Gill Pinner

Attitudes of patients with mild dementia and their carers towards disclosure of the diagnosis.

OBJECTIVE To determine the attitudes of patients with mild dementia and their carers towards the disclosure of diagnoses of cancer and dementia and whether there are differences between these groups. To determine whether any major adverse events occurred following disclosure of diagnosis of dementia. DESIGN A prospective study followed by a retrospective case-note study after 1 year. PARTICIPANTS AND SETTING Fifty patients with mild dementia and their carers were recruited from a memory clinic in Nottingham, UK, and an old-age psychiatrist, using a semistructured questionnaire, separately interviewed them. MEASURES Questions related to whether patients and their carers wished to be told diagnoses of cancer and dementia; the reasons for this; whether they would want treatment and make use of predictive testing if available were included. At 1-year follow-up whether antidepressants had been prescribed and whether any catastrophic reaction had occurred following disclosure of diagnosis. RESULTS An overwhelming majority of patients with mild dementia wished to be informed of their diagnosis (92%); even more (98%) of the same patient sample reported wanting disclosure of a hypothetical diagnosis of cancer. A higher proportion (98%) of carers would wish to be told if they were to develop either dementia or cancer. All patients and carers would like access to treatment for dementia. A total of 88% of patients and 86% of carers would make use of a predictive test of Alzheimers disease. Before receiving a diagnosis of mild dementia, only 28% of patients had insight that they may have dementia. Only a quarter (26%) of carers did not want the doctor to disclose the diagnosis of dementia to the patient. After 1 year, 6% of the original sample had developed a depressive illness requiring antidepressant treatment and no major incidents had occurred following disclosure of diagnosis. CONCLUSIONS There has been a growing debate on whether patients with dementia should be informed of their diagnosis. Until recently, little was known about the views of patients themselves. This study shows that the vast majority of patients with mild dementia wish to be fully informed. Despite increasing awareness, a quarter of carers still do not wish their relative to be informed, though wish to be informed if they themselves were to develop the illness. This is significantly lower, however, than previously reported, suggesting a shift in attitudes. This study adds support to the weight of evidence that disclosure of a diagnosis of dementia does not cause depression or any irreversible harm to the patient.

On telling the truth in Alzheimer's disease: a pilot study of current practice and attitudes.

Research suggests there has been a cultural change in the disclosure of diagnosis; most evidence relates to cancer and there is little knowledge of attitudes towards disclosing the diagnosis of Alzheimers disease. A questionnaire was used to survey the current practice and attitudes of old-age psychiatrists and geriatricians in Nottingham, UK. The results of this pilot study suggest that only 40% of respondents regularly tell patients the diagnosis. Although physicians are aware of many benefits in disclosing, they have concerns regarding the certainty of diagnosis, the patients insight, and potential detrimental effects. The advantages of disclosure and the ethical issues involved are discussed.

Enhancing Objective Structured Clinical Examinations through visualisation of checklist scores and global rating scale

The effective use of Objective Structured Clinical Examinations (OSCEs) requires careful attention to the construction of clinical scenarios and to the way clinical performance will be measured. Assessments must be aligned to the curriculum and when constructing an OSCE station, care must be taken to set a scenario which tests the desired learning outcomes. This allows the student to demonstrate evidence of the clinical skills they have acquired and will create measures of clinical performance likely to be valid and reliable. However, there are many sources of variability which can influence the generalizability and reliability of the scores. For example, the assessors’ judgment, students’ ability and the sampling of tasks/cases. An assessor’s judgment may lose validity if the tasks within a station are so complex and extensive, resulting in examiner ‘demoralization and fatigue’.1 Such errors will distort the measurement and threaten the validity and reliability of the examination. If an OSCE is planned prudently with ‘a common conceptualization’ of performance among assessors when they are observing and rating students,2 performance ratings will accurately reflect the clinical skills of students. The judgments that assessors make may also be prone to measurement error leading to issues of inter-rater reliability i.e., would the student have achieved the same performance rating in the “abdominal” station if a different examiner had rated them? Performance ratings are susceptible to the various types of internal and external errors which can contribute to less reliable OSCE performance ratings. An example of an internal error would be variations in the level of student’s interest and motivation, whereas differences in circuits, student gender, the ethnicity of standardised patients and students,3 sites and examiners are considered as external error factors.4 Since the subjectivity in ratings can be a major source of error in OSCEs, one solution to minimise such error and improve OSCE-rating reliability is to obtain ratings from multiple assessors in each station (i.e., each student is rated by two or more independently judging assessors). However, to improve the reliability of an OSCE it is better to increase the number of stations with one examiner present than have fewer stations with two examiners assessing5 and is more feasible in practice. In addition, using multiple assessors may not affect systematic errors (for example, a combination of lenient and stringent examiners). Psychometric methods such as Generalizability Theory and Item Response Theory enable us to identify, isolate and estimate sources of measurement error, which has been extensively covered elsewhere.6 To standard set an OSCE it is common place to use a borderline group or regression method. This requires assessors to both score the task set in a station, using either a checklist or domain-based marking scheme and also to give their global rating of the overall students’ performance. It is particularly important that the assessors are aware of the standard of the set of students being examined, especially that of a “borderline candidate”. More than a century ago, Thorndike and Hagan concluded that “the ideal rater is the person who has had a great deal of opportunity to observe the person being rated in situations in which he would be likely to show the qualities on which ratings are desired”.7 The discrepancy between assessor checklist scores and their global ratings is likely to have an impact on borderline pass/fail decisions and become a significant source of measurement error. Studies have raised concerns about common errors in judging student performance ratings.8,9 Indeed, placing students correctly into a specific category (e.g., fail, borderline, good and excellent) can be an indicator of rating accuracy and is necessary for providing effective observational feedback, a key area to address for assessment leads.2