Prodigy, a computer assisted prescribing schemeBMJ 1996; 313 doi: https://doi.org/10.1136/bmj.313.7071.1549b (Published 14 December 1996) Cite this as: BMJ 1996;313:1549
Interim data show that it is worth taking the scheme further
- Ian Purves
EDITOR,—I wish to address the criticisms of the computer assisted prescribing scheme Prodigy that Iain E Buchan and colleagues raise.1 Prodigy is an innovative project that seeks to involve practising general practitioners in evaluating and refining a product intended for use by general practitioners. The primary role of the Sowerby Unit for Primary Care Informatics at the University of Newcastle is to evaluate rigorously and objectively the results of the various phases of testing and to highlight their implication for improving the prototypes. I believe that the collaborative methodology that has been used is new to the NHS and has merit (I N Purves et al, annual conference of primary health care specialist group of British Computer Society, 1996). The Prodigy team has received, sought out, and welcomed criticism from health service researchers and people working in health informatics around the world.
An interim report on Prodigy was written to inform the community in a politically charged atmosphere.2 The report is clearly a snapshot of currently available information, with conclusions based on a number of evaluations (I N Purves et al, annual conference of primary health care specialist group of British Computer Society, 1996).2 Quotes by Buchan and colleagues originate from the conclusions of the report but were not immediately linked to the data in the way that their letter suggests. Buchan and colleagues juxtapose the quoted conclusion of “‘confirming [the] desirability’” of Prodigy to the responses to the question “How much would you want to continue with Prodigy?,”1 but the text that was juxtaposed in the report was “94%… consider Prodigy to be a concept worth developing (16% being extremely happy, and 78% endorsing its development as long as either ‘some improvement’ or ‘significant improvement’ is made).”2 This is not confirming desirability but supporting a concept. Similarly, the interim report does not claim that Prodigy's “‘effectiveness’ is confirmed”1 because of “‘relative reduction in the rise of expenditure of 1.1%.’” In fact it states, “At this early stage we have insufficient data… to make any statistical inferences.”2 Finally, Prodigy has been “thrust” on no one: all 137 sites using it volunteered.
Prodigy is a combination of active and passive systems. It is not ready to be rolled out: the data in the interim report suggest only that the concept is acceptable and worth taking further. Prodigy is an innovative yet practical step in clinical computer systems, and one has to start somewhere. Reliable and statistically robust conclusions will be possible once the study is completed; only then will we be in a position to judge the merit or otherwise of Prodigy.