Decision support applications and technology

There are a myriad of applications for computerised decision support systems (CDSS) in a hospital:

• implementing consensus based protocols and recommendations (for example for thrombophrophylaxis, falls prevention, antibiotic and other prescribing etc)
• pre-procedural recommendations customised by procedure, patient condition and  proceduralist
• standardised and customised admission and discharge evaluation batteries
• investigation result actioning
• pharmacological decision support
• triage to any service (from activating a MET call, to categorising out-patient appointment, allied health and specialist medical referrals)
• provision of targeted patient education at all phases of patient care
• information processing (filtering, sorting, clarifying, encoding - this banal job is possibly one of the most immediate applications)

Sharing knowledge bases (that capture the protocol logic pathways and idea classifications) between institutions requires shared syntax.

And after over 40 years of research into CDSSs the necessary standards are starting to emerge - in a sensible and usable format. GELLO looks like one of the best candidate languages for writing decision support rules^{[1] [2]}. Combined with the recently released usable sets of clinical terms arranged in a hierarchy it starts to look possible to start implementing decision support without being locked in to proprietary formats.

For the IT department, the questions become: How to evolve towards systems that are capable of :

• feeding all the necessary data, in a suitably structured format into the decision support engine that is aware of all we know about the patient? as well as
• providing the necessary hooks, triggers and output mechanisms to implement (or at least communicate) automated conclusions as alerts, reminders and even automated actions...

For the entire organisation, the challenge is to organise the efforts so that they are engineered for the future.



The potential is enormous... and close.