Patients Can Squawk, But We Have Little To Crow About Open Data

One of the biggest disappointments at this year’s Health Datapalooza (which I found disappointing overall) was the continued impasse presented to patients who, bolstered by the best thinking in health care as well as Federal laws and regulations, ask for health data stored about them by doctors and other institutions.

Activists such as Regina Holliday and e-Patient Dave proved years ago that giving patients information and involving them in decisions will save lives. The Society for Participatory Medicine enshrines the principle. But the best witnesses for patient empowerment are the thousands of anonymous patients, spouses, parents, and children quietly trundling folders with their own records through the halls of hospitals, building up more knowledge of their chronic conditions than any professional clinician possesses, and calmly but inflexibly insisting on being equal partners with those who treat them.

There were plenty of high-minded words at the Datapalooza about patient rights to data. It was recognized as a key element of patient empowerment (or “activation,” as the more timid speakers liked to say) as well as an aid to better care. An online petition backed by an impressive array of health reformers is collecting signatures (whom someone will presumably look at) and encourages activists to speak up about this topic on July 4. HHS announced that anyone denied access to data to which the law gives her a right can submit an informal report to

Although occasional mention was made of personal health records (PHRs), most of the constant discussion about interoperability stayed on the safe topic of provider-to-provider data exchange. Keeping data with health care providers leads to all sorts of contorted practices. For instance, patient matching and obtaining consent are some of the most difficult challenges facing health IT in the U.S., all caused by keeping data with providers instead of the patients themselves.

The industry’s slowness to appreciate patient-generated data is also frustrating. Certainly, the health IT field needs to do a lot more to prepare data for use: consumer device manufacturers must assure clinicians of the devices’ accuracy, and researchers need to provide useful analytics that clinicians can plug in to their electronic systems. Still, doctors are demonstrating a disappointing lack of creativity in the face of this revolutionary source of information. It’s all to easy to carp about accuracy (after all, lab tests have limited accuracy as well) or just to state that you don’t know what to do with the data.

I heard about recent progress at the UK’s National Health Service from Brian Ahier, who is the only person I know who can explain the nuances of extensions to FHIR resources while actively using both his laptop and his cell phone at the same time. Ahier heard at a UK-US Bootcamp before the Datapalooza that the NHS has given 97% of its patients access to their records.

But there’s a bit of a caution around that statistic: only one-fifth of the patients have taken advantage of this right. This doesn’t bother me. First of all, one-fifth of the population with access to their personal records would be a dizzying accomplishment for most countries, including the U.S. Second, few people need access to records until some major problem arises, such as the need to see a specialist. They probably feel relieved to know the records will be there when needed.

Another aspect of patient control over data is research. The standard researcher-centered model is seen as increasingly paternalistic, driving patients away. They’re not impressed with being told that some study will benefit people like them–they want to tell researchers what really matters to them as sufferers, and hear more about the study as it goes along. Researchers are frantic to reverse a situation where most studies fail simply because they can’t sign up enough subjects.

The Patient-Centered Outcomes Research Institute (PCORI) is one of the progressive institutions in health care who understand that giving patients more of a say will be increasingly important for signing up patients in the first place, as well doing research of value to them. Its PCORnet combines traditional research databases with databases maintained by patient advocacy groups. Each member network can create its own policies for getting consent, which allows researchers to bend with the needs of their research subjects.

OpenClinica, the open source clinical research platform, just announced the release of an app that may contribute to the goals of taking input from patients and binding them closer to the research endeavor.

Public health officials also recognize the sensibilities of the people they monitor. At a panel on data about low-income people, speakers stressed the importance of collecting data in a respectful way that doesn’t make people feel they’re being spied on or could be punished for their behavior.

Let’s talk a minute about health care costs, if only because doctors and insurers don’t want to. (Some doctors are prohibited by their employers from telling patients how much a recommended procedure will cost, supposedly because they don’t want costs to intrude on what should ideally be a clinical decision. This is changing with the increase in deductibles, but often the doctors don’t even know what the final cost will be after insurance.)

One app so admired by the Datapalooza team that they allowed the company to demonstrate its product on the main stage during keynote time was Sensentia. This product everybody is so impressed with takes in information from health plans to allow patients as well as the staff at health care providers to quickly find the health plan benefits for a procedure. (I recently covered another company doing similar work with insurance and costs.)

Sensentia is a neat product, I am willing to aver. It accepts natural language queries, crunches the data about health plans and insurers, and returns the actual health plan benefits for a treatment. Of course, I know the cost of flying from Boston to San Francisco after six clicks in my browser, even though the calculations that go into offering me a price are at least as complicated as those run by health plans. One may be shocked to hear that that current phone calls to an insurer cost $3-$10. This is the state of health care–it costs more than five bucks on average for a doctor just to find out how much it will cost to offer his own service.

A panel on patient-generated data reported more barriers than successes in getting doctors to work with data from patient devices and reports from everyday life. Another panel about improving quality measures culminated in the moderator admitting that more patients use Yelp than anything else to choose providers–and that it works pretty well for them.

For me that was the conference’s low point, and a moment of despairing cynicism that doesn’t reflect the mood of the conference or the health care field as a whole. Truly, if Yelp could solve our quality problems, we wouldn’t need a Datapalooza or the richness of data analysis it highlights. But I think reformers need more strategies to leap the hurdles we’re facing and implement the vision we all share.

About the author

Andy Oram

Andy Oram

Andy Oram writes and edits documents about many aspects of computing, ranging in size from blog postings to full-length books. Topics cover a wide range of computer technologies: data science and machine learning, programming languages, Web performance, Internet of Things, databases, free and open source software, and more. My editorial output at O'Reilly Media included the first books ever published commercially in the United States on Linux, the 2001 title Peer-to-Peer (frequently cited in connection with those technologies), and the 2007 title Beautiful Code. He is a regular correspondent on health IT and health policy for He also contributes to other publications about policy issues related to the Internet and about trends affecting technical innovation and its effects on society. Print publications where his work has appeared include The Economist, Communications of the ACM, Copyright World, the Journal of Information Technology & Politics, Vanguardia Dossier, and Internet Law and Business.


  • Andy,

    Thank you for a lucid, thoughtful analysis. When it comes to any complex data there are many issues of ownership and access. We live in a health data world that’s not only byzantine and constantly changing, but also one that is rule bound by many different actors.

    As you point out, we need to look at several related aspects of secondary users and uses. These raise the constant question of who gets to define health data, originate it, modify it, share it, use it and look at it. We tend to lump these together, so I was happy to see how you’ve parsed a number of them into more chewable form.

  • This article really highlights the chasm of using healthcare data in meaningful way in affecting processes. You can clamor about the quality of data but its not just about the quality as much as willingness to be data driven by the providers that is the hurdle. They can keep blaming HIT for things such as interoperability, which is a challenge, however what it takes away from the need for provider attitudes to change.

Click here to post a comment