Effectiveness of a multidisciplinary rehabilitation program for persons with acquired brain injury and executive dysfunction.

PURPOSE
The purpose of this study is to explore the effects of a multidisciplinary acquired brain injury rehabilitation out-patient program (5 d/week for 7 weeks) on improvements to participants' activity and participation outcomes related to meal preparation and to determine whether gains are maintained at 3 and 6 months post program.


METHODS
A single case experimental design with repeated measures pre- and post-intervention with 7 adult participants with ABI and executive dysfunction (4 females, mean age 38 ± 10.1 years) was used.


RESULTS
A strong improvement effect between pre and post phases was found for number of errors on the Cooking Task for 6/7 participants; four participants showed significant improvement immediately after the program and at 3 and 6 months post. Six out of seven participants improved significantly on the Instrumental Activities of Daily Living Profile and four participants improved between the post and 6 month follow-up. Four out of seven participants showed significantly improved Life Habits scores pre- versus post-program.


CONCLUSIONS
Significant improvements were observed in activity and participation outcomes related to preparing a meal in adults with ABI and executive dysfunction who participated in a 7-week multidisciplinary rehabilitation out-patient program. Treatment gains were maintained for the majority of participants at 3 and 6 months following the program. Implication of Rehabilitation A 7-week multidisciplinary rehabilitation out-patient program appears to improve activities and participation; the effects are sustainable after 6 months. A detailed description of the therapeutic interventions provided during the cooking activity should help clinicians better understand what specific functions are solicited or required during a particular activity. Knowledge from this study may help guide clinicians in their work within this complex area of rehabilitation.


Introduction
Acquired brain injury (ABI) can result in neurological, cognitive and behavioral impairments. Cognitive and behavioral impairments include but are not limited to memory and attention difficulties, slowing of information processing and executive dysfunction often responsible for severe and longstanding disabilities in daily life activities. Persons with executive dysfunction can demonstrate difficulties in taking initiatives and being in control, in changing organizational strategies, conceptualizing, or planning, leading to activity limitations (i.e., difficulties in executing activities) and participation restrictions (i.e., problems in involvement in life situations). In fact, combined with changes in behavior and personality, many persons with ABI remain dependent in accomplishing complex daily activities at home and in their community [1][2][3][4]. This dependency justifies the need for rehabilitation services with the ultimate goal to optimize a person's executive function and subsequently his/her return to community life at home, school, and work or with respect to leisure.
Some ABI rehabilitation interventions focus on improving body functions as defined by the WHO [5] and reducing impairments, while others may use a more holistic approach to improve activity and participation levels while considering the individual's environment and personal factors. Interventions focusing on body function improvements (such as EF) are relatively well described (e.g., working memory [6,7], problem solving [8,9]) and treatments have been shown to be effective in improving body function with the assumption that improving the functions are likely to be transferred to activity improvements. For example, Vallat et al. [6] demonstrated three 1-h working memory training sessions per week for 6 months could significantly improve the storage and processing components of verbal working memory in one person with stroke. Their therapy improved the subject's ability to carry out activities such as have a phone conversation or shop for groceries assessed using a verbal communication questionnaire (Echelle de communication verbale de Bordeaux: ECVB [10]).
In contrast, more holistic interventions (i.e., multidisciplinary programs) focusing on improving activities are relatively poorly described; they are treated as a black box, many of their components are unknown and the role of the therapist is undefined. For example, a home rehabilitation program for adults with stroke focusing on cognitive functioning using mixed training treatment of cognitive remediation therapy, story retelling, cognitive enhancing games, and aerobic exercise [11], is underdeveloped thus limiting its reproducibility by others. Moreover, compared with specific treatment interventions, fewer studies have examined the effectiveness of holistic rehabilitation programs [12]. Moreover, no study appears to have evaluated the impact of interventions targeting the improvement of executive dysfunction in patients' activity and participation as defined by the International Classification of Function (ICF) [5] using reliable and valid standardized measures.
In this context, we chose to evaluate the effectiveness of an intensive client-centered and multidisciplinary out-patient rehabilitation program for ABI patients developed in 1988 at the la Piti e-Salpêtri ere Hospital in Paris, France, and to document this program using the ICF. This program is offered to groups of four persons at a time who have an ABI and are living at home. It aims to optimize the ability of persons with ABI and executive dysfunction to execute activities and improve their participation using individual and group interventions involving ecologically valid activities outside of the hospital. With respect to the ICF, these activities include grocery shopping (ICF code d6200), meal preparation (d630), cleaning cooking area and utensils (d6401), taking public transport (d470), and visiting a public space (d9202). This program has never been described in detail nor has its effectiveness ever been formally investigated.
In this first article, we focus on the effects of the cooking activity within the program on participants' ability to prepare a meal (d630). Activities relating to domestic/household life (d630-649) are important components of the program and preparing a meal is an important basic need of persons living alone. Being able to prepare/cook a meal for one's family may be an important personal goal of patients with ABI enabling them to resume this role within their family. Moreover, clinicians need to know whether a person with executive dysfunction can perform this complex activity independently and safely. Specifically, our principal objective was to determine whether the rehabilitation program leads to improvements in activity and participation outcomes related to preparing a meal, and whether treatment gains are maintained at 3 and 6 months following program completion. We expected improvements in participants' activity and performance levels and hypothesized outcomes related to body function would remain stable because program participants are typically seen more than 9 months post-ABI [13], and because the program does not specifically target improvements in body function. Our secondary objectives were to explore the effects of the program on other situational activities (e.g., cleaning up after meals, obtaining information) and expected improvements in trained activities as opposed to untrained ones (e.g., preparing a budget).

Materials and methods
Design A single case experimental design (SCED) with repeated measures pre-and post-intervention was used because of the heterogeneity of the program participants (i.e., varied health conditions and disabilities) [14][15][16][17]. Moreover, the clinical environment within which this research was conducted (i.e., four new participants every two months) did not lend itself well to large group analyses.

Setting
This study was conducted within the Physical Medicine Rehabilitation Department of the Piti e-Salpêtri ere Hospital (Paris, France) where many of the interventions take place (e.g., gymnasium, kitchen) and in the surrounding community (e.g., Paris museums, and grocery stores and within the public transport system). At the time of the study, the program team was composed of eight health professionals: one clinical psychologist, two occupational therapists (OT), one physiotherapist, one speech therapist, one nurse, one attendant, and one physiatrist. A neuropsychologist evaluates the cognitive functions of potential participants before entering the program to help identify where interventions in occupational therapy, speech therapy and physiotherapy should focus, and a social worker can intervene when social or financial issues arise. The intensive out-patient day program is offered to four persons with ABI at a time from 9 a.m. to 4 p.m. each weekday over seven weeks. A new program begins with new participants every 8 weeks. The complete description of the Cooking Activity Intervention over the 7-week program including the overall treatment goals for the group and participants' individual goals related to this activity are presented in the supplement file.

Subjects
Study subjects were persons with ABI receiving the program described above. Participants are adults with either a traumatic brain injury or stroke, who have lived at home at least 2 months before the beginning the program and have difficulties managing multi-step instrumental activities of daily living (i.e., have executive dysfunction). They must also have received traditional outpatient multidisciplinary services from the same hospital (e.g., speech therapy, etc.) at least twice a week for 1 month before the program. This way the team members get to know relatively well the participants before entering the program and are able to determine with each participant, his or her personal goal(s) while in the program. Participants must (1) be medically stable and have no severe psychiatric illness, (2) show rehabilitation potentialbe able to fully participate in the program (i.e., score at least 1 on the Boston Diagnostic Aphasia Examination, BDAE [18]), (3) not have important spatial neglect preventing them to find their way within the hospital, (4) be able to walk safely 800 m (evaluated by physiotherapists), (5) not suffer from fatigue serious enough to limit their ability to participate in program activities, and (6) be motivated to participate in the program by having personal goals linked to improved participation (e.g., learn to use public transport more safely, prepare family meals). When they enter the program, some participants will have recovered basic functions in the management of daily living tasks (e.g., washing, dressing), and others may require physical assistance or need verbal reminders to initiate tasks. All study participants provided written informed consent and this study was approved by the hospital's ethics committee.

Intervention under study
The main aim of the program under study was to encourage participation at home and in the community. Figure 1 illustrates a typical week's schedule for one participant (subject FA). Activities are based on the experience of the therapists, evidence-based practice and/or theoretical models.
The therapeutic cooking activity (provided on Tuesday mornings) was based on the models of EF proposed by Lezak [19,20] and Shallice [21,22]. For example, each participant chooses the recipe(s) for the part of the meal the group prepares together and shares with invited guests and two OTs that day, the choices corresponding to the volition component of an activity as proposed by Lezak [20]. Each participant then works with a therapist to decide how best to plan the activity while focusing on the skills they need to develop (e.g., checking off steps in the recipe when completed). Following the meal, time is taken with each participant to discuss how they felt the activity went and what might need to be worked on the next week. A complete description of the Cooking Activity (i.e., "meal preparing" d630) is in the supplement material.
The Cooking Activity has a general structure (and template) based on theoretical models of EF (i.e., Lezak [20] & Shallice [21,23,24]). It integrates cognitive interventions based on principles of holistic neuropsychological rehabilitation [25,26] and of errorless learning, problem solving and metacognitive rehabilitation [27,28] while considering the individual's personal goals. The OT interventions target participants' individual abilities, and in many cases, propose compensatory aides or accommodations like internal strategy (e.g., visual imagery of the steps of the recipe) and external memory compensation (e.g., program an alarm) [28][29][30][31][32] to improve activity performance (see the supplement material). They also aim to help participants better control potentially inappropriate social behaviors, while considering environmental facilitators and obstacles.

Procedure
We attempted to recruit all participants of two groups of four people, enrolled in the program from January to February 2010 (Group 1) and from November to December 2010 (Group 2). Eight patients accepted, but one female in Group 2 developed medical complications just before starting the program and was replaced by another woman ineligible to participate in this study because she joined the group too late to be assessed preprogram.
Each participant was evaluated every 2 weeks during a 6-week period (T1, T2, and T3) before the intervention began to describe the subjects and to ensure a stable baseline. During the treatment phase, no assessments were carried out to ensure the program was not modified in any way. In the week immediately following the treatment phase, a second series of measures were collected at (T4) and again at 3 months (T5) and 6 months (T6) post-treatment. Each subject participated in this study for 9 months (i.e., 6 weeks before the program, 7 weeks of treatment and 6 months after the program).

Outcome measures
The capacity (or ability) to cook or prepare a meal is the primary outcome, which was assessed using two measures (the Cooking Task and the Instrumental Activities of Daily Living Profile (IADL Profile)). Subjects' performance (i.e., participation) was assessed using the IADL Profile and the Assessment of Life Habits, both chosen because of their strong psychometric properties for the patient populations under study.
The Cooking Task [33][34][35] assesses EF in ecological situations in an OT kitchen. Subjects must make a chocolate cake with a recipe and an omelet. The Cooking Task measures the success of the activity, the execution time, and the number of errors made during the activity. For reasons of feasibility, the Cooking Task was administered and evaluated by an OT from the program, 15% of the evaluations were also assessed by a second OT independent from the program and their inter-rater agreement was 98%. The concurrent validity of the Cooking Task showed that the total number of errors on the Cooking Task is significantly correlated with two dysexecutive questionnaires: the Behavioural Scale (rho ¼ 0.77; p ¼ 0.02) and the Dysexecutive Questionnaire (DEX) (rho ¼ 0.74; p < 0.03) [34]. Moreover, the Cooking Task showed good internal consistency (Cronbach alpha ¼ 0.74) but a low testretest reliability in a group of 11 persons with ABI (ICC ¼ 0.36) [35].
The IADL Profile determines whether a subject's main difficulties pertain to goal formation, planning, carrying out the task and/or attaining the initial task goal and thus serves to assess the impact of executive dysfunction on independence [36]. Eight tasks are performed in the home and community environments (e.g., "prepare a hot meal for guests", "go to the grocery store", and "make a budget"). Subjects must prepare a meal in their homes for themselves and two guests and the evaluators do not intervene during the activities, unless for safety reasons. Recipes used in these evaluations were chosen by participants each time and were different from those used during the 7-week therapeutic cooking activity. The IADL Profile was administered by a trained OT from the program and all evaluations were filmed by a research assistant. These filmed performances (i.e., n ¼ 28 films) were scored in random order. Two OTs, trained in the evaluation but who were not aware of the purpose of the research project (and blind to the assessment), objectively evaluated subjects' performance to optimize the internal validity of the study.
The Assessment of Life Habits (LIFE H) [37,38] can be used to determine subjects' level of participation in 77 life habits spread across 12 domains or life habit categories. The LIFE-H self-report questionnaire assesses accomplishments rated across the degree of difficulty experienced and the type of assistance required (help, technical assistance, and physical arrangements). Participants were interviewed by an OT from the program to help control for the subjects' cognitive disorders (e.g., attention disorder) during testing. In this study, only the nutrition domain (including "preparing a meal") was assessed. The nutrition domain of the Life-H has demonstrated test-retest reliability (ICC ¼ 0.72) for a stroke population [38]; the reliability of the tool with the traumatic brain injury (TBI) population however, seems not to have been tested.

Secondary measures
The secondary measures were evaluated in the most part by independent assessors from the rehabilitation program who were therapists trained in the use of the measurement tools. However, the assessment of mood was assessed by the program psychologist for reasons of clinical feasibility.

Measures of body and structure functions
Four measures of body function were assessed.
Functions of position and positional sense; functions of balance of the body and movement (b235). Balance was measured with the Berg Balance Scale (BBS) [39][40][41][42], which assesses subjects' ability to perform 14 static and dynamic balance activities of varying difficulty, several of which include time and distance requirements. Item-level scores range from 0 to 4 determined by the ability to perform the assessed activity, and the maximum score is 56. Studies have shown test-retest reliability to be excellent among persons with stroke (i.e., Intra-Class Coefficient (ICC) ¼ 0.98 [41] and persons with TBI (i.e., ICC¼ 0.986 [42]. Neuromusculoskeletal and movement-related functions (b710-789). The Six Minute Walk test (6MWT) was used to assess the distance subjects could walk for 6 min [43][44][45][46]. The subject must perform at the fastest speed possible with or without assistive devices. Studies have shown test-retest reliability to be excellent among persons with stroke (i.e., ICC ¼ 0.99 distance in meters [44,45]) and persons with TBI (i.e., ICC ¼ 0.986) [42].
Walking speed was measured by the 10 Metre Walk Test (10MWT) [45,47] which assesses walking speed in meters per second for a distance of 10 m. The subject can perform the test with or without assistive devices. Studies have shown an excellent test-retest reliability for comfortable (ICC ¼ 0.94 time in sec.) and fast (ICC ¼ 0.97) gait speeds for persons with stroke [45], and for between day reliability for persons with TBI at comfortable (ICC ¼ 0.95) and fast speeds (ICC ¼ 0.96) [47].
Coordination (b760); dexterity (d430-d445); upper extremity function. The Box and Blocks Test [48][49][50] was used to measure subjects' upper extremity function. Individuals are seated at a table, facing a rectangular box divided into two square compartments of equal dimension by means of a partition. One hundred and fifty wooden cubes are placed in one compartment. The test requires the subjects to move as many blocks as possible, one at a time, from one compartment to the other during 60 s [48,49]. For persons with stroke, the Box and Blocks Test demonstrated excellent test-retest reliability when tested on the more affected (r ¼ 0.98) and the less affected hand (r ¼ 0.93) [50].
Mood (psychic stability, b1263). Subjects' mood was assessed to understand the subject's mood state before and after the program. The one preprogram evaluation was conducted at T3. Although there is no tool to specifically to assess mood among persons with ABI, two tools identified by the program's psychologist were used: the Montgomery Asberg Depression Rating Scale [51] and the Hamilton Depression Rating Scale (HDRS) [52]. Both tools were administered by a psychologist during an interview with the patient. The Montgomery Asberg Depression Rating Scale includes 10 items (e.g., difficulty concentrating, sadness), where each item has a broad definition and a severity of 0, 2, 4, and 6.
Specific mental functions (b140-b189). For this study, subjects' attention functions (b140) and memory functions (b144) were assessed. The Evaluation Test of Attention was used to assess processing speed or alertness and sustained attention [53]. With regard to processing speed, only simple reaction was measured requiring the subject to press a key as quickly as possible when a cross appeared on the monitor at randomly varying intervals. Sustained attention is required in tasks with very different cognitive demands, ranging from simple stimulus detection tasks to tasks with a high cognitive load. This involves focusing attention on a mentally demanding activity for a sustained period of time. In this test, a sequence of stimuli is presented on the monitor. The stimuli vary in a range of feature dimensions: color, shape, size, and filling. A target stimulus occurs whenever it corresponds to one or the other of two predetermined stimulus dimensions with the preceding stimulus (e.g., for this study the same shape but with a different color, size, and filling was used) [53].
Memory was assessed using the Rey's 15-word Auditory Verbal Learning Test [54]. For this task, the subject has to memorize a series of 15 words in five repetitions. The subject must repeat the words after learning them the first time (free recall), then after repetitions (total free recall) and after a 20 min delay (delayed free recall) [54]. When the subject could not read the words (i.e., had significant aphasia), a visual memory test was used [55]. EF were assessed during the Cooking Activity using Cooking Task [33][34][35].

Data analysis
In small-N designs, the individual participant is the unit of analysis with each person serving as his or her own control. Thus for each subject in the present study, data were recorded for each variable of interest for each assessment over the 9-month period.
Data analysis followed a descriptive approach. Raw data were graphed using Microsoft Excel 97-2003 for each primary outcome measure and changes were visually examined from pre-to postprogram for improvement in activity or participation. This method to determine treatment effects has been used by Ottenbacher (1986). However, to increase the study's internal validity, two statistical tests which consider subjects' scores differently were performed. The "non-overlap of all pairs" (NAP) method is a nonregression-based approach that summarizes data overlap in the pre and post intervention periods as the percentage of data points in the post intervention that are better than the data points in the pre-intervention [56]. The NAP is a non-parametric statistic indicating the strength of an effect between the pre and each post treatment phase. A NAP of 0.93-1.00 suggests a strong effect of performance change whereas 0.66-0.92 suggests medium effects.
To specify where the differences are located and the level of significance of a treatment effect, the two standard deviation band method (SDB) was also used. It relies on means and detect a change in response level between the baseline and treatment phases. A significant treatment effect is detected to have occurred if at least two consecutive data points during the intervention phase lie outside the two standard deviation band [57,58]. The rationale is that the probability of such an occurrence is less than p 0.05 [59]. In the present study, we modified the use of this method because no data were collected during the program for obvious reasons.

Results
The results are presented according to the study's primary and secondary objectives and within ICF categories. Subjects' characteristics with their personal factors and environmental factors are presented in Table 1. Health condition, body function, and structure data were collected from participants' medical health records or from the subjects themselves (see Table 2). In the preprogram phase, all subjects demonstrated problems with EFs. Most had amnesic difficulties and two subjects had aphasic sequela (i.e., CH & FA) (see Table 2). Four subjects had hemiplegia-related motor problems with a severe deficit of the upper limb. These descriptive data serve to help explain the effects or non-effects of treatment on the activity under study. Subject's raw data for each body outcome measure and for each activity/ participation measure are presented in Tables 3 and 4

Primary objective
Subjects' Cooking Task scores for each of the five assessments (T2-T6) are presented in four graphs, two indicating the number of errors (NAP and SDB methods) and two indicating execution time (NAP and SDB methods) allowing visual and statistical analyses of the effect of the rehabilitation program over time ( Figure 2). Two subjects (MN and CH) could not perform the Cooking Task preprogram, but could do so following the program. For the number of errors, visual analysis demonstrates improvement between the pre and the post phase for six out of seven subjects (i.e., MN, CO, FA, CH, NI, and MO), which is confirmed by a NAP index of 1.0 indicating strong improvement. Four participants (i.e., MN, CO, FA, and MO) showed significant improvement immediately after the program and at 3 and 6 months post. For the Cooking Task execution time, visual analysis demonstrates improvement between the pre and the post phases for three subjects (MN, CO, and FA), confirmed by the NAP index.
IADL Profile scores are presented in Table 4. For "preparing a hot meal for guests", visual analysis demonstrates improvement between the pre and the post phases for six out of seven subjects (i.e., MN, CO, FA, CH, NI, and MO), which is confirmed by NAP indices (see Figure 3). For CH, considering the small number of measurement points for this subject (i.e., no data at T4), these data should be interpreted with caution. Visual analysis shows an improvement between the post and 3 month post for three out of seven subjects (NI, CO and MO) and an improvement between the post and 6 month follow-up for four out of seven subjects (FA, CH, NI, and MO). At 6 months, three subjects (i.e., MN, MI, and MO) had a score greater than or equal to 3 (i.e., independence with difficulty), and three subjects (FA, CH, and NI) reached the maximum score of 4 (i.e., independent without difficulty) which is considered clinically important. Only CO's score decreased slightly at 6 months which may be explained by the participant's choice of a more complicated menu/meal (unpublished analysis). The meals prepared by the subjects in their homes each time they were evaluated with the IADL Profile were selected by the subjects themselves and thus varied greatly among subjects and over time. For example, FA chose to make far more complicated dishes than MI and MN did. Moreover, the results of the IADL Profile for FA were similar at T4 and T5 and the observed improvement (i.e., independence) between T5 and T6 was likely because he chose to prepare a less complicated menu at T5 and T6 compared with T4 (unpublished data). Over time, FA may also have become more conscious of his disabilities (tires easily and has motor difficulties related to hemiplegia) and in fact may consciously have chosen a less difficult menu to better suit his abilities.
The LIFE-H scores are presented for each subject for the Nutrition domain and for meal preparation specifically. For Nutrition, visual analysis demonstrates improvement between the pre-and the post-phase for four out of seven subjects (i.e., MN, FA, CH, and MI), which is confirmed by NAP indices ( Figure  4). CO and NI scored very high on this measure even before the intervention, thus only a small improvement effect was observed. Using the SDB method, only FA showed improvement at T4, while the program had a significant effect on MI at T5 and T6. For meal preparation, visual analysis demonstrates improvement between the pre and the post phases for three out of seven subjects (i.e., MN, FA, and CH) (NAP index ¼ 1.0). For MI, meal preparation indicates a medium improvement effect of the intervention. For NI and CO, the NAP indices suggest a lack of change or a ceiling effect. However, using the SDB method, the intervention was effective for MN but for MI, improvement was observed only at T4.

Secondary objective
Scores for the other activity/participation measures are presented in Table 4. Concerning "putting on outdoor clothes", visual analysis, confirmed by the NAP, shows a strong effect between the pre and the post phases for MO and a medium effect for FA. For the "going to the grocery store" task, visual analysis and NAP indicate a strong effect between the pre-and the post-phases for two subjects (i.e., FA and CH), relating to an improvement in walking. The NAP indicated a medium effect for NI and can be explained by the environment; in fact, NI lives far from the store, he no longer has driving privileges since his injury, and he is dependent on others to get there by car. Concerning "shopping for groceries", visual analysis demonstrates improvement between the pre and the post phases for five out of seven subjects (i.e., MN, CH, NI, MI, and MO), which is confirmed by the NAP index. Visually CO enhances his task performance between pre and post phases but poorly performed at T5.
Visual analysis of "having a meal with guests" scores indicates improvement between the pre and the post phases for three out of seven subjects (i.e., MN, CO and CH), which is also confirmed by the NAP indices. The intervention shows medium improvement effects for two subjects (FA and MO). The effect of the intervention is minimal for MI, however (NAP of 0.33), possibly explained by this subject's visual spatial neglect (e.g., forgot to serve the person on the left). For "cleaning up after meal" improvement was noted between the pre and the post phases for two out of seven subjects (i.e., CO and MO) (NAP ¼ 1.0). Three subjects (i.e., MN, CH, and MI) did not improve.
Concerning "obtaining information", improvement is noted between the pre and the post phases for three out of seven subjects (i.e., CO, FA, and MO) (NAP ¼ 1.0). Moreover, visual analysis confirmed by the NAP index demonstrated a medium effect for two subjects (i.e., MN and NI). Only for CH and MI, did the intervention not demonstrate an effect, which can be explained, respectively, in these two participants by aphasia (poor reading ability) and visual spatial neglect.
Lastly, for "making a budget" NAP indices show a strong effect of the program for MN, and a medium effect for FA and Ni, but their maximal scores indicated a need for verbal or physical assistance. In other words, the seven subjects are dependent or strongly dependent for this task before and after the intervention and in the long term.
Scores for the secondary outcome measures are also presented in Table 3. The majority of the scores associated with measures of body function did not improve over time. For the time to complete the 6-min walk test, visual analysis (not presented here) demonstrated improvement between the pre-and the postphases for two out of seven subjects (i.e., FA and NI) (NAP of 1.0 indicating strong improvement). However, using the SDB method, one subject (FA) showed significant improvement following the program (i.e., at T4, T5, and T6), while two subjects (MI and MO) showed significant improvement at T6. For the fast-paced 10-m walk test, the NAP indices suggest moderate improvement for MN (0.67), FA (0.92), and for CH (0.83). However, using the SDB method, two subjects (i.e., FA and CH) showed significant improvement (p 0.05) following the program (i.e., at T6), while MN showed significant improvement following the program at T5. In summary, six out of seven participants improved on at least one parameter of walking.

Activities and participation
Cooking Task   the memory test (i.e., without indices) for three participants: FA at T4, T5, and T6, and for the delayed free recall (after 20 min) for CH at T4, T5, and T6, and for NI at T5 and T6.

Discussion
The purpose of this study was to explore the effects of an intensive client-centered and multidisciplinary out-patient rehabilitation program on participants' activities and participation. We report on the effects of the program, specifically the cooking activity component, on participants' ability to prepare a meal (ICF code d630). Meal preparation is an important aspect of domestic/household life (d630-649), particularly in France, and can be an important basic need of persons living alone or with children. French stroke patients (31%) report difficulty preparing meals independently [60]. Three valid and reliable tools were used to evaluate the effects of the program's cooking activity on participants' ability to prepare a meal (d630), two of which used a situational approach while accounting for executive dysfunction. Globally, six of the seven participants improved their scores significantly on the two situational tools (i.e., Cooking Task and IADL Profile) following the program and in follow-up thus providing supporting evidence of the program's effectiveness. Only one participant did not show significant improvement on these measures.
One might argue that improvements in the scores/performances reflect subjects' learning with the repetition of the assessment tasks. This is, however, unlikely for several reasons. For the Cooking Task, two subjects were unable to complete the task twice before the program, yet, all subjects were able to do so post program. For each assessment with the IADL Profile, subjects chose and prepared a completely different meal/menu thus reducing a potential learning effect; improved scores thus truly reflect improved capacity to perform. Furthermore, subjects did not improve their performance in the budget related task, which was not taught during the program, despite repeating the same task during multiple assessments. CT execution time did not, however, significantly improve for subjects, suggesting this variable is not a good indicator of improved performance. For example, some subjects took more time to complete the task because their monitoring of the task improved. In other cases, subjects were better organized and committed fewer errors and reduced execution time.
The results obtained using the LIFE-H were less suggestive of the program's effectiveness and do not corroborate the results discussed above. The LIFE H results may reflect some of the problems when using a self-report questionnaire among persons with anosognosia who are less aware or deny/ignore their problems, and among those who try to impress the evaluator by overestimating their abilities (i.e., desirability bias). For example, CO and NI overestimated their abilities to prepare a meal when responding to the LIFE-H preprogram (T3) and their results were not corroborated by their scores on the IADL Profile recorded at the same time. MI, who also suffered from anosognosia, overestimated her ability to prepare a meal post program. Some thought should be given to the choice of measurement tools used when examining the effectiveness of rehabilitation programs for persons with ABI and anosognosia.
Although the article focuses on the impact of the rehabilitation program on "preparing a meal", it is interesting to consider these results in relation to the performance of other tasks included in the IADL Profile tool but not specifically/directly taught in the program. Indeed throughout the program, participants learn how to obtain different types of information (e.g., ask for information about museum schedules) and the effects of this training can be observed in the specific "obtaining information" task of the IADL Profile.
Conversely, creating and managing a budget is not taught by the program; the results showed all subjects remained dependent with respect to this task. In other words, the activities trained by the program appear to develop capabilities that are generalizable to the same activities in a different environment (e.g., preparing a meal in the OT kitchen versus at home). But these abilities do not seem to be transferable to other activities. These results support the principles of metacognitive rehabilitation of EF [27].
Given the goals of the rehabilitation program, we did not necessarily expect improvements in participants' function as defined by the ICF. In fact, with the tools used in this study, very few improvements in function were noted. However, certain subjects significantly improved specific body functions (e.g., FA improved walking speed) and mental functions post program (e.g., CH and NA significantly improved their scores on the Rey verbal memory test despite using parallel test versions over time). Improvements in memory may have been partly responsible for improved activity performance during the preparation of a recipe and of a meal (i.e., not forgetting the dish in the oven). Improvement of participants' EF was only evaluated by the Cooking Task where the goal is to assess EF during an activity [39,40,61] via measures of type and quantity of errors made during the test. The type of neuropsychological errors was not explored in this study but the number of errors committed during the test was examined, and improvements are suggestive of EF during an activity. The reduction in the number of errors during the Cooking Task could reflect an improvement in executive dysfunction and the participants' ability to plan, monitor and execute the activity [62]. Classic paper and pencil neuropsychological testing of EF was not appropriate in this study (repeated measures) because we would not have been able to control for subjects' improved performance over time due to learning.
Moreover, as a result of the seven weeks, it is possible that participants may have improved their physical condition (e.g., walking perimeter) and decreased their fatigue. It would have been interesting to have assessed fatigue after each cooking activity intervention.
Finally, corroborative positive results were shown using subjects' data from two different groups of program participants. To increase the internal validity, observer bias was limited by using measures with established inter-rater reliability and raters independent from the intervention team. Using two statistical methods (NAP and SDB) to analyze the results further increases the internal validity of our research. To our knowledge, these two statistics have never been used together before to provide complementary information about whether and when significant improvements occurred.
In light of these results, we feel there is convincing evidence about the effectiveness of the program under study. Indeed it is difficult to study the effects of a multi-faceted holistic multidisciplinary rehabilitation program, particularly one that provides services to a small group of individuals at a time [36,37].

Limits
One of the limitations of this study is the small number of subjects. However, given the individual approach used during the Cooking Activity, we feel the SCED used in this study to be most appropriate. Indeed, randomized clinical trials (RCTs) are powerful techniques for determining the efficacy of interventions. However, they may have practical limitations when applied to many rehabilitation settings and research questions [61]. Besides being difficult to use with groups of heterogeneous patients, clinical trials are often difficult to conduct while respecting realities of clinical settings, especially when time-consuming situational assessment tools are necessary to clearly establish a link between an intervention and its impact on patients' activity and participation. We acknowledge that small-N research methods can only directly determine the best treatments for persons involved in a specific study [61]. Indeed SCED methodology has limited generalisability compared with RCTs. The generalisability of our study using SCED was however strengthened by the use of multiple baseline across many subjects (n ¼ 7) and because the results were positive for six of the seven subjects (85.7% of the group). The second limitation is the absence of a control group; however, Perdices [15] supports the use of SCED as it builds into a control condition with the multiple baseline measures across behaviors pre-and post-program to determine treatment efficacy. Moreover, Tate et al. [63] advocate for the inclusion of the randomized n-of-1 trial (such as SCED) as Level 1 evidence likely to have major implications for what constitutes the evidence base of health interventions [63]. We believe the present study further supports the use of this experimental design in rehabilitations efficacy studies.
We acknowledge we did not evaluate all functions or activities intrinsic to the cooking activity (e.g., communicating withreceivingwritten messages -d325); the program involves multiple components and evaluating its impact on all aspects of the participants' participation at home and in the community would be an enormous task and almost impossible.
The detailed description of the therapeutic interventions provided during the cooking activity should help clinicians better understand what specific functions are solicited or required during a particular activity (e.g., reception of written language (b16701), problem solving (b1646) when planning and executing a recipe). It also could be useful for future research aimed at corroborating the findings of this study.

Conclusions
A single case experimental design was used to demonstrate significant clinical and statistical improvements in activity and participation outcomes related to preparing a meal, in adults with ABI and executive dysfunction who participated in a 7week multidisciplinary rehabilitation out-patient program. Treatment gains were maintained for the majority of participants at 3 and 6 months following the completion of the program. These findings are encouraging and provide evidence to support the effectiveness of the rehabilitation program and these can guide clinicians in their work within this complex area of rehabilitation.

Disclosure statement
Three of the authors participated in the rehabilitation program; however, assessors were not in the program.