© 2005 by the American Diabetes Association, Inc.
Guidelines for Computer Modeling of Diabetes and Its ComplicationsResponse to American Diabetes Association Consensus Panel
1 Economics Division, University of Liverpool Management School, Liverpool, U.K Address correspondence to Prof. A. Bagust, Economics Division, University of Liverpool Management School, Chatham Street, Liverpool, U.K. L69 7ZH . E-mail: a.bagust@liv.ac.uk
The ADA consensus panel guidelines for computer modeling (1) may create a misleading and unduly optimistic view of modeling for general readers. Repeated use of the terms "accurate" and "reliable" suggests that models are capable of generating precise predictions of long-term clinical outcomes and costs. Sadly, models do no more than encompass our current understanding (and ignorance) of a disease, its treatment , and its natural history. Despite powerful computers and sophisticated software, no model generates new datait merely combines existing findings within a
|
| |||||||||||||||||||||||||||