantly slower than the group without an injury (p<0.01) on the simple reaction time test. Results from the multiple choice reaction time (MCRT), indicated that regardless of the severity of the injury (mild concussion to severe TBI) those with a TBI had a slower reaction time than the controls. Azouvi et al. (2004) found that those with a brain injury were slower than those without when looking at the results of tasks the groups were asked to perform. Results of the visual analogue scale also indicated that mental effort was higher for those with a TBI than for the controls. The results of this study confirmed what previous studies had found: those with a TBI have greater difficulty when dealing with 2 simultaneous tasks (Azouvi et al., 2004).
There is Level 3 evidence that reaction times of those with a TBI are slower than the reaction times of those without.
Reaction times are slower in post-TBI individuals.