|Home | About | Journals | Submit | Contact Us | Français|
Many patients with schizophrenia have negative symptoms, but their evaluation is a challenge. Thus, standardized assessments are needed to facilitate identification of these symptoms. Many tools have been developed, but most are based on observer ratings. Self-evaluation can provide an additional outcome measure and allow patients to be more engaged in their treatment. The aim of this study was to present a novel tool, Self-evaluation of Negative Symptoms (SNS), and demonstrate its validity. Forty-nine patients with schizophrenia and schizoaffective disorders according to DSM-5 were evaluated. Cronbach’s coefficient (α = 0.867) showed good internal consistency. Factor analysis extracted 2 factors (apathy and emotional) that accounted for 75.2% of the variance. The SNS significantly correlated with the Scale of Assessment of Negative Symptoms (r = 0.628) and the Clinician Global Impression on the severity of negative symptoms (r = 0.599), supporting good convergent validity. SNS scores did not correlate with level of insight (r = 0.008), Parkinsonism (r = 0.175) or Brief Psychiatric Rating Scale positive subscores (r = 0.253), which indicates good discriminant validity. The intrasubject reliability of the SNS revealed excellent intraclass correlation coefficients (ICC = 0.942). Taken together, the results show that the SNS has good psychometric properties and satisfactory acceptance by patients. The study also demonstrates the ability of patients with schizophrenia to accurately report their own experiences. Self-assessments of negative symptoms should be more widely employed in clinical practice because they may allow patients with schizophrenia to develop appropriate coping strategies.
Negative symptoms of schizophrenia reflect a loss or reduction of certain behaviors or normal functioning. Severe negative symptoms are found in many patients with schizophrenia (28–36%).1 The low or lack of efficacy of antipsychotics,2 the limitations of psychosocial interventions,3 and the slow and insidious progression of negative symptoms explain why these symptoms are overlooked compared to positive symptoms. Moreover, the assessment of negative symptoms is challenging due to the low inter-rater reliability and the interrelations with other dimensions, such as depression, extrapyramidal symptoms, social withdrawal secondary to positive symptoms, and institutionalization.4,5 Moreover, these symptoms are responsible for impaired social functioning6,7 and have a deleterious impact on the quality of life.8,9 In this context, the use of standardized assessment tools may be pertinent10 to improve the identification of negative symptoms and their treatment. A recent review of the literature included nearly 20 tools used to assess the negative dimension of schizophrenia.11 Recently, new scales have appeared, the most well-known being the Brief Negative Symptoms Scale (BNSS)12 and the Clinical Assessment Interview for Negative Symptoms (CAINS),13 which meet the criteria presented at a consensus conference on the agreement of negative symptoms.14
An analysis of the tools highlighted the predominance of assessments based on observer ratings (heteroassessments) over self-assessments, as well as the need to evaluate 5 negative dimensions (asociality, blunted affect, avolition, anhedonia, and alogia). Three scales were previously developed to assess the subjective experience of deficit or negative symptoms.15–17 However, these 3 older scales cannot be considered self-assessments, and 2 of them were not specifically based on negative symptoms.15,16 The recently developed Motivation and Pleasure Scale Self-Report (MAPSR)18 is a version of the CAINS self-report19 that fails to cover the 5 negative dimensions required.
Some authors have found that a majority of severely ill patients were unable to accurately report negative symptoms,20,21 whereas others have underlined that the subjective evaluation by schizophrenia patients is an outcome measure complementary to heteroevaluations.22,23 Self-assessment is pertinent as it allows the patients to evaluate their overall functioning and requires their participation and analysis of their own symptoms. Moreover, self-assessment is a time-efficient method for the initial identification of negative symptoms and could be useful for detecting negative symptoms in the early stages of schizophrenic disease. In addition to heteroevaluation, self-evaluation also provides clinical information not necessarily detected by caregivers or medical staff in a standard interview24 and can provide some information on the symptoms recognized by the patients themselves. This kind of assessment can constitute a basis for cognitive or social therapy and/or allow medical staff to provide recommendations to patients with respect to the management of their negative symptoms.25
This study aimed to present a novel tool, Self-evaluation of Negative Symptoms (SNS), and demonstrate its validity. The SNS is simple in design and content, easy to complete, and covers 5 subdomains (social withdrawal, diminished emotional range, avolition, anhedonia, and alogia).
Participants were individuals with schizophrenia (n = 23) or schizoaffective disorders (n = 26) recruited from the clinic affiliated with the University of Medicine of Caen (France). The inclusion criteria were: diagnosis of schizophrenia or schizoaffective disorders (DSM-5), age >18 years, and consent to participate. Exclusion criteria were substance abuse within the past month or mental retardation.
The demographic and clinical characteristics of the sample are provided in table 1. Overall, the sample was 79% male with a mean age of 36.6 years (SD = 11.6). All patients received antipsychotic drugs.
All participants provided informed consent. Participants attended a session of ~45–60min in which they completed all study measures in an identical sequence.
Because this study was observational, the scales were those routinely used in clinical practice. Negative symptoms were assessed by the Scale of Assessment of Negative Symptoms (SANS),26 the Clinician Global Impression on severity of negative symptoms (CGI-S negative), and the Brief Psychiatric Rating Scale (BPRS).27 Depression, positive symptoms, insight, and extrapyramidal symptoms were evaluated based on the Calgary Depression Scale for Schizophrenia (CDSS),28 the BPRS positive cluster, the insight scale (IS),29 and the Clinicians Global Impression of the Extrapyramidal Symptom Rating Scale-Abbreviated version (CGI-Parkinsonism),30 respectively.
The SANS had 25 items divided into 5 subscales: Affective Flattening or Blunting, Alogia, Avolition-Apathy, Anhedonia-Asociality, and Attention. Each of the subscales had a global severity item for that symptom domain. Each item is rated from 0 to 5, with higher scores for the more severe psychopathology. The SANS was chosen for this study because it has been widely used in clinical practice.
The CGI-S reflects the severity of negative symptoms on a 7-point scale ranging from 1 (no symptoms) to 7 (very severe).
The BPRS is an 18-item clinician-rated measure that assesses clinical psychiatric symptoms. Items are rated on 7-point Likert scale ranging from 1 (not present) to 7 (extremely severe). The total score reflects the severity of symptoms, whereas subscores reflect positive components (conceptual disorganization, suspiciousness, hallucinatory behavior, unusual thought content) and negative components (emotional withdrawal, motor retardation, and blunted affect).
The CDSS is a 9-item semi-structured interview that assesses depressive symptoms in schizophrenia. Each item is rated on a 4-point scale ranging from 0 (absent) to 3 (severe).
The IS is a self-report of 8 items to indicate the awareness of illness, the need for treatment, and the attribution of symptoms. The scores range from 0 (no insight) to 16 (excellent insight). A score < 9 reflects bad insight.
CGI Parkinsonism is a scale with scores ranging from 0 (absent) to 7 (extreme).
The principles for the design of the SNS included: that items be concise and easily understandable; that 5 domains of negative symptoms be considered with a separate subscore for each; that items be provided verbatim from patients with schizophrenia as much as possible; that items be focused on internal experiences in order to be complementary with other scales based on observer ratings; and that anticipatory and consummatory aspects of pleasure be evaluated because consummatory pleasure may be preserved in schizophrenia.31
The most original aspect of this scale is the reliance of the majority of items on verbatim reports on negative symptoms by patients with schizophrenia during focus groups. In this study, 5 groups from France involved 28 patients with schizophrenia. Each group session lasted 1h, 30min of which were devoted to a structured interview of 4–5 questions to explore the perception of negative symptoms.
This scale was designed primarily for clinical evaluations to guide practitioners in the treatment, such as cognitive or social therapy, but also useful during the prodromal and early phases of the disease to identify negative symptoms or in therapeutic trials.
The SNS includes a form for patients with 20 items and a score sheet (see supplementary data). The patient puts a cross in the box next to the response that best corresponds to his/her current feelings based on the previous week, scoring 2 (strongly agree), 1 (somewhat agree), or 0 (strongly disagree). The number of responses was voluntarily limited to 3 in order to simplify completion and avoid random responses when score ranges are too broad. The total score is the sum of the 20 items, ranging from 0 (no negative symptoms) to 40 (severe negative symptoms).
The current version of the SNS evaluates 5 subdomains (social withdrawal, diminished emotional range, avolition, anhedonia, and alogia), presenting 5 subscores comprising the sum of 4 items each:
Social withdrawal assesses social, family, and friend relationships, as well as the patient’s desire to establish new relationships (example item 1: I prefer to be alone in my corner). Diminished emotional range evaluates happiness or sadness as perceived by the patient in situations in which happiness or sadness is usually felt (example item 6: There are many happy or sad things in life but I don’t feel concerned by them).
Avolition is assessed by the patient’s difficulty with the goals he/she has set with respect to consistency in activities of daily life, his/her desire, his/her motivation, and energy (example item 13: I find it difficult to meet the objectives I set myself). Anhedonia evaluates the pleasure perceived by the patient with those around them, consummatory and anticipatory pleasure (example item 19: When I imagine doing one thing or another, I don’t feel any particular pleasure in the idea).
Alogia is assessed by his (her) patient’s perception (sample item 9: I don’t have as much to talk about as most people) and the efforts required by the patient to interact.
Patients generally completed the questionnaire in less than 5min.
The SNS was completed after the IS and before the hetero-assessments.
A second SNS was scheduled 4–8 weeks (T1) after the first (T0) but only for stable patients characterized by no increase in symptoms and no change in doses of psychotropic drugs or social therapy since the first assessment.
Analyses were performed to examine the reliability and validity of the SNS.
First, item-level statistics were examined to determine internal consistency. Cronbach’s alpha was calculated for the 20 items and the 5 subscores at both times: baseline and T1.
Next, construct validity was evaluated by principal component factor analysis (PCA) with varimax rotation on the 5 subscores in order to examine the factor structure of the SNS. The factors retained had an eigen value > 0.8. Bartlett’s sphericity test was used to determine if the PCA can compress the available information only if the null hypothesis is rejected (H0: the variables are orthogonal). This test is recommended only if the ratio (ie, number of instances divided by the number of variables) is < 5. The Kaiser-Meyer-Olkin (KMO) index was used to compare the correlations between variables and the values of partial correlations. If the KMO index is high (≈ 1), the PCA is pertinent; if the KMO index is low (≈ 0), the PCA is not relevant. KMO index values are unacceptable if <0.5, intermediate between 0.6 and 0.7, good between 0.7 and 0.8, very good between 0.8 and 0.9, and excellent between 0.9 and 1.0.
In addition, correlation analyses were performed to examine the convergent validity of the SNS with other negative scales (ie, SANS, CGI-S negative, and BPRS negative subscore) and to examine the discriminant validity of the SNS with measures of insight, psychotic, depressive, and extrapyramidal symptoms (ie, IS, BPRS positive subscore, CDSS, and CGI-Parkinsonism, respectively). When significant correlations were observed, additional correlations were examined with the SNS subscores or items. Correlations between SNS subscores and SANS global evaluations were also performed. Convergent and discriminant validities were tested with Pearson’s correlations.
Finally, the intrasubject reliability of the SNS was tested by intraclass correlation coefficients (ICCs) in 33 stabilized patients in whom no change in treatment, drugs or social therapy had been made.
Type I error was 5%. All analyses were performed on IBM SPSS 22.0 software.
Patient characteristics are given in table 1.
Cronbach’s alpha (αc) for the 20 items of the SNS at baseline (N = 49; αc = 0.867) and T1 (N = 33; αc = 0.897) and for the 5 subscores at baseline (N = 49; αc = 0.784) and T1 (N = 33; αc = 0.872) showed excellent results.
The KMO index was determined to be 0.73 in the PCA, and the Barlett test value was χ2 = 87.926 (P < .001). Thus, the data collected are suited for PCA.
Factor analysis indicated a 2-factor solution with factor 1 accounting for 54.2% (eigenvalue = 2.709) of variance and factor 2 accounting for 21.0% (eigenvalue = 1.052) (table 2). Factor 1 includes anhedonia, avolition, asociality, and alogia subscores, and factor 2 encompasses diminished emotional range. Therefore, both factors suggest apathy and emotional components, as previously found in other factorial analyses of negative symptom scales.32,33
Significant positive correlations were observed between the total scores of the SNS and the total scores of SANS global evaluations (r = 0.628, P < .0001), CGI-S negative (r = 0.599, P < .0001), and BPRS negative subscores (r = 0.298, P = .037). Moreover, alogia, avolition, and social withdrawal SNS subscores correlated significantly with alogia (r = 0.43, P < .01), avolition/apathy (r = 0.39, P < .01), and anhedonia/asociality (r = 0.40, P < .01) in SANS. Diminished emotional range and anhedonia from the SNS subscores failed to correlate significantly with blunted affect and anhedonia/asociality from the SANS global evaluations (r = 0.23, P = .10 and r = 0.24, P = .09 respectively). Among the 4 items constituting diminished emotional range in the SNS, only 1 item (item 8: “It is difficult for people to know how I feel”) correlated with blunted affect in SANS (r = 0.37, P = .0093).
SNS scores did not correlate with the level of insight (r = −0.008, P = .957), Parkinsonism (r = 0.175, P = .230), or the BPRS positive subscores (r = 0.253, P = .079). However, a positive correlation was found between the SNS and CDSS (r = 0.495, P < .0001). A positive correlation was also observed between the CDSS and the SANS (r = 0.384, P = 0.006). Though the apathy factor extracted from the factorial analysis correlated significantly with CDSS scores (r = 0.55, P < .0001), the emotional component did not (r = 0.031, P = .81). Moreover, no significant correlation was observed between the diminished emotional range subscores from the SNS and depressed mood (item 1 of CDSS) (r = 0.13).
The SNS is a remarkably understandable instrument for patients with schizophrenia as it allows them to readily complete it without assistance, providing information with respect to their own perception of negative symptoms. The SNS is a self-assessment that permits patients to evaluate themselves in 5 dimensions of negative symptoms. This first validation study for the SNS revealed good psychometric properties alongside satisfactory acceptance by patients. In addition, this study has demonstrated the ability of patients with schizophrenia to accurately report their own experience.
Our results provide strong evidence supporting the conclusion that patients with schizophrenia are able to complete the SNS reliably and consistently; they support some studies that indicate that self-report measures may be valid for most personality and symptom domains.34 The accuracy and completeness of answers to questionnaires may depend on the questionnaire’s relevance, difficulty, and length. The SNS used in the present study is a well-validated assessment based exclusively on the patients’ perspectives to ensure highly relevant content. One original aspect of this self-assessment tool is that most items are verbatim. Focus groups have been demonstrated to enhance the validity of existing questionnaires by highlighting subjects’ concerns that may otherwise be neglected.35 Moreover, the items of the SNS refer to the present with 3 simple response options, making it easy to complete for schizophrenia patients. Finally, due to its short format, the SNS appears to be well adapted to schizophrenia populations, with notable concentration difficulties and cognitive impairments. The ICC was very high and a marked internal consistency was found in the second assessment, confirming the reliability of both measures by each individual. The test-retest reliability ranges were satisfactory compared to those observed with heteroassessments.13,36,37 In the present study, the self-assessment was proposed to stabilized patients for whom no changes occurred between observations. The results demonstrate the ability of schizophrenic patients to assess themselves, even in the negative dimension effectively known to be associated with cognitive impairment,38 which can dull their own perception of symptoms. Moreover, this ability to provide reliable or valid information by a self-report has been observed in patients with executive dysfunction.39
In contrast to previous scales, which report subjective experiences of deficit or negative symptoms,15–17 the SNS is a self-assessment without an interview and specifically evaluates 5 negative dimensions. For example, the Subjective Deficit Syndrome Scale (SDSS)16 contains 19 items focused on patient complaints, with only 3 related to negative symptoms. The Subjective Experience of Deficits in Schizophrenia (SEDS)15 and Subjective Experience of Negative Symptoms (SENS)17 require a semistandardized interview, and the SENS might take up to 45min to complete. In contrast to the MAP-SR,18 the present self-assessment includes diminished emotional range, which is independent from the dimension of apathy. The assessment of emotional feeling is a strength of the SNS. The MAP-SR is a self-report derived from the CAINS, from which the expression items (blunted affect/decreased emotional expression and alogia/reduced verbal expression) were removed because of their poor reliability and validity. In the SNS, 4 items that reflect the individual’s capacity to express and feel emotions are used to evaluate diminished emotional range separately from that of depression. Consequently, the SNS allows patients to express their deficit in motivation and pleasure, as well as their loss of emotion independent of depressed mood. Moreover, the MAP-SR does not cover the 5 dimensions required and may be considered sophisticated by some patients for whom the concepts of consummatory and anticipatory pleasure (as they are formulated in this scale) may be challenging to self-evaluate. The evaluations also regard the past week, and the questions use the terms “how often” and “how much”, requiring the patients to remember what feelings or events happened over the stipulated period. This requirement can be difficult for many patients with impaired memory.
The good internal consistencies found for all 20 items and on all 5 subscores confirm that all items measure negative symptoms. Factor analysis identified 2 factors, the first covering an “apathy” dimension and the second an “emotional” dimension, which is in keeping with the literature which supports a 2-factor structure for negative symptoms.32 These 2 factors were replicated previously with the BNSS.33 These 2 factors are often termed avolition/apathy (including apathy, amotivation, asociality, and anhedonia) and expressivity of emotion (including blunted affect and alogia). Further studies with a larger cohort will allow us to identify whether the 4 items constituting alogia in the present self-assessment are restricted to a single dimension (apathy). Interestingly, alogia can also be correlated with inattention depending on the scales that are used.32 Moreover, the possibility that patients report this symptom less often than the others could also be raised. Indeed, Liraud et al 40 showed that, regardless of the level of insight, schizophrenia patients are able to report their negative symptoms accurately, with the exception of alogia. Bottlender et al 41 also found that clinicians and patients do not converge on assessments of alogia.
Significant correlations between SNS scores and SANS, CGI-S negative, or BPRS negative scores confer a tight convergent validity of the SNS and, thus, its significant capability to effectively evaluate negative symptoms. This convergent validity is reinforced by the strong correlations observed between alogia, avolition, and social withdrawal SNS subscores and the corresponding global heteroevaluations of the SANS. The absence of correlation between the SNS subscores diminished emotional range and the corresponding global heteroevaluations of the SANS can be explained by the blunted affect of the SANS being based only on observations that specifically evaluate unchanging facial expression, decreased spontaneous movements, paucity of expressive gestures, poor eye contact, lack of vocal inflection, and affective non-responsiveness. In contrast, the diminished emotional range of the SNS is based solely on emotions felt by the patients. This result underlines the interest in taking into account subjective experiences of the patients that are not rated in heteroassessments, as well as the fact that the blunted affect rated by an observer and emotion felt by a patient may be 2 distinct dimensions that should be distinguished in scales orientated towards negative symptoms. Among the 4 SNS items that assess diminished emotional range, only one correlated with blunted affect in the SANS. Thus, this item (“It is difficult for people to know how I feel”) reflects the heteroevaluation of blunted affect. The strong correlation between anhedonia/asociality in the SANS and social withdrawal, but not anhedonia, in the SNS is due to anhedonia/asociality containing more items in the SANS regarding social interactions than items related to consummatory and anticipatory pleasures.
The lack of correlation between the SNS scores and those of the IS, CGI-Parkinsonism, and BPRS positive speaks in favor of a considerable discriminant validity. The absence of correlation between the SNS and IS shows that the level of insight was not related to the self-assessment of negative symptoms, which can support the notion that a patient is able to understand and evaluate their own negative symptoms regardless of their level of insight. The IS in particular assesses recognition of the disease, attribution of symptoms to the disease, and the need for treatment. Our results suggest that subjects with schizophrenia may be able to report some symptoms accurately, even if they do not believe that they are ill, need treatment, or are suffering because of the consequences of their illness. Our results counter a number of publications and ideas which state that the level of insight correlates better with lesser severity of negative symptoms or state that patients with negative symptoms cannot properly be self-assessed due to their poor insight.42,43
The SNS did not correlate with CGI-Parkinsonism or BPRS positive subscores. In contrast, we and others44,45 have found a significant correlation between the SANS and CGI-Parkinsonism (r = 0.37, P = .008). Therefore, in contrast to these reports showing a strong correlation between the negative dimension evaluated by heteroassessments and the severity of extrapyramidal syndromes,44,45 our results support the idea that self-assessment with the SNS can help distinguish negative symptoms secondary to antipsychotic treatment from primary negative symptoms. The lack of correlation between the SNS and positive symptoms also concurs with the fact that the SNS identifies primary negative symptoms as defined elsewhere.46,47
In contrast to expectation, we observed a correlation between the SNS and the CDSS. A lack of correlation between the CDSS and negative symptoms was reported previously,48,49 but weak or significant correlations were also found with hetero-assessments13,28,50 or self-assessments19 and in the present study with the SANS or SNS. Of the 2 factors extracted from the factorial analysis, the emotion component failed to correlate with CDSS scores, emphasizing the importance of this component to differentiate depressive and negative domains in schizophrenia. Moreover, the lack of a significant correlation between diminished emotional range of SNS and depressed mood (the first item of the CDSS) demonstrates that the patient is able to distinguish a loss of emotion from a depressive mood. Nevertheless, the significant correlation between the apathy component and CDSS scores needs additional investigation in patients with schizophrenic and schizoaffective disorders. This result was unexpected because the CDSS does not contain any items regarding avolition, alogia, anhedonia, or social withdrawal. However, we can stipulate that, in some patients, depressed mood can induce or increase the level of negative symptoms (so-called secondary negative symptoms) or patients with a high level of awareness of emotional deficit may be more prone to feeling depressed.
The present study has some limitations, including a small sample size and few scales used. The size of the present sample allowed us to conduct a factor analysis on the 5 subscores but not on the 20 items of the SNS. Using the Positive And Negative Syndrome Scale (PANSS)37 would allow us to more accurately evaluate the psychopathology. Another limitation is that the present study did not include an assessment of functioning which might be impaired in patients with severe negative symptoms.13 The relationship between the CDSS and the apathy component of the SNS should be explored in future studies.
The next steps are to use the SNS to assess negative symptoms in different countries and cultures and compare the SNS to recently developed scales based on observer ratings, such as the BNSS and the CAINS. Future research is also needed to address issues such as the sensitivity and specificity of the SNS for screening negative symptoms or associations between the SNS and the kind of negative symptoms (ie, primary or secondary). Another challenge is to demonstrate that the SNS is also able to detect negative symptoms at the beginning of illness and even in subjects with a high mental risk of psychosis. Primary negative symptoms are observed in 7–25% of patients at the first episode of schizophrenia51,52 and are often the first symptoms to appear before the emergence of positive symptoms. Self-reports have been developed to evaluate psychotic symptoms in schizophrenia53 and prodromal psychotic symptoms54,55 but not to screen negative symptoms. Nevertheless, negative symptoms appear long before positive symptoms; they have a high prevalence compared to attenuated positive symptoms; might go unnoticed, and consequently untreated, leading to poorer functioning; and finally they are reportedly predictive of transition to psychosis.56 Thus, the use of a self-assessment that specifically emphasizes negative symptoms may be particularly useful in high-risk subjects.57 Though the SNS demonstrated temporal stability for a short period in stabilized patients, the question of its sensitivity to change is also crucial in patients receiving treatment for negative symptoms.
In conclusion, this novel self-assessment can contribute to a better management of patients with schizophrenia. Overall, because of excellent psychometric properties, the SNS shows promise as a self-report measure of negative symptoms in schizophrenia. In addressing the symptoms directly, self-assessment may lead patients to improve their knowledge about the disease and its treatment, decreasing its stigma and improving quality of life.
Supplementary material is available at http://schizophreniabulletin.oxfordjournals.org.
We would like to extend our thanks to the clinicians Professor Perrine Brazo, Mr Laurent Lecardeur and Doctor Sophie Meunier who gave advice on the form and contents of the text and tested the first version in some patients. We also thank Mr Laurent Marty and Professor Pierre-Michel Llorca who gave us their permission to include some verbatim notes from the patients involved in focus groups on negative symptoms in the SNS. We thank Mr Eric Mac Kensie for reading and correcting the English language.