Do Personal Health Records Improve Care?

In the July issue of the JAMIA, Peggy Wagner et al provide an interesting report on the use of personal health records (PHR) in an effort to improve hypertension control. In this well-designed study, 446 patients were randomized to either receive access to a PHR connected to the patient’s electronic medical record or usual care (i.e. no PHR). Physicians were from ambulatory practices in either family medicine or internal medicine. Metrics included blood pressure, body mass index, the Patient Activation Measure (PAM), Patient Empowerment Scale, Patient Assessment of Chronic Illness Care (PACIC), Consumer Assessment of Healthcare Providers and Systems Clinician & Group Survey (CAHPS) and utilization measures (e.g. hospital days, emergency department visits, and outpatient visits). The PHR functions available included secure messaging, access to educational materials, medication interaction checking, recording/monitoring of health measures such as blood pressure, viewing of some EMR data, and some goal setting and health diaries.

So… what were the results?

  • Patients in family medicine were more likely to use the PHR
  • As age increased, PHR usage decreased
  • PHR use increased with self-reported Internet use, higher average diastolic blood pressure, and higher provider scores on CAHPS communication measures
  • A greater number of inpatient days was associated with lower PHR use

But… only 26% of the patients used the PHR frequently. There were no statistically significant changes regarding patient activation, improvement in outcomes, satisfaction with care, or change in the use of medical services. The authors write:

This underscores the necessity to consider the immediate health outcome potential of subsequent stage 2 implementation plans for meaningful use criteria which require that patients are provided with information access after inpatient and clinical encounters with expectations that 20% will use a portal at least once. Our results suggest stage 2 implementation may not produce changes in patient health outcomes. Whether this is due to limited likelihood of patient use or limited impact when used is not yet understood.

Of course, this is only one study and whether these findings can be generalized is difficult to determine. The authors do cite other studies that also noted infrequent use of PHRs. They also point out that patients who were the most frequent PHR users did reduce their diastolic blood pressure by 5.25 points and almost 4 points in the systolic blood pressure — but they could not determine whether these results were related to PHR use, greater motivation for change, or other factors. As for the overall results, it isn’t clear whether the lack of PHR usage was the reason for the lack of benefit, but the fact that 54% of patients used the PHR less than two times in 9–12 months (and 20% accessed it only 3–5 times) might indicate a gap between the enthusiasm for PHRs expressed by some and the actual interest level of patients.

Should these findings influence the implementation of Stages 2 and 3 of Meaningful Use? There are four policy priorities for Meaningful Use:

  1. Improve quality, safety, efficiency, and reducing health disparities
  2. Engage patients and families in their care
  3. Improve care coordination
  4. Improve population and public health

Use of portals, PHRs, and other strategies related to patient/family interaction with health information and eligible professionals are in priority #2. The proposed stage 2 Meaningful Use metrics for patient engagement — presumably through a patient portal or connected PHR — include the following core objectives and measures:

  1. Provide patients the ability to view online, download, and transmit their health information within four business days of the information being available to the eligible professional. The measures are: a) more than 50% of all unique patients seen by the EP during the reporting period are provided timely online access; b) more than 10% of all unique patients seen by the EP during the reporting period view, download, or transmit to a third party their health information.
  2. Use secure electronic messaging to communicate with patients on relevant health information. The measure is: a secure message was sent using the electronic messaging function of certified EHR technology by more than 10% of unique patients seen during the EHR reporting period.

The current working document for Meaningful Use stage 3 from the Health Information Technology Policy Committee Meaningful Use Workgroup (of which I am a member) includes the following patient engagement related objectives and measures:

  1. View/Download/Transmit/Report, including ability to upload or report data to primary place of choice by patient (PHR, Portal, care team member, EHR, etc.); EP to choose one or more objectives according to what is most appropriate to their practice/facility. Some options listed: provide patients with the ability to self-report information such as: a) family history; caregiver status and role; functional status; patient-created health goals; medical device information (e.g., glucose levels; blood pressure).
  2. Provide 50% of patients the ability to designate to whom and when a summary of care document is sent to specific care team members (across settings/providers) and the ability of providers to review/accept updates.
  3. Create the capacity to accept pre-visit prep tools into the EHR (without a use requirement).
  4. Offer 10% of patients the ability to reconcile information — correct errors, add addenda (without requiring usage — just the ability to offer this option; no threshold for percentage of patients that take action or use it).
  5. More than 15% of patients use secure electronic messaging to communicate with EPs.

Most physicians and practices would likely agree with the need to better engage patients and families in their care — and that many of these objectives are worthy of attention and implementation. The question is one of balancing these priorities with the finite capacity of practices to implement and satisfy a growing number of objectives, escalating expectations of performance on earlier measures, and a number of other “distractions” (e.g. e-Prescribing penalties, physician quality reporting system, ICD-10 implementation, and payer-specific performance-based compensation programs).

Is the effort to drive patient engagement through Meaningful Use criteria ahead of the evidence? Should more testing and experimentation take place before adding new criteria and/or raising thresholds to satisfy Meaningful Use objectives for patient/family engagement?

This post is the personal opinion of the author and does not necessarily reflect the official policy or position of the American College of Physicians (ACP). ACP does not endorse a specific EHR brand or product and ACP makes no representations, warranties, or assurances as to the accuracy or completeness of the information provided herein.


9 responses to "Do Personal Health Records Improve Care?"
  • August 1, 2012
    Sherry Reynolds @cascadia

    Thoughtful analysis but you are asking the wrong question.. A patient portal is a tool not the goal and it is fundamental to redesigning healthcare around patients needs vs providers profits.

    It is always interesting to me when providers want to see clinical data to include patients in their own care? Did they ask for this same level or results when we first added phones to a practice?

    How many tests are performed on patients each year that we know have no impact on outcomes?

    There is a fundamental shift in the expectations we have for patients being responsible for their own health and yet we seem to want to block them from having the data in real time (not a 4 day delay).

    Can you imagine a health care team where the doctor gets the results immediately, her nurse gets it after a 2 hour delay, the medial assistant gets it a day later and finally the rest of the care team (which includes the patient) gets it four days later???

    Where is the evidence to show that NOT including the patient is better? This is simply a question of money and how much time it might take the providers to actually learn how to use a new tool. At places like Group Health 30% to 50% of all primary care visits now happen via email or the phone and both patient and provider satisfaction have gone up and outcomes as well..

    It wasn’t because of the EHR – it was just the tool that allowed them to redesign their practice around the patients workflows..

    • August 1, 2012
      Michael S. Barr, MD, MBA, FACP

      Thanks, Sherry for your great comment. You are correct – PHRs, portals, etc are all tools. Without better workflows and redesign of practice, no tool will likely achieve its potential. I am very supportive of sharing data with patients (and families when appropriate) and support many of the goals of the objectives. The concern I have is the push for more objectives and higher thresholds for meeting these objectives when we haven’t really figured out what works and what doesn’t work – and the potential unintended consequences of devoting a growing share of the limited capacity of many practices to focus on implementing these tools versus trying to get fundamental workflow and EHR implementation right. While Group Health and other health systems can deploy resources, design schedules, create incentives and payment models to promote patient portal use and email in support of these MU objectives, the majority of practices in the U.S. are not as fortunate. We need to get there – but at what pace and what potential cost with respect to other key elements of appropriate EHR implementation for the benefit of patient care and quality?

  • August 2, 2012
    Sherry Reynolds @cascadia

    Thanks Micheal – I appreciate the thoughtful work that you all are doing on the HIT committee and agree that the objectives are running ahead of the science.

    Although I am focused on patient centered design in healthcare as a former EHR implementation project manager and consultant I realize but often forget to also be aware of the lack of “provider centric” design in most of the major EHR’s.

    What might an EHR look like if the key stakeholder in the EHR was the provider (most were designed with the payer – hospital systems) in mind? I trained OB’s who delivered at 3 different hospitals who had to learn 3 different user interfaces on 3 different EHR’s in the same 2 month time period for example.

    I also apologize – without tone of voice I sounded far more cynical than I am about how the financing drives the process. This was meant to acknowledge the power that the payers – in this case CMS yield in driving the process.

    Half of all providers are now salaried but very few are in systems that are able to capture the savings from changes in workflows that integrated systems are able by leveraging EHR’s.

    We increasingly have evidence that what matters to patients and what also drives quality and safe is the patient experience and that isn’t even hinted at in the MU regs yet. Perhaps a broader measure like that might be more useful vs 15% of this or that?

    • August 6, 2012
      Michael S. Barr, MD, MBA, FACP


      The last line of your comment is spot on…”We increasingly have evidence that what matters to patients and what also drives quality and safe[ty] is the patient experience and that isn’t even hinted at in the MU regs yet.” I would add that to get the greatest buy-in and support for meaningful implementation of EHRs and other health information technology, the technology needs to address the concerns patients/families and clinicians. When those interests are aligned – and they should be as much as possible – good things will happen. My main concern is that the expectations being proposed for MU will out pace the ability of vendors to build/test and the capacity of many health care professionals and health systems to meaningfully implement.

      While the certification process is (and should be) pushing the development of new features/functions into Certified EHR Technology, the current construct of MU has eligible professionals (EPs) and hospitals (EHs) implementing these new features/functions on the same time line as the certification requirements. This is comparable to writing new flight control software for airplanes and taking off with passengers before the software has been tested and the pilots familiarized with the new programming. The delay in MU Stage 2 requirements was, in part, a recognition of this challenge and was designed to give EPs/EHs more time to achieve Stage 1 and lay the foundation for moving forward with Stage 2 of Meaningful Use. However, the Stage 2 delay was met with calls by some on the HITPC MU Workgroup for using the one year delay to raise the bar for Stage 2 even higher than was originally being considered.

      What gets added to the long list of MU requirements should build upon firmly established and vetted technology standards and implementation guidelines, have been tested in practice, extend previously established capabilities, have a reasonable expectation of being achieved by the majority of EPs/EHs, and be responsive to both the challenges faced by EPs/EHs and the needs of patients/families.

  • August 3, 2012
    Pieter Van Gorp, PhD

    Dear Michael, thanks for opening a very interesting debate on prioritizing development efforts at large and small care institutions. Thanks also to Sherry for raising excellent points such as “Where is the evidence to show that NOT including the patient is better?”.

    I am an Information Systems scholar from Eindhoven University of Technology and I have visited the Clinical Informatics R&D dept. of Partners HC for the last 5 weeks. I have had highly productive meetings with PHR pioneers and visionaries such as Pete Solovitz and Ken Mandl but I have also had many discussions where conservative questions were raised that challenged my enthusiasm as an innovator. Finally, I have thoroughly analyzed the literature on PHR barriers and facilitators.

    I will comment on the paper by Wagner et al. later. For now, I would like to suggest that we should NOT questions the MU L2 requirement of providing patients the right to receive a digital copy of their EMR data. It is very cheap to implement a patient-specific data export. There is a large enough community of empowered patients and engaged scholars and entrepreneurs to expect that this data would be used meaningfully. Personally, I am building MyPHRMachines [] as a trustworthy platform for crowdsourcing the development of apps that add value to heterogenous and longitudinal patient data.

    We should however question expectations related to patients interacting with provider portals or sending messages to their providers. That would be too costly compared to likely benefits. Please focus on getting data outside of institutional boundaries and feel free to give providers more time to build fancy software. Once the data is out, the community may well take care of building that for them…

    Pieter Van Gorp

    • August 6, 2012
      Michael S. Barr, MD, MBA, FACP


      Thanks for your comment. The concept of providing copies of records to patients is not being questioned. It’s already a requirement (pre-Meaningful Use) under HIPAA and Stage 1 of Meaningful Use included a Core Objective to provide an electronic copy of the record upon request within 3 days. The Veterans Administration introduced the Blue Button for record download and the White House initiated “Blue Button for America.”

      I look forward to your additional comments on the Wagner paper.

  • Pingback: Saturday morning coffee [August 4 2012] « Jerry Fahrni

  • August 6, 2012
    Michael S. Barr, MD, MBA, FACP

    Just came across a blog post by Dr. Rob Lamberts – a long time EHR user (16 years) and his observations of how MU has affected his practice. The title is: “My dream of universal acceptance of EHR has turned sour.” Worth a read.

  • August 9, 2012
    Vince Dubravec, MD

    The “big picture” seems to me that the initially referenced study supports that, unlike those of us in the medical cyber/IT community, the vast majority patients we docs “in the trenches” see tend not to be engaged in even a small part of their medical management, despite a lot of encouragment by a lot of us, and will never meet the utopian goals set forth by the promoters of the IT solution to our healtcare whoes. I suspect 20% “buy in” to this ideal would not stand up to a larger study with a longer timeline. Although there may be a subset of patient (like health IT professionals) that emedicine will be particularly helpful/utilized, I think a personal connection (through face to face follow-ups with patients) will continue to provide the least complicated solution and best outcomes.

Leave a Reply

Your email address will not be published. Required fields are marked *