Paper: The accuracy of a single question in screening for depression in a cohort of post-stroke patients

By C Watkins et al (Manchester)

There are a number of points relating to the design and methodology that need to be addressed by the authors. The data appear to have been analysed appropriately, but the presentation of results could be improved.

Design and Methods:

  1. The patients included in this study appear to have been part of a randomised controlled trial (Page 3). Which trial? Were these patients in a control or intervention group? More details should be given.
  2. "Between April and November 1999, 242 stroke patients were registered. Of 110 in hospital at week 2, 79…….were included in the study" (Page 3).


    3. It is unclear why information is given on the number of patients registered in an 8-month period, and why patients in hospital during ‘week 2’ were chosen.
     

Statistical Analysis and Presentation of Results:
     
  1. "36(47%) of patients were classified as clinically depressed, of whom 37 screened positive……." (Page 3).

    2. The numbers of patients given in this sentence and the following sentence are incorrect. Phrases in one sentence appear to have been wrongly placed in the other sentence.
     
  2. What cut-off point on the Montgomery Asberg scale has been used to diagnose depression ?
5. 95% confidence intervals for the predictive values would be useful.

Julie Morris