Current document formats tend to prioritize ‘discrete’ data elements that are easy to store and understand individually over longer format narrative information that better captures the ‘story’ of the patient. Improving our ability to send and include that information in ways that are easily digestible by receiving organizations and clinicians can significantly improve patient care.
Reply to this post with your answers to these questions:
If it is available, should a clinical narrative always be included when primarily discrete information is shared (e.g. automated summaries of care)
1. Are there specific scenarios where this is more useful?
2. In what ways can context between narrative and discrete data be improved in external summaries/documents to easily tell the patient’s story, integrate and support clinical decision making within workflow?
The USCDI has the right idea in promoting summary documents–the discharge summary for inpatient care and “progress note” for outpatient care. These two documents best convey the patient story. I will again point out that “progress note” is a terrible label for the outpatient note–very few people in the clinical arena consider “progress note” and “outpatient note” to be synonymous, and this will cause confusion. That aside, these two summary notes are probably the most useful, and the ones in which an un-parsed narrative would be most useful. Discrete data can be sprinkled in the narrative where appropriate, lending it useful context.
One barrier in creating C-CDA compliant notes, however, is that most EHRs are not capable of putting them in the fully-compliant state. Another problem is that attempts to do so may distort the relationships between parts of the narrative. LOINC coding of the documents at a granular level is vital, but attempts to parse narrative and make it more “discrete” (along with LOINC coding the parts) beyond that may cause more problems than they solve.
Dr. Bronken, can you elucidate more on EHRs non compliant notes? Also why are you trying to parse notes to make them more discrete? Are the discrete data elements used to create the C-CDAs not available with appropriate coding (i.e. lab results LOINC, meds, RxNorm, ICD/SCT Problems)? For labs that send discrete results encoded with LOINC (and SCT for specimen, organism, qualitative values), is this information being lost/translated, etc. resulting in errors of omission and commission you are seeing on the EHR side? (Some of us are working on national lab interoperability initiatives and understanding these barriers would be helpful as we work on solutions.)
I will make one more plea for a document that permits narrative as well as discrete data without the constraints imposed by the C-CDA templates. The Unstructured Document C-CDA template does not allow structured data. Most of the other C-CDA templates have mandatory required sections that may not be germane or easily created by the user or the vendor. I urge consideration of a CDA document that allows an unstructured narrative and as many or as few structured sections as appropriate. That would be better than an Unstructured C-CDA Document (with no structured sections), and gets us closer to the ideal–a fully-structured C-CDA document. Let’s not let perfect be the enemy of good.