Testing of PRODIGY continuesBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7133.776 (Published 07 March 1998) Cite this as: BMJ 1998;316:776
- Ian Purves, Head,
- Gary Nestor, Research manager,
- Katherine Williams, PRODIGY communications coordinator
- Sowerby Centre for Health Informatics at Newcastle, University of Newcastle, Primary Care Development Centre, Newcastle General Hospital, Newcastle upon Tyne NE4 6BE
EDITOR—Walton et al correctly state that “both CAPSULE and PRODIGY [computer support systems for prescribing] represent attempts to bring the benefits of computer support … into primary care” and highlight the lack of relevant research at the primary level.1 Their paper provides evidence that computers influence prescribing in general practice, with the potential economic benefits identified by the Audit Commission. Analysis of the cost effectiveness of prescribing interventions needs, however, to be presented within a wider assessment of overall healthcare costs, and the notions of rational and quality prescribing should be broadly defined.
PRODIGY's clinical recommendations are not restricted to drug prescribing but cover the full range of management issues for over 200 clinical conditions common to general practice. Options presented include referral or investigation, non-drug treatment, and patient information leaflets. Recommendations are produced by a multidisciplinary team after a structured literature search and review of the evidence; they are rigorously validated by a panel drawn from the Royal Pharmaceutical Society, Royal College of General Practitioners, and General Medical Services Committee. The PRODIGY project team, funded by the NHS Executive, is doing pioneering work at the Sowerby Centre towards developing a robust method for integrating evidence into rational, educational, and interventional guidance. Phase two of the trial also included a large amount of patient related data from the electronic medical records to enable the system to present information tailored adequately to the individual.
As Walton et al state, “it is important that the effects of decision support systems for prescribing are carefully evaluated both during their development cycle and in real life.” PRODIGY has been available for use in consultations with real patients to over 750 general practitioners nationally, with eight different clinical systems being used. The project has achieved the delivery of prescribing support on current practice computer systems, as requested by the general practitioners interviewed in Walton et al's study, and the experience of users of PRODIGY has contributed to the innovative development process. The rigorous “real life” evaluation of PRODIGY represents a clear advance on the testing performed on a laptop computer by 42 local doctors, using simulated cases, as described in the paper.
Laboratory evaluation is an important aspect of the development of health information systems, but such systems should undergo two further types of trial if they are to be fully evaluated before implementation.2 PRODIGY has been submitted to type 2 (real life) testing; the COGENT project (information available at http://www.ncl.ac.uk/~ncenthsr/research/cogent/index.html), which uses a common technology and is synergistic with PRODIGY, is a type 3 trial.