Dear All
There is an issue that I have been thinking about for some time and which has come up again around an openEHR implementation site - it is this. Should a clinician be able to mark Entries as unsafe for automatic processing? The idea is that information might appear in the record - it is not an error in the usual sense (it is not wrong) - but it is wrong from the point of view of queries or decision support.
The example that made me think this was worthwhile was a patient who seeks opioid medications who insists that she is allergic to Paracetamol - but takes Mersyndol regularly (a compound product with 500mg of Paracetamol). She insists on having the allergy to Paracetamol in her record - and I have discovered that if I remove it, she informs a member of staff who is not known to her and it gets in there again. I would like to leave it in but mark it as unsafe for automatic processing.
Another example that springs to mind is if a health professional with a different model (e.g. Naturopathy) writes a useful communication about a patient that I want to incorporate into the record. In the health record this person has entered 'Hypercalcaemia' in the problem list derived from some observation - this does not conform to Medicine's definition of this term and I would like to mark it as unsuitable for automatic processing.
In the implementation trial the EHR is a repository receiving information from many sites - they have some errors of spelling of information that is important and is misleading - as they are unable to edit this they want to be able to mark it as unsafe for processing.
The criteria for using this marker on an entry are:
1. It is not an error - ie it should not be corrected.
2. It will lead to errors of automatic processing - ie it applies a concept to the subject of care that is not appropriate and does not add - or detracts from the safety of their health care.
I am interested in what others think.
Sam Heard