Tuesday, April 14, 2009

The Health Information Legacy Problem

“No one’s talking about the legacy problem with Information Technology systems,” he said.

“What do you mean?” I asked.

“While every one’s talking about developing a fully-integrated nationwide electronic medical record, no one give a damn about the old systems that exist out there. They’re not worth supporting. No one cares about the data they contain, even though for the doctor, they contain critical documentation about patient’s prior health care and are vital documentation to prevent litigation.

He continued, “I was in an eight man group in 1995. We were very cutting-edge. We bought the latest and greatest software that printed out fully legible prescriptions in the patient’s exam room using personalized prescription pads. It organized our clinic visits – helped with scheduling and documentation of the examination – the works. Our first problem was when Windows NT shifted to Windows XP, we found the had to pay for the upgrade because Microsoft stopped supporting the old system. We quickly discovered that the printers we used to print our prescriptions didn’t work. Why? Because printer drivers for our older specialized printers weren’t available for the new operating system. The printers became useless. Who could I pay to develop these drivers?

But it didn’t end there. For other reasons, our group split up. Our eight-man group became four, and eventually over the next several years, two. We simply could no longer afford the maintenance on the fancier medical record system, so we bought a simpler one that was much less expensive to maintain but found that the data from the older system could not be imported into the new system.

What to do?

We couldn’t get rid of the old system for it contained all of the documentation on patients from the recent past. We had to pay to have it maintained while working to re-create patient data on a new system. Eventually, as time went on, fewer and fewer people knew how to support the old system, the hardware became obsolete, and finally I just stopped paying the maintenance fee. I went “naked” without archived medical records. What choice did I have?

So I ask you, when the government decides on a single electronic medical record system that suits their needs and is all-encompassing, what’s going to happen to all the data on the other systems? I bet every one's going to experience the same problems I did.”


Image reference: Andrew* on Flickr via a Creative Commons license.


Keith Sarpolis said...

You reference all the reasons that we docs in smallr groups are scared to death of adopting a specific system. Many of the big players are judged to run on platforms that may not be abl to talk to other systems in the future. Many have not been time tested to assure that they work the ways they are supposed to and are works in progess. You don't know this till you start using the software. Many are really esigned more for the business side of medicine to deal with all the coding and billing instead of fulfilling the stated objective of better patient safety. Your prior post about e-scribing hit home with me in that it clearly has its failing, yet we are now being mandated to use these systems. Where is the reseach showing that this improves anything? Why are we blindly being told that this will improve patient safety? One has to be suspicious that the health IT lobby is selling our politicians a pipe dream!

Unknown said...

The upgrade path for any software project can be problematic. But I simply would not call Health IT a pipe dream. There are many functioning EHR systems out there. Patients do not appear to have any preference for or against physicians who use EHR. There are studies that show that EHR improves some things like comprehensiveness.

Use of an Electronic Medical Record Improves the Quality of Urban Pediatric Primary Care

DrWes said...


But I simply would not call Health IT a pipe dream. There are many functioning EHR systems out there. Patients do not appear to have any preference for or against physicians who use EHR. There are studies that show that EHR improves some things like comprehensiveness.I am not calling Health Care IT a pipe dream. On the contrary, it is the platform upon which we are building our health care future. And while there are many functioning systems out there, how many communicate with each other. Answer: none. Worse, even hospital systems who own the same system do not permit information to flow between competing centers. Why is this? Answer: competition.

I would further disagree with your assessment that patients don't care who is on an electronic system or not. My experience has been that fully integrated EMR's are sought after by patients who have experienced them.

Regarding "Comprehensiveness:" what does that mean? Good at filling out Medicare requirements for billing? Then I'd say they're perfect. If it means better interphysician communication? I'd say "yes" to that too. Does it mean better health care? That has not been proven yet.

Unknown said...

Dr. Wes,

Yes, I was not addressing your comment I was addressing Keith Sarpolis who said, "The health IT lobby is selling our politicians a pipe dream!"

But you're right, I have yet to see a study that says that patient outcomes are improved through the use of EMR or even that clinical decision support systems makes an improvement to patient outcomes. There are other measures that CDSS has improved - like physicians follow better processes - like evidence based guidelines - more often. But better processes do not necessarily lead to better health outcomes.

That said, I would prefer my physicians follow best practices regardless of the effect on my health outcome.

But, if you read the Pediatrics article you will see that the "comprehensiveness" they refer to is the physician-patient communication process and not inter-physician communication. Not that inter-physician communication is not affected but that is not the focus of the article.

I have read many studies about patient satisfaction before and after an EMR system is in place in a practice and every study I have read has shown no effect in patient satisfaction. These satisfaction surveys have usually been just before implementation, just after and 6-12 after.

I've asked patients who have and have not experienced EMR about whether they would choose a doctor based on EMR use and everyone simply laughed and said, "no." I am currently doing a quick study about EMR and I decided not to even ask the question because the pretest just showed that no one chooses their based on EMR use. That said, there certainly could be bias in my sampling. But I've looked for a study to indicate whether people choose a physician based on EMR and found nothing.

If you know of a study that says that patients seek EMR I would love to have the reference.

JAMA Volume 12 Number 4 Jul/Aug 2005
Health Information Technology and Physician-Patient
Interactions: Impact of Computers on Communication
during Outpatient Primary Care Visits


Computer-Based Documentation:
Effect on Parent and Physician Satisfaction During a Pediatric Health Maintenance Encounter
Arch Pediatr Adolesc Med. 2005;159:250-254.

Keith Sarpolis said...

Dear Csibona,

First off, the link you gave does not seem to work. Having searched the literature on a few ocasions in the area, there are indeed a spackling of studies suggesting improved care or "quality", however there have been recent studies intoducing EMR at a pediatric hospital showing that the mortality rate actually went up after the EMR was brought on line. More recently, JACHO has distributed a sentinel event warning in regards to EMRs cautioning hospital to be careful about the process in how they are developed and set up. They have had some major issues with errors at the much touted VA Vista system held out often as IT at its best.

I am sure we could go on and on quoting studies showing both sides. The bottom line is we are about to spend enormous amounts of money on technology that has not been fully tested and tried in the field and may not fully integrate with other systems. CCHIT, which presumably qualifies systems as interoperable is nothing more than a front for the IT industry and I would be very careful about accepting their recommendations.

Maybe pipe dream is too strong of language, since there are many promising benefits of EMR, but please do some real world testing before these systems are disemminated more widely at significant cost. Prove to me that quality indeed is improved with EMR adoption instead of giving me platitudes.

The hospital I spend my time at has a fully functioning EMR and there are many wonderful aspects of it, but disappointingly, they have done nothing to show increased quality since the EMR adoption and the only signicant improvement that has been touted is to the bottom line since every test and supply now gets automatically billed by the system. This has improved collections for the hospital, but unless you argue that better billing allows deployment of more resources, there is no way to know whether the EMR has done anything for quality. Since notes get populated with all the labs history and other asosciated minutiae, you have to search around for the meat and I fear that mnay of these notes are being generated (again for mainly billing purposes) with little review by the physician.

Finally, although there are many function EMRs out ther, how many implentations have crashed and burned? I keep seeing many mentions of situations where an EMR was shut down or simply not used.

Anonymous said...

As long as businesses are financially rewarded by creating "exclusive" programs that are outdated by the upgrades they later create, we won't ever have systems that interface and communicate with each other. Planned obsolescence almost killed the American auto industry, and Microsoft is filling landfills with our private data and hardware...

Anonymous said...

PS Keith, those redundant EMR notes full of useless copy DRIVE ME CRAZY!

I love it when I see an "old timer" doc who writes in narrative style. "Ah, a story I can sink my teeth into!"

These are the notes actually worth reading.


Jay said...

ER doc pages me.

Patient in with chest pain. Thinks she had a cath at another hospital, but no records available. ER doc wants to admit.

I hop on my laptop. In a couple of minutes I have report. "Normal Cors" at hospital X.

Patient goes home.

I'm happy. ER doc is happy. Patient is happy. We just saved a ton of money for the health care system.

Does anyone need a study to prove that this is improving quality??? How do you get a P value for this sort of thing.

True story -- several times over.


P.S. I could go on and on. If medicine were a "real" business, competitive forces would have settled this issue long ago. EMR's are a good thing.

Anonymous said...

Well this healthcare provider, who has been using an EHR for more than a decade, literally walked out during a consult with a newly minted sub-specialty doc today. He introduced himself and said he had reviewed my record before coming in to meet me (I have chronic medical conditions, had major surgery last fall at an institution other than where this consultant was located). I made sure my tertiary level surgical info was provided to my PCP - actually saw a link to the outside records as the LVN was entering my VS. As I discussed my current sx with the consultant it was patently clear that he did NOT even know that I had undergone major surgery to treat a condition related to the reason for the referral, despite the data being in the EHR. Three times he said he wasn't sure how he could help me as he stared at his computer screen looking at data rather than LISTEN to me as I gave my hx. I walked out of his office and went to file a complaint with the clinic manager. Called patient relations too.
Too many clinicians have forgotten that the PATIENT should be the focus of the visit, not the blasted computer.

Unknown said...


First, which link does not work? I checked them both twice here and they both work on my computer without special access. Your mileage may vary.

Second - You say, "we could go on and on quoting studies showing both sides." The funny thing is I actually put the links to the studies or the detailed information on how to get to the studies. I posted three. You posted, well, 0.

We are about to spend $19 billion dollars on IT. That is an enormous amount of money. Of course the entire health care system takes about 15% of our GDP and is $2.1 trillion. As a ratio the amount of money the federal government is about to spend on Health IT is minimal. (http://en.wikipedia.org/wiki/Health_care_in_the_United_States)

I would not say that Health IT is untested. NEJM says that 4% of ambulatory care physicians had a fully functioning EMR system at 13% had a basic system. That's 17% of physicians in the national survey. See NEJM Electronic Health Records in Ambulatory Care — A National Survey of Physicians. N Engl J Med 2008;359:50-60. (http://content.nejm.org/cgi/content/full/NEJMsa0802005). While still a minority I would not call those numbers "untested." The NEJM article goes on to say that 16% of physicians reported that their practice purchased but had not yet implemented the systems. An additional 26% of respondents said that they planned to purchase a EMR in the next two years.

I have read numerous studies about EMR system implementations going badly. In particular I enjoy this article:
I'm not saying that all EMR system implementations go well. I am suggesting that many are implemented and working.

So, I know you ask this question: how many implementations have crashed and burned?
I don't know. Do you?

Unknown said...

Cat said:

As long as businesses are financially rewarded by creating "exclusive" programs that are outdated by the upgrades they later create, we won't ever have systems that interface and communicate with each other.

Companies are rewarded by creating new features to their software that companies want to use and are willing to purchase. The software companies are not forcing anyone to buy anything. People buy the software they want from what is available at the time. The software company may make upgrades to enhance features, work with environments, etc. What is the alternative that the Health Information Technology companies never update their software or work with new operating systems/technologies/environments? I suppose some may think if only the systems were written correctly the first time we would never need to upgrade. That simply is not the case. There are new capabilities available in hardware that people want to use that software writers could not interface before the devices are built.

Keith Sarpolis said...

Dear csiboana,

You only gave one link in your first post to pediatrics .org. I can't get it to work!

As to your other comments, usually when someone proposes a new treatment or technique, the onus is on that person to prove what the benefit of their technique; not for others to disprove that it doesn't work. The fact that many practices are using IT systems does not prove they are beneficial. In fact one might offer that they are not being adopted for the reasons purported (increased quality) but simply to better manage data collection for billing and to overpopulate physicians notes so they can upcode their visits. One can easily turn this argument around and say the other 80% of MDs not using EMRs speak to some underlying problems with EMR. Just maybe we have all learned our lesson that it is not wise to be the first adopter of an unproven system, especially with the example of the IT industry in general which is notorious for rolling out clunky software and then charging you every few years for the new version that corrects all the problems with the first version

I offer no references because it is not up to me to prove your claims of the superiority of EMRs. It is the burden of those who are pushing these systems to offer up the proof!

Would you willingly take the medication I would give you on the basis of my anecdotal and untested experience with the medication? Then please don't ask the medical community to swallow your huge EMR pill without proof of either cost savings, improved quality, or better communication without demonstrating such. Especially when EMR companies want to charge 20-40 grand per MD for such technology.

The hospital where I practice has installed s systemwide EMR that has been in place for 4 years. It has no proof of any increase in quality or cost savings that I am aware of as a result of this implementation. It does state it improved its collections because now they could bill insurers for every item down to toothpicks because they all have to be ordered through the system. Here it is 4 years later in this grand experiment and no one has meaasured what the true benefits are from this EMR. In fact , this hospital was brought to trial by the FTC for violation of antitrust laws for jacking up its prices after merging with a close competitor. They claimed in that trial that they had brought great improvements in quality at the merged hospital by virtue of adding their EMR system at this merged hospital, but could offer no proof or statistics to substantiate their claims! They lost this case as a result.

As to the number of system failures, once agin, my information is anecdotal, but it would seem someone wants to look at this before systemwide implementation. And once again, it is the seller of the technologies job to prove it's product works; not that of the clinicians who will have to deal with this mess if it clearly does not reach the level of functionality that are frequently claimed but rarely substantiated.

Unknown said...

Hmm, the pediatrics link works for me - I just cut and paste into a browser. You can try:
It sounds to me like you should have access to an online medical library anyway. I posted the title:
Use of an Electronic Medical Record Improves the Quality of Urban Pediatric Primary Care
It shows up in Google search.

There is evidence of better process outcomes with clinical decision support systems. There is evidence of a reduction of duplicate or redundant tests with EMR. There is not evidence of better health outcomes or cost savings overall (as far as I know). The Congressional Budget Office's report on Health IT states that cost reductions are probable in the future (but predictions of the future are a little difficult to prove or disprove). The process outcomes and reduction of duplicate tests are in peer-reviewed journals (sorry I don't have the references handy at the moment).

I understand your unwillingness to offer references. But remember these words are yours, "I am sure we could go on and on quoting studies showing both sides." But you didn't "quote" a single study. Now your position is NOT that we could both quote studies that contradict each other but that the EMR advocates must prove their systems worth in terms of health outcomes, cost reductions or better communication. You're changing your position. And I offered a study on better physician-patient communication.

Your counterargument is good - that the 80% of docs not using EMR may indicate an underlying problem. If that was truly the case then we would not see a rapid increase in the adoption practices by physicians. Again, the NEJM article indicates that in 2008 17% of physicians were using EMR and that an additional 16% had purchased and but not yet implemented and an additional 26% planned to purchase in the next two years. This is before the stimulus in ARRA. So while I do believe there is an early adopter difference among physicians I believe EMR is in the early part of the product lifecycle - that is the adoption rate is increasing.

As for difference among adoption rates see:

Technology Acceptance among Physicians: A New Take on TAM
Amy K. Yarbrough and Todd B. Smith
Med Care Res Rev 2007; 64; 650 originally published online Aug 23, 2007
DOI: 10.1177/1077558707305942

A survey of factors affecting clinician acceptance of clinical decision support
BMC Medical Informatics and Decision Making 2006, 6:6 doi:10.1186/1472-6947-6-6

You make several claims but back down after a little argument. First you said "pipe dream" then you made this statement:
"Maybe pipe dream is too strong of language, since there are many promising benefits of EMR"

Then you say:
"I am sure we could go on and on quoting studies showing both sides."
But follow up with:
"I offer no references."

and, well finally you say:
"Finally, although there are many function EMRs out ther, how many implentations have crashed and burned? I keep seeing many mentions of situations where an EMR was shut down or simply not used."
and follow up with:
"once agin, my information is anecdotal."

I mean, well, just read your statements yourself and come to your own conclusions.

Keith Sarpolis said...


Why don't you work your way through this bunch of references(at the bottom of this post) and let me know what you think.

I don't know why you find my statements so confusing, but for the sake of argument let me simplify.

1. EMR is a complex new technology that has only been introduced in the medical field in the past 10-20 years
2. Like most new things in medicine, we like to prove they work before spending large sums of money to only find out later they really don't work. (Actually this is not always true in medicine which is why we have such an expensive system; witness the fact that a whole generation of kids had their tonsils removed only to find out the procedure was worthless)
3. There are woefully few studies looking at the impact of EMRs, despite the significant deployment you reference. Some of this is that apparently companies that design these systems don't like other people researching them and finding out that there are some serious issues, so they routinely put clauses in their licensing banning such activity.
4. It is the responsibility of the seller of such innovative technology to prove its worth; not for the buyer to prove its lack of worth.

Does this clarify my seemingly contradictory statements?