Assessment of competence to complete advance directives

BMJ 1999; 318 doi: https://doi.org/10.1136/bmj.318.7200.1765 (Published 26 June 1999) Cite this as: BMJ 1999;318:1765

Standardised instruments are needed to guide the measurement of competence

  1. Saskia Teunise, Senior psychologist, psychology department, community mental health unit (Steunisse{at}boreas.clara.net),
  2. Sarah Eagger, Senior lecturer in old age psychiatry,
  3. James Warner, Senior lecturer in old age psychiatry
  1. St Charles Hospital, Exmoor Street, London W10 6DZ
  2. Section of Old Age Psychiatry, University of Oxford Department of Psychiatry, Warneford Hospital, Oxford OX3 7JX

    EDITOR—Fazel et al described an instrument for the assessment of competency to complete advance directives that seems to be reliable, valid, and ready for use in clinical practice.1 We believe, however, that certain aspects require clarification before the clinical usefulness of the test is assured.

    Firstly, who is this test meant for? Different groups are suggested in different parts of the text. Is it meant to differentiate between competency and incompetency in demented patients, in the broader group of patient with cognitive impairment, or in elderly volunteers? If the test aims to assess competency in dementia, test-retest analysis should include only data on demented patients, because fluctuations in functioning in this group are more common. The reported test-retest correlation was mainly based on data on non-demented subjects, and there is a likelihood of overestimation of the reliability of the test for demented subjects.

    The test discriminated between non-demented volunteers and demented participants, but the influence of education on the results was not considered. Volunteers are often well educated, so the results may reflect differences in educational background instead of differences in competence.

    Finally, and probably most importantly, the gold standard and test result were not independently obtained in the present study, and therefore the test could be considered as having only face validity. Use of this gold standard for the calculation of a cut-off score and the fact that it was based on only 19 (including eight non-demented) subjects necessitate further development considering the huge consequences that the application of this test could have in clinical practice. This may also apply to shortening the test by using two instead of three vignettes without checking its effect on the reliability of the test result.

    In conclusion, there is a critical need for standardised instruments to guide the measurement of competence in many domains, including consent to treatment, engaging in research, and making choices about living arrangements. Fazel et al's study is a welcome initial step.


    1. 1.

    Authors' reply

    1. Seena Fazel, Wellcome research registrar (seena.fazel{at}psychiatry.oxford.ac.uk),
    2. Tony Hope, Reader in medicine,
    3. Robin Jacoby, Professor of old age psychiatry
    1. St Charles Hospital, Exmoor Street, London W10 6DZ
    2. Section of Old Age Psychiatry, University of Oxford Department of Psychiatry, Warneford Hospital, Oxford OX3 7JX

      EDITOR—We believe that the instrument we have described to assess competence to complete advance directives can be used on anyone, including healthy elderly people and those with mild dementia. We argue in the paper that it would seem worthwhile for advance directives to be completed at a time when people already have some disease or disability, and those with dementia could therefore be the largest group for whom an advance directive is relevant. Nevertheless, in palliative care and oncology there will be adults for whom assessments of competence to complete advance directives are required.

      Teunisse et al ask whether we considered the influence of education. It was tested with the national adult reading test, a valid and reliable index of premorbid intelligence.1 We found no significant difference in the mean scores on the national adult reading test between the elderly volunteers (predicted IQ 101.0 (SD 14.5)) and patients with dementia (predicted IQ 96.8 (17.2)).

      The authors' final point concerns the validity of the instrument. Our results go beyond face validity in that the gold standard was conducted by two old age psychiatrists; they had not conducted the initial interview and therefore did not know the score that the participant had received on the competence instrument. As the authors say, further work refining this and other instruments would be worthwhile.


      1. 1.
      View Abstract

      Sign in

      Log in through your institution

      Free trial

      Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
      Sign up for a free trial