Psychosis is a complex disorder that expresses itself in a variety of different ways. The clinical assessment of the experience of psychosis is a challenge for health care professionals to communicate precisely amongst one another. It is also difficult to accurately measure the efficacy of treatments. Different scales have been developed to measure the presence and severity of symptoms of the illness. These tools, such as the Positive and Negative Syndrome Scale (PANSS) [1], the Scale for the Assessment of Positive Symptoms (SAPS) [2] or the Brief Psychiatric Rating Scale (BPRS) [3], measure a broad range of experiences and behaviours, something which restricts the detailed measurement of specific symptoms [4]. These scales have few items dedicated to the measurement of delusions or hallucinations, both of which are key symptoms of the psychotic experience. These symptoms are summarized as single broad scores without taking into account the multidimensional features of the experience. The various dimensions of the psychotic experience include the degree of conviction with which the delusional belief or beliefs about auditory hallucinations are held, the level of preoccupation it engenders, the degree of distress experienced as well as the behavioural responses used to cope with the experience [4]. These various dimensions may respond differentially to pharmacological [5–7] and psychological treatments [8–10].

Although there is some inconsistency in the literature about the chosen dimensions of delusions or hallucinations, assessment tools limited to a single item for key symptoms will miss large parts of the complex nature of such symptoms.

Haddock et al.[11] developed the Psychotic Symptom Rating Scales (PSYRATS) in order to improve the measurement of various dimensions of psychotic symptoms. The PSYRATS contains separate scales for auditory hallucinations and delusions. The PSYRATS auditory hallucination scale includes eleven items and the delusion scale includes six, all rated from zero to four. The psychometrics qualities of the PSYRATS indicate good internal consistency, sensitivity to changes, relationships with other measures of symptoms, inter-rater and retest reliability. The properties of the PSYRATS have been studied with various populations: hospitalized patients or day-treatment patients with first episode of schizophrenia or related disorders [12], recently relapsed patients with schizophrenia and schizoaffective disorders [4], hospitalized patients [13] and a mixed population of hospitalized patients and outpatients with schizophrenia and schizoaffective disorders [11], and patients with mental retardation [14]. The auditory hallucination scale has also been validated in Spanish [15].

In the initial study, the factor analysis yielded three independent factors for the auditory hallucination scale: emotional characteristics, physical characteristics and cognitive interpretation [11]. Two factors were reported in relation to the delusion scale, designated as emotional characteristics and cognitive interpretation (Haddock et al. 1999). The PSYRATS has been used as outcome measures in a number of studies aimed at evaluating the effectiveness of psychological interventions for psychosis [16–19]. It has also been used to quantify the impact of psychotic symptoms on the cognitive process [20, 21]. However, the PSYRATS factor subscales have not been consistently used. This may be due to the reported lack of consistency in replications of the first factor solution [4, 11, 13, 12]. Whilst the PSYRATS scales provide a total cumulative score for each symptom, the multidimensional nature of the symptoms suggests that the total score should not be presented alone. This has driven some researchers to report their use of PSYRATS on the basis of single items [22]. The use of single items has also been recommended to assess specific changes during therapy [4]. The aim of this study was to examine the psychometric properties of the French version of the PSYRATS in a sample of patients suffering from schizophrenia spectrum disorders with persistent psychotic symptoms. The two scales of the PSYRATS were evaluated for inter-rater reliability and inter-relationships of the items. In addition, construct and concurrent validity were examined.

The fifty-five participants with auditory hallucinations had a mean score of 26.5 (SD = 7.6; range = 8–38) on the auditory hallucination scale. The 94 participants with delusions had a mean score of 15.1 (SD = 3.6; range: 3–24) on the delusions scale. High scores are indicative of more severe symptoms.

The current study presents the factor structure and the relationship between items of the PSYRATS and items of the PANSS with a population of French-speaking outpatients having persistent psychotic symptoms. The participants showed a total score close to that of recently relapsed patients [4]. The factor structure of the delusions scale of the PSYRATS is the same as in the original study [11], the German version study [13] and the first episode psychosis study [12]. The factor structure of the hallucinations scale shows a four- factor structure. Three of them—emotional characteristics factor, cognitive interpretation factor and physical characteristics factor—appear as in previously published studies [4, 11, 13]. The factor analysis for our study encountered the same difficulties as previous studies did when replicating the factorial structure of the original version of the hallucinations scale. As in previously published studies, an emotional factor appears clearly. [4, 11–13]. The frequency and duration items of auditory hallucination load on a physical factor as in previous studies. The belief about voice items load on a cognitive interpretation factor as in previous studies. Also, similar to previously published studies, the location factor loads on a cognitive factor [4, 12, 13]. The disruption to life item and the loudness item load on a disruption factor as in the Steel et al. study [4]. The controllability item load on a cognitive factor close to the Drake et al. study [12]. As suggested by Steel et al.[4], we probably lack a clear understanding of the dimensions of hallucinatory experience. Moreover, the different evaluations of the PSYRATS have been carried out with dissimilar participant populations with psychotic symptoms. The difficulty to replicate the factor structure of the auditory hallucinations scale of the PSYRATS with different populations may be explained by a differential response between hallucination and delusion to psychiatric treatment. For example, Schneider et al.[7] have shown with patients suffering from schizophrenia and auditory verbal hallucinations that in the longitudinal course, a general symptomatic decrease became apparent only for auditory hallucinations but not for delusions. Loudness of the voice and distress associated with auditory hallucination decreased early compared to other aspects of hallucination that took more time to fade. Chang et al.[28] have described that cluster structure stabilizes after 6 months of antipsychotic treatment, and the changes in physical characteristics may not be the sole sign of clinical improvement. The difficulty in replicating the factor structure may also be explained by the heterogeneous nature and complex nature of auditory hallucinations [29].

For the concurrent validity, the total score of the PSYRATS hallucinations significantly correlate with the P3 item (hallucination) of the PANSS. The total score of the PSYRATS delusions significantly correlate with the P1 and P6 of the PANSS. Some individual items of the PSYRATS hallucinations or delusions scales do not respectively correlate with the P3 Hallucination item and P1 Delusion item of the PANSS, suggesting that this last item reflects selected aspects of psychotic symptomatology as already demonstrated by Steel et al.[4] on a larger sample. In the current study, the P1 item of the PANSS fails to assess the amount of distress and the intensity of the distress associated with delusions. These items are more related with the G6 depression item of the PANSS. Inversely, the PSYRATS delusions scale does not appear to take into account the elements associated with delusions of grandiosity as measured with the P5 item of the PANSS. The amount of distress and intensity of distress items are negatively associated with grandiosity. A more detailed assessment of emotional and behavioural reactions to grandiosity would be worthwhile, in order to measure the effects of new psychological interventions in this area. Grandiosity delusions are associated with poor medication compliance [30]. The same comment could apply to benevolent voices and the auditory hallucination scale of the PSYRATS, which does not take into account the particular emotional and behavioural reactions to that kind of voice [31].

The present study does not replicate the Steel et al. failure of the P3 item of the PANSS to assess the negative content and distress associated with hallucinations [4]. This might be explained by a population difference and the fact that our sample of participants with hallucinations who were also assessed with the PANSS was smaller.

The disruption to life items of the PSYRATS scales need refinement. The anchor point of the items have varied between studies [4, 11] and there is a difference in precision of the description of the clinical observations in comparison to the other items which are described more precisely. In the present study, the lower inter-rater ICCs were obtained for these two items. This may lead to differences between the different factor analyses published until yet.

A first limitation of the current study was the relatively small sample of participants, something which may have prevented the replication of previously published results. The fact that participants were administered a French version of the PANSS may also have contributed to some of the unique findings of this study. We selected participants with persistent psychotic symptoms, an often neglected population that is not easy to reach. Beyond the specific interest for this subpopulation, this was also a limitation of the study since results may not be replicated in a normally distributed population of persons suffering from schizophrenia. For the external validity, the smaller number of participants with persistent hallucinations compared to participants with persistent delusions might also have been a limitation of the present study.

In conclusion, the French version of the PSYRATS partially replicates previous evaluations of the scale in English or German. As in previous studies the factor structure of the auditory hallucination scale appears less stable. The longitudinal course of auditory hallucinations need further research to better understand how the factor structure of the scale varies across time. The disruption to life items of both scales should be more detailed to better tape the specific behavioural consequences of psychotic symptoms. The future development of the scale should take into account the emotional and behavioural consequences of delusions associated with grandiosity and benevolent voices in order to capture better important parts of the psychotic experience.