Re: Are graded clinical signs more reliable than dichotomized?

Roland wrote:
Clinical diagnosis is often based on the presence of clinical signs,
like tenderness. These signs present with varying intensity and the
judgement if a sign is present or not is a subjective assessment with
no objective definition.

Instead of dichotomising these clinical signs I think that a grading
of the intensity of the sign must retain more of the diagnostic
information, ie instead of deciding on if tenderness is present or not
the tendderness is graded absent, slight, moderate or intense.

One problem is of course that there is no definition for the grades of
the variable. However the dichotomised variable has the same problem
and I assume that the agreement between two examiners will be larger
if graded variables are used instead of dichotomised.

I have not found any previous research on this subject and would be
very happy for your input.


Roland Andersson
Surgeon
Sweden


Here is one article you might find interesting from a recent issue of Statistics in Medicine.

-------------------------------------------------------------------
STATISTICS IN MEDICINE
Statist. Med. 2006; 25:127–141
Published online 11 October 2005 in Wiley InterScience (www.interscience.wiley.com)

Dichotomizing continuous predictors in multiple regression: a bad idea

Patrick Royston; Douglas G. Altman and Willi Sauerbrei

SUMMARY
In medical research, continuous variables are often converted into categorical variables by grouping values into two or more categories. We consider in detail issues pertaining to creating just two groups,
a common approach in clinical research. We argue that the simplicity achieved is gained at a cost; dichotomization may create rather than avoid problems, notably a considerable loss of power and residual confounding. In addition, the use of a data-derived "optimal" cutpoint leads to serious bias. We illustrate the impact of dichotomization of continuous predictor variables using as a detailed case study a randomized trial in primary biliary cirrhosis. Dichotomization of continuous data is unnecessary for statistical analysis and in particular should not be applied to explanatory variables in regression models.

Copyright ? 2005 John Wiley & Sons, Ltd.
KEY WORDS: continuous covariates; dichotomization; categorization; regression; efficency; clinical research
-------------------------------------------------------------------


--
Bruce Weaver
bweaver@xxxxxxxxxxxx
www.angelfire.com/wv/bwhomedir
.