The effect of Cognitive Adaptation Training on social functioning: Helpful or harmful?
Milka Meijer
Name: Milka Meijer Student number: 6102565
Supervisor: Prof. dr. A. Scholing Second supervisor: dr. E. M. Wekking Date: may 2016
Abstract
Patients with a severe mental illness (SMI) often have serious deficits in social functioning caused by the cognitive impairments that are associated with their illness. To improve social functioning an intervention is needed to bypass the cognitive impairments. Cognitive adaptation training (CAT) is such an intervention that aims to work around the cognitive deficits through reminders and environmental cues to support specific adaptive behaviors in patients with an SMI. In the current single blind randomized controlled trial the effect of six months of TAU+CAT on social functioning in patients with an SMI was compared to the effects of TAU only. In addition, it was tested whether verbal memory and executive functioning could predict the outcome of TAU+CAT. It was found that worse verbal memory counter-indicates TAU+CAT as patients seem to deteriorate on social functioning, whereas better verbal memory indicates a good progression on social functioning after six months of TAU+CAT in comparison to patients receiving TAU only.
Introduction
Patients with a severe mental illness (SMI) such as schizophrenia or bipolar disorder, have a chronic and severe course of disorder, and often suffer from comorbid disorders. Generally, they have serious deficits in social and societal functioning and need mental health care (Delespaul & Consensusgroep EPA, 2013). Their social deficits are not only a result of the mental disorder, they also maintain the disorder (Delespaul & Consensusgroep EPA, 2013). In addition, cognitive deficits -which are central in SMI’s- form a challenge for the treatment of SMI’s as they limit the benefits of psychosocial rehabilitation and predict poor performance of daily activities (Eckert et al., 1996; Velligan et al., 1997). Thus, to improve social functioning an intervention is needed which also overcomes the cognitive deficits. There are broadly two types of treatments that focus on cognitive deficits. The first are restorative approaches, such as cognitive remediation. Here the focus is on tasks that aim to directly improve cognitive deficits, such as computer tasks to improve memory. The second are compensatory approaches, which aim to work around the cognitive deficits, with the organization of belongings, reminders and environmental cues to support specific behaviors (Draper, Stutes, Maples, & Velligan, 2009). In the general population such cues are also often used, for example the use of alarm clocks, to-do lists and organizers.
Because of the chronic course of an SMI, focus should not be on full recovery but on minimizing the consequences of the SMI and thereby improving daily and social functioning (van den Bosch, 2000). Cognitive Adaptation Training (CAT) is such a compensatory intervention, aiming at reducing the functional problems caused by the cognitive deficits (Velligan, Mahurin, True, Lefton, & Flores, 1996). This is done by environmental cues, antecedent control and behavior reinforcement (Draper et al., 2009). The theory behind CAT is that impairments in cognitive functioning in patients with an SMI lead to a characteristic behavioral style which becomes evident during the execution of daily tasks. There are three of such behavior styles: 1. Apathy; this is characterized by a poverty of speech and movements, and the inability to initiate or follow through behaviors which require multiple steps. 2. Disinhibition; this is characterized by disorganization, easily being distracted, and being highly driven by environmental cues. 3. Mixed; this is a combination of the apathy style and disinhibition style, characterized
by both having trouble initiating behavior and being easily distracted (Draper et al., 2009). The behavior style is determined by means of neuropsychological tests, an assessment of daily functioning and a structured evaluation of the observed behavior of the patient (Velligan & Bow- Thomas, 2000). Based on the behavioral style, executive functioning, everyday functioning and needs and wishes of the patient an individually tailored intervention is composed (Stiekema et al., 2015). During the execution of CAT, compensatory strategies and environmental supports are used to bypass the cognitive deficits and improve daily functioning of patients with an SMI (Velligan et al., 2000; Quee et al., 2014). Those strategies and environmental supports are based on the level of executive functioning of the patients. Strategies and supports for patients with poor executive functioning are more easily notable, more specific and more numerous than those for patients with fair executive functioning (Stiekema et al., 2015).
In the United States, CAT has been shown to improve adaptive and social functioning of outpatients. In a randomized controlled trial it was found that outpatients with either schizophrenia or schizoaffective disorders receiving CAT by a trained therapist showed improvements on psychotic symptoms, motivation and adaptive and social functioning (Velligan et al., 2000; Velligan et al., 2008). In the Netherlands, it was found in a pilot study that in comparison to treatment as usual (TAU), patients receiving both TAU and CAT by psychiatric nurses improved more on their level of functioning in terms of community adjustment (Quee et al., 2014). TAU consists of pharmacotherapy in combination with different therapies and activities, such as sport groups, music therapy, psycho-education or creative activities that match the needs and goals of the patient (Stiekema et al., 2015). As mentioned, in the United States CAT was given by trained therapists to outpatients. To date, it is unclear if long-term inpatients would also benefit from CAT when given by psychiatric nurses instead of trained therapists. In addition to the distinction between in- and outpatients and therapists and nurses, patient factors such as age or gender might also play a role in the effectiveness of CAT. In a meta-analysis on social cognitive trainings in patients with schizophrenia it was found that the age of the patient influences the outcome of the training, i.e. younger patients improved more on social cognitive functioning than older patients (Kurtz & Richardson, 2011). CAT focuses, just as social cognitive trainings, on reducing the problems caused by the
cognitive deficits, it might therefore be possible that younger patients profit more from CAT than older patients.
In patients with an SMI different factors have found to be related to social functioning. For example, it was found that measures of executive functioning and verbal memory predict multiple domains of functional outcome, such as occupational functioning, degree of independent living, social problem solving and psychosocial skill acquisition (Green, 1996; Green, Kern, Braff, & Mintz, 2000; Velligan et al., 2000). Furthermore, social functioning was found to be positively related to executive functioning and verbal fluency in patients with schizophrenia (Breier, Schreiber, Dyer, & Pickar, 1991; Kern et al., 1992; Green et al., 2000). As to gender, it was found that impaired memory is related to poorer social skills in women but not in men (Mueser, Blachard, & Bellack, 1995). This finding has not been replicated yet though. Gender differences in SMI’s have also been further studied. For example, schizophrenia has a later onset in women than in men and women with schizophrenia also seem to be more socially adapted early in life than their male counterparts. Moreover, over both short- and long-term recovery, social functions generally recover better in women than in men with schizophrenia (Childers & Harding, 1990). These findings could therefore suggest that women possibly profit more from CAT than men as their social functions generally recover better than those of men. In sum, even though TAU+CAT has been shown effective in the United States it remains unclear if it will also be effective in improving social functioning in the Netherlands as a nursing intervention, instead of an intervention given by trained therapists. Therefore, the primary goal of this study was to investigate whether TAU+CAT as a nursing intervention improves social functioning for long-term inpatients with an SMI in the Netherlands. In addition, as social functioning in SMI’s has been found to be related to age, gender, executive functioning and verbal memory, the secondary goal of this study was to investigate whether these four patient factors might also influence the effect of TAU+CAT on social functioning.
In the current single blind randomized controlled trial, 89 patients with an SMI from three different mental health institutions in the Netherlands were given either TAU or TAU+CAT. After six months it was tested whether they improved in social functioning, and whether certain patient factors, i.e. age, gender, executive functioning, verbal memory, might be predictive of their outcome. It was expected that after six months of
TAU+CAT, patients with an SMI would have an improvement on social functioning, as compared to patients receiving TAU. Furthermore, it was expected that after six months of TAU+CAT, patients with an SMI who are women or younger would improve more on social functioning than their male and older counterparts. Finally, it was expected that after six months of TAU+CAT, executive functioning and verbal memory would predict the outcome of TAU+CAT. The direction however was not predicted, as to date no studies have reported on this topic and the theoretical justification for formulating a hypothesis about the direction was insufficient.
Methods
Participants
Participants (N = 89) were inpatients with a diagnosis of Severe Mental Illness (SMI). Participants lived at the long-stay departments of three different institutions: Lentis (Zuidlaren), GGZ Drenthe (Assen) and Dijk en Duin (Castricum). Participants were allocated to either the TAU-condition (n = 47) or the TAU+CAT condition (n = 42). At Lentis (n = 40) allocation to either TAU+CAT or TAU was randomly done, at GGZ Drenthe (n = 22) and Dijk en Duin (n = 27) allocation was according to the sub-departments where the patients stayed within the institution. For example, at Dijk en Duin participants lived in a building with three different floors and a separate building. Patients living on the first and third floor were allocated to the TAU+CAT condition and patients living on the second floor and in the separate building were allocated to the TAU condition. Allocating the floors and building in which the patients lived to either the TAU+CAT or TAU condition was randomly done.
The participants were on average 53.18 years old (SD = 10.35), with the youngest participant of age 22 and the oldest participant of age 72. Of all participants at baseline 66.3 percent (n = 59) were male, and 33.7 percent (n = 30) were female. Of all participants 55.1 percent (n = 49) had the diagnosis schizophrenia, 9 percent (n = 8) schizo-affective disorder, 10.1 percent (n = 9) had bipolar disorder, 5.6 percent (n = 5) borderline personality disorder, and 5.6 percent (n = 5) depression. The diagnosis of one participant was unknown, and the other participants (n = 13) were diagnosed with other disorders. All participants received five euro for participation at T0 and another five euro
for participation at T1.
Materials
Social Functioning
Social functioning was measured with the Multnomah Community Ability Scale (MCAS; Barker, Barron, McFerland, & Bigelow, 1994). The MCAS is a 17-item semi-structured interview, consisting of a total scale and four subscales: Interference with Functioning, Adjustment to Living in the Community, Social Competence, and Behavioral Problems. All items are filled in on a five-point Likert scale. Higher scores reflect better functioning and less impairment (Dickerson et al., 2003). The instrument has shown to be valid and reliable (α = 0.86, rxy = 0.83) and has also been used in previous CAT-studies (Corbière et al., 2002).
Social functioning was also measured with the Social Functioning Scale (SFS; Birchwoord, 1983). The SFS is a questionnaire filled in by the nurse concerning the patients’ behavior over the previous month. It consists of seven subscales: Withdrawal, Interpersonal, Pro-Social, Recreation, Independence Performance, Independence Competence, and Employment Occupation (Birchwood, Smith, Cochrane, Wetton, & Copestake, 1990). Higher scores reflect better competence in the specific area. The SFS has good psychometric qualities (α = 0.80) (Birchwood et al., 1990).
Executive Functioning
Executive functioning was measured by means of the perseveration and categories scores of the Modified Card Sorting Test (MCST; Nelson, 1976) and by means of the Dutch Controlled Oral Word Association Test (COWAT; Schmand, Groenink, & Van den Dungen, 2008). The COWAT measures letter fluency by taking the sum of the correct responses for each trial, with higher scores reflecting better letter fluency. The COWAT has shown to be valid and reliable (α = 0.82) (Schmand, Groenink & Van den Dungen, 2008).
Verbal Memory
Verbal memory was measured with the Word Task 15 (15WT-A; Saan & Deelman, 1986). For each trial of remembering words the scores range from 0 to 15, with higher
scores reflecting better verbal memory. The 15WT-A has shown to be valid and reliable (rxy = 0.80) (Bouma, Mulder, Lindeboom, & Schmand, 2012).
Procedure
A CAT training was given to the clinical case managers and nurses of each TAU+CAT department. Thereafter, clinical case managers and nurses approached patients from their own department for participation. Patients received an information folder about the study and CAT. When they decided to participate, they signed the informed consent given by their nurse or clinical case manager. Data were collected at baseline (T0) and six months after baseline (T1). Psychology students with a bachelor degree and trained for the testing purpose tested the participant. Students were blind for the condition to which the participants were assigned. Clinical case managers and nurses were also asked to fill in questionnaires concerning their participating patients.
Statistics
To test whether participants who dropped out significantly differed from participants who continued with CAT or TAU on demographic variables, t-tests were executed for interval data and χ2
for ordinal or nominal data. To test whether participants from the TAU condition differed from participants from the TAU+CAT condition at baseline on social functioning, executive functioning and verbal memory, t-tests were executed.
Linear mixed models (LMM) were used to test whether there was an effect of condition and time on social functioning as measured by the MCAS and SFS. Furthermore it was also tested if baseline measures of gender, age, executive functioning or verbal memory were of influence on the effect of CAT. Because LMM analyses allow inclusion of participants with incomplete data (Snijders & Bosker, 1999), all patients who participated at baseline were included. In this way the requirement of intention to treat was met. In LMM when measurements are done twice, for every ten participants one predictor can be taken in the model (Wilson VanVoorhis & Morgan, 2007). In the current model eleven possible main effects and interaction effects were identified at start. For the model selection process, firstly a maximum model was specified, including all main and interaction effects. When an interaction effect was non-significant, it was taken out of the model by backward elimination. To be parsimonious the variables ‘executive
functioning’ and ‘verbal memory’ were chosen to be taken out as main-effects in case there would be no interaction effects found in which they were included. The choice for these two was made because there was no direction on forehand hypothesized.
To interpret the results, all continuous predictor variables were centered by subtracting the average score of the group from the individual score of the participant (Aiken & West, 1991). Gender was coded as 0 = male, and 1 = female, condition was coded as 0 = TAU, and 1 = TAU+CAT, and time as 0 = T0 and 1 = T1. In case there was a significant condition × time effect it could be concluded that TAU+CAT showed a significant improvement at T1 compared to TAU. To interpret 3-way interaction effects, the results were plotted. Because LMM shifts estimates toward each other by centering continues predictor variables no Bonferroni correction needed to be applied (Gelman, Hill, & Yajima, 2012).
To calculate clinically relevant change of scores from the MCAS and SFS, the Reliable Change Index (RCI) was used. The RCI is derived from the equation: RCI = (x1 –
x2)/Sdiff. The baseline score is x1 and the follow-up score is x2. Sdiff is the standard error of
the difference. Sdiff is derived from the equation: Sdiff = √2(Se)2, with Se being the standard
error of the measurement. The standard error of the measurement is derived from the equation: Se = SD √(1-rxx) (Jacobson & Truax, 1991).
Results
Comparisons of drop-outs
Of the 89 participants 16 dropped-out between T0 and T1, six of them refused further participation without a given reason, two moved to some place else, one participant could not further participate because of an addiction, one participant found the idea of being interviewed and tested too stressful, one did not want to talk about feelings, one died, and the other four did not show up anymore.
Participants who dropped-out did not significantly differ on age from participants who continued participation, t(85) = .928, p = .356. The mean age of participants who continued was 53.53 (SD = 9.99), the mean age of participants who dropped-out was 50.80 (SD = 12.03). Participants who did drop out did not differ in allocation to condition from participants who continued participation χ2
(1,89) = .001, p = .978. Nor did the proportion of males that out differ from the proportion of females that
dropped-out, χ2
(1,89) = .000, p = 1. Finally, the participants who dropped-out did not differ in diagnosis from the participants who continued participation χ2(88) = 5.084, p = .341.
Severity of symptoms was not included in the analyses because this variable was only measured for those with a diagnosis of schizophrenia, therefore of 35.9 percent of the participants the severity of symptoms was not known, which precluded reliable conclusions.
Baseline comparisons
At baseline participants in the TAU+CAT condition did not differ from participants in the TAU condition on total MCAS score. Nor did they differ on the MCAS subscales: Interference with Functioning, Adjustment to Living in the Community, Social Competence, and Behavioral Problems (Table 1). Nor did participants in the TAU+CAT condition differ from participants in the TAU condition on the SFS subscales: Withdrawal, Interpersonal, Pro-Social, Recreation, Employment Occupation. They did differ on the SFS subscales: Independence Performance, and Independence Competence (Table 1). Finally, at baseline participants in the CAT condition did not differ from participants in the TAU condition in executive functioning, or in verbal memory (Table 1).
A one-way ANOVA was used to test whether participants in the TAU+CAT condition from the three institutions differed at T0 MCAS scores and SFS scores. Participants from the three institutions did not differ at T0 on total MCAS score (Table 2). Nor did they differ on the MCAS subscales: Interference with Functioning, Adjustment to Living in the Community, Social Competence, and Behavioral Problems (Table 2). Participants from the three institutions did not differ at T0 on the SFS subscales: Withdrawal, Interpersonal, Pro-Social, Recreation, Independence Performance, Independence Competence, and Employment Occupation (Table 2).
Table 1.
Baseline Mean Scores of MCAS- and SFS Scales, Executive Functioning and Verbal Memory of CAT and TAU CAT TAU Scale n M (SD) n M (SD) F p MCAS Total 37 58.84 (6.49) 42 57.93 (6.97) .357 (1,77) .552 MCAS Interference Functioning 38 18.63 (2.39) 43 18.23 (2.35) .573 (1,79) .451 MCAS Adjustment to Community 39 8.62 (2.01) 43 8.21 (2.09) .802 (1,80) .373 MCAS Social Competence MCAS Behavioral Problems 39 37 16.38 (3.39) 14.89 (2.01) 43 42 16.65 (3.42) 14.89 (2.18) .125 (1,80) .032 (1,77) .724 .859 SFS Withdrawal 42 102.79 (12.52) 39 103.08 (11.26) .012 (1,79) .913 SFS Interpersonal 37 112.92 (20.72) 33 118.03 (19.75) 1.109 (1,68) .296 SFS Pro-social 42 91.19 (10.68) 39 94.60 (12.38) 1.772 (1,79) .187 SFS Recreation 41 95.59 (12.82) 41 101.31 (14.94) 3.461 (1,80) .067 SFS Independence Performance 41 88.45 (13.70) 39 98.92 (12.65) 12.583 (1,78) .001* SFS Independence Competence 42 93.44 (9.60) 40 97.95 (8.48) 5.064 (1,80) .027* SFS Employment Occupation 42 95.81 (8.16) 40 98.11 (9.84) 1.336 (1,80) .251 Executive Functioning 37 1.32 (.32) 41 1.27 (.25) .384 (1,76) .537 Verbal Memory 24 26.13 (12.06) 35 22.20 (10.68) 1.731 (1,57) .194 * p < .05
Table 2.
Baseline Mean Scores and Differences between Institutions on Social Functioning
Lentis Dijk en Duin GGZ Drenthe
Scale M (SD) M (SD) M (SD) F (df) p MCAS Total 58.74 (7.19) 57.74 (6.20) 58.33 (6.65) .155 (2, 76) .857
MCAS Interference with Functioning 18.45 (2.23) 18.09 (2.37) 18.75 (2.65) .421 (2, 78) .658 MCAS Adjustment to Living in the Community
8.79 (2.12) 7.83 (2.10) 8.33 (1.77) 1.626 (2, 97) .203 MCAS Social Competence 16.24 (3.83) 16.78 (3.25) 16.76 (2.72) .251 (2, 79) .779 MCAS Behavioral Problems 15.26 (2.24) 15.04 (1.89) 14.11 (1.88) 1.944 (2, 76) .150 SFS Withdrawal 100.95 (10.21) 105.25 (14.19) 103.93 (11.92) 1.018 (2, 78) .366 SFS Interpersonal 118.14 (20.36) 117.25 (19.27) 108 (20.29) 1.621 (2, 67) .205 SFS Pro-social 94.74 (11.61) 93.74 (10.58) 88.68 (11.90) 2.042 (2, 78) .137 SFS Recreation 100.07 (13.12) 94.07 (10.81) 100.02 (17.96) 1.459 (2, 79) .239 SFS Independence Performance 95.27 (13.91) 90.38 (15.91) 93.41 (12.83) .790 (2, 77) .457 SFS Independence Competence 97.33 (8.99) 91.60 (9.58) 96.50 (8.77) 2.855 (2, 79) .064 SFS Employment Occupation 96.99 (9.11) 98.98 (10.16) 94.89 (7.63) 1.104 (2, 79) .337
Confirmatory LMM analyses
Based on LMM analyses for the MCAS total scale and subscales and SFS subscales, no main effect of time was found and no interaction effect of condition × time was found. For all main effects -significant and non-significant- see Table 3.
Based on LMM analyses for the MCAS total scale main effects of gender and age were found (Table 4), with women showing better social functioning than men, and younger participants showing better social functioning than older participants.
A three way-interaction effect for condition, time and executive functioning on the MCAS total scale was also found (Table 4). Participants in the TAU+CAT condition with better executive functioning improved more on social functioning than participants with worse executive functioning in the TAU+CAT condition and more than all of the participants in the TAU condition (Figure 1). In addition, a three way-interaction effect for condition, time and verbal memory on the MCAS total scale was found (Table 4). Participants in the TAU+CAT condition with better verbal memory improved more on social functioning than participants with better verbal memory from the TAU condition. In contrast, participants from the TAU+CAT condition with worse verbal memory deteriorated on social functioning, while participants with worse verbal memory from the TAU condition retained the same level of social functioning (Figure 2). With the MCAS Interference with Functioning subscale a main effect of gender was found (Table 4), with men having better social functioning.
With the MCAS Adjustment to Living in the Community scale a three way-interaction effect for condition, time and verbal memory was found (Table 4). Participants in the
Figure 2. Predictive MCAS Total scores. Three way-interaction Figure 1. Predictive MCAS Total scores. Three
way-TAU+CAT condition with better verbal memory improved more on social functioning than participants with better verbal memory from the TAU condition. In contrast, participants from the TAU+CAT condition with worse verbal memory deteriorated on social functioning while participants with worse verbal memory from the TAU condition retained the same level of social functioning (Figure 3).
With the MCAS Social Competence scale a main effect of gender was found (Table 4), with women having better social functioning. A main effect for age was also found (Table 4), with younger participants showing better social functioning than older participants. A three way-interaction effect for condition, time and executive functioning on the MCAS Social Competence scale was also found (Table 4). Participants in the TAU+CAT condition with better executive functioning improved more on social functioning than participants with worse executive functioning in the TAU+CAT condition and more than all participants from the TAU condition (Figure 4). A three way-interaction effect for condition, time and verbal memory on the MCAS Social Competence scale was also found (Table 4). Participants in the TAU+CAT condition with better verbal memory improved more on social functioning than participants with better verbal memory from the TAU condition. In contrast, participants from the TAU+CAT condition with worse verbal memory deteriorated on social functioning while participants with worse verbal memory from the TAU condition retained the same level
Figure 3. Predictive MCAS Adjustment to Living in the Community scores. Three way-interaction effect for condition, time, and verbal memory, t(95) = 2.504, p = .014.
Figure 4. Predictive MCAS Social Competence scores. Three way-interaction effect for condition, time, and executive functioning, t(52) = 3.307, p = .002.
of social functioning (Figure 5).
With the MCAS Behavioral Problems scale a main effect of gender was found (Table 4), with women showing better social functioning.
With the SFS Withdrawal scale a main effect of gender was found (Table 4), with women showing better social functioning. For the SFS Interpersonal scale a main effect of gender was found (Table 4), with women showing better social functioning. With the SFS Recreation scale a main effect of executive functioning was found (Table 4), with participants with worse executive functioning showing better social functioning. With the SFS Independence Performance scale a main effect of condition was found (Table 4), with participants in TAU condition showing better social functioning. A three way-interaction effect for condition, time and verbal memory on the SFS Independence Performance scale was also found (Table 4). Participants in the TAU+CAT condition with better verbal memory improved more on social functioning as measured by the SFS Independence Performance scale than participants with better verbal memory from the TAU condition. In contrast, participants from the TAU+CAT condition with worse verbal memory deteriorated on social functioning while participants with worse verbal memory from the TAU condition retained the same level of social functioning (Figure 6).
With the SFS Independence Competence scale a main effect of verbal memory was found (Table 4), participants with better verbal memory showing better social functioning.
Figure 5. Predictive MCAS Social Competence scores. Three way-interaction effect for condition, time, and verbal memory,
t(53) = 3.555, p = .001.
Figure 6. Predictive SFS Independence Performance scores. Three way-interaction effect for condition, time, and verbal memory, t(90) = 3.312, p = .001.
Table 3.
LMM Main Effects of Predictors on MCAS- and SFS Scales
Condition Time Gender Age Verbal memory Executive functioning Scale p p p p p p MCAS Total .501 .568 .034* .040* .174 .883 MCAS Interference with Functioning .878 .087 .022* .623 - - MCAS Adjustment to Living in the Community .264 .406 .543 .060 .800 - MCAS Social Competence .424 .689 <.001* <.001* .075 .960 MCAS Behavioral Problems .750 .737 .028* .373 - - SFS Withdrawal .549 .607 ,042* .147 - - SFS Interpersonal .958 .371 .002* .519 - - SFS Pro-social .661 .815 .406 .179 - - SFS Recreation .174 .239 .135 .370 - .026* SFS Independence Competence .097 .774 .252 .419 .001* - SFS Independence Performance .007* .823 .852 .892 .781 .088 SFS Employment Occupation .226 .510 .357 .096 - - * p < .05 Table 4.
Summary of MCAS and SFS Significant Main- and Interaction Effects
estimates p MCAS Total intercept 58.13 <.001** gender 2.89 .034* age -.12 .040* condition × time × EF 15.02 .039* condition × time × VM .44 .008*
MCAS Interference with Functioning
intercept 19.00 <.001**
gender -1.02 .022*
MCAS Adjustment to Living in the Community
intercept 8.12 <.001**
condition × time × VM .11 .014*
MCAS Social Competence
intercept 15.89 <.001**
gender 2.58 <.001**
age -.11 <.001**
condition × time × EF 11.50 .002*
condition × time × VM .27 .001*
MCAS Behavioral Problems
intercept 15.14 <.001** gender .90 .028* SFS Withdrawal intercept 101.30 <.001** gender 4.76 .042* SFS Interpersonal intercept 111.57 <.001** gender 13.99 .002* SFS Recreation intercept 100.46 <.001** EF -13.14 .026* SFS Independence Performance intercept 98.76 <.001** condition -9.85 .007* condition × time × VM .89 .001* SFS Independence Competence intercept 98.41 <.001** VM .28 .001*
Note. VM (verbal memory) and EF (executive functioning) * p < .05
Exploratory analyses
Even though no treatment effect of TAU+CAT was found, it remains important to test whether participants from the TAU+CAT condition received as much attention from their personal nurse as participants from the TAU condition. This is because CAT is a treatment that should be implemented in daily care and should therefore not result in extra attention given to the patient. A t-test was used to test this. Attention was measured by time spent by the personal nurse with the participant over the previous three months. Participants from the TAU+CAT condition did not differ at T1 in attention received from their personal nurse from participants in the TAU condition, F(1, 66) = .654, p = .422. The three different institutions might have different treatment results. If this is the case, then more effective results from one institution could be obscured by less effective results from the other institutions. Therefore, a one-way ANOVA was used to test whether participants in the TAU+CAT condition from the three institutions differed at T1 MCAS scores and SFS scores. Participants from the three institutions did not differ at T1 on total MCAS score (Table 5). Nor did they differ on the MCAS subscales: Interference with Functioning, Adjustment to Living in the Community, Social Competence, and Behavioral Problems (Table 5). Participants from the three institutions did not differ at T1 on the SFS subscales: Withdrawal, Interpersonal, Pro-Social, Independence Performance, Independence Competence, and Employment Occupation (Table 5). They did differ on the SFS subscale Recreation (Table 5). Post hoc comparisons using Bonferroni correction indicated that the mean SFS subscale Recreation score of participants in the TAU+CAT condition from Dijk en Duin (M = 83.81, SD = 11.56) did significantly differ from the mean score of participants in the TAU+CAT condition from Lentis (M = 97.31, SD = 11.85) and GGZ Drenthe (M = 94.54, SD = 14.10), p = .044. The RCI was used to calculate clinically relevant change on the MCAS total scale, MCAS Social Competence subscale, MCAS Adjustment to Living in the Community subscale and the SFS Independence Performance subscale. Participants from the CAT+TAU were -based on their 15WT-A scores- allocated to either the below average verbal memory group or the above average verbal memory group (Table 6). Changes in scores between T0 and T1 were not greater than the reliable change criterion, thus no clinically relevant change of scores in the two groups were found (Table 6). However,
T1-T0 scores of patients with an above average verbal memory were generally higher than scores of patients with a below average verbal memory (Table 6).
Table 5.
Scores for Differences between Institutions on Social Functioning at T1
Lentis Dijk en Duin GGZ Drenthe
Scale M (SD) M (SD) M (SD) F (df) p MCAS Total 60.56 (6.79) 57.86 (7.36) 59.00 (4.00) .496 (2, 29) .614
MCAS Interference with Functioning 19.41 (2.15) 18.43 (2.37) 19.00 (1.32) .603 (2, 30) .554 MCAS Adjustment to Living in the Community
8.18 (2.24) 8.00 (2.07) 7.89 (1.36) .065 (2, 31) .938 MCAS Social Competence 17.35 (3.24) 15.88 (4.02) 16.67 (2.65) .469 (2, 30) .630 MCAS Behavioral Problems 15.25 (2.18) 14.63 (1.77) 15.44 (1.33) .441 (2, 30) .648 SFS Withdrawal 103.82 (12.38) 93.86 (7.82) 103.32 (8.31) 2.425 (2, 34) .104 SFS Interpersonal 106.36 (10.03) 118.00 (14.56) 112.44 (21.45) 1.430 (2, 27) .257 SFS Pro-social 90.25 (15.38) 88.94 (9.08) 88.73 (11.48) .055 (2, 34) .946 SFS Recreation 97.32 (11.85) 83.81 (11.56) 94.55 (14.10) 3.332 (2, 35) .047* SFS Independence Performance 92.33 (14.68) 88.81 (18.57) 92.25 (17.65) .342 (2, 33) .713 SFS Independence Competence 95.42 (9.55) 90.06 (12.05) 95.60 (11.50) .815 (2, 34) .451 SFS Employment Occupation 98.26 (10.75) 95.75 (7.61) 93.45 (7.18) .901 (2, 34) .416 * p < .05
Table 6.
Mean Scores and Reliable Change Criterion of Groups with Below versus Above Average Verbal Memory
Verbal memory T0 M T1 M SD T1-T0 RCI - criterion MCAS Total (α = .86) Below average 58.30 58.40 5.63 0.10 6.43 Above average 61.40 62.60 4.63 1.20 5.29 MCAS Social Competence (α = .79) Below average 16.30 16.10 3.43 -0.20 4.35 Above average 17.55 18.00 2.81 0.45 3.57 MCAS Adjustment to Living in the Community (α = .63) Below average 7.90 6.60 2.12 -1.30 3.58 Above average 9.36 9.45 1.53 0.09 2.58 SFS Independence Performance (α = .85) Below average 83.55 85.14 15.89 1.59 17.06 Above average 94.10 101.50 10.63 7.40 11.41
Discussion
In the present study the effectiveness of six months of cognitive adaptation training (CAT) in addition to treatment as usual (TAU) on social functioning in patients with a severe mental illness (SMI) was investigated. Social functioning was measured by means of the Multnomah Community Ability Scale (MCAS) and the Social Functioning Scale (SFS). Both scales contain multiple subscales, all related to social functioning. It was found that TAU+CAT did not outperform TAU on social functioning, disconfirming hypothesis one. Patients in both conditions remained stable in level of social functioning. However, when taking a closer look at patient factors there were indeed some effects, which could explain the above finding.
First, as expected, younger patients generally showed better social functioning than older patients (Kurtz & Richardson, 2011). However, the level of functioning in both groups remained stable, younger patients did not respond better to TAU+CAT than older patients, disconfirming hypothesis two. Second, as expected, women generally showed better social functioning than men (Childers & Harding, 1990). But again, women did not respond better to TAU+CAT than men on social functioning, disconfirming hypothesis two. Third, concerning verbal memory it was found that patients with worse verbal memory generally showed equal social functioning to those with better verbal memory. For patients in the TAU group this was the same after six months. However, for the patients in the TAU+CAT condition baseline verbal memory was a predictor of social functioning six months later. This was found for the MCAS, the MCAS Adjustment to Living in the Community subscale, MCAS Social Competence subscale and the SFS Interpersonal Performance subscale. Patients with worse verbal memory actually deteriorated in terms of social functioning, as measured by above mentioned scales, after six months of TAU+CAT, whereas patients with better verbal memory improved in terms of social functioning after six months of TAU+CAT.
Thus, it seems that worse verbal memory predicts lower social functioning after six months of TAU+CAT and might therefore be a counter-indicator for TAU+CAT, whereas better verbal memory predicts better social functioning after six months of TAU+CAT and is therefore a good indicator for TAU+CAT. The same was found for executive functioning, however this effect was less pronounced. An explanation of this
finding could be that comprehension of the signs and signals used in CAT requires better verbal memory than some patients have, which could make these particular patients confused resulting in doing less activities and having less social contacts thus lowering social functioning. Whereas for patients with better verbal memory, the signs and signals are helpful, resulting in better social functioning. In spite of this differential result between patients with worse or better verbal memory in the TAU+CAT condition, the group as a whole showed no difference, explaining the result and disconfirmation of hypothesis one, of no effect of TAU+CAT over six months in comparison to solely TAU. Considering this effect of verbal memory on social functioning it is suggested to replicate this study in future CAT interventions to be sure of its validity. If again the same results show up, indicating that the level of verbal memory is predictive for the result of the CAT intervention, it is important to not only measure verbal memory at the start of the intervention to determine which behavioral type the patient has and tailor the intervention for the patient but also to decide whether or not a patient will profit from the intervention. If the patient has a below expected average verbal memory, the CAT intervention could be harmful instead of profitable, and should thus not be given to the patient.
A possible limitation of the current study is the number of participants. The number of participants for the LMM showed to be critical, concerning the number of predictors in relation to the number of participants. Even though ten participants per predictor are appropriate, a number of 30 participants per predictor would be ideal (Wilson VanVoorhis & Morgan, 2007). A higher number of participants would make the findings more robust with better power to detect a small effect size (Wilson VanVoorhis & Morgan, 2007). However, other studies on CAT have included less participants, or both in- and outpatients (Kidd et al., 2014; Velligan et al., 2000; Quee et al., 2014). Even though all nurses received the same CAT training, it still remains possible that there were subtle differences in implementing CAT in the three different institutions, as the institutions do not work together and thus possibly have different management strategies. It is therefore possible that an effective treatment result in one institution could be obscured by less effective treatment results of the other institutions. To test this, exploratory analyses were done (see Section Exploratory Analyses), where it was found that participants in the TAU+CAT condition from the three different institutions did not
differ in social functioning after six months of TAU+CAT. Thus possible differences between institutions did not influence the effect of TAU+CAT.
Concerning treatment integrity, it was examined whether CAT was given as intended. Even though CAT should not result in extra time from the nurses spent with their patients, nurses still might have given extra attention to their patients receiving CAT due to unintentional motivation and enthusiasm of the nurses. Exploratory analyses on how much time the nurses spent with their patients disconfirmed this (see Section Exploratory
Analyses). However, this result should be interpreted with care as nurses mentioned that
they found it hard to gauge how much time they had spent with their patients.
Despite the found effects of verbal memory on several measures of social functioning no clinically relevant change was found when the participants from the TAU+CAT condition were divided in either below average or above average verbal memory, even though participants from with better verbal memory generally showed better and more improvement on social functioning. To find a clinically relevant change an even larger effect is needed as participants are no longer compared to a control group but are compared to themselves. However, six months of an intervention is relatively short, especially for patients with a severe mental illness. Typically, the length of treatments given in randomized controlled trials is nine months and previous studies on CAT were also conducted over a period of nine months of CAT (Draper et al., 2009; Velligan et al., 2000; Velligan et al., 2008). As the current study is part of a longitudinal study of 12 months, it might be possible that a clinical relevant change can be found at the end of the longitudinal study.
Summarizing the results of the present study, it can be concluded that TAU+CAT as an intervention for patients with an SMI can be both helpful and harmful for their social functioning, depending on their verbal memory capacities. It is therefore highly suggested to take verbal memory into account before starting a CAT intervention.
Literature
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.
Barker, S., Barron, N., McFarland, B. H., & Bigelow, D. A. (1994). A community ability scale for chronically mentally ill consumers: Part I. Reliability and validity.
Community Mental Health Journal, 30, 363-383.
Birchwood, M., Smith, J., Cochrane, R., Wetton, S., & Copestake, S. (1990). The Social Functioning Scale: The development and validation of a new scale of social adjustment for use in family intervention programmes with schizophrenic patients. British Journal of Psychiatry, 157, 853-859.
Bosch, van den, R. J. (2000). Schizofrenie en andere psychotische stoornissen. In W. VanderEycken, C. A. L. Hoogduin, & P. M. G. Emmelkamp (Eds.), Handboek
psychopathologie deel 1 Basisbegrippen (pp. 115-163). Houten: Bohn Stafleu van
Loghum.
Bouma, A., Mulder, J., Lindeboom, J., & Schmand, B. (2012). Handboek
neuropsychologische diagnostiek. Amsterdam: Pearson Assessment and
Information B.V.
Breier, A., Schreiber, J. L., Dyer, J., & Pickar, D. (1991). National Institutes of Mental Health longitudinal study of chronic schizophrenia: Prognosis and predictors of outcome. Archives of General Psychiatry, 48, 239-246.
Childers, S. E., & Harding, C. M. (1990). Gender, premorbid social functioning, and long-term outcome in DSM-III schizophrenia. Schizophrenia Bulletin, 16, 309-318.
Corbiere, M., Crocker, A. G., Lesage, A. D., Latimer, E., Ricard, N., & Mercier, C. (2002). Factor structure of the Multnomah Community Ability Scale. Journal of
Nervous and Mental Disease, 190, 399-406.
Delespaul, P. H., & de Consensusgroep EPA (2013). Consensus over de definitie van mensen met een ernstige psychische aandoening (EPA) en hun aantal in Nederland. Tijdschrift voor psychiatrie, 55, 427-438.
Dickerson, F. B., Origoni, A. E., Pater, A., Friedman, B. K., & Kordonski, W. M. (2003). An expanded version of the Multnomah Community Ability Scale: Anchors and interview probes for the assessment of adults with serious mental illness.
Community Mental Health Journal, 39, 131-137.
Draper, M. L., Stutes, D. S., Maples, N. J., & Velligan, D. I. (2009). Cognitive adaptation training for outpatients with schizophrenia. Journal of Clinical Psychology, 65, 842-853.
Eckert, S. L., Diamond, P. M., Miller, A. L., Velligan, D. I., Funderburg, L. G., & True, J. E. (1996). A comparison of instrument sensitivity to negative symptom change.
Psychiatry Research, 63, 67-75.
Gelman, A., Hill, J., Yajima, M. (2012). Why we (usually) don’t have to worry about multiple comparisons. Journal of Research on Educational Effectiveness, 5, 189- 211.
Green, M. F. (1996). What are the functional consequences of neurocognitive deficits in schizophrenia? Am J Psychiatry, 153, 321-330.
Green, M. F., Kern, R. S., Braff, D. L., & Mintz, J. (2000). Neurocognitive deficits and functional outcome in schizophrenia: Are we measuring the “right stuff”?
Schizophrenia Bulletin, 26, 119-136.
Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology. 59, 12-19.
Kern, R. S., Green, M. F., & Satz, P. (1992). Neuropsychological predictors of skills training for chronic psychiatric patients. Psychiatry Research, 43, 223-230. Kidd, S. A., Herman, Y., Barbic, S., Ganguli, R., George, T. P., Hasan, S., et al. (2014).
Testing a modification of cognitive adaptation training: Streamlining the model for broader implementation. Schizophrenia Research, 156, 46-50.
Kurtz, M. M., & Richardson, C. L. (2011). Social cognitive training for schizophrenia: A meta-analytic investigation of controlled research. Schizophrenia Bulletin, 1-13. Mueser, K. T., Blachard, J. J., & Bellack, A. S. (1995). Memory and social skill in
schizophrenia: The role of gender. Psychiatry Research, 57, 141-153. Nelson, H. E. (1976). A modified card sorting test sensitive to frontal lobe defects.
Cortex, 12, 313-324.
Quee, P. J., Stiekema, A. P. M., Wigman, J. T. W., Schneider, H., van der Meer, L., Maples, N. J., et al. (2014). Improving functional outcomes for schizophrenia patients in the Netherlands using cognitive adaptation training as a nursing intervention – A pilot study. Schizophrenia Research, 158, 120-125.
Saan, R., Deelman, D. (1986) In: A. Bauma, J. Mulder, & J. Lindeboom (Eds),
Neuropsychologische diagnostiek handbook (pp. 13-28). Lisse: Swets &
Zeitlinger Publishers.
Psychometrische eigenschapen en Nederlandse normen. Tijdschr Gerontol Geriat,
39, 64-76.
Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to basic
and advanced multilevel modeling. London: Sage Publishers.
Stiekema, A. P. M., Quee, P. J., Dethmers, M., van den Heuvel, E. R., Redmeijer, J. E., Rietberg, K., et al. (2015). Effectiveness and cost-effectiveness of cognitive adaptation training as a nursing intervention in long-term residential patients with severe mental illness: Study protocol for a randomized controlled trial. Trials, 16, 1-10.
Velligan, D. I., & Bow-Thomas, C. C. (2000). Two case studies of cognitive adaptation training for outpatients with schizophrenia. Psychiatric Services, 51, 25-29. Velligan, D. I., Bow-Thomas, C. C., Mahurin, R. K., Miller, A. L., & Halgunseth, L. C.
(2000). Do specific neurocognitive deficits predict specific domains of community function in schizophrenia? Journal of Nervous & Mental Disease,
188, 518-524.
Velligan, D. I., Bow-Thomas, C., Huntzinger, C., Ritch, J., Ledbetter, N., Prihoda, T. J., et al. (2000). Randomized controlled trial of the use of compensatory strategies to enhance adaptive functioning in outpatients with schizophrenia. The American
Journal of Psychiatry, 157, 1317-1328.
Velligan, D. I., Diamond, P. M., Mintz, J., Maples, N., Li, X., Zeber, J., Ereshefsky, L. (2008). The use of individually tailored environmental supports to improve medication adherence and outcomes in schizophrenia. Schizophrenia Bulletin, 34, 483-493.
Velligan, D. I., Mahurin, R. K., Diamond, P. L., Hazleton, B. C., Eckert, S. L., & Miller, A. L. (1997). The functional significance of symptomatology and cognitive function in schizophrenia. Schizophrenia Research, 25, 21-31.
Velligan, D. I., Mahurin, R. K., True, J., Lefton, R. Flores, C. V. (1996). Preliminary evaluation of cognitive adaptation training to compensate for cognitive deficits in schizophrenia. Psychiatry Serv., 47, 415-417.
Wilson VanVoorhis, D. R., & Morgan, B. L. (2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for
Psychology, 3, 43-50.
