Next, we address several other gaps in the PHR standards arena that have not been discussed in other sections.
Data Content Standards
Donald Mon, Ph.D.suggests one of the major areas that the standards community will need to revisit is data content standards. While HL7, DICOM, SNOMED, and LOINC are useful for PHRs, health care organizations still record data according to their own current data set. Thus, interoperability challenges are still present. Two examples illustrate the need for greater standardization with respect to data content.
The first example addresses gender data. Gender is currently collected in multiple ways across EHRs and PHRs (e.g., male = 1, female = 2; male = M, female = F; etc). Health care organizations may have systems to recognize various types of codes for male and female during data exchange. However, other organizations may also utilize additional codes for gender equal to hermaphrodite or unknown. During data exchange, this additional information would not be recognizable to a system that does not have hermaphrodite or unknown in the current data set.
The second example relates to medication use habits. David Lansky, Ph.D. asserted that there are a number of medical history data elements that are not standardized across PHRs, leading to interoperability challenges. One example is medication use habits (e.g., does the patient split pills in half; what are the patient’s reasons for not taking the prescribed pills; does the patient have post-procedure complications, etc.). The industry does not currently have a method to make this information part of the patient’s PHR, though it is critical data to include in the patient’s health record.
Until the industry has greater data standardization or potentially a standard data set, it will be impossible to achieve semantic interoperability for certain data elements. Dr. Archelle Georgiou, an independent consultant, commented on the difficulty of achieving consistency around semantics: ‘Achieving consistency around semantics – beyond the semantics for very specific or objective data points – is a much broader objective than establishing a technology based resource for holding health related information. Semantic interoperability would require cultural change that begins during the medical education process. Since this is unlikely, another approach is to recognize the inevitability of semantic variability and to create systems, support, and processes that work around them.’
Areas of Overlap in Standards Development
A number of standards organizations and work groups have been tasked to address similar issues related to PHR privacy, security, interoperability, and portability. While coordination between initiatives is ideal, initiatives sometimes move along different paths – with different missions, charges and time frames. Thus, areas of overlap do exist in the current standards development arena.
AHIMA is exploring the issues of overlap in standards development initiatives through a new project focused on the development of Regional Health Information Organization (RHIO) best practices for the Office of the National Coordinator for Health Information Technology (ONCHIT). As part of this project, AHIMA will look at how health information is exchanged in a health information exchange (HIE) environment, whereby the PHR could be one node in any HIE. Donald Mon, Ph.D. indicated that the ONCHIT project will require the standards community to think critically about the gaps and inconsistencies with respect to PHR standards and standards initiatives. Currently, a number of different initiatives have addressed PHR standards – including but not limited to the HISPC project, HITSP, IHE, and CCHIT. The project will explore whether the different initiatives, and their areas of overlap, are affecting the HIE (when the PHR is one of the nodes that information can be exchange to and from). Several areas of overlap to date are discussed below.
Consents. One specific example where there is overlap among standards development efforts is for consents. David Lansky, Ph.D. suggests that currently consents are done in different ways. One method is to use a consent directive, which essentially enables the provider to acquire all of the account holder’s consents up front. The benefit is that when a health care episode occurs, the provider does not need to acquire consent again. This process reduces burden on clinicians and improves the clinician’s workflow. However, from a policy and privacy perspective, there are problems with using a consent directive. Consumer advocates have asserted that consumers will be better protected by a policy that requires providers to ask for consent during each health care episode. Consent directives are one key area where the EHR-PHR standards community lacks a consistent policy. HISPC, HITSP, and other groups have addressed consent directives in a variety of ways, creating overlap in the standards development space.
Conditions and Diagnoses. According to HITSP, there is also an overlap of standards for conditions and diagnoses – namely SNOMED-CT and ICD-9/10. In 2007, HITSP recommended that both types of codes could be requested, and thus, whatever form the information is available in would be provided.
Portability. In 2007, HITSP also explored PHR portability standards and determined that that IHE-XDM and X12 are areas of overlap.