There is no generally accepted metaphor to describe the process of standardizing electronic health records, journalist Stephen Strauss says. (iStock)Who would have thought that researching the contorted birthing of electronic health records would make an ordinarily smiley journalist feel like an extraordinarily frustrated poet?

The poetical question I keep asking people in the EHR field is: What's the metaphor? What can the process of medical information going digital be likened to so that it has some kind of emotional resonance in the life of Joe or Jane Everybody?

The query comes because without a metaphor it seems almost impossible to capture the complexity of the situation. The current system has individual doctors, nurses, pharmacists, clinics, hospitals and so on deciding what should be written about and saved when it comes to patient records, and the goal is to move to a standardized, computerized national/international system of medical communication.

Dennis Giokas is chief technology officer for Canada Health Infoway, the federally funded, provincially- and territorially-guided agency trying to engineer the change to EHR. Here's a metaphor he recently described to me.

"The most common metaphor that we have used for years is that it is like the banks. You can go to any bank in the world and [make a] deposit, withdraw money or check a bank balance. What makes that possible is a common language of debit and credit and money movement."

True enough, banks did digitize their print records and automate their transfer of information. That's similar to what hospitals hope to do.

Except that when you take a closer look at banking, any comparison seems more akin to likening a three-year-old's stick figure self-portrait to all the paintings on the ceiling of the Sistine Chapel. Most of the transactions you do in banking are mundane — deposit, withdraw, pay bills, see what your financial history looked like last month. They were things banks had been doing for hundreds of years and everyone knew how they worked. It's also information which is going to have little general use after a couple of years.

Thomas Beale, chief information officer with Ocean Informatics — a London-based company trying to foster open-architecture EHR — also points out in a 2005 paper entitled The Health Record — why is it so hard? that the customers in the bank metaphor are "grossly simplified abstract versions of a person." The efforts to standardize EHR communication show how grossly complicated the "patient" version of a person is.

Consider SNOMED CT, short for Systematized Nomenclature of Medicine - Clinical Terms, which was created in 2002 by pathologists in the United States and the U.K.'s national health service. The idea was to come up with some standard way that all the world's medical concepts can be understood by everyone who uses them. Think of it like this: a heart attack to a paramedic means the same thing as myocardial infarction to a heart surgeon.

Today SNOMED has about 315,000 unique concepts in it. These are described in 805,000 ways in English, and there are more than 1.3 million descriptors of the relationship between things. By way of linguistic comparison, The second edition of the Oxford English Dictionary contains full entries for 171,476 words in current use, and 47,156 obsolete words.

But SNOMED's verbosity doesn't even cover all of medicine. That's why there is LOINC — you have to love the word-nerdiness of health IT geekdom — which officially stands for Logical Observation Identifiers Names and Codes.

LOINC aims to standardize the ways of describing lab results and observations. It now includes 58,000 terms, not to mention six different categories into which they might be differentiated.

And let's not forget HL7, officially Health Level Seven, an international effort to create some uniformity in how health care information is exchanged, shared, retrieved and integrated. This would allow companies, theoretically, to create standardized templates that would permit all the hardware and software in hospital systems around the world to speak one language.

How complex is this? Well, here's just a taste — HL7 requires a uniform way of informing everyone that a therapy might: (1) interact with foods; (2) interact with other drugs; (3) interact with a natural product; (4) interact with a non-prescription substance; (5) be affected by other conditions; (6) be affected by age; (7) be affected by height or body surface; (8) be affected by weight; (9) be affected by breast feeding; (10) be affected by gender and at least 30 more potential therapy caveats. (For a more detailed description of the problems EHR standards cause see Beale's views.)

And on top of all of this, any EHR system has to convey the different routes of medical reasoning. As Giokas points out, "we might be diagnosed with the same disease, but the path to get to that diagnosis might be completely different from one doctor to another."

So what else besides banking can this effort to standardize a very complex system like EHR be likened to? Beale argues that because the various standardization formats may be in conflict with one another, the situation today is akin to wanting to buy a car but instead being offered a bunch of car parts. And this batch of parts may or may not come together to produce an operating vehicle.

Better metaphor?

But all of this misses the point that we are also moving from an existing way of doing something — paper records — to a new way of doing something — digital records. So in the absence of a generally accepted metaphor, let me suggest one.

Imagine it is 1902. Imagine you are watching the first mass-produced, low cost cars roll off the Oldsmobile assembly line in Lansing, Mich. And then imagine some transit czar says: "To deal with the revolutionary possibilities that cars and trucks and buses are bringing, we have to put in place a 21st century road system."

This means we need to create: Overpasses, underpasses, superhighways, toll booths, bridges, lane divisions, speed signs, parking meters, trolley tracks, billboards, off ramps, warning lights, no parking between 4 and 6 p.m. signs, handicap parking, bus stops, taxi stands, stop lights with turn signals, and crossing warnings for pedestrians.

We also have to engineer room in the system for traffic cops, speed traps, tractor trailers, motorcycles, highway narrowing signs, break down lanes, sidewalks, car dealerships, bike paths, gas stations, garages, rest stops, lane markers, tow trucks, scenic vista widenings, snow plows, road salting machines, lane sizes, how-far-you-are-from-somewhere signs, and indications of what is going to be found in this turnoff.

But more than anything, you have to engineer a system to deal with speed. A horse drawn buggy moves at about 20 kilometres per hour and the horse has to rest or be changed after a couple of hours. Meanwhile, 180 kilometres an hour is nothing for a modern car. So you have to figure out how fast people are allowed to go in cities and highways and alleys, and then figure what to do with traffic jams which spring up when speed inevitably leads to worse accidents.

And oh, yes, you have to teach people whose driving experience is limited to horse-driven buggies, mule trains and oxen wagons to immediately understand how everything works in the new road system and how to use it effectively.

And perhaps most importantly, you must do everything without stopping all existing traffic while you build the new infrastructure.

Doing all this to build the current road infrastructure took about 100 years. But in the case of EHR, which has similar challenges, you must do it all on the federal government's timetable: Within 20 years.

The metaphor has traction with people in the EHR field, while the banking metaphor doesn't.

Beale tells me, "I think it is an excellent analogy because it makes it easy for people to see how many things have to work together. You can easily imagine any one thing not working and causing chaos."

Sam Marafioti, chief technology officer at Sunnybrook Hospital in Toronto, agrees. "And that metaphor is especially good because we often used to call what we are building 'the information highway.'"

But can we do everything necessary for EHR in the 20 years or so the provinces and the federal government keep telling us it will be done in?

"A century is probably extreme, but it is decades," he responds. Then Marafioti goes over everything that is being attempted and is needed for EHR, and he reconsiders. "Maybe that is 100 years worth of work."

And therein is both an image and an explanation.

Want to know why you continually read about hiccups, and stoppages, and scandals and recrimination when it comes to electronic health records? Governments everywhere are telling doctors, hospitals, nurses, and pharmacists that they have to compress maybe a century's worth of medical information changeover into 20 years.

And the transit system metaphor tells us that this time scale is not just hard. It may well be impossible.