Sam Wiebe, Ronald Ingle, Steve Simon & Victor Montori debate on:
'How to judge the minimal clinically relevant difference of a
symptom measurement on an ordinal scale, and how to present such
differences in a sensible way without all too much ignoring the
statistical drawbacks of summing and dividing values on non-linear
scales?'
We did publish on this subject based on several studies on the
responsiveness of the WONCA/COOP scales, a 5-point Likert scale.
Unfortunately it is in Dutch:
Van Duijn NP, Groenier K, Schuling J, van Weert H, Meyboom-de Jong
B. De gevoeligheid van de COOP-kaarten voor verandering van de
klinische toestand. Huisarts Wet 1995; 38: 139-44.
An version in English, extended upto 8 studies from 3 countries has
been drafted, but died on the road of international cooperation, on
the desk of one of the UK authors who promised to polish the
language, but never did. I gave up in april 1997. [Never write an
article with authors from other countries (or other institutions for
that matter].
I think I gave a sensible solution to present differences and to
judge the clinical relevance of differences on a 5-point scale.
At the time I am almost drowning in my own succes as at
the moment 2 major projects in palliative care (my subject at the
moment) are granted; possible 2 minor ones coming up. 20 % of the
week meant for research.
Sam, I hesitate to pick it up, as I promised my boss to stick to my
core bussiness. But I am interested in your opinion on the draft
article. So, I will send it to you in a attached file in WORD, but
keep it a secret.
Nico van Duijn, general practiontioner-epidemiologist
Depart. General Practice
Division Public Health
Academic Medical Centre
University of Amsterdam
the Netherlands
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|