Assess reliability of ICD and DSM (20) Flashcards
Reliability Definition
A diagnosis is considered reliable if more than one psychologist gives the same diagnosis to the same individual. Clinicians diagnosis of disorders must be consistent with each other or issues of reliability will arise the clinicians disagree over a diagnosis and treatment may not work.
Goldstein strength
Goldstein re-diagnosed 169 out of 199 using DSM III when they had originally been diagnosed with DSM II. Therefore, this suggests that the DSM does have reliability in the form of inter-rater reliability and patients will have consistency of diagnosis where clinicians use different versions of the same classification system within the same culture. Equally Galeazzi et al arranged for two researchers to conduct a joint interview to assess 100 consecutive clients for psychosomatic symptoms and showed a very high agreement when using ICD-10. Therefore, this suggests that for some disorders there is a good level of consistency and so reliability in between clinicians.
ICD reliability
Other studies have also shown that new editions of the ICD have improved in reliability. For example, Ponizovsky et al carried out a large scale, longitudinal study and found that the proportion of people who retain the same diagnosis when reassessed increased by 26% for SZ, 16% for mood disorders and 8% for anxiety disorders. Therefore, this shows improved reliability and suggests that the increased number of disorders included from ICD-9 to ICD-10 has not detracted from the reliability of these diagnoses. Although the study showed less satisfactory results for childhood disorders. Therefore, this suggests that the ICD may not be reliable for some specific disorders.
Cooper however
However, research by Cooper et al suggests that the DSM lacks cross-cultural reliability. Cooper showed UK and US psychiatrists the same videotaped interviews and ask them to make a diagnosis. NY psychiatrists diagnosed SZ twice as often whereas London psychiatrist diagnosed depression twice as often. Therefore, this suggests that the DSM is not reliable even when the clinicians come from very similar cultures and are sing the same classification system. This may be due to different cultural interpretations of the DSM for the same set of symptoms.
Clinicians differences
The unstructured nature of the clinical interview can lead some clinicians to focus on different symptom presentation. This can lead to different information being gathered about a patient and results in a different diagnosis. For example, the majority of those suffering from depression also have anxiety disorders. However, there have been attempts to improve reliability, focusing on developing standardised interview schedules which specify the sets of symptoms which must be inquired about and defining them precisely giving instruction on rating their severity. Therefore, more recent editions of classification systems should show greater reliability.
Pies however
However, Pies suggested that the DSM (including DSM 5) does not offer a ‘comprehensive understanding of the patient’ which is what clinicians want. Equally, Cooper suggested that just under half of clinicians had moved over to the DSM 5 a year after if had been released, commenting that they were concerned about it’s reliability.
Patient differences
Patients may vary in the detail, emphasis and type of information that they present which may affect the diagnosis. For example, a patient may go to their GP reporting symptoms of depression, such as a lack of motivation and crying all the time. Once referred tp a psychiatrist, the patient may then emphasise other symptoms, such as sleeping problems. Therefore, an issue may not be with the classification system itself but with patients presenting differently which could result in a different and unreliable diagnosis.
Rosenhan reliability
Rosenhan (1973) found that the DSM had good reliability for the same set of symptoms as 8 pseudo-patients presented at 12 different hospitals with the same single symptom of SZ and all but one (diagnosed as bipolar) were diagnosed with SZ. Therefore, this shows reliability as clinicians across different hospitals and 5 states in the US had consistency of diagnosis for the same symptoms presented.
Reliability differences
There seems to be differences in the reliability for some mental disorders and not others using DSM 5. For example, ASD and ADHD in children and PTSD and binge-eating in adults had good or very good reliability according to trials, whereas ,major depressive disorder and generalised anxiety disorder both had low reliability scores, perhaps due to issues such as disorders having symptoms that vary a lot over weeks. However, some have argued that the reliability of classification systems is good as the reliability for medical disorders. For example, Falek and Moser found a 66% agreement rate between what doctors indicated was the cause of death without the aid of any tests, and post-mortem findings.
Conclusion
In conclusion, classification systems, such as the DSM, are largely reliable as more recent versions have added more detailed lists of symptoms, which used alongside the 4D’s and more structured interviews, should mean that different clinicians are able to diagnose a patient with the same disorder. However, there are still variations in reliability due to issues such as clinician and patient factors.
