Issues affecting the reliability and validity of Schizophrenia diagnosis

Hello everyone, sorry it's been a while! Thought I'd finally finish off the last schizophrenia post - diagnostic criteria, and issues affecting diagnostic validity. As usual:

Black: AO1 - Description
Blue: AO2 - Evaluation - studies
Red: AO2 - Evaluation - evaluative points/IDAs

Diagnostic Criteria

The diagnostic criteria for schizophrenia were separated by Schneider into two sets: "First-rank" and "Second-rank" symptoms, and it was suggested that having one of the 1st rank symptoms means that you are likely to have schizophrenia. The ICD (International Classification of Diseases) still focuses on 1st-rank symptoms, but the DSM (Diagnostic and Statistical Manual of mental disorders) has moved away from them.

First-rank symptoms include delusions (e.g of control, or persecution), auditory hallucinations, and thought disturbances (the belief that either your thoughts are being broadcast for others to hear, or that others are inserting thoughts into your head.)

Second-rank symptoms include disturbances of speech such as fragmentation, interruption, and incoherency, catatonic symptoms such as stupor, mutism and repeated movements, and "negative" symptoms such as apathy, avolition, a flattened range of emotional expression, and a lack of communication.

However, an issue with the concept of first-rank symptoms is that many other conditions such as bipolar disorder share these symptoms, threatening validity of diagnosis - the presence of just one first-rank symptom without any second-ranks to help make a more specific diagnosis could lead to a bipolar patient being incorrectly diagnosed with schizophrenia instead, under the ICD's diagnostic criteria.

Diagnostic Issues

Due to the two different classifications for diagnosis, criterion validity is an issue - the extent to which two different diagnostic systems agree. If schizophrenia can be diagnosed using one but not the other, either could be a potentially invalid diagnosis - one must be incorrect.

The DSM-V diagnoses schizophrenia on 5 axis: 1 and 2 for symptoms, 3 for medical conditions, 4 for social conditions, 5 for state of function. It no longer differentiates between subtypes of schizophrenia, as these are unreliable due to symptom overlap between subtypes and prominent symptom changes. To be diagnosed, a patient must have two of the criteria, one of which must be first-rank. The symptoms must be present for the last 6 months, and active for at least one.

The ICD-10 differentiates between 7 subtypes of schizophrenia, such as catatonic, paranoid, residual and undifferentiated, based on most prominent symptoms. To be diagnosed under the ICD, the patient must display one first-rank symptom, or two second-rank symptoms. It places much more emphasis on first-rank symptoms by allowing a patient to be diagnosed based on the presence of just one of them.

Comorbidity is another potential problem with schizophrenia diagnosis - the presence of one or more additional disorders occurring alongside schizophrenia makes it difficult to identify which disorder is the cause of a specific symptom, making it harder to diagnoses the correct disorders and treat them accordingly and reducing diagnostic validity.

Buckley et al (2009) analysed medical records of their schizophrenic patients, and found that 50% had depression, 47% had substance abuse disorders, 29% had PTSD, and 15% had panic or anxiety disorders. Bottes (2009) carried out a similar analysis for the Psychiatric Times, and found that 26% of schizophrenic patients had OCD, 52% had obsessive compulsive symptoms. These studies suggest that many schizophrenics have multiple disorders, and this should be taken into consideration when diagnosing patients in order to avoid an invalid diagnosis.

Contrastingly, symptom overlap can also invalidate diagnosis - many disorders share symptoms with schizophrenia, meaning that the wrong disorder could be diagnosed based on a specific, shared symptom.

Zorumski and Rubin (2013) found that bipolar disorder's most prominent symptom, severe episodes of mania, often includes delusions, hallucinations and catatonia - all symptoms that could lead to a bipolar sufferer being incorrectly diagnosed with schizophrenia.

Marsha et al (1995) found that 32% of bipolar sufferers showed 1st-rank symptoms of schizophrenia, and Carpenter (1974) found that 16% of depression sufferers showed them, suggesting that there is a clear symptom overlap between the conditions that could affect validity of diagnosis. 

Ross (1998) found that the more 1st-rank symptoms a patient displays, the more likely they are to be diagnosed with multiple personality disorder rather than schizophrenia, suggesting that 1st-rank symptoms are an indicator of MPD rather than schizophrenia specifically.

Reliability

Reliability is the consistency of diagnosis, measured by inter-rater reliability, internal consistency (if multiple patients with the same symptoms will be diagnosed in the same way) and test-retest consistency.

Several factors can reduce reliability of schizophrenia diagnosis. The main diagnostic method is the clinical interview, and individual differences between clinicians will mean different responses according to the age, personality, aptitude of the rater, as well as the bond of trust between the psychologist and the patient. These all reduce inter-rater reliability. The severity of symptoms at the time of diagnosis can affect the test-retest consistency, and deception can affect general reliability, as evidenced by Rosenhan's 1973 study.

In Rosenhan's 1973 study, 8 psychologists turned up at mental hospitals claiming to have experienced auditory hallucinations. Upon admission for schizophrenia, they stopped presenting symptoms, and were kept in for 7-52 days despite no further schizophrenic symptoms, and were given over 2000 doses of antipsychotic drugs between them. A followup on this told a mental hospital to expect pseudopatients over the next 3 months - 83 out of 193 patients were suspected of being pseudopatients by at least 1 medical professional, but they were all actually genuine.

Although based on the DSM-II criteria at the time of at least 1 auditory hallucination, the diagnoses were both reliable and valid, the admissions only showing an inability to recognise a lie, Rosenhan claimed from these results that psychiatrists could not make a consistent and accurate diagnosis. He suggested that patients' behaviour was viewed through their label of mental illness - the original set of pseudopatient psychologists' behaviour was pathologised, with note-taking observations being pathologised as "writing behaviour".

Rosenhan's conclusions lack temporal validity - the modern version of the DSM has been revised to have more stringent diagnostic criteria. Symptoms must now be present for 6 months and active for 1 month in order for schizophrenia to be diagnosed.

Once institutionalised, the participants were passive, not normal - passively keeping up the deception rather than admitting to their lie, and the relative speed of release in light of this could actually have suggested competent and efficient mental health staff.

Rosenhan's suggestion that psychiatrists cannot reliably and accurately diagnose has been challenged by: 

Jakobsen (2005) who tested 100 Danish schizophrenic patients and assessed them based on their case notes, coming to a correct diagnosis for 98% of them.

Hollis (2000) who used the DSM-IV and case notes to correctly assess and diagnose 100% of a sample of schizophrenic patients. 

These results suggest that mental health professionals are much better at diagnosing schizophrenia nowadays, and this part of Rosenhan's conclusion lacks temporal validity.

Cultural bias in diagnosis

Emic constructs are behaviours or norms that only apply to a certain number of cultures, whereas etic constructs apply globally. When an etic construct is considered to be a universal norm, this is imposing an etic. Eurocentrism leads to imposed etics in the diagnosis of schizophrenia, as the imposed etic of the DSM is used to apply a western, subjective idea of perfect mental health to non-western cultures. If people from one culture are assessing people from another, behaviour can be misconstrued, leading to an invalid diagnosis.


Cultural difference could help to explain the higher rates of schizophrenia diagnosis in some ethnic minorities. If someone is uneasy talking to a psychologist of a different ethnicity to them, they may show withdrawal, alogia, and a lack of eye contact - all of which could be interpreted as symptoms of schizophrenia, reducing diagnostic validity.

Cochrane (1977) found that schizophrenia rates in the UK and in the West Indies are very similar, and close to 1%, but people in the UK of Afro-Caribbean origin are 7 times more likely to be diagnosed with schizophrenia than those of white European ethnicity. Migration stress and socioeconomic factors were ruled out, as other ethnic groups such as South Asian that migrated at a similar time are no more likely to be diagnosed, suggesting that there is a bias that leads to racial overdiagnosis of Afro-Caribbean people.

Harrison (1997) supports the temporal validity of Cochrane's earlier research, finding that 20 years later, the gap in racial diagnosis rates had widened - Afro-Caribbean patients were now 8 times more likely to be diagnosed with schizophrenia. 

Stowell-Smith and McKeown (1999) carried out a discourse analysis of psychiatrists' reports on 18 white and 18 black psychopaths, and found that with black psychopaths there was more emphasis on aggression and potential threat to society, compared to a greater emphasis on trauma and emotional state with white psychopaths. This suggests a racial bias in the area of psychopathology that leads to biased reporting of symptoms, calling diagnostic reliability into question. 

Read (1970) gave 194 UK and 134 US psychiatrists a case report and asked them to come to a diagnosis from it. 69% of US and 2% of UK psychiatrists diagnosed schizophrenia from it, suggesting large cultural differences in behaviour interpretation and diagnostic criteria. 

Neki (1973) studied the prevalence of catatonic schizophrenia among schizophrenics in the UK and India, and found the rate was 44% in India but only 4% in the UK, again suggesting large cultural variations in behaviour interpretation and classification.