Validity and Reliability of the Tunisian Arabic Functioning Assessment Short Test (FAST) and Functional Outcome Factors in a Sample of Patients with Bipolar Disorder

2.1. Design and Participants

This study has a cross-sectional design and was conducted between February 2022 and June 2022.

Sixty outpatients were consecutively recruited at Razi Psychiatric Hospital in Tunis/ Tunisia. Patients who fulfilled the following criteria were included: (i) clinical diagnosis of Bipolar Disorder type I or type II according to DSM 5, (ii) 18 years or older, (iii) euthymic at the time of the study with scores < 8 on the Hamilton Depression Rating Scale and < 6 on the Young Mania Rating scale), and (iv) no history of a manic or depressive episode nor psychiatric inpatient admission within the three months prior to study intake. Patients with comorbid intellectual deficiency, major cognitive impairment, or physical disability that could significantly impact their functioning were not included. The sample size was determined based on previous validation studies of the FAST.

Sixty controls were recruited among the visitors of Kassab Orthopedics Hospital in Tunis/Tunisia and were matched for sex and age. Individuals who fulfilled the following criteria were included: (i) no current depressive or (hypo)manic episode according to a screening with the Structured Clinical Interview for Diagnosis and Statistical Manual of Mental Disorders-Fourth Edition (SCID), (ii) no personal or first-degree relatives’ history of psychiatric disorders, (iii) no personal history of intellectual deficiency or major cognitive impairment, and (iv) no physical disability that could significantly impact their functioning.

2.2. Assessments and Procedures

Information was gathered on socio-demographic and clinical variables, including age, sex, marital status, education, employment status, bipolar disorder type and polarity, age of onset, number of depressive and manic episodes, treatment resistance, prior psychotic features, number and duration of hospitalizations, number of suicide attempts, current treatment, treatment compliance and comorbid psychiatric disorders, such as substance use and anxiety disorders.

The Global Assessment of Functioning Scale (GAF) and the brief version of the World Health Organization Quality of Life scale (WHOQOL-BREF) were chosen to assess concurrent validity. The first researcher recorded socio-demographic and clinical variables and administered the FAST. A second researcher administered the GAF and WHOQOL-BREF. The two interviewers were blinded to each other. Test-retest reliability was checked two weeks after the original interview in 25 patients.

2.3. The Functioning Assessment Short Test (FAST)

The FAST is an interview-administered instrument designed to be used by a trained clinician [12]. The studied time frame refers to the last two weeks before the assessment. It comprises 24 items divided into 6 specific areas of functioning: autonomy, occupational functioning, cognitive functioning, financial issues, interpersonal relationships, and leisure activities. All items are rated on a four-point Likert scale. The global score is obtained by adding the scores of each item. The higher the score, the more severe the functional impairment. The original Spanish version of the FAST has been translated into English, Italian, Portuguese, Chinese, Finnish, and Turkish.

2.4. World Health Organization Quality of Life-brief Version (WHOQOL-BREF)

The ‘WHOQOL-BREF’ was developed by the World Health Organization in 1998 to provide a succinct and ergonomic scale for assessing quality of life [15]. It is derived from the World Health Organization Quality of Life-100 questionnaire (WHOQOL-100), which was considered too long for clinical use [16].

The WHOQOL-BREF consists of 26 items. It includes two global items and 24 items divided into four general domains: physical health, psychological health, social relationships and quality of the environment [15]. Item 1 and item 2 assess global appreciation of quality of life and global health satisfaction, respectively.

The WHOQOL-BREF has four types of response scales rated from 1 to 5, allowing assessment of intensity (Not at all-Extremely), ability (Not at all-Completely), frequency (Never-Always), or assessment of satisfaction (Very dissatisfied/Very poor-Very satisfied/Very good) [17].

The WHOQOL-BREF has been translated and validated in more than ten languages, including Arabic [18]. We used the Tunisian version as the external reference questionnaire for the study of the convergent validity of the FAST.

2.5. Global Assessment of Functioning (GAF)

Global Assessment of Functioning (GAF) is a single measure of overall psychosocial impairment caused by mental factors ranging from 0 to 100. It constitutes Axis V of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Revised Text DSM-IV-TR. The clinician chooses to give the patient a score from 0 to 100, which corresponds to his or her overall functioning (psychological, professional, and social). The higher the score, the more satisfactory the patient's functioning [8].

2.6. Validation Procedure

After obtaining authorization from the original authors, the English version was translated into Tunisian Arabic and back-translated to English according to Brislin’s back-translation method.

The preliminary version was evaluated by a committee of experts to analyze the clarity, relevance, and discrimination of each item. The revised version was later presented to a pre-test population of twenty patients with bipolar disorders. Changes were made based on the recommendations of the experts and participants.

2.7. Statistical Analysis

Statistical analysis was performed using SPSS for Windows- Version 26.0. Internal consistency was measured using Cronbach’s alpha and the item-scale correlation using the Spearman correlation coefficient. Test-retest reliability was assessed using the intra-class correlation coefficient and the Wilcoxon rank test using two matched samples.

Concurrent validity was measured by Spearman’s correlation coefficient to examine correlations between GAF and FAST (total score and dimension scores) and WHOQOL-BREF and FAST (total score).

Discriminant validity was measured by the Mann-Whitney coefficient to compare the FAST scores between patients and control subjects. The cut-off value between patient and control groups was determined by the Receiver Operating Characteristic (ROC) curve analysis.

For the univariate analysis, we examined the correlations of socio-demographic and clinical variables with impaired functioning in the patient group according to the cut-off score of the FAST (dependent variable). The following statistical tests were used: Fisher's exact test, Chi-squared test, and Mann-Whitney test. The significance level was set at 0.05.

Multivariate binary logistic regression was used to identify independent factors that might influence patient functioning according to the FAST score (dependent variable). We included variables with an acceptable significance level in a multivariate logistic regression model as well as variables identified in the literature as predictors of impaired functioning in people with BD [19, 20]. The degree of dependence between impaired functioning and the various variables was expressed by the Odds Ratio (OR). The confidence interval of the Odds Ratio was set at 95%.

3.1. Sociodemographic and Clinical Characteristics

Sixty patients with BD (BP I: fifty-three, BP II: seven) participated in the study. Males comprised 58% of the study population, with a sex ratio of 1.4. The mean age of participants was 45.43 ± 12.03 (ranging from 23 to 70 years). Table 1 shows the sociodemographic and clinical features of both patient and control groups.

3.2. Reliability Analysis

The internal consistency obtained was high, with a Cronbach’s alpha of 0.785. The Cronbach’s Alpha in each functioning domain also indicated good internal consistency, ranging from 0.815 to 0.735, except for leisure time (0.590) (Table 2).

All six FAST domains had significant correlations with each other and with the total score (p<0.001). The statistical correlation was strongest between the first two functioning domains: autonomy and professional activity (Table 3).

The intraclass correlation coefficient was 0.998, indicating high test-retest reliability (Table 2).

Following the comparison of the patients’ mean scores between baseline and week 2 using the Wilcoxon test, we did not find any statistically significant difference in the mean values between T0 and T1 for the total FAST score as well as for the scores of the different FAST domains (p>0.05).

3.3. Construct Validity Analysis

Concurrent validity based on functional impairment according to the GAF scale showed a highly significant, negative, and linear correlation with the FAST scale (rho= -0,788, p<0.001). These results indicate that patients with impaired functioning assessed using the FAST obtained lower scores on the GAF scale. The GAF score also had negative correlations to each specific FAST dimension (Table 4).

The correlation between the total FAST score and domains of the WHOQOL-BREF was significant and negative: FAST and item 1 on quality of life (rho=-0.533, p<0.001), FAST and item 2 on general health (rho=-0.521, p<0.001), FAST and physical health dimension (rho=-0.635, p<0.001), FAST and psychological health dimension (rho=-0.663, p<0.001), FAST and social relationships dimension (rho=-0.582, p<0.001), and FAST and environmental health dimension (rho=-0.489, p<0.001).

To evaluate the discriminative effect of the FAST, the total scores of patient and control groups were compared. The total score of patients with BD (35.98) was significantly higher than controls (16.5) (p<0.001). The mean FAST dimensions scores of patients were approximately twice the controls’ scores (Table 5).

The optimal cut-off value was calculated using the Receiving Operating Characteristic (ROC) curve. The cut-off value of the FAST was 26 (sensitivity of 73% and specificity of 92%) to differentiate between patients and control subjects. The area under the ROC curve (AUC) was 0.886 (p<0.001).

To construct the multivariate logistic regression model, we included the six variables correlated with impaired functioning according to the univariate analysis, as well as the following variables identified in the literature: total number of episodes, number of depressive episodes, and number of hospitalizations. The results of the multivariate regression are detailed in Table 6. A lower educational level, living alone, age of onset, number of depressive episodes, and treatment association of a mood stabilizer and antipsychotics were all independently correlated with functional impairment.

4.1. Strengths and Limitations

Our study is one of the very few assessing functional outcomes in patients with bipolar disorder in the MENA region, highlighting the importance of acknowledging cultural specificities in understanding and evaluating functioning.

Data from the patient group were collected by two different clinicians. The first evaluator administered the FAST, and the second administered both the GAF and the WHOQOL-BREF, thus guaranteeing the independence of the measures and further supporting the validity of the Tunisian Arabic FAST.

The inclusion of a control group in addition to BD patients enabled the assessment of the FAST discriminant validity.

The main limitation of our study is the inclusion of patients consulting at a tertiary care center with a history of one or multiple hospitalizations, and, hence, more severely affected by BD. Therefore, the sample is not representative of the whole spectrum of bipolar disorders with various clinical severities. Moreover, internal validity was not investigated due to the relatively small sample size.

We did not include an assessment of psychotherapy interventions in the survey and their effect on functioning, given the lack of access to psychotherapy in Tunisia.