Will an EHR Ever Last as Long as a Patient?

I have this notion that my chart notes contain information that will be useful in guiding patients' care in the future. Obviously, this can only happen if the charts, and my notes, are still around when the future arrives. A couple of recent experiences lead me to believe that today's computer technology puts this notion at risk.

In trying to brush up on my freshman linear algebra recently, I found a tutorial on the Internet that uses a mathematical programming language called ISETL that was produced about 10 years ago by a couple of college professors who have now either retired or have moved on to other things. There are versions of the old software for Windows and Mac. I would prefer to run ISETL on my Mac but I can't. It's not compatible with current Macs. Apple has never worried about the past. In fact the faster they can make the past obsolete, the more new product they can sell. If you bought an EHR that ran on the Apple II, those records are toast. If you saved data on your Apple Lisa, that data is toast, unless you printed copies and saved the paper. If you bought an EHR to run on a Mac in the late 80s it became useless when they switched to the PowerPC chip. If you bought one for the PowerPC chip it became useless recently because Apple abandoned support for that class of applications; they won't run on today's Macs unless rewritten.

Microsoft, whatever it may have done wrong, has so far maintained good application compatibility throughout the past 30 years. Most DOS programs, including a DOS-based EHR that I developed in 1982, will run on a new Windows 7 PC. The days of long-term application viability may be coming to an end. The rumor mill has it that Microsoft may be about to emulate Apple by not providing a way to run old apps on future versions of Windows. Your PC-based apps and data may follow your old Apple apps into oblivion. So at least for now, I'll be running ISETL on Windows, not on my Mac.

On the heels of that discovery, I went to see my physician and guess what? A couple of years ago his practice group implemented an EHR. He has been faithfully typing notes and entering data into that application at every visit. At this last visit he was complaining that the practice had just switched to a different EHR. He was sitting there with the paper chart copying data from it into the new system. He took my past medical history over again and re-entered all the information into the new EHR. There is apparently no viable means of transferring information from the old EHR to the new one so they had to start from scratch. So much for that computerized longitudinal record.

I won't bore you with the details of the situation at the hospital where I practice. Suffice it to say that it's the same story.

So, what chance is there of maintaining a computerized longitudinal medical record when the computer part becomes obsolete every couple of years? This lifespan of computer systems averages five years to 10 years. The required retention period for most of the clinical data that originates in hospital and critical care settings is seven years. This is far short of the lifetime that a medical record should remain accessible and unaltered, but as I recently discussed, most of this stuff has no enduring value anyway. Clearly the information included in the longitudinal record should be carefully selected. Information without long-term value is best kept in the systems where it was generated. Purging is facilitated and the damage that will be done when the system becomes obsolete will be minimized.

Whatever you choose to do about EHR, please remember that most EHRs are just gadgets, not real longitudinal medical records. Gadgets are destined to be replaced but the medical chart ought to be irreplaceable. If your goal is to keep records on your patients available for a long time, you had better make specific arrangements for that. Paper is definitely still in the race.

Find out more about Dan Essin and our other Practice Notes bloggers.