Several questions can arise when considering whether to parse the data of a CCD document from a remote facility and merge it into the local EHR. HIEs are spurring increased emphasis on the use of CCD documents to exchange information among facilities, so it would be ideal to have consistent practices regarding the merging of the received data into the existing EHR. With the CCD seen as the vehicle for EHR-to-EHR communication, the questions below raise some concerns as to how this vehicle should be utilized.
1. Is having the CCD available as an attachment sufficient?
With some EHR implementations, the receiving system may not be able to support the import of the CCD. Or, the EHR vendor may chose to optionally import only some of the data, such as allergies and medications, but not the rest of the data. The provider may also be given the option to select what data is imported during the initial implementation of the EHR system.
Having a human readable CCD available to the physician is obviously preferable to not having any data available at all. But is simply having the human readable CCD sufficient? It may be sufficient in some cases, but it is definitely not optimal. With the use of clinical decision support systems and quality of care analytics, having level 3 coded entries imported into the EHR is the only way to take advantage of such healthcare tools that utilize the stored data.
2. Who is liable that the data is correct?
Physicians can be hesitant to accept other provider’s data into their EHR because they assume they will become liable for it. This opens a large legal question as to who is responsible if the patient is treated incorrectly based on bad data that was not even collected by the physician’s medical staff. Receiving providers do not like the idea of supporting the imported data because they only truly trust what was directly input into their EHR by their staff.
Separating the CCD as a human readable attachment draws a clear distinction as to what information was entered in-house, and what is relayed from an outside source. If the CCD is parsed and merged, it would be preferable from the physician’s point-of-view to have this data clearly distinguished as received from an outside source with the ability to identify and contact that source.
3. What audit trail exists for the data?
The ONC has specific guidelines for audit tracking in its EHR Certification process. But how do those audit trail standards apply to imported data? A CCD does not include all the details about who entered every entry, and whether the entry was ever edited. From a legal perspective, some states also have specific laws pertaining to re-disclosure. These laws define how imported information must be tracked and protected.
When data is imported, the EHR system should be as detailed as possible about the source of all the data. The CCD audit trail is limited to the participant information included in the header of the document. At a minimum, this participant information must be tagged to each entry as it is parsed and merged with existing data.
4. What if there is conflicting data?
The reconciliation of the data can be a difficult routine for vendors to implement as part of the process for importing CCD data into existing EHR data. Sorting out duplicate data is the first logical step, but resolving conflicting data or displaying it in an effective manner can be a difficult task to perform.
At a minimum, if the resolution of the conflict cannot be determined, the physician must be made aware of the conflict so they can do their own analysis of the situation. Ideally, during import an alert would be sent to the staff so the patient, and/or responsible parties, could be contacted to ensure the correct data is represented in the system. This puts a large burden on the EHR vendor to accommodate a unique workflow that can be utilized for resolution.
5. Is someone responsible if the data is not merged and a bad decision is made because of the lack of data?
Clinical decision support systems will not work if the data is not in the EHR. What if a medication interaction was not flagged because the medication section of a CCD was not imported into the EMR, and that interaction resulted in harm to the patient?
Clinical decision support systems and quality-driven analytics are important tools in the changing healthcare landscape. Level 3 CCD data, in the form of coded entries, support the use of these tools, which aim to provide better patient care through the use of more available data. The decision to only use the CCD in a human readable format, rather than importing the data, directly impacts the effectiveness of the tools.
6. What is best for patient care?
The final question is undoubtedly the most important. The quality of care for the patient should ultimately be the driving force in whether the CCD is imported into the receiving EHR system. On one hand, if the imported data simply clouds the existing data because of poor reconciliation or a failure to clearly indicate the source, then the “easy” approach of simply making a human readable format would be superior. But, given the capability to cleanly import the data and clearly mark its source, there is little doubt that importing and merging data would lead to superior clinical outcomes.
Forget the complexity of the CDA format and the difficulties of parsing, mapping, and reconciliation. The politics behind actually importing the data for real patient care is the big hurdle that needs to be overcome. Issues behind liability and re-disclosure need to be addressed so the tools that enable a higher quality of care, based on coded data, can be utilized to their full extent. Maybe future stages of Meaningful Use will provide the guidance.
This article originally appeared in the January 2012 issue of HL7 News, the newsletter of Health Level Seven International. Republished with permission.
Latest posts by Rob Brull (see all)
- Opinions about HL7 FHIR in the JASON Report - January 22, 2015
- What the JASON Report says about CDA - January 8, 2015
- The CCD/CCR Fork in the Road. Did Health IT Take the Wrong Path? - July 30, 2014