§1. Big data analysis

Health care is an information- and communication-intensive industry that creates large collections of data (called “big data”) for many purposes. Analyzing big data can reveal hidden patterns that can (i) improve understandings of disease; (ii) assess how well treatments work; (iii) measure the quality of care and clinician performance within their care practices; (iv) analyze costs and the use of resources; and (iv) monitor the public’s health.

The increased use of health IT by providers and by patients will produce even larger amounts of data:

  1. EHRs will be one source of information to feed into the combined data repositories.
  2. Other sources include patient registries and insurance claims from private payers and public programs including Medicare and Medicaid.
  3. Patient-generated health data (from, e.g., mobile apps, wearable and implanted sensors, and other forms of telemedicine) can also be collected through an EHR or directly submitted to researchers, to a public or private health information exchange, or to another information system. (See Patient generated health data.) Patients could, for example, use e-tools to record drug side effects or observations from their treatment or clinical trial settings.
  • How can health IT support the flow of information from patients to researchers through various channels, including health care providers, payers, and patients themselves?
  • Would consumers expect to be compensated for donating data?
  • How can the federal government best spur innovation and support the flow of information that makes big data analysis possible?

§2. Patient-centered outcomes research (PCOR)

Bringing together these data sources for analysis can enable low-cost, pragmatic clinical trials that are person-centered, incorporate real-world settings, and focus on highly targeted population groups. The aim of “patient-centered outcomes research” (PCOR) is to organize scientific evidence into a national network of databases and use the evidence to identify the best treatment options for specific populations. In this way, care can be individualized when the research is combined with a provider’s knowledge of a specific patient’s demographics, biological or genetic makeup, treatment history, and health goals – for example, avoiding treatments that may not work well, or improving the chances that a new drug could provide better outcomes for one patient with diabetes but not another.

This information allows providers to benchmark their work to best practices and improve patient care. In this way, the practice of health care can become more evidence-based, leading to faster, more accurate diagnoses and highly personalized care plans.

Health IT can also improve patient outcomes by enabling researchers to share the results of studies more broadly than before, through electronic alerts, new decision support tools and similar services. Providers then can apply these findings to their patients, identifying who will benefit most from new treatments or how to screen for certain health issues.

  • How can the federal government make the dissemination of research and study results more efficient and effective at improving patient health outcomes?
  • What policies and guidance will maximize consumers’ and providers’ ability to translate PCOR results into clinical and personal practice patterns?
  • What policies and guidance will encourage providers to use PCOR results, combined with individuals’ personal and medical histories and preferences, to establish highly personalized care plans?
  • How can developers create apps and eHealth tools that combine personal health information with scientific evidence on how well treatments and services work, as well as with potential insurance coverage issues?

§3. Privacy & security

ONC recognizes that privacy and security concerns may impact the willingness of consumers and providers to use health IT to increase access to health information. Privacy and security are the topic of a separate section of the Strategic Plan (Goal 3), since these concerns cut across all areas of health IT. Although this public comment period is not targeted at updating Goal 3, ONC will be revising that section of the Plan and your discussion of these concerns, and ideas about addressing them, will help the federal government decide on appropriate future steps.

(You can read the current version of Goal 3 by clicking the Background Documents tab at the top of this page, and selecting 2011-2015 Federal Health IT Strategic Plan).

People's Comments

(12)
Comments RSS Feed
This discussion is now closed.
show all (12)
There are no comments. Click the text to your left to make a new comment.
March 27, 2013 11:16 am

In my opinion, one of the keys to realizing the full potential of the growing health information database will be to introduce through legislation or rule making a provision that very clearly defines de-identified data as no longer belonging to or being in control of the patient. This will remove the uncertainty around the use of aggregate data in the research process. While HIPAA alludes to this in provisions around Healthcare Operations, I think a more specific ruling would be beneficial.

Use these buttons to endorse, share, or reply to the preceding comment by jherbert.
    March 28, 2013 9:57 am

    jherbert, do you think that patients will accept that their de-identified data will no longer belong to them? What could ONC do to help patients accept this?

    Use these buttons to endorse, share, or reply to the preceding comment by Moderator.
      March 28, 2013 10:04 am

      It will be a challenge but I think attitudes can change. I think education will be the key. Helping patients and providers to understand the societal benefit of being able to study this data in the aggregate would be a good start. Also, insuring that once this data goes into the cloud that it cannot ever be traced back to the patient will be important – that’s an area where ONC could help a great deal by setting standards.

      Use these buttons to endorse, share, or reply to the preceding comment by jherbert.
      March 30, 2013 1:32 am

      In order to meet jherbert’s challenge to convince patients that they should have no control over their “de-identified” health records, ONC could attempt to convince patients that all the scientific literature on the ease with which de-identified information can be re-identified is fallacious.
      Compensating patients who wish to contribute de-identified information to the research community, which is well compensated for their efforts, might prompt some patients to agree to this. ONC could conduct an economic analysis on this approach to determine whether the price that researchers are willing to pay is sufficient for patient participation. However, arguing that patients should have an altruistic interest in furthering research, which isn’t demonstrably altruistic in… more »

      …its contributions to improving overall health of their communities, may be a “hard sale”.
      Unfortunately, in many cases, even getting on the candidate list for a clinical trial, which may not even exist yet, a patient has to agree to disclose all health information (no guarantees of de-identification) into the future to all entities involved in the research community in which a current (or future) clinical trial is being conducted.
      So the question of how to convince patients that they don’t own their de-identified health information may be moot due to the research community’s “contract of adhesion” approach to forcing authorized disclosure of the patient’s health information. Interesting that this approach is similar to that used by insurers who are not governed by HIPAA. « less
      Use these buttons to endorse, share, or reply to the preceding comment by curious.
April 2, 2013 1:39 pm

There seems to be a possible correlation between exercising and decreased PTSD symptoms. When soldiers are discharged from the military they are usually given a physical and a last PT test (physical training test).
Giving discharged soldiers an app for recommended exercise monitoring and routines for after they are discharged would be relatively easy as many of these kinds of apps are online already. Connecting the app to AKO (Army knowledge online) and connecting it to DTMS (which is where soldiers can see their physical training exercise scores online) would give a discharged soldier a continuation of their exercise records and possibly routine after they were discharged.
If soldiers during the discharge process were sat down in front of a computer to download a personal app which… more »

…was connected to their previous physical exercise achievements and linked to their personal email and ultimately to the medical records “blue button” veterans and reservists could continue their exercise routines. The app would also be a link to soldiers/veteran “ blue button” which when visiting their veterans clinic could be downloaded (giving the information, the individual soldier/veteran inputted) to the clinician for possible evaluation.
I believe this process would enable researchers and help veterans.
As we are talking only about exercise routines having this information online would not be controversial and relatively easy to pass.
« less
Use these buttons to endorse, share, or reply to the preceding comment by eddie morgan.
    April 2, 2013 3:33 pm

    Thank you for your suggestion, eddie morgan. It sounds like you have some background in armed services health care. Do you have any thoughts on how the government could facilitate sharing the data you discuss beyond the armed services context?

    Use these buttons to endorse, share, or reply to the preceding comment by Moderator.
April 16, 2013 4:53 pm

In regards to the third bullet, ONC should continue to promote interoperability standards and certification criteria across the continuum of care. This will allow important information to flow freely for research, quality measurement and other ‘big data’ analysis purposes.

Use these buttons to endorse, share, or reply to the preceding comment by willr.
May 1, 2013 6:19 pm

PCORI has the right thinking for doing patient outcome research. I’d just like to add on it. From the success of Internet consumer businesses like Amazon and Google, I know we have the technologies required to create cloud-based patient services with real time research capabilities in super-large scale. To me, this really means that it’s possible now to build real-world scalable systems that can guarantee reaching the goals of improving patient outcomes across a large population. Such new research approach will not suffer from hit-or-miss and hard-to-scale issues facing the traditional research approaches. If HHS could encourage and fund more research projects that apply this new approach, we may see more bang for the buck.

There are several requirements for this new research approach… more »

…to work. I want to mention three here:

(1) The proposed solution must be designed to be scalable right from the start. Scalability here means that you can deploy the solution from one clinical setting to another or across the whole population with no or minimal additional cost. This upfront design requirement is absolutely critical because the technologies you choose to build a large scale system vs. a small scale system can be very very different!

(2) The system must enable real-time patient outcome research on real time data, lots of data. Automation is a key. Readily available big data analytics makes this possible.

(3) The system must also be able to deploy the research results rapidly so that the system can be iterated and improved rapidly.

In a nutshell, I think the healthcare reform is transforming healthcare to a patient-centered and data-driven learning health system. Government funding policy would need to shift accordingly in order to see more research going to that direction. « less

Use these buttons to endorse, share, or reply to the preceding comment by ajchen.

(Q#1) ONC should require the capability for EHRs to identify relevant clinical trial opportunities for individual patients as a way to introduce efficiencies for both providers and patients in terms of identifying potential trials for patient participation. Facilitating communication between health practitioners, the research community, and patients and their caregivers is critical to advancing cures and treatments, and ensuring patient access to them.
o If patients are providing health information directly to a researcher, it is essential that there are privacy and security protocols in place that govern the transmission of health information.

Use these buttons to endorse, share, or reply to the preceding comment by national partnership for women families.
May 8, 2013 8:16 pm

Health IT has the potential to vastly expand data available for health services research and quality improvement by making available more comprehensive, real-time, population-level and community-level clinical data available for analyses. Aggregation of data from all local providers in a geographic service area is possible with health information exchange capacities. In order for health IT to enable research and inform clinical practice improvement that reduces disparities, all quality data must be stratified by patient race, ethnicity, language and other factors such as disability status, sexual orientation, and gender identity. These demographic data elements must continue to be included in any aggregation of the data from multiple providers and sources. In terms of race and ethnicity,… more »

…further granular disaggregation, as recommended by the Office of Minority Health under its Affordable Care Act section 4302 standards, and further recommended by the Institute of Medicine in its 2009 report “Race, Ethnicity, and Language Data: Standardization for Health Care Quality Improvement” should be implemented through health IT, including all ONC-certified EHRs and all health information exchanges.
-California Pan-Ethnic Health Network « less
Use these buttons to endorse, share, or reply to the preceding comment by csanders.
May 9, 2013 4:18 pm

To: Office of the National Coordinator of Health Information Technology
From: The Fenway Institute at Fenway Health
Re: Comments on Federal HIT Strategic Plan submitted at http://www.planningroom.org
Date: May 9, 2013
Topic 10: Using health IT to enable research and inform practice
§1. Big data analysis
• How can health IT support the flow of information from patients to researchers through various channels, including health care providers, payers, and patients themselves?
• Would consumers expect to be compensated for donating data?
• How can the federal government best spur innovation and support the flow of information that makes big data analysis possible?
The Office of the National Coordinator of Health Information Technology… more »

…should mandate sexual orientation and gender identity data collection in EHR as part of Stage 3 meaningful use guidelines and provide a functionality to track this information in the core demographic variables section of EHR. This would allow systemic data collection that could dramatically improve research science on LGBT health disparities.
– Sean Cahill, Ph.D.; Harvey Makadon, M.D> « less
Use these buttons to endorse, share, or reply to the preceding comment by sean cahill.
May 9, 2013 4:21 pm

To: Office of the National Coordinator of Health Information Technology
From: The Fenway Institute at Fenway Health
Re: Comments on Federal HIT Strategic Plan submitted at http://www.planningroom.org
Date: May 9, 2013
Topic 10: Using health IT to enable research and inform practice

§2. Patient-centered outcomes research (PCOR)

We know that Automatic Computer Assisted Self-Interview technology dramatically increases response rates regarding homosexual and bisexual identity, behavior and attraction (see Villaroel et al, Public Opinion Quarterly, 2006). It is logical to assume, therefore, that Patient Reported Outcomes tablets may also lead to higher and more accurate response rates to questions about LGBT identity, as well as other sensitive… more »

…issues such as health risk behaviors, substance use and mental health issues. The Office of the National Coordinator of Health Information Technology should mandate sexual orientation and gender identity data collection in EHR as part of Stage 3 meaningful use guidelines and provide a functionality to track this information in the core demographic variables section of EHR. This would allow systemic data collection that could dramatically improve research science on LGBT health disparities. Reminding providers about LGBT health disparities—highlighted in Healthy People 2020 and in recent years by the Institute of Medicine—and of the connection between data collection and addressing these disparities, could help educate providers about the importance of data collection and how it can help them to provide better care to their LGBT patients.
– Sean Cahill, Ph.D.; Harvey Makadon, M.D. « less
Use these buttons to endorse, share, or reply to the preceding comment by sean cahill.