Interesting discussion, and Andrew Jull's illustration is very informative.
Much like indices and ratios, letter grade schemes can obscure as much as
they reveal. In his book How to Write & Publish a Scientific Paper, Robert
Day objects to "handwaving reference" in which work is cited but not
adequately described.
The strength-of-recommendation category (A-E) and type-of-evidence grade
(I-III) system put forward in Macpherson DW, CCDR 1994:20:145-7 coupled with
decent description (in an evidence table or well-crafted review section
text) of the nature of studies cited seems to serve well. CATMAT uses that
scheme for their tropical/travel medicine guidelines, as do various authors
of nosocomial infection control guidelines.
Their category/grade rating alone doesn't convey sufficient information
about strengths & weaknesses of individual studies, but is an efficient way
to label individual guideline statements; at the same time, review text
(preceding the guideline statements) is informative but not as concise. I
suggest that a spirit of effective communication, not advocacy of any
particular "one size fits all" single evidence grading scheme, is what we
seek here to fulfill Robert Day's instructions. Perhaps we all agree that
these grading schemes are useful to screen published evidence, but that
letter grades alone don't convey enough information to describe the state of
knowledge underlying conclusions in evidence-based guideline documents.
David Birnbaum, PhD, MPH
Clinical Assistant Professor
Dept. of Health Care & Epidemiology
University of British Columbia, Canada
David Birnbaum, PhD, MPH
Clinical Assistant Professor
Dept. of Health Care & Epidemiology
University of British Columbia, Canada
|