Hello all,
I've been testing the AKI algorithm in our version of APEX, and have seen some slightly odd behaviour. If multiple creatinines exist for a patient without an AKI result calculated for them, the algorithm appears to be looking at ALL these creatinines simultaneously and applying the calculation for the final creatinine to all the previous empty fields. Hence a perfectly normal creatinine can get reported as AKI, IF there is a later result which is consistent with AKI.
I'll concede this is largely an artefact of how rapidly I've been entering results in whilst testing - the delay whilst APEX gets around to calculating a value is only 10-30 seconds, and in routine practice there would rarely be more than one creatinine result on a patient authorised in that time, although it certainly is possible (e.g. a backlog of work, released as a batch from validation).
I can't think of a scenario where this would cause a significant problem, but it does imply APEX isn't behaving exactly as per the NHS England algorithm (no mention of including later results in the calculation there!).
Has anyone else with APEX seen this or are they able to replicate it? Would this behaviour worry anyone?
CSC have been contacted, we're still awaiting their response...
Alan.
Dr Alan Dodd, Senior Clinical Scientist
Biochemistry, William Harvey Hospital
Ashford, Kent.
|