Reliability and Validity of Diagnosis Flashcards
nReliability of diagnosis
Reliability refers to consistency of diagnosis
inter-rater reliability
If a patient is given the same diagnosis from two or more psychiatrists then there is inter-rater reliability.
Why might diagnosis not be reliable
Symptoms can overlap meaning the same symptom can occur across different disorders such as 2 clinicians might see the same symptom but assign their cause to different disorders for e.g. those with PTSD schizophrenia and bipolar, all can suffer with hallucinations, meaning different psychiatrists diagnosis different disorders. suggesting that the diagnosis is unreliable.
Why might diagnosis not be reliable in terms of patient factors (2)
Information provided by patient to the clinician may be inaccurate due to problems with memory, denial or shame. Different information could therefore be given to different psychiatrists leading to different diagnosis.
Patient’s may also provide different information on different days due to differences in their mood, or how they are experiencing symptoms, leading to different diagnosis on different occasions. Thus making diagnosis unreliable.
another reason why diagnosis might not be reliable in terms of clinician factors ( 3)
The unstructured nature of the clinical interview means that clinicians may ask different questions, and may focus on different symptoms presentation e.g. one may focus on a patient’s nightmares, while others may follow a different cause of questioning for e.g a traumatic past event, so this can lead to different information being gathered leading to different diagnosis. the first clinician may diagnose a depressive disorder and the second a post-traumatic stress disorder.
Clinicians use their subjective judgment about how they interpret the symptoms a patient presents. This is dependent on their training and experience for e.g. if they have had psychodynamic training they may focus on early childhood and attribute hallucinations to a past trauma, whereas a more medically trained clinician might see hallucinations as due to excess dopamine.
Clinicians gather qualitative data through interviews - what the patient is saying can be interpreted - clinicians use their one judgement and may interpret information differently, so the process is subjective, leading to different diagnosis.
Research evidence for reliability : Ward et al ‘62
Studied two psychiatrists diagnosing the same patient and found that disagreement occurred because of inconsistency of the information provided by the patient (5%), inconsistency of the psychiatrists interpretation of symptoms (32.5%) and inadequacy of the classification system (62.5%).
Main reason diagnosis is unreliable is due to problems with the classification systems.
Research evidence for reliability : Brown 2001
Reliability of DSMIV for mood and anxiety disorders is good to excellent.
However, disorders where symptoms overlap to a high degree have lower reliability e.g. PTSD
Reliability depends on the disorder being diagnosed.
Research evidence for reliability : Cooper ‘72
American and British psychiatrists saw the same videotaped interview and asked to diagnose. New York psychiatrists said it was schizophrenia twice as often; London psychiatrists said it was depression twice as often.
Cultural factors may lower reliability of diagnosis.
Research evidence for reliability : Beck ‘54
Same set of symptoms only diagnosed as the same disorder in about half of cases.
Highlights low reliability.
Research evidence for reliability : Rosenhan ‘73
All 8 pseudo-patients were admitted to hospital; 7 were diagnosed with schizophrenia.
High reliability as all diagnosed with a MH disorder and ⅞ the same disorder.
Validity of diagnosis
Validity refers to whether the diagnosis is true; in other words is it a genuine reflection of the underlying disorder. to an extent the treatment must work for the diagnosis to be valid.
Aetiological validity
is achieved when the diagnosis reflects known causes e.g. a family history in a disorder with a genetic component.
Concurrent validity
refers to agreement between different tools. So if the same diagnosis was given using the DSM and the ICD, there would be concurrent validity.
Predictive validity
is when the diagnosis can predict the prognosis of the disorder; i.e. the future outcomes or response to treatment.
Why might diagnosis not be valid in terms of Implicit bias:
An attitude held at an unconscious level which can affect a clinician’s interpretation e.g. more likley to diagnose a female with depression as depression is more common in females. Can lead to an invalid diagnosis.
