Introduction
Over the last years, a number of studies have highlighted the prevalence, severity and persistence of the psychosocial and occupational impairment often associated with bipolar disorder (Sanchez-Moreno et al. Reference Sanchez-Moreno, Martinez-Aran, Tabares-Seisdedos, Torrent, Vieta and Ayuso-Mateos2009).
Epidemiological data suggest that, even after remission of mood symptoms, many patients fail to reach premorbid social and occupational status (MacQueen et al. Reference MacQueen, Young and Joffe2001). Long-term outcome studies show that only 25–40% of patients achieve full functional recovery (Huxley & Baldessarini, Reference Huxley and Baldessarini2007).
Any treatment aimed to enhance function to the highest possible level should stem from an individualized and personalized plan (Ruggeri & Tansella, Reference Ruggeri and Tansella2012), based on an assessment of the ability to efficiently and effectively function in different areas of life and throughout different phases of the illness.
Therefore, reliable assessment tools are needed to study the factors associated with disability and to evaluate functioning in clinical practice. Wide variations can be observed across studies in methods to assess social disability in bipolar patients and only a few instruments have been used in more than two studies (Sanchez-Moreno et al. Reference Sanchez-Moreno, Martinez-Aran, Tabares-Seisdedos, Torrent, Vieta and Ayuso-Mateos2009). Moreover, most instruments, such as the Life Skills Profile (Rosen et al. Reference Rosen, Hadzi-Pavlovic and Parker1989), the Social Adjustment Scale (Schooler et al. Reference Schooler, Hogarty, Weissman, Hargreaves, Attkisson and Sorenson1979) and the Short Form-36 (Ware & Sherbourne, Reference Ware and Sherbourne1992), have a broad focus or have been developed for patients with schizophrenia and are therefore not suitable to address specific areas of impairment in bipolar disorder.
The aim of filling this gap prompted a research team from the Barcelona Bipolar Disorder Program to design a simple instrument, easy to apply, requiring a short time to be administered. This led to the development of Functioning Assessment Short Test (FAST) for the clinical evaluation of functional impairment presented by people suffering from bipolar disorders (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007). The growing importance of functional assessment in cross-cultural comparative studies and multinational trials requires instruments validated in several languages (Swaine-Verdier et al. Reference Swaine-Verdier, Doward, Hagell, Thorsen and McKenna2004). We present here data on the Italian version of FAST, to address its validity and reliability. This study was carried out in the framework of the European Network of Bipolar Research Expert Centre (ENBREC, http://www.chusa.upmc.fr/ENBREC), a project funded by the European Union designed to foster multinational collaboration among centres with expertise in the clinical management of and/or research in bipolar disorders.
Methods
Participants
The study sample included 132 adult patients with a clinical diagnosis of Diagnostic and Statistical Manual of Mental Disorders-IV bipolar disorder, recruited between 1 January 2010 and 31 May 2011 from the caseloads of the Department of Mental Health and Neuroscience, University of Siena and the Department of Mental Health, San Carlo Hospital Trust of Milan and 132 healthy controls recruited from the general population in various community settings. The patients were referred by the treating clinicians to the validation study. Patients under 18 and over 65 years were excluded.
Trained clinicians confirmed the clinical diagnosis and assessed the bipolar patients by a package of instruments developed in the framework of the ENBREC project including, in addition to FAST, demographic data, clinical information on psychiatric history, comorbidity and current symptoms, the Montgomery–Asberg Depression Rating Scale (MADRS) (Montgomery & Asberg, Reference Montgomery and Asberg1979), the Young Mania Rating Scale (YMRS) (Young et al. Reference Young, Biggs, Ziegler and Meyer1978). MADRS and YMRS were administered at the same time of FAST. Healthy controls were assessed by FAST and provided socio-demographic data. All subjects gave informed consent to the study participation.
The instrument
The FAST is an instrument designed to be administered by trained clinicians. Its time frame refers to the last 15 days before assessment. It is a quite simple instrument, which usually requires less than 10 min to apply (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007). It had been validated in Spanish (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007) and Portuguese (Cacilhas et al. Reference Cacilhas, Vieira Da Silva Magalhães, Cereser, Walz, Weyne, Rosa, Vieta and Kapczinski2009). An English version has been used, but not validated (Smith et al. Reference Smith, Griffith, Poole, di Florio, Barnes, Kelly, Craddock and Hood2011). A recent longitudinal study found FAST to be useful in both multiple-episode and first-episode patients, showing that it is sensitive to detect small changes in functioning over 1-year follow-up (Rosa et al. Reference Rosa, Gonzalez-Ortega, Gonzalez-Pinto, Echeburua, Comes, Martinez-Aran, Ugarte, Fernandez and Vieta2012).
The Italian version was developed using the original Spanish version (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, van Riel, Bonnin Roig, Ayuso-Mateos, Ferrier, Kapczinski and Vieta2008). After translation and back translation, items were discussed by the investigators' team and a final version was agreed. The manual was translated as well and it was used to train a group of clinicians to administer the instrument.
FAST comprises 24 items, which cover six specific areas of functioning:
(1) Autonomy refers to the capacity of doing things alone and making one's own decisions.
(2) Occupational functioning refers to the capacity to maintain a paid job, efficiency of performing tasks at work, working in the field in which one was educated and earning according to the level of the employment position.
(3) Cognitive functioning refers to the ability to concentrate, perform simple mental calculations, solve problems, learn new information and remember learned information.
(4) Financial issues involve the capacity of managing the finances and spending in a balanced way.
(5) Interpersonal relationships refer to relations with friends, family, involvement in social activities, sexual relations and the ability to defend ideas and opinions.
(6) Leisure time refers to the capacity of performing physical activities (sport, exercise) and the enjoyment of hobbies.
All items are rated using a 4-point scale, 0 = no difficulty, 1 = mild difficulty, 2 = moderate difficulty and 3 = severe difficulty. The total score is obtained by adding up the scores of each item. The higher the score, the higher the impairment.
Statistical analyses
Psychometrics
We examined the construct validity, the discriminant validity and the inter-rater reliability of the FAST.
Construct validity
A principal component analysis (PCA) with varimax and oblique rotation was carried out on the 24 FAST items to examine the construct validity of the instrument (Jolliffe, Reference Jolliffe1986). The number of factors was selected by inspecting the screen plot and using components with an eigenvalue >1. Cronbach's alpha was used to evaluate the internal consistency of the factors.
Discriminant validity
Discriminant validity was examined by comparing the total FAST score between patients with bipolar disorders and controls, and by comparing the total and factor scores among euthymic patients and those currently in a manic, depressive or mixed state. Multiple linear regression models were used to examine the relationship between total FAST score and diagnosis (bipolar v. control) or disorder phase (euthymic, manic, depressed and mixed) after adjusting for age, gender, education (in years), marital status (married/single) and work status (employed/other).
Receiver operating characteristic (ROC) analysis was carried out to determine the optimal cut-off on the total FAST score discriminating patients from controls. The ROC curve is a plot of the FAST total score sensitivity v. one minus specificity at each cut-off score.
Sensitivity refers to the scale's ability to correctly discriminate individuals with bipolar disorder from controls, and is calculated as the number of individuals with the disorder who score positive on the test divided by the total number of individuals with the disorder.
Specificity refers to the scale's ability to identify people with bipolar disorder, and is calculated as the number of individuals without the disorder who score negative on the test divided by the total number of individuals without the disorder. One minus specificity is the false-positive rate, i.e., the number of individuals without the disorder who score positive on the test divided by the total number of individuals without the disorder.
The area under the ROC curve may range from 0.5 if the instrument performs no better than chance in discriminating patients with bipolar disorder from controls, up to 1, in case of perfect discrimination.
Inter-rater reliability
Ten study clinicians were given each 10 clinical vignettes comprising relevant clinical information regarding fictitious patients with bipolar disorder and were asked to rate their functioning.
Inter-rater reliability was examined using the intraclass correlation coefficient (ICC). A two-way mixed effect model was used to calculate the coefficient, in which patient effects are random and rater effects are fixed. All statistical analyses were carried out using SPSS, version 20.
Results
The characteristics of 132 subjects with bipolar disorders and 132 controls are shown in Table 1.
Fig. 1. ROC curve depicting the relationship between sensitivity and one minus specificity (false-positive rate) at each cut-off of total FAST score.
Table 1. Characteristics of patients and control groups. Data are presented as mean (s.d.) or as frequency (%) or as median (interquartile range, IQR)
The controls were 60.6% female and had a mean age of 41.2 (s.d. 14.4), the patients were 62.1% female and had a mean age of 47.7 years (s.d. 12.8). Eighty-seven patients met criteria for bipolar I disorder, 32 for bipolar II and 13 for bipolar disorder Not Otherwise Specified.
The patients' current state was euthymic (N = 44), manic (N = 11), depressed (N = 59) or mixed (N = 17) and undefined for 1 subject who was not rated with the MADRS and the YMRS. Depressive state was defined by a MADRS score ≥10, manic state by YMRS >7, mixed state by MADRS score ≥10 and YMRS >7. We used the cut-offs first proposed by Chengappa et al. (Reference Chengappa, Baker, Shao, Yatham, Tohen, Gershon and Kupfer2003) and Hawley et al. (Reference Hawley, Gale and Sivakumaran2002), and later suggested by Berk et al. (Reference Berk, Ng, Wang, Calabrese, Mitchell, Malhi and Tohen2008) in their empirical redefinition of psychometric criteria for remission in bipolar disorders to identify patients who were not in remission and showed identifiable symptoms. All patients not meeting these criteria were defined as euthymic.
YMRS scores for manic, depressed, mixed and euthymic patients were 15.8 ± 12.4, 2.2 ± 2.2, 12 ± 4.7 and 1.7 ± 2.5, respectively. For the same patient groups MADRS scores were 5 ± 2.8, 20.3 ± 8.4, 19.6 ± 9 and 3 ± 2.9, respectively. Scores differed significantly among groups (YMRS: F = 52.7, df = 127, p < 0.001; MADRS: F = 62.8, df = 127, p < 0.001). Post-hoc pairwise comparisons using the Tamhane test indicate that manic and mixed patients had significantly higher scores on YMRS than euthymic and depressed patients (p < 0.001) and that depressed and mixed patients had significantly higher scores on MADRS than manic and euthymic patients.
FAST administration was straightforward and took on average 8 min. The time frame of administration was one week. 98% of patients answered all items. For the seven patients who did not answer one item, the total FAST score was calculated by replacing the missing item with the mean.
Construct validity
An exploratory PCA was carried out on the patient group with oblique (promax) rotation. The oblique rotation proved to be more suitable than orthogonal rotation because factors were assumed to be correlated.
Five factors were extracted accounting overall for 73.7% of the variance of the items. The first factor identified was ‘work functioning’ that included five items, followed by ‘interpersonal relationships’ (seven items), ‘cognitive functioning’ (six items), ‘autonomy’ (four items) and ‘finances’ (two items). All item loadings were greater than 0.40 and no item had a cross-loading, as shown in Table 2.
Table 2. Results of PCA with oblique rotation
Extraction method: principal component analysis.
Rotation method: promax with Kaiser normalization.
The five factors extracted, except for ‘finances’, exhibited a fair to moderate correlation with each other, as shown in Table 3. Cronbach's alpha for the five components was 0.96, 0.88, 0.88, 0.91, 0.92, respectively, and for the total was 0.93.
Table 3. Factor correlation matrix
Discriminant validity
Means and standard deviations of the total FAST and of the factors scores for patients and controls are provided in Table 4. The optimal cut-off on FAST total score discriminating patients from controls was 15, with a sensitivity of 0.79 and a specificity of 0.80. The ROC area was 0.88 (95% CI 0.84–0.92), denoting a good discriminant validity of the scale, as shown in Fig. 1.
Table 4. Total FAST and factor scores by diagnostic status and bipolar disorder phase. Results are presented as unadjusted means (s.d.) and adjusted means (s.e.m.)
One patient has missing phase. Linear regression models by bipolar disorder phase were carried out on the subset of participants with complete demographic data (118 patients, 125 controls).
* All differences between patients and controls were significant at p < 0.001.
$ Significant differences: euthymic v. depressed, p = 0.037.
# Significant differences: euthymic v. depressed, p < 0.001.
§ Significant differences: euthymic v. depressed, p < 0.001; euthymic v. mixed, p = 0.003.
° Significant differences: euthymic v. depressed, p < 0.001; euthymic v. mixed, p < 0.001; euthymic v. manic, p = 0.005.
^ Significant differences: euthymic v. depressed, p < 0.001; euthymic v. mixed, p = 0.002; euthymic v. manic, p = 0.033.
The total FAST and factor scores were significantly higher in patients than in controls after controlling for the effect of age, gender, education (years), marital status (married/single) and working status (employed/other) in linear regression models (Table 4).
We then compared the FAST total and factor scores among bipolar disorder phases.
Four factors and the FAST total score proved to discriminate euthymic patients from symptomatic patients in multiple linear regression models, after controlling for demographic characteristics, as shown in Table 4.
In particular, compared with euthymic patients, impairment in ‘interpersonal relationships’ was significantly higher in patients with depression, mania or in a mixed state. ‘Cognitive functioning’ and overall functioning (total FAST) were significantly poorer in depressed and mixed patients, and ‘autonomy’ and ‘work functioning’ were significantly impaired only in depressed patients. No significant differences were found on factors and FAST total scores among patients with depressive, manic and mixed states. Then, given the variability in the cut-off scores on YMRS for mania and on MADRS for depression in the literature, we replicated linear regression analyses with FAST total score as the dependent variable using a more conservative criterion for euthymia based on a score of 0–1 on Clinical Global Impression-depression and Clinical Global Impression-mania. ‘Depression’ was defined by a score >1 on Clinical Global Impression-depression, ‘mania’ by a score >1 on Clinical Global Impression-mania and ‘mixed state’ by a score >1 on both scales. Using this categorization, 33 patients were euthymic, 33 depressed, 15 manic and 50 mixed. After adjusting for demographic characteristics, estimated mean (s.e.) were 18.9 (2.7) for euthymia, 35.2 (2.9) for depression, 25.4 (4.0) for mania and 34.7 (2.1) for mixed states. Pairwise significant differences at p < 0.05 were found for euthymia v. depression and mixed states, and mania v. mixed states and depression.
Inter-rater reliability
The ICC for the total FAST score was 0.983 (95% CI 0.963–0.995).
Conclusions
The FAST Italian version confirmed in the present study the user-friendly characteristics and strong psychometric properties of the instrument, already shown by previous studies (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007). The PCA identified the same five factors found in the validation studies of the Spanish and Portuguese versions (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007; Cacilhas et al. Reference Cacilhas, Vieira Da Silva Magalhães, Cereser, Walz, Weyne, Rosa, Vieta and Kapczinski2009), but incorporated the two items of original ‘leisure time’ factor into the ‘interpersonal relationships’ and ‘cognitive function’ factors. Factor loading components were to a large extent consistent with the a priori structure of the instrument. Internal consistency was excellent for the five components and the whole scale. The ability of FAST in discriminating between patients and controls, and between patients in different mood states was confirmed. The score of euthymic patients was significantly higher than controls, thus showing once more that people with bipolar disorders can be functionally impaired, even when they are in symptomatic remission. It is worth noting that the mean score of euthymic patients in our sample and in the Spanish and Portuguese samples was similar (15.6 ± 12.3 v. 18.6 ± 13.2 v. 16).
The higher score of our patient group with respect to the original Spanish study (29.9 ± 17.2 v. 25.4 ± 16.3) was due to the higher proportion of euthymic patients in the latter (70% v. 33%). This is also reflected in the slightly higher cut-off discriminating patients from controls in our study (15 v. 11).
The study has some weaknesses. First, no additional scale to assess the concurrent validity of FAST was administered. However, concurrent validity with Global Assessment of Functioning was previously demonstrated by Rosa et al. (Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van Riel, Ayuso-Mateos, Kapczinski and Vieta2007) for the original Spanish version. Second, we found that the ability of the instrument to discriminate between bipolar disorder phases partly depended on the criteria used. This could be considered as a limitation of the instrument. However, the functional assessment cuts across diagnostic categories. Functioning assessment tools are not clinical diagnostic tools and are not necessarily aimed at discriminating between diagnostic groups. Our capacity to identify significant differences was also limited by the small sample size of patients with mania symptoms or the combination of manic and depressive symptoms.
In conclusion, the psychometric validity and reliability of FAST in the Italian population will provide the Italian clinicians and researchers with a suitable instrument to assess functional impairment in bipolar disorder, useful not only in everyday practice, but also in clinical trials, cross-cultural research and multicentre studies.
Declaration of Interests
Funding for this study was provided by the European Union's seventh framework programme FP7/2007–2013 under grant agreement 223102 and by additional funding from the European College of Neuropsychopharmacology.
Members of the ENBREC Steering Committee: C Henry (Paris, Project Coordinator), O Andreassen (Oslo), A Barbato (Milan), M Bauer (Dresden), G Goodwin (Oxford), M Leboyer (Paris), E Vieta (Barcelona).
Supplementary material
The supplementary material referred to in this article can be found online at http://journals.cambridge.org/eps
