Just in case some readers might consider this a conflict of interest, I’ll say it up front: I covered the world of PDF for nine years at PDFzone.com, a site affiliated with eWeek. PDFzone’s mission was to evaluate PDF document technologies and the market in which they lived, in a vendor-agnostic forum. As such, Adobe never liked me enough to add me to its Christmas card list.
Over the last week, PDF — let’s not forget, it’s only a document format — took a couple sharp body blows in the court of health care IT opinion.
Last week at the federal HIT Policy Committee’s Adoption and Certification Workgroup meeting, Dr. Christine Sinsky, a Dubuque, Iowa internist, called it “death by PDF” when there’s “50 pages of outside records scanned in PDF format [that] are technically ‘in’ the record, but functionally the information is not readily accessible to the clinicians,” viewable a few lines at a time through a window in the electronic health record (EHR).
Something similar came up during a convocation of the National eHealth Collaborative’s Consumer Consortium meeting in Washington D.C. yesterday:
“The word PDF is such a dirty word in patient data access that I do not tend to use it in polite company,” patient advocate Regina Holliday live Tweeted. That tripped off a three-way Twitter discussion between me, Regina and Dave deBronkart (ePatientDave) on the usability of PDFs in personal health records (PHRs).
I wanted to know, was Holliday talking about bad PDF implementation, or PDF itself? “I was remarking how hard a PDF is for usage. Not easy to input into a PHR,” she said. Later, she added, “We want data with usable fields. Easy to amend, add notes and patient input. A PDF is a dead document.”
deBronkart, who has some experience working with print and computer graphics, said that PDFs do possess three virtues: They’re “printable, readable on any smart device…[and] can carry data….Of course a limit is it’s no good to someone w/no smart device.”
All valid points. When Adobe co-founder John Warnock came up with the idea for PDF in the early 1990s, he simply wanted a good electronic rendering of paper documents, that’s it. It was the savvy technologists around him who foresaw PDF as a pretty good way to send documents — even graphics-intensive documents — around the Web in a lightweight “container.” Over the years, Adobe grew PDF’s capabilities to support forms, comments and annotations, and shrewdly gave away the Reader so it’s on pretty much everyone’s computer. Several government agencies have legislated it as the official document exchange format, and the Internal Revenue Service uses it for all America’s tax forms.
The upside for PDF in health care is how it smashes big graphics files such as an MRI “picture” into a small (the “P” in PDF stands for “portable,” the “DF” being “document format”) file. It’s not that tough to embed Flash into PDF either; imagine the ability to send a cardiac video study across town en route to getting a second opinion. The free Reader opens them all, on Mac, Windows, Linux, smartphones, whatever — eliminating the need for proprietary viewers.
Yet, when it comes to medical records, most implementations of PDF are, as Holliday points out, static documents. Even so, people who create the PDFs have the ability to enable end users to comment and annotate in Reader. There’s also a substandard of PDF — PDF/H, the “H” standing for “health care” — around which the Center for Health IT has created an instructional video; the Association for Information and Image Management also devised some guidance materials and sample documents for using the PDF/H standard.
PDF industry expert Michael Jahn pointed out to me in a Facebook exchange that for PDF to work in health care — legally, in a HIPAA compliant way — those who implement PDF really should take advantage of deep security capabilities embedded in the format to verify the identity of people viewing a document. Furthermore, he said, using forms capabilities built into PDF would go a long way in making those documents more interactive.
The trouble is, these are far down the road from the simple, static implementations that most organizations can afford, or are capable of imagining for EHR and PHR systems. It’s one thing to know about PDF’s interactive capabilities, and quite another to possess — and expend — tech resources on making them more usable (read: Don’t blame PDF for an EHR vendor’s making them viewable a few lines at a time, as Sinsky put it).
Adobe must be held accountable for a piece of this, too: While the company did turn over its intellectual property to ISO to make the PDF file format an international, open standard, the company still does things like this: Yes, one can make a PDF open for comments and annotations in the free Reader, but to do that, you have to do it one document at a time, running Acrobat, a $449 piece of software. Holliday’s problem would be solved if Adobe would just provide some simple markup tools in the free Reader.
Seriously, if Adobe could fork over the whole file format to ISO, is a few more tools in Reader too much to ask, especially if it promoted the free flow of crucial health information between patients and providers and saved lives in the process? Really? (Uh oh, I did it again. One more year I won’t be on the Adobe Christmas-card list.)
[Update 4/28: Adobe recently added some basic PDF markup — highlighting and commenting — tools for those who download the free Reader X.]
The other problems outlined above, however, look to me like implementation issues outside of Adobe’s control. EHR vendors and companies who are digitizing paper documents have the power to make PDFs more interactive. Instead, they choose to use PDF as a static, or dead format as Holliday put it. It’s kind of like John Warnock’s original vision of “e-paper.”
Data exchange is hard; in fact, this blog entry wasn’t simple to write. It was a hassle concatenating content from Twitter, Facebook, HTML Web pages and yes, a couple PDFs. Had I known about the quantity of cutting, pasting and reformatting, I’d probably have written about something else.
But just as I don’t blame the HTML, or Twitter, for the time it took, we shouldn’t blame PDF for weak implementations. To paraphrase Elton John, don’t shoot PDF, it’s only a document format. And don’t go making an ICD-10 code for “death by PDF” quite yet.