Cross-sectional assessment of the factors associated with occupational functioning in patients with schizophrenia.

BACKGROUND: Studies from other countries indicate that occupational skills training can improve the social functioning and the quality of life of patients with schizophrenia but there is little research about the relationship of occupational skills and the functional status of patients in China. AIM: Use a translated Chinese version of the Comprehensive Occupational Therapy Evaluation scale to evaluate occupational functining in inpatients and recently discharged patients with schizophrenia and assess the relationship of occupational functioning to demographic, clinical and cognitive measures.
METHODS: Thirty-five inpatiens and 29 recently discharged outpatients with schizophrenia were evaluated by trained clinicals using the COTE, the Positive and Negative Syndrome Scale (PANSS) and a neuropsychological battery that included the Wisconsin Card Sorting Test (WCST), the Continuous Perfomance Test (CPT), the digit symbol-coding subtest from the Wechsler Adult Intelligence Scale (WAIS), and Trail Making Test Parts A and B (TMT-A, TMT-B).
RESULTS: The total score on the COTE and the three COTE dimensional scores (evaluating general behavior, interpersonal communication and task behavior) were all strongly correlated with the PANSS total score and the PANSS positive symptom, negative symptom and general pathology subscale scores (ranked correlation coefficients range from 0.40 to 0.90). The correlationship of the COTE measures was significantly greater with the PANSS negative symptom score than with the PANSS positive symptom score. The COTE scores were also significantly correlated with the Continuous Performance Test measures, the WAIS digit symbol-coding test scores and some, but not all, of the measures derived from the TMT-A, the TMT-B, and the WCST. Mutiple regression analyses found that the four COTE measures of occupational functioning were most strongly associated with either the PANSS total score or the PANSS negative symptom score and secondarily associated with time to complete TMT-B, the CPT measure on number of omissions, and the respondent's years of education.
CONCLUSION: Occupational functioning measures of inpatients and recently discharged outpatients with schizophrenia are closely related to the severity of psychiatric symptoms and, to a lesser extent, with cognitive functioning measures and duration of education.

Introduction

Schizophrenia is associated with a significant decrease in occupational functioning. Less than 20% of individuals with schizophrenia can maintain regular employment.[1] Overcoming this problem is one of the main tasks facing mental health professionals. The combination of antipsychotic medication, psychotherapy and occupation skills training yield the best rehabilitation outcomes for patients with schizophrenia.[2] One controlled trial[3] reported a 65% re-employment rate in patients with schizophrenia who received occupational skills training (versus 26% in the control group) and several studies[4]–[6] report that skills training also results in improvements in social functioning and quality of life. Some studies suggest that outcomes of occupational skills training for indiviudals with schizophrenia are better in women than in men.[7]

Few studies in China have systematically assessed the effects of occupational skills training. The studies that have been done[8] are largely focused on chronic patients and only indirectly assess occupational functioning by evaluating changes in psychiatric symptoms. This cross-sectional study uses a translated version of the Comprehensive Occupational Therapy Evaluation Scale[9] (COTE) to evaluate the occupation skills of patients with schizophrenia and assesses the demographic and clinical factors that are associated with occupational functioning.

Subjects and Methods

Subjects

The enrollment of patients is shown in Figure 1. Current inpatients and outpatients discharged within the last month from one ward at the Mental Health Center of the West China Hospital at Sichuan University from May 2010 to June 2010 were potential participants. Inclusion criteria were a) meeting the diagnostic criteria of schizophrenia listed in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR);[10] b) 16-45 years of age; c) a least three years of formal education and able to understand the testing materials; d) currently being treated with a second-generation antipsychotic medication; e) no serious physical illnesses or disabilities; f) no history of a serious head injury (unconscious for > 2 hours); and g) no mental retardation, dementia, substance abuse or other co-morbid mental disorder.

Figure 1.

Flowchart for the study

All subjects and/or their guardians signed informed consent forms.The study was approved by the Ethics Committee of the West China Hospital at Sichuan University.

Measures

Demographic and psychiatric history variables

A standardized form was developed to obtain basic demographic information and information about patients' duration of illness, number of relapses, medication use, and primary symptoms.

Assessment using the Comprehensive Occupational Therapy Evaluation Scale (COTE)

COTE is a 26-item scale that has a total score and three dimension scores: ‘general behavior’ (7 items); ‘interpersonal communication’ (6 items); and ‘task behavior’ including focused attention, coordination, learning, planning and so forth (13 items). Each item is coded on a 5-point scale (0=no problem, 4=severe problem). The total score (range 0-104) and dimension scores are the sum of the respective item scores; higher scores represent poorer functioning. The instrument has been translated into Chinese. [The translated version has not yet been published; it is available from the author on request.] The internal consistency of the total score of the Chinese version of the COTE was excellent (alpha=0.95) and the internal consistency of the three COTE dimensional scores are good (alpha values all >0.79). The inter-rater reliability for the total COTE score of the two clinicians trained in the use of the scale for the current study (based on simultaneous evaluation of 23 patients) was excellent (ICC=0.91).

Assessment of psychiatric symptoms

The severity of psychiatric symptoms was assessed using The Positive and Negative Symptom Scale[11] (PANSS) which has a total of 33 items, including 7 items that assess positive symptoms, 7 items that assess negative symptoms, 16 items that assess general psychopathology, and 3 supplementary items.

Neuropsychological battery

Cognitive functioning was assessed using a neuropsychological battery[12] that included the Wisconsin Card Sorting Test[13] (WCST); the Continuous Performance Test[14] (CPT); the digit symbol-coding subtest from the 3rd Edition of the Wechsler Adult Intelligence Scale[15] (WAIS-III); and the Trail Making Test[16] Parts A and B (TMT-A, TMT-B). The WCST evaluates working memory, flexibility and executive functioning; the CPT evaluates general learning, perceptual discrimination, flexibility, and motivation; and WAIS digit symbol-coding test assessed processing speed; and the TMT-A and TMT-B evaluate visual attention and task shifting.

Administration of instruments

The different measures were administered by one of two trained psychiatrists in a single 40-60 minute session that was held in a special room for psychological testing.

Statistical analysis

The data were analyzed using SPSS16.0. The total and dimensional COTE scores were correlated with demographic, clinical and cognitive variables using Spearman correlation coefficients. More than 20 variables were correlated with the occupational functioning scores so the level of statistical significance was set at p<0.002. Multivariate regression analysis was used to identify factors that are independently associated with the total COTE score and with each of the three COTE dimensional scores. Six independent variables were forced into each of the four regression models; the variables selected were the most strongly correlated demographic, PANSS, CPT, WCST, and TMT test variables from the univariate analysis and the WAIS digit symbol-coding test result.

Results

The demographic characteristics of the 64 patients included in the study and their basic results for all of the instruments employed in the study are presented in Table 1. The 64 patients include 33 men and 31 with a mean age of 25 and a median duration of illness of 3 years. There were no statistically significant differences in demographic, occupational functioning, clinical, or cognitive functioning variables by gender.

Table 1.

Characteristics and basic results for all subjects and by gender

Variables	All subjects(n=64)	Males (n=33)	Females (n=31)	statistic	p-value
Inpatient (n, %)	35 (55%)	18(55%)	17(55%)	Chi=0.001	0.981
Age in years (mean,sd)	24.8 (9.2)	25.1 (1.7)	24.5 (1.7)	t=0.26	0.796
Months duration of illness (median,IQR)	36 (6-51)	36 (60-10)	26 (48-5)	Ua=-0.13	0.899
Years of education (mean,sd)	11.6 (3.2)	12.2 (3.1)	11.0 (3.4)	t=1.39	0.170
Number of episodes (mean,sd)	1.8 (0.9)	1.8 (0.9)	1.8 (0.9)	t=-0.11	0.910
COTE results
Total score (median,IQR)	10.0 (6.0-19.0)	10.0 (5.5-24.5)	9.0 ( 6.0-17.5 )	Ua=0.16	0.870
—general behavior (median,IQR)	1.0 (0.0-2.0)	1.0 ( 0.0-2.0 )	1.0 ( 0.0-2.0 )	Ua=0.51	0.607
—interpersonal communication (median,IQR)	1.0 (0.0-3.0)	1.0 ( 0.0-3.5)	2.0 ( 0.5-3.0 )	Ua=0.36	0.721
—task behavior (median,IQR)	8.0 (4.0-16.3)	8.0 ( 4.0-20.0 )	8.0 ( 4.0-15.0 )	Ua=0.60	0.548
PANSS results
Total score (mean,sd)	100.3 (26.1)	100.7 (27.2)	99.8 (25.3)	t=1.38	0.891
—positive syndrome score (mean,sd)	10.9 (3.71)	11.3 (4.4)	10.2 (3.4)	t=1.07	0.290
—negative syndrome score (mean, sd)	12.3 (5.3)	12.2 (5.1)	12.4 (5.5)	t=-1.12	0.906
—general psychopathology score (mean,sd)	24.6 (7.1)	25.1 (2.9)	24.1 (7.3)	t=0.59	0.557
WAIS-III results
Digit symbol-coding test (mean,sd)	58.2 (20.1)	54.6 (17.0)	61.8 (22.7)	t=-1.40	0.168
TMT-A, TMT-B results
Seconds to compute TMT-A (mean,sd)	1.0 (0.9)	0.9 (1.1)	1.0 (0.8)	t=-1.45	0.883
Errors during TMT-A (median,IQR)	0.0 (0.0-0.0)	0.0 (0.0-0.0)	0.0 (0.0-0.0)	Ua=-1.32	0.895
Times pen leaves paper during TMT-A (median,IQR)	0.0(0.0-1.0)	0.0 (0.0-1.0))	0.0 (0.0-1.3)	Ua=-0.77	0.444
Seconds to compute TMT-Bb (mean,sd)	2.6 (2.6)	2.3 (2.1)	2.9 (3.1)	t=-0.91	0.367
Errors during TMT-Bb (median,IQR)	0.0 (0.0-1.0)	0.0 (0.0-1.0)	0.0 (0.0-1.0)	Ua=-0.83	0.408
Times pen leaves paper during TMT-Bb (median,IQR)	1.0 (0.0-3.0)	1.0 (0.0-4.0)	1.0 (0.0-3.0)	Ua=-0.55	0.586
CPT results
Number of errors (mean, sd)	7.8 (17.4)	8.9 (15.8)	6.6 (19.1)	t=0.50	0.623
Number of omissions (mean,sd)	7.7 (9.8)	9.6 (10.8)	5.9 (8.5)	t=1.49	0.141
Reaction time in seconds (mean,sd)	699.0 (147.4)	725.2 (149.5)	672.3 (142.9)	t=1.39	0.170
WCST results
Number of categories completed (mean,sd)	4.9 (1.0)	4.7 (1.1)	5.1 (1.0)	t=-1.48	0.144
Total time of complete (mean,sd)	530.7 (376.4)	582.6 (472.1)	471.5 (259.1)	t=1.13	0.263
Response time (mean,sd)	216.5 (150.2)	223.3 (179.7)	208.7 (122.3)	t=0.37	0.716
Correct response time (mean,sd)	314.2 (240.1)	359.3 (302.0)	262.8 (155.5)	t=1.56	0.125

COTE=Comprehensive Occupational Therapy Evaluation scale; PANSS=Positive and Negative Syndrome Scale

WAIS-III=3rd edition of Wechsler Adult Intelligence Scale; CPT=Continuous Performance Test; TMT-A, TMT-B=Trail Making Test Parts A and B

WCST=Wisconsin Card Sort Test IQR=Interquartile Range (25%-75% percentiles)

aU-value from Mann-Whitney rank test

bTwo patients were unable to complete the trail making Test B so the ‘n’ for these measures is 62, not 64

Correlation analysis of occupation skill scores with demographic, clinical and cognitive variables

The ranked correlations of the total COTE score and the three COTE dimension scores with the demographic, clinical and cognitive measures are presented in Table 2. Using the value of p<0.002 as the cut-off for significance (see methods), years of education was the only demographic and illness history variable that was significantly related with the occupational functioning measures. Individuals with higher levels of education had better (lower) scores on the occupational functioning scales.

Table 2.

Ranked correlation of occupational functioning parameters with other variables in 64 patients with schizophreniaa

Variables	Total COTE score r (p-value)	COTE dimension scores
		general behavior r (p-value)	interpersonal communication r (p-value)	task behavior r (p-value)
Inpatient (inpatient=1; outpatient=0))	0.12 (0.381)	0.17 (0.204)	0.26 (0.052)	0.23 ( 0.081)
Age in years	-0.18 (0.049)	-0.26 (0.049)	-0.27 (0.043)	-0.10 (0.450)
Male (male=1; female=0)	0.17 (0.208)	0.09 (0.518)	0.10 (0.479)	0.02 (0.886)
Months duration of illness	-0.03 (0.804)	-0.03 (0.804)	-0.10 (0.479)	0.02 (0.886)
Years of education	-0.62 (<0.001)	-0.57 (<0.001)	-0.52 (<0.001)	-0.67 (<0.001)
Number of episodes	0.31 (0.017)	0.34 (0.009)	0.24 (0.075)	0.31 (0.017)
PANSS results
Total score	0.82 (<0.001)	0.75 (<0.001)	0.81 (<0.001)	0.77 (<0.001)
—positive symptom score	0.59 (<0.001)	0.50 (<0.001)	0.52 (<0.001)	0.40 (0.002)
—negative symptom score	0.90 (<0.001)	0.74 (<0.001)	0.80 (<0.001)	0.84 (<0.001)
—general psychopathology score	0.82 (<0.001)	0.78 (<0.001)	0.84 (<0.001)	0.71 (<0.001)
WAIS-III results
Digit symbol-coding test	-0.52 (<0.001)	-0.59 (<0.001)	-0.47 (<0.001)	-0.55 (<0.001)
TMT-A, TMT-B results
Seconds to compute TMT-A	0.45 (<0.001)	0.55 (<0.001)	0.48 (<0.001)	0.48 (<0.001)
Errors during TMT-A	0.35 (0.008)	0.24 (0.075)	0.23 (0.084)	0.48 (<0.001)
Times pen leaves paper during TMT-A	0.37 (<0.001)	0.32 (0.016)	0.33 (0.013)	0.57 (<0.001)
Seconds to compute TMT-Bb	0.60 (<0.001)	0.56 (<0.001)	0.44 (<0.001)	0.61 (<0.001)
Errors during TMT-Bb	0.38 (0.001)	0.36 (0.005)	0.28 (0.033)	0.58 (<0.001)
Times pen leaves paper during TMT-Bb	0.54 (<0.001)	0.47 (<0.001)	0.41 (<0.001)	0.60 (<0.001)
CPT results
Number of errors	0.51 (<0.001)	0.42 (0.001)	0.44 (0.001)	0.51 (<0.001)
Number of omissions	0.42 (<0.001)	0.72 (<0.001)	0.60 (<0.001)	0.59 (<0.001)
Reaction time in seconds	0.42 (<0.001)	0.52 (<0.001)	0.32 (0.002)	0.44 (0.001)
WCST results
Number of categories completed	-0.24 (0.014)	-0.47 (<0.001)	-0.25 (0.059)	-0.31 (0.019)
Total time of complete	0.44 (<0.001)	0.47 (<0.001)	0.46 (<0.001)	0.38 (0.004)
Response time	0.40 (0.010)	0.52 (<0.001)	0.48 (<0.001)	0.38 (0.004)
Correct response time	0.35 (0.005)	0.33 ( 0.010)	0.37 (0.005)	0.35 (0.009)

COTE=Comprehensive Occupational Therapy Evaluation scale; PANSS=Positive and Negative Syndrome Scale

WAIS-III=3rd edition of Wechsler Adult Intelligence Scale; CPT=Continuous Performance Test; TMT-A, TMT-B=Trail Making Test Parts A and B

WCST=Wisconsin Card Sorting Test

a Given the large number of correlations considered, p<0.002 was set as the level of statistical significance; significant results shown in bold

bTwo patients were unable to complete the trail making Test B so the ‘n’ for these measures is 62, not 64

The PANSS total score and each of the three PANSS subscale scores were all strongly positively correlated with the COTE total score and the three COTE dimension scores. With only one exception (the correlation of the COTE task behavior score with the PANSS positive symptom score) all the correlation coefficients were > 0.50; this indicates a very close relationship between the severity of clinical symptoms and the degree of occupational dysfunction. The correlation of the PANSS negative symptom score with the COTE total score was significantly stronger than the correlation of the PANSS positive symptom score with the COTE total score (t=0.439, p<0.001); this was also true for all three COTE dimensional scores.

The total COTE score and the three COTE dimension scores were all significantly positively correlated (at approximately the same level) with the WAIS digit symbol-coding test results.

The pattern of results for the TMT-A and TMT-B varied by the type of occupational functioning that the test results were being correlated with. The three COTE dimension scores and the total COTE score were all significantly positively correlated with the time it took to complete TMT-A and TMT-B. But for the TMT-A and TMT-B error rates and rates of lifting the pen from the paper the correlations with the COTE task behavior dimension were much stronger than the correlations with the COTE general behavior and interpersonal communication dimensions, though these differences did not reach statistical significance.

The three measures of the CPT test were all significantly positively correlated with the total COTE score and with the three COTE dimension scores. More errors, more omissions and a longer performance time on the continuous performance test were closely associated with worse occupational functioning.

The correlation between the WCST measure on number of categories completed was strongest with the COTE general behavior dimension and the correlation between the WCST measures on total time and response time was stronger for the COTE general behavior and interpersonal communication dimension than for the task behavior dimension, though these differences did not reach statistical significance.

Multivariate regression analysis of factors associated with occupation skills

Table 3 shows the multivariate regression analyses of the factors that are independently related to the COTE total score and each of the three COTE dimensions scores, entering six independent variables into each of the four models as described in the statistical methods section. The COTE total score is significantly associated with the PANSS negative symptoms score and with the time it took to complete the TMT-B. The COTE General Behavior dimension score is significantly associated with the PANSS total score, the number of omissions on the CPT, and the respondent's years of education. The COTE Interpersonal Communication dimension score is significantly associated with the PANSS total score. And the COTE Task Behavior dimension score is significantly associated with the PANSS negative symptom score and the time it took to complete the TMT-B.

Table 3.

Multivariate regression analysis of demographic, clinical and cognitive factors associated with overall occupational functioning (the total COTE score) and with different components of occupational functioning in patients with schizophrenia in China

Variables		Non-standard-ized coefficient B	Standard error of B	Standard -ized coefficient B	t	p	95%CI
Total COTE score
PANSS negative symptom score		1.97	0.22	0.77	8.87	<0.001	1.52-2.42
Seconds to complete TMT-B		2.18	0.75	0.43	2.92	0.006	0.68-3.69
WAIS digit symbol-coding test score		-0.07	0.06	-0.11	-1.12	0.268	-0.19-0.05
Number of errors on CPT		-0.08	0.08	-0.09	-1.04	0.303	-0.23-0.07
Years of education		-0.33	0.34	-0.09	-0.99	0.327	-1.01-0.34
Time to complete WCST		0.00	0.00	0.03	0.47	0.641	0.00-0.00
COTE General Behavior score
Total PANSS score		0.03	0.01	0.32	2.84	0.007	0.01-0.05
Number of omissions on CPT		0.08	0.03	0.31	2.49	0.017	0.01-0.14
Years of education		-0.16	0.08	-0.24	-2.16	0.036	-0.32- -0.01
WAIS digit symbol-coding test score		-0.03	0.02	-0.26	-1.72	0.093	-0.07-0.01
WCST Response time		0.00	0.00	0.16	1.61	0.114	0.00-0.00
Seconds to complete TMT-A		-0.52	0.47	-0.16	-1.10	0.278	-1.46-0.43
COTE Interpersonal Communication score
Total PANSS score		0.06	0.01	0.55	4.58	<0.001	0.03-0.08
Years of education		-0.17	0.09	-0.23	-1.95	0.057	-0.35-0.01
WCST Response time		0.00	0.00	0.16	1.47	0.150	0.00-0.00
Number of omissions on CPT		0.04	0.04	0.14	1.08	0.286	-0.03-0.11
Seconds to complete TMT-A		-0.49	0.55	-0.14	-0.90	0.376	-1.60-0.62
WAIS digit symbol-coding test score		-0.01	0.02	-0.06	-0.38	0.704	-0.05-0.03
COTE Task Behavior score
PANSS negative symptom score		1.11	0.16	0.63	6.84	<0.001	0.79-1.44
Seconds to complete TMT-B		0.41	0.19	0.21	2.22	0.031	0.04-0.79
Years of education		-0.44	0.23	-0.17	-1.87	0.069	-0.91-0.04
WAIS digit symbol-coding test score		-0.02	0.04	-0.04	-0.42	0.676	-0.10-0.07
WCST Response time		0.00	0.00	0.11	-0.12	0.908	0.00-0.00
Number of omissions on CPT		0.00	0.09	0.00	0.01	0.998	-0.17-0.17

COTE=Comprehensive Occupational Therapy Evaluation scale; PANSS=Positive and Negative Syndrome Scale

WAIS-III=3rd edition of Wechsler Adult Intelligence Scale; CPT=Continuous Performance Test; WCST=Wisconsin Card Sorting Test

Discussion

Main findings

We found that occupational functioning — as assessed by the COTE — in inpatients and recently discharged outpatients with schizophrenia with a median duration of illness of 3 years are closely related to the severity of psychiatric symptoms and, to a lesser extent, with cognitive functioning measures and duration of education.

In this group of patients the occupational functioning measures were not related to gender, age, inpatient versus outpatient status or duration of illness, though they were weakly correlated with the number of prior episodes. In the univariate analysis occupational functioning measures are correlated with all the measures derived from the CPT, with the WAIS digit symbol-coding test, and with most of the measures derived from the TMT-A, the TMT-B, and the WCST (with some variation in results for the different dimensions of occupational functioning); but most of these associations become insignificant after adjustment for the PANSS clinical symptom scores in the multivariate analyses. Negative symptoms were more closely associated with the occupational functioning measures than positive symptoms, a finding that is consistent with previous reports that find improvements in the negative symptoms of patients who participate in occupational skills-training programs.[17],[18]

Limitations

The main instrument used to assess occupational functioning in the study, the COTE, has not previously been used in China so additional work will be needed to demonstrate its reliability, validity and appropriateness for different types of patients. In particular, the validity of using the same instrument to assess occupational functioning in both inpatients and outpatients needs to be carefully evaluated.

This was a cross-sectional study in which the evaluation of occupational functioning and the assessment of clinical symptoms and cognitive functioning was made by the same clinician at the same point in time. This methodology increases the likelihood of correlated ratings and makes it impossible to determine the causal direction of the identified associations. (Do negative symptoms lead to poor occupational skills or do poor occupatonal skills magnify negative symptoms?) Future studies need to have the evaluations of occupational functioning made independently from the evaluations of clinical and cognitive measures and, more importantly, need to follow the different measures over time to clarify the causal relationship of the variables.

These results in current inpatients and recently discharged patients with a relatively short duration of illness may not be relevant for patients with chronic illness or for those who have been clinically stable for a period of time. In particular, the much stronger relationship of occupatinal functioning measures to clincal symptoms than to cognitive functioning measures may not be the case among patients with schizophrenia whose symptoms are stable or in remission.

Significance

There is, as yet, no widely accepted method of assessing rehabilitative outcomes for patients with severe mental illnesses in China. As the emphasis on community-based services with a strong rehabilitative component increases in the country, the need for such measures will become more and more evident. We believe that the Chinese version of the COTE measure used in the current study is a good candidate measure for this purpose that deserves further evaluation. It is easy to use and provides information on different components of occupational functioning (general behavior, interpersonal communication, and task-oriented behavior) that can be used in the design and evaluation of individualized rehabilitation plans. For example, individuals with deficits in general behavior could be helped in improving their self-care and basic activities of daily living,[19] individuals with deficits in interpersonal commnication could be trained to proactively initiate interactions with others, and individuals with deficits in task-oriented behaviors could be encouraged to focus on time management and on onging self-monitoring of the effectiveness of their goal-oriented behaviors.