There are two particularly important changes in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) [1] related to psychotic disorders. First, special treatment for bizarre delusions (BDs) and other Schneiderian first-rank symptoms (FRSs) was removed for schizophrenia (SZ); a statement was added indicating that at least one of the two symptoms required to meet criterion A must be delusions, hallucinations, or disorganised thinking. Second, psychotic disorder not otherwise specified (PNOS), listed in the Diagnostic and Statistical Manual of Mental Disorders-Forth Edition-Text-Revised (DSM-IV-TR) [2], was divided into two categories: Other Specified Schizophrenia spectrum and Other psychotic disorders (OSSOs), and unspecified SZ spectrum and other psychotic disorders. The first change prompted us to explore how many individuals would no longer be diagnosed with SZ if special treatment of BDs or FRSs were eliminated. Several trials have demonstrated that the percentages of individuals diagnosed with SZ solely based on the presence of BDs or FRSs are in the range of 0.4%-2.1% [3-6].

However, in a study of 196 first-admission patients with psychosis, 7 patients (7.5%) received a diagnosis of SZ solely based on BDs [7]. Goldman et al. [3] reported that 7 of 152 patients with SZ (4.6%) would have failed to receive a diagnosis of SZ if BDs had not been a “sufficient” psychotic symptom in the DSM-III-R criteria. Importantly, 5%-8% is not a negligible proportion of patients; a more careful diagnostic assessment should be performed in patients with first-episode psychosis (FEP). A key problem concerning the DSM is that the term “bizarreness” is not clearly defined except ‘‘involving a phenomenon that the person’s culture is totally implausible’’ in the DSM-III-R and ‘‘clearly implausible and not understandable and not derived from ordinary life experiences’’ in the DSM-IV and DSM-IV-TR. Additionally, some examples in the DSM-III-R (thought broadcasting and being controlled by a dead person) were derived from FRSs. A diagnosis of SZ can be made in the International Classification of Diseases (ICD) system if the individual has a single symptom corresponding to any of the six FRSs ([made-acts, made-feelings, and made-impulses], thought insertion, thought broadcasting, thought withdrawal, voices arguing, or voices commenting) in the ICD-9 [8] or eight FRSs ([made acts, made feelings, made impulses], thought insertion, thought broadcasting, thought withdrawal, somatic passivity, delusional perception, voices arguing, and voices commenting) in the ICD-10 [9]. Considering the proximity of BDs to FRSs, the results would have differed depending on the number of FRSs evaluated in previous studies. Unfortunately, most of those studies did not specify the number of FRSs; one study clearly stated that somatic passivity, audible thoughts, and the perception of delusions were not assessed [5]. Thus, we hypothesised that if we clearly defined BDs and the numbers of FRSs undergoing evaluation, the proportion of patients with DSM-IV-based diagnosis of SZ who no longer met the DSM-5 criteria for SZ would be substantial or higher than in previous studies. Next, we considered what DSM-5-based diagnosis would be made for individuals with only one BD or FRS. Notably, five of seven patients in the work of Goldman et al. [3] met the DSM-III-R criteria for PNOS. The first two OSSO examples are persistent auditory hallucinations (AHs) and delusions with significant overlapping mood episodes. Therefore, we hypothesised that individuals with single BDs or first-rank auditory hallucinations (FRAHs; e.g., voices arguing and voices commenting) would qualify for the OSSOs category.

In the present study, we investigated the proportion of patients with first-episode schizophrenia spectrum disorders (FE-SSDs) who were diagnosed based on the single presence of BDs or FRAHs. Moreover, we attempted to establish alternative diagnoses and subtypes, and then identify their diagnostic stability over a 1-year period.

There were some missing data regarding the PANSS and diagnostic stability (n=3). The missing PANSS data were mainly from the Japanese study, which was a retrospective analysis. Forty individuals (7.89%) were diagnosed with SZ, schizoaffective disorder, or schizophreniform disorder by the DSM-IV-TR criteria solely based on the presence of BDs or FRAHs. All patients met the OSSO criteria and were subsequently classified according to the presence of pure delusions (n=22), delusions with attenuated AHs (n=5), pure AHs (n=3), or AHs with attenuated delusion (n=10). No differences in age, gender, education, and CPZ dose equivalent were observed among patients with FE-SSDs according to the DSM-IV-TR and FE-OSSOs according to the DSM-5, except a longer DI and higher PANSS scores among patients with FE-SSDs. There were no significant differences between individuals with FE-OSSOs whose diagnosis remained stable and individuals who received a new diagnosis (Table 1).

The overall percentage of affected patients was 7.89% (40/507). However, there was broad variation in percentages across the hospitals (2.27%-16.48%). Twenty-seven of the 40 patients had a single BD. The BDs fulfilled mostly strict definition (n=13), narrowly defined (n=2), and broadly defined (n=12; mostly somatic passivity and delusional perception). The diagnostic stability of FE-OSSOs and its subgroups (pure delusion and delusion with attenuated AHs) over the 1-year period was 70.27% and 84% (21/25; 2 patients with missing data on stability), respectively (Table 2).

The effects of changes from DSM-IV to DSM-5 regarding the definition of the criterion A symptom on SZ caseness have been forgotten because several studies reported that those changes had minimal impact, the position of OSSOs/PNOS was always unclear, and its importance was neglected. We evaluated how many patients with FE-SSDs were diagnosed based on single BDs or FRAHs, what alternative diagnoses might have been given, and their diagnostic stability over a 1-year period. The results differed from the findings in most previous studies, and the implications should be considered.

The proportion of patients with FE-SSDs who had been diagnosed with single BDs or FRAHs was 7.89%. This result is substantially higher than in most previous studies but similar to the proportion (7.45%) of first-admission schizophrenic patients in the work of Tanenberg-Karant et al. [7]. Six of seven patients no longer considered schizophrenic had experienced their first psychotic episode during the previous 12 months [3]. These findings suggest that it is more common for patients to present with a single criterion A symptom in the early stages of SSDs or SZ; a more careful diagnostic evaluation of FEP is required. Among the patients diagnosed with single BDs, most diagnoses utilised either strictly defined BDs or two FRSs (somatic passivity and delusional perception; i.e., loosely defined BDs). This phenomenon was not surprising, considering that the frequencies of somatic passivity (33%) and delusional perception (77%) are high in FEP. A possible explanation for the low percentage of affected patients (0.5%- 2.1%) after the removal of special treatment of FRAHs and BDs in the work of Shinn et al. [5] may be that they did not assess the “made” symptoms (made feelings, made impulses, and made volitional acts), somatic passivity, delusional perception, or audible thoughts. Other factors related to the low percentage of affected patients according to changes in the DSM-5 may be that the SZ diagnosis is often made with delusion plus disorganised behaviours or negative symptoms, which are sometimes overrated by clinicians. The user’s guide for the DSM-IV-TR Structured Clinical Interview [16] clearly emphasises that apparently disorganised or bizarre behaviour may have a goal; it should not be rated as disorganised behaviour and the main problem with diagnoses involving negative symptoms is overdiagnosis. In the regular Zoom meetings, more cases were debated with regard to disorganised behaviours or negative symptoms, rather than BDs.

In the present study, all patients diagnosed with a single BD or FRAH met the criteria for OSSOs. Thus far, limited research is available regarding OSSOs/PNOS because it has often been used as a temporary diagnosis or as a last resort when the symptomatology does not conform to any other distinct type of functional psychosis. We previously reported that patients with PNOS had lower symptom severity, a better treatment response at 2 months, and higher remission rates at 12 months compared with patients displaying SZ [17]. Additionally, PNOS and SZ exhibited shared and distinct changes in neuronal connectivity, suggesting that PNOS should be treated as a separate clinical syndrome with distinct neural connectomics [18]. In particular, some of the affected patients exhibited manifestations classified as delusional with attenuated AHs or AHs with attenuated delusions. Although the current DSM-5 does not specify these subtypes, we are confident that they exist. To detect these subtypes, clinicians must be trained to accurately evaluate attenuated psychotic symptoms. Further classification of OSSOs with these subtypes should be considered in the next DSM revision. Notably, among the 10 affected patients that converted to SZ later, seven were patients with pure delusions. This finding indicates that patients who initially have single delusions can experience worsening via development of AHs during the course of illness. Therefore, prevention or early detection of AHs should be the primary therapeutic goal for patients with FE-OSSOs. Finally, the diagnostic stability of the OSSOs in the present study was very high compared with most previous studies [19] but similar to our previous study [17]. A critical limitation of most previous studies is that, with one exception [20], they did not specify the PNOS subtype. The high diagnostic stability result of OSSOs suggests that OSSOs constitute a separate and distinct clinical syndrome, rather than a placeholder for SZ. In the clinical staging model proposed by McGorry et al. [21], FEP is regarded as stage 2; no detailed diagnostic features are listed. Based on the current results and our previous findings, stage 2 can be further divided into stages 2a and 2b, which correspond to FE-OSSOs and FE-SZ, respectively.

Several limitations of this study should be discussed. First, this was a retrospective study. However, because most patients were from the prospective KEPS multicentre study, the data are generally accurate and reliable. Second, the percentages of affected patients were highly variable across institutes. This variability may have arisen because of differences in patient composition or diagnostic practices between hospitals, suggesting that the regular Zoom meetings were insufficient to secure good inter-rater reliability between hospitals regarding the diagnosis of OSSOs and SZ; more strict procedures should be considered in future studies. Third, a 1-year follow-up period may be insufficient to monitor the complete trajectory of OSSOs. Considering the lower diagnostic congruence for PNOS at the 2-year follow-up (25%-51.5%) [22-24] compared with the 1-year follow-up (44%) [25], a longer follow-up duration is warranted. Despite these caveats, one strength of this study is that it constitutes the first international multicentre study regarding the impact of SZ-related changes in DSM-5 criteria on Asian patients with FE-SSDs.

In conclusion, 7.89% of patients with OSSOs were affected by SZ-related changes in the DSM-5. The diagnostic stability of OSSOs at the 1-year follow-up was 70.27%. These findings suggest that more rigorous diagnostic assessments should be performed, particularly to distinguish OSSOs from SZ in patients with FE-SSDs. A more refined classification of OSSOs subtypes should be under consideration for the next DSM revision.