For any diagnostic system to work effectively, it must possess reliability and validity. Reliability means that there is good consistency over time and between different people’s diagnosis of the same patient; the latter is known as inter-judge (or inter-rater) reliability. If a diagnosis of schizophrenia is valid then patients who are diagnosed as suffering from schizophrenia must have the disorder. If a diagnostic system is to be valid, it must also have high reliability. Clearly if a disorder cannot be agreed upon (so low reliability) then all of the different views cannot be correct (so low validity). Whereas a diagnostic system can be reliable but not valid—it can produce consistently wrong diagnoses.
In terms of classification, DSM-IV and ICD-10 take a categorical approach, which assumes that all mental disorders are distinct from each other, and that patients can be categorised with a disorder based on their having particular symptoms. However, diagnosing abnormality is not as straightforward as this approach suggests.
Issues in the classification and diagnosis of schizophrenia: Comorbidity
This refers to a patient who suffers from two or more mental disorders at the same time. Patients with schizophrenia often have other disorders such as major depressive disorder or bipolar disorder or an anxiety disorder. Comorbidity happens because some symptoms are the same across disorders. This creates problems of reliability as there can be inconsistency in which disorder is diagnosed.
- Sim et al.’s (2006) study of 142 hospitalised schizophrenic patients, 32% of whom had an additional mental disorder.
Issues in the classification and diagnosis of schizophrenia: The continuity approach
This suggests that there is a continuum between schizophrenia and normality.
- Chapman et al. (1994, see A2 Level Psychology page 381) have found evidence for schizotypy, which is a proneness to developing psychosis (especially schizophrenia). This supports the continuity hypothesis and reduces the reliability with which schizophrenia is diagnosed.
The five types of schizophrenia
The five types raise issues of reliability and validity:
- Some of the types of schizophrenia have no symptoms in common, which questions if they can be the same disorder.
- The undifferentiated schizophrenia type raises issues of reliability and validity because it is basically a “rag bag” category for all those patients with schizophrenia who are hard to classify. There is such diversity of symptoms that two patients with undifferentiated schizophrenia might have no common symptoms.
- Another issue is residual schizophrenia. There can be inconsistency in judgements as to whether schizophrenia cases have reduced sufficiently to be categorised as residual. It can also be difficult to decide if the patient has recovered or still has residual schizophrenia.
Rosenhan’s (1973) research on “being sane in insane places”
- Rosenhan (1973, see A2 Level Psychology pages 382–383) found hospital staff could not distinguish between the sane and the insane. This was because they failed to detect pseudo-patients who had faked symptoms (reported hearing indistinct voices saying “empty”, “hollow”, and “thud”) to gain admittance to 12 different psychiatric hospitals and once admitted behaved in a normal manner and said they had no further symptoms to the hospital staff. Seven out of eight participants were diagnosed with schizophrenia. This clearly questions the reliability and validity of diagnosis.
- Errors made in diagnosis may not be representative of diagnosis under more typical circumstances as psychiatrists do not expect people to fake mental illness. Kety (1974, see A2 Level Psychology pages 382–383) has countered this by pointing out that if a patient faked physical symptoms, then they may similarly be mistakenly diagnosed with a physical illness.
- The psychiatrists may not have been completely convinced. The pseudo-patients were given a very rare diagnosis of “schizophrenia in remission” and most were released in a few days, which suggests the psychiatrists were unconvinced that the patients had really suffered from schizophrenia.
This refers to the extent to which an assessment measure covers the range of symptoms of schizophrenia. Thus, the diagnostic manuals have content validity if they provide detailed information regarding all of the symptoms of schizophrenia.
- Jakobsen et al. (2005, see A2 Level Psychology pages 383–384) used the Operational Criteria Checklist (OPCRIT), a symptom checklist with a glossary providing clear and explicit descriptions of the symptoms, to study patients with a history of psychosis. There was good agreement on the diagnosis of schizophrenia when the diagnoses of OPCRIT were compared against those of ICD-10, indicating a high level of reliability.
- There was also good agreement (and thus high reliability) when ICD-10 and DSM-IV diagnoses were compared. Both of these findings support content validity as the findings suggest they have sufficient detail of symptoms for accurate diagnosis.
- Measures such as standard semi-structured interviews or the Operational Criteria Checklist possess good content validity. This is because they involve working through all of the symptoms associated with schizophrenia and with other related mental disorders.
Any form of assessment for schizophrenia possesses good criterion validity if those diagnosed as having schizophrenia differ in predictable ways from those not diagnosed as schizophrenic.
- Comer (2001, see A2 Level Psychology page 384) found that people with schizophrenia are much less likely than non-schizophrenics to be in full-time employment and to have a strong social network.
- This provides some support for criterion validity but it is not especially convincing. This is because individuals suffering from almost any mental disorder are more likely than healthy individuals to experience social, relationship, and job problems, and so this doesn’t distinguish schizophrenics from patients with other mental disorders.