March 2015

Volume 30 Number 3

Don't Get Me Started - Hitting a New Wall

By David Platt | March 2015

David PlattI really hate talking to a doctor’s back. But that’s what I did as this doctor took my history. “Any allergies?” he asked, back firmly turned toward me. “Have you had this before?” (back turned)—on  and on, through his entire questionnaire. It would have been ludicrous if it weren’t so sad.

The reason the doctor did this, as you’ve probably guessed, is that he had to enter all of my responses into his computer, to become part of my electronic medical record. U.S. federal incentives call for hospitals to be doing something with electronic medical records (“meaningful use”), though they don’t specify exactly what. This priority inversion, valuing data input over patient contact, is one of the results.

Even the best usage of PCs by primary care physicians is highly intrusive. My daughters’ pediatrician uses a laptop and faces her patients. But the device still creates a barrier, consumes the physician’s time, diverts her clinical attention, forces her to repeatedly break eye and hand contact; not to mention transferring germs between patients via impossible-to-sterilize keyboards.

Computerizing operations was supposed to make users more efficient. In primary care medicine, we’ve accomplished exactly the opposite. As Katie Haffner wrote in The New York Times: “For decades, physicians pinned their hopes on computers to help them manage the overwhelming demands of office visits. Instead, electronic health records have become a disease in need of a cure, as physicians do their best to diagnose and treat patients while continuously feeding the data-hungry computer” (

These demands rapidly consume the physician-user’s hassle budget. As I wrote in April 2013 ( and July 2011 (, the hassle budget is the amount of extraneous effort that users will tolerate in order to get their computing jobs done. If an operation exceeds a user’s hassle budget, he’ll either toss that program or figure out a workaround. Writing down ever-changing passwords on sticky notes is the classic example.

Some physicians have created a workaround by dumping the data entry task onto another person, the medical scribe. The doctor faces and interacts directly with the patient (what a concept!). The scribe sits in the same room but apart, listens to the conversation, and enters the resultant data into the computer (see

The scribe dresses in black to signal her invisibility, like the kuroko stagehands in traditional Japanese theater ( She does not speak, except to answer a doctor’s question or request a clarification. Her sole function is to enter data into a computer so the doctor doesn’t have to. I’d call her a liveware analog-to-digital converter.

Kathleen Myers is an emergency physician and founder of Scribes STAT, a company that provides medical scribe services to hospitals. Meyers says of working with a scribe: “I get to sit down and look at [my patient], and really focus on what they’re saying. I feel like I miss less information—the patients have a greater bond with me and are able to share more information.” (See a video about the service at

Scribes only earn $10 to $15 per hour. Considering the physician’s time they save, scribes would seem to quickly pay for themselves. Although, as I recall, that’s what computer systems were supposed to do, and we’re seeing exactly the opposite.

We could certainly build better medical apps. Involving the users from the project start would go a long way; instead of writing what we geeks think they need and cramming it down their protesting throats, as usually happens.

But I wonder: Have we reached the limits of what computer programs can accomplish? Quantum mechanics reached its limit with Heisenberg’s uncertainty principle, and mathematical logic with Gödel’s incompleteness theorem. Have we hit such a wall with medical software?

As my primary care physician (not the back-turner in the opening paragraph) Peter Zuromskis, M.D., likes to say, “Medicine is an analog process. Those bean counters are trying to make it digital, but at its core, it simply isn’t.”

David S. Platt teaches programming .NET at Harvard University Extension School and at companies all over the world. He’s the author of 11 programming books, including “Why Software Sucks” (Addison-Wesley Professional, 2006) and “Introducing Microsoft .NET” (Microsoft Press, 2002). Microsoft named him a Software Legend in 2002. He wonders whether he should tape down two of his daughter’s fingers so she learns how to count in octal. You can contact him at