EHR Interoperability: The Structured Data Capture Initiative

Last week, we kicked-off the latest S&I framework initiative called “Structured Data Capture.”  In this week’s blog, I’d like to describe why this initiative is a fundamental and important addition to our portfolio of standards to support electronic health record (EHR) interoperability.

What are the Practical Applications of EHR Interoperability?

Meaningful Use (MU) Stage 1 and MU Stage 2 established the initial portfolio of interoperability standards to support exchanging data from one EHR to another EHR.

As more and more people recognize the value of standardized data, people are looking to take advantage of the power EHR’s offer to support: MU, quality measure reporting, and other important activities to aid in patient care.

Electronic, standardized, and sharable data means that there are many other uses that we might consider for the data contained within an EHR. For instance:

  • Patient safety advocates may want to use EHR systems to collect patient safety information, leveraging existing standards like the AHRQ “common format” for patient safety reporting
  • Providers and researchers may want to use the EHR systems to collect data for clinical research, including patient-centered outcomes research, and to identify patients who could benefit from participating in a research study
  • Providers may want to give referrals to their patients for community services, like smoking cessation or weight management programs, after discussing these topics with them during an office visit
  • Providers working with disease surveillance case report forms may wish to collect additional information about reportable conditions, such as infectious diseases
  • Provider’s office staff can use EHR’s to gain pre-authorization of certain kinds of medical devices where health payers may want to leverage clinical information collected in EHRs to support additional review of expensive medical equipment.

Although health care providers use various sources and methods to capture and synthesize patient-level data, EHRs are a data source with tremendous potential to provide timely and relevant data in a form that is:

  • Usable for quality and safety improvement
  • Population health
  • Research (sometimes labeled “secondary” use) [1]

EHR Interoperability: The Benefits of Structured Data Capture

It would be challenging to include every possible data element (as important as it may be) in the core MU data elements. If we do so, we risk overwhelming providers, vendors, or others with the complexity and scope of the standardized data that EHRs would be required to collect.

A solution to this problem is being explored in the latest new initiative in the S&I Framework – the structured data capture (SDC) initiative.

The goal of the initiative is to identify how EHR interoperability technology can be used to:

  1. Access a template that contains structured data, which is sometimes called common data elements (CDEs)
  2. Automatically populate the template with the correct CDEs from existing EHR data
  3. Store or transmit the completed template to the appropriate organization or researcher

Diagram of Structured Data Conceptual Workflow

Thankfully, there has been significant work on CDEs already, so like other S&I framework initiatives, we expect that the SDC Initiative will leverage existing EHR interoperability standards, harmonize them, and agree on a common approach to support structured data capture.

A list of the standards that already exist is available on the SDC Initiative wiki page here External Links Disclaimer.

ONC’s Work on EHR Interoperability Standards

ONC is working with the National Library of Medicine and the NIH to:

  • Identify an initial core research set of CDEs that could be used by the SDC project
  • Pilot how different agencies and organizations can coordinate and govern this core set of CDEs

We are also working with AHRQ to leverage their work on the common format for safety reporting.

Given the significant investments that have been made in EHR adoption and interoperability in the last 4 years, enabling structured data capture within EHRs is poised to be a critical way to integrate EHR data into a variety of health services and clinical research activities.

Future stages of MU will put a “down payment” on learning the health system to support health care quality, research, and overall public and population health. The SDC initiative will support the secure, trusted, patient-centered integration of the EHR into other parts of the learning health care system such as:

  • Research consortia
  • Registries
  • Bio repositories

Success in the SDC project will give us four important interoperability results:

  • A standard definition (and structure) for what a data element (CDE) is
  • A standard way to collect these data elements into templates
  • A standard way to automatically populate the CDEs with data from an EHR
  • A standard way to access, display, and store the data.

Ultimately these results will improve access to standardized electronic versions of data collection instruments for use in a variety of research and clinical reporting functions.

Our intent here is not to increase the data collection burden on providers or those entities charged with gathering this type of information.

Our focus is enabling the integration of these instruments into EHRs so that duplicate data entry is reduced throughout the workflow, and most importantly, the data collected is comparable and more useful across multiple groups—from researchers, clinicians, payers, and public health agencies to patients and their caretakers.

I hope that you will consider joining and participating in this exciting initiative!

Next week: Interoperability Is Not a “One-Size-Fits-All” Concept.

Read other blogs by Dr. Fridsma on standards development and harmonization, coordination of federal and private efforts toward interoperability and health information exchange, and health IT innovations.

[1] Kahn, M., Raebel, M., Glanz, J. M., Riedlinger, K., & Steiner, J. (July 2012). A Pragmatic Framework for Single-site and Multisite Data Quality Assessment in Electronic Health Record-based Clinical Research. Medical Care, S21-S29.


  1. Nick van Terheyden, MD says:

    Interesting outline on structured data entry and capturing this using standard templates that are pre-populated – that’s dead on (how often does the family history change in a clinical note…?)
    What I don’t see but believe will be central to capturing this information without burdening the physician with additional administrative tasks that detract from their focus on the patient is Natural Language Processing or as we call it Clinical Language Understanding (CLU)
    We like narrative because it captures all the information and leaves nothing out – technology is able to convert this into structured data, today we can do some elements really well at an accuracy in excess of 90%. This could flow seamlessly into the S&I framework creating structured data for sharing that is tagged against the relevant controlled medical vocabulary. This data is readily available for registries, quality reporting etc, but more importantly it provides a level of granularity on clinical information that can be used for multiple other purposes especially verifying the delivery of high quality care.

  2. Dr. Barry Robson says:

    As an e-epidemiologist data mining medical records for epidemiological and EBM/CER metrics and probabilistic CDS rules, I’m dubious that a further structuralizing Event-Attribute-Value type of layer will allow us to tackle and respond fast to say 250 million EHRs, and changing constantly with the wanted data rather spread around the XML. That is certainly more of a problem if access, as it ideally ethically should do, captures some spirit of fine grained consent where the patient dictates by notification, observation by observation etc in the EHR, what can and can and cannot be available for access and to whom, with what resolution of value and timestamp, and what combinations. I retired moderately recently as CSO IBM Global Healthcare Pharmaceutical and Life Sciences, and whilst these views are certainly mine and not necessarily IBM’s, I don’t see much in my personal experience there, nor in my current job as an epidemiologist, to alleviate these concerns as the EHR stands at present. As an epidemiologist I would of course be very happy to be proven wrong.
    My views on how to tackle this, and experiments with a pro-PCAST style of approach, can be seen in the following recording of the Johns Hopkins Grand Rounds Lecture that I gave on Feb 2 2013:-

  3. Adrienne Tannenbaum says:

    Thank you, thank you, thank you. I I am very excited that the US govt is now the place with the power to deploy standardization as a way of integrating data so that it can be “shared”, and the healthcare industry is the beneficiary that will be impacted almost immediately. I am looking forward to an organized approach that will begin here, but expand to other areas outside of healthcare.

  4. Andrew says:

    I applaud the government for going in this direction. As a very brief means of background, we have been performing precisely this programs ideals for the last 7 years and have perfected this structure in several industries. The explanation of how templated forms are selected and filled out and then autopopulates and is stored and sent back to the end user is really far too many steps to then diagram out how the data is then made “interoperable” and through what mechanism the repository is made available on demand. The very last thing I wish to note is that while I absolutely applaud the powers that be for going in this direction, I wish to caution them on having the “repository” be a private sector thing, and not a government owned and controlled entity. The macro direction is wonderful but please understand that Nasdaq, NYSE and other entities are where centralized data goes in the financial world. Enable the medical records to move in that same direction such that access to the repository or repsitories is 100% private sector and it will turn out much better in the end. More important than anything I have written thus far, let me share the most profound aspect of whats in the way of sharing information beyond interoperability, its simply put “clean data capture” and you have successfully addressed this concept here. Junk in equals junk out so it pays not to share junk. Thus, we must truly start by parsing data based upon when and by what means it was collected. Electronic form data with biometric or signature based sign offs is by far the most accurate and clean data capture in the medical world, where quality in =quality out and thus, is worth sharing!

Leave a ReplyComment Policy