A Tale of Two C-CDAs: A Revolution in Usable Clinical Data
C-CDAs Still Leave Much to Desire in Terms of Usability; We’ve Got a Plan for That
Christopher Vitale, Pharm.D, R.Ph, Senior Clinical Informaticist
Much has been written about the cognitive and workflow burdens of today’s electronic health records including the oft-quoted Fortune/Kaiser March 2019 article “Death by a Thousand Clicks: Where Electronic Health Records Went Wrong.” 1 Standards for exporting more usable data from these systems for transitions of care such as the CCD also fall short. Diameter Health has addressed these issues by offering a configurable stylesheet that, coupled with our ability to normalize and enrich data elements for ingestion by software systems, improves access to what might be called a “digital chart” – the human-readable rendering of key healthcare data.
In software development, the concept of usability focuses on a user’s ability to perform tasks safely, effectively and efficiently within the constraints of an electronic system. This concept has been studied within the clinical realm for some time and has guided the evolution of existing clinical systems, such as electronic health records and clinical decision support. For example, Zhang and colleagues studied usability in the context of clinical workflows for over two decades at The University of Texas Health Science Center at Houston School of Biomedical Informatics. Through efforts like these, we better understand the potential impact of usability on patient care, such as:
- Fragmented patient records requiring users to know where to look to find all relevant information and require intra-team conversation to close information gaps, which leaves problem recognition left to chance.2
- User interfaces intended to reduce errors that do not reflect human factors and safety design, which can yield new errors by creating cognitive burden of absorbing information and an environment where important information and trends are easily overlooked. 2
- Core cognitive contributors to medical errors including poor individual-technology interaction and lack of standardization of work process and communications. 3
Concepts developed through research like this have been so impactful that they were incorporated into the SAFER Guides developed by The Office of the National Coordinator (ONC) which provide recommended practices to optimize the safety and safe use of EHRs.5 For example, the guide for Computerized Provider Order Entry with Decision Support states that functionality should be “supported by usability testing based on best practices from human factors engineering,” and that risks of untested usability include “decreased clinician efficiency and clinician dissatisfaction, as well as errors and adverse events.”
Standardized document types defined by the Consolidated Clinical Document Architecture (C-CDA) were developed to assist in communicating relevant information between clinical settings and reducing the cognitive burden of absorbing information. But these standards present their own set of issues in terms of consistency and usability.
Usability concepts studied within the context of EHR workflows should also be applied to clinical data curation and presentation. Health Level 7 International (HL7) offers guidance on “improving the relevance and pertinence of C-CDA documents as experienced by the clinician, which means are displayed or ‘rendered’” with the Clinical Summary Relevant and Pertinent Data Implementation Guide.6 The importance of such guidance becomes obvious when considering examples of clinical data from C-CDA documents rendered through an XML stylesheet. The examples visualized below are testing artifacts from fictional patient records scripted by ONC and created by various health information technologies as part of Meaningful Use certification.3 These documents have been rendered via a standardized XML stylesheet that is publicly available through HL7 (CDA.xsl).7
Figure 1: Three examples of identical patient records captured through various certified health information technologies and viewed with a single stylesheet. Note the difference in available data, value rounding and units, and overall organization.
Figure 2: Example of a patient record where allergy information is captured across different sections of the clinical document (i.e., ALLERGIES and HEALTH CONCERN).
At Diameter Health, our core application, Fusion, was designed to refine healthcare data by normalizing, enriching, reorganizing, deduplicating, and summarizing all clinical data for a patient regardless of source system or format. By harmonizing the use of coded terminologies (i.e., normalizing and enriching), moving and removing redundant entries within the appropriate sections of a clinical document (i.e., reorganizing and deduplicating), and summarizing all clinical data into a single consolidated view, we make clinical data easier for computers to process.
Additional steps must be taken to present this enhanced data in a clear, consistent, clinically-relevant, and configurable fashion for human users. XML stylesheets are designed to operate on the human-readable sections of clinical documents; content in sections must be organized to support a consistent view. Included with Fusion is a configuration file called config-ccda.js, which enables data ordering within sections of a clinical document and configurable styling of the human-readable content. Configurable styling includes which values to expose, table and data formatting, and element path formatting, as well as many other options to meet the needs of various clinical and non-clinical end users. Also supplied with the Fusion application is the Diameter Health XML stylesheet (DH.xsl) which offers visual and organizational benefits over the standard HL7-developed stylesheet for viewing CDA documents. Additional information regarding the numerous configuration features that are available through the use of config-ccda.js, and how to use DH.xsl can be found here and here on the Diameter Health customer portal.
Below is a view of the same clinical document used to generate two of the images above but after Fusion processing with the default config-ccda.js configuration settings and rendering the resulting human-readable portions via DH.xsl. All allergy information is now presented in the Allergies section, and vital signs now include all clinically-relevant information; data across document sections is presented in a consistent fashion.
Figure 3: Example of a sample clinical document after processing through Fusion with default config-ccda.js configuration settings and rendering via DH.xsl.
In summary, Diameter Health solutions incorporate key software usability principles to address data fragmentation, cognitive burden of absorbing information, and standardization of work process and communications, which can lead to an increase in productivity and decrease in human error. Configuration and data rendering capabilities enabled by Fusion offer the ability to organize data in ways most meaningful to relevant use cases while supporting key usability principals.
1) khn.org. https://khn.org/news/death-by-a-thousand-clicks. Accessed May 2, 2020.
2) Zhang J, Walji M. Better EHR. National Center for Cognitive Informatics & Decision Making in Healthcare; 2014. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.703.7379&rep=rep1&type=pdf. Accessed May 2, 2020.
3) Zhang J, Patel V, Johnson T. Medical Error: Is the Solution Medical or Cognitive? JAMIA. 2002;9(6):S75-77.
4) healthit.gov. https://github.com/onc-healthit/2015-ccda-certification-samples. Published 2015. Accessed May 2, 2020.
5) healthit.gov. https://www.healthit.gov/topic/safety/safer-guides. Published November 28, 2018. Accessed May 2, 2020.
6) HL7 CDA® R2 Implementation Guide: Clinical Summary Relevant and Pertinent Data, Release 1. Hl7.org. http://www.hl7.org/implement/standards/product_brief.cfm?product_id=453. Published April 13, 2017. Accessed May 2, 2020.
7) cda-core-xsl. HL7.org. https://github.com/HL7/cda-core-xsl/releases/tag/v4.0.1. Last published 2019. Accessed May 2, 2020.