What you might have seen in a GP record circa 1980, if you could have read the handwriting!

Early GP computing was fighting the hardware. How could we get 6000 patient records on a  pair of 8 inch floppy disks (well under 1Mb)?  The terse style of GP records gave us a chance, particularly if you could compress this rich clinical narrative into a couple of 4 byte Read codes   “H14, e151″)

Things got easier you could get a 10Mb hard disk for under £5k and we could afford a bit of free text – “Tonsillitis (probably) Pen .V and thus was born the computerised GP record that survives to this day. A structure coded, record, with text used occasionally to clarify or maybe confuse what w e meant by the code “Carcinoma of the Broncos (B192) Excluded (Text)” an approach driven by the limitations of the hardware but one that fitted with the GP approach to record keeping and led us to discover the power and flexibility of a structure record built on a terminology – What to record some new stuff, no need to change the software just add a few codes to the terminology.

This paradigm has served GPs well and continues to do so, codes first, text an afterthought. However, trying to extend this approach into other areas of healthcare where there is a need to record a richer narrative has been a challenge, spawning ever more complex terminologies (SNOMED) and complex structures (ENV 13606, HL7 V3, OpenEHR……) the holy grail has been semantic interoperability, computable, interoperable records. In my view this remains a noble pursuit but one that fits poorly with the alternate paradigm in much of secondary care where the clinical narrative (text) comes first and coding is an afterthought, probably done by someone else for purpose not directly related to the delivery of care.

Today I detect a new zeitgeist, which might provide a way to combine these two worlds. The priority now seems to be to put the right information in front of the right person at the right time in an eye-readable form. This takes us to a document-centric view where all you need is some computable metadata to route the document. However, while this approach offers many benefits over paper, it perpetuates records based on forms, scales, assessments and unstructured clinical correspondence which hardly unleashes the power of computerised records. Those leading the development of records in secondary care understand this and we are now seeing the development of records standards which apply some high level structure (headings) to create semi-structured clinical narrative. For those in secondary care this is useful progress and will hopefully facilitate the electronic sharing of clinical records but also, and more excitingly, help them understand the clinical benefits of coding parts of this clinical narrative.

What I hope to see as the next phase in the development of electronic records are records which support a rich clinical narrative, structured so that we can get the right information, in front of the right person at the right time, but also coded, where this is useful and practical, so that we can start to explore the possibilities that flow from computable semantically interoperable records.