To evaluate the efficacy of this particular system of cognitive rehabilitation, we have carried out a comprehensive controlled trial that took over 3 years to complete. The aim of the study was to compare two groups of the very severely head-injured (defined as a P.T.A. of greater than 7 days) that differ only in that one group had access to the computer system. As is our usual practice, the patients in the 'treatment' group were loaned a computer system for the duration of their participation in the trial. The primary hypothesis was that the 'treatment' group would show a faster and greater cognitive recovery than the untreated control group. There were 25 patients in each group. There was also a third group of 25 undamaged individuals to provide baseline data for some of the assessments for which there were insufficient norms.

The study involved serial assessments over the first year post injury. For both groups of the head injured there were 5 assessment points. These were as soon as possible after emergence from P.T.A. and then at 3 monthly intervals. The normal control group was assessed at 3 points, 6 months apart.

The outcome measures were extensive and can be subdivided into clinical, behavioural and electrophysiological.

Clinical Assessment Measures

 Figure 1 - The clinical assessment checklist of signs and symptoms that was used in this study. This was performed at each of the 5 assessment points for both groups of head-injured.

Fig. 1

Figure 2 - The disability assessment scale used in this study to evaluate the level of 'real-life' functioning. This was administered at each of the 5 assessment points for both groups of head-injured.

Fig. 2

 The clinical assessment was carried out by one of the two neuropsychiatrists involved in the project at each of the 5 time points. In addition to interview the assessment included a checklist of signs and symptoms (figure 16) and a rating on a disability scale. This scale (DAS) was a modification of the Rappaport Disability Assessment Scale and was designed to evaluate the degree of disability produced by the head injury in the wider context of social and personal functioning (figure 2).

Neuropsychological Assessment

A fairly comprehensive neuropsychological assessment was performed at each of the assessment points (5 for both groups of head-injured and 3 for the normal control group). The assessment consisted of the following:

  Frenchay Orientation Test (FOAT) - To confirm end of PTA and assess severity of any gross memory impairments.

Neurophysiological Assessment

 A small and highly selective neurophysiological assessment was performed at each of the assessment points (5 for both groups of head-injured and 3 for the normal control group). The assessment consisted of the following:

Sensory

   (a)    Brainstem Auditory Evoked Responses

   (b)    'Checkerboard' Visual Evoked Responses

Cognitive

 7 condition reaction time task designed as a mental chronometer using identical stimuli ('L' and 'R') in all conditions:

   Condition 1 - Passive

      Just watch

   Condition 2 - Simple RT

     Press to all events

   Condition 3 - Spatial

     Press RIGHT to RVF stimuli

      Press LEFT to LVF stimuli

   Condition 4 - Spatial Crossed

     Press RIGHT to LVF stimuli

      Press LEFT to RVF stimuli

   Condition 5 - Verbal

      Press RIGHT to the letter 'R'

      Press LEFT to the letter 'L'

   Condition 6 - Verbal Crossed

      Press RIGHT to the letter 'L'

      Press LEFT to the letter 'R'

   Condition 7 - Combination (3 response condition)

      Press RIGHT to the letter 'R' in the RVF

      Press LEFT to the letter 'L' in the LVF

      Press CENTRE to the letter 'R' in the LVF

      Press CENTRE to the letter 'L' in the RVF

Results and Conclusions

 It is clear that a study of this size will produce an enormous number of results. Instead of detailing individual findings we will summarize the major findings and conclusions. An examination of figures 3, 4, 5 and 6 reveals the major findings of the study which are that:

Figure 3 -- Some of the results from the neuropsychological assessments. Each graph illustrates the data for one measure as indicated above the graph. In each graph the 3 subject types are plotted on the Z-axis in the order of normal control, head injury control and head injury treatment (front-to-back). On the X-axis are the assessment points - on entry to the study and then at 3, 6, 9 and 12 months later.
 

1) Despite rigorous randomization procedures for assigning the head-injured patients to 1 of the 2 groups, there was a significant difference between the groups in terms of the degree of disability produced by the head injury. The treatment group was substantially more disabled than the supposedly matched control group. This difference was present from the outset and carried on through the year.

2) Both groups of head injury patients showed marked recovery over the 1st year after injury as evidenced by the increasing proximity to the normal control values.

3) As a group the patients who were using the cognitive rehabilitation system did not generally recover faster or to a greater level than their control group. Therefore the primary hypothesis was not supported.

 Despite the clear rejection of the primary hypothesis it is difficult to accept its full implication at face value because of the known confounding factors in the study design and execution. There are 3 major confounds. The first has already been mentioned. This is the initial significant difference between the groups in terms of the degree of disability produced by the head injury. Secondly there was a tremendous variability amongst the treatment group in the degree to which they made use of the computer system. Some patients worked on the rehabilitation material as directed - some used it more than specified - others barely turned the computer on. By the way of an analogy, if this had been a drug trial the treatment group would consist of a mixture of subjects - some of whom would have had the trial dose, some would have had more than the trial dose and others would only have held the bottle. Clearly this is not sensible. The third major confound is essentially unavoidable in this type of study. The control HI group - for ethical and humanitarian reasons - could not be denied access to whatever rehabilitation service that was required. Some went to specialized rehabilitation centres, some had speech therapy, some had occupational therapy and some became regular attendees at something called Headway House ( a non-professional respite and rehabilitation facility). Unfortunately - from the point of view of the study - most of the available rehabilitation facilities use the Burden cognitive rehabilitation software to some extent. Therefore the control group was probably no better a 'control' group than the others were a 'treatment' group.

Figure 4 -- Some of the results from the neuropsychological assessments. Each graph illustrates the data for one measure as indicated above the graph. In each graph the 3 subject types are plotted on the Z-axis in the order of normal control, head injury control and head injury treatment (front-to-back). On the X-axis are the assessment points - on entry to the study and then at 3, 6, 9 and 12 months later.

Figure 5 - This figure illustrates a major discrepancy between reaction time (RT) and P300 latency in 7 different reaction time tasks that varied in complexity. These data are from the normal control group. The P300 latency (triangles) increases only over the first 5 tasks and then plateaus. The RT by comparison linearly and dramatically increases over the full set of tasks.
 

Fig. 5

Figure 6 - This figure illustrates the difference in RT between the two patient groups and the normal control group for each of the conditions (X-axis) and each of the 5 sessions (Z-axis). It is apparent that 1) the treatment group was initially much more impaired than the control HI group, 2) both groups improved dramatically over the year and 3) at the end of the year the treatment group was actually performing better than both groups of controls.

Fig. 6