Which EMR Product Are You Using? Part Deux

[RDGelzer]

Started at the beginning of the thread and read forward, interesting and enjoyable. I was in practice as a General Internist in a rural area and, like many, became increasingly frustrated with a payment scheme that rewarded for volume. In pursuit of better options, I ended up working for a small company about to roll out another (!) EHR to the market. During during medical school I had done standards of care research for medical/legal defense attorneys and had learned a bit about basic principles of medical records as legal documents. To make a long story short, I found that the EHR product of the company had never been reviewed for whether it would qualify as a legal record. Compressing the next few years into a sentence-I have since found that few EHRs have been designed with basic design requirements to qualify them as meeting, for example, Federal Rules of Evidence requirements. When I ask vendors why their products have never been qualified as meeting such requirements, the uniform answer is, "nobody has asked for it."

My first suggestion to anyone shopping for an EHR (or who has an EHR) is to ask, "Has your system been demonstrated to meet the requirements of a legal record?" If the answer is "No", then I think you will agree this is a very easy test for excluding a given EHR from consideration.

These rules are actually quite straightforward. They are:
1. It has to meet the well-established requirements of a medical record
2. It has to meet the well-established requirement of a computerized business record.

Here is a reference for a nice write-up of these requirements:

"Maintaining a Legally Sound Health Record", a Practice Brief from the professional association of health information managers (medical records professionals) AHIMA

http://library.ahima.org/xpedio/groups/public/documents/ahima/pub_bok1_014041.html

By the way, you can join AHIMA as an associate member for, I believe, $140 per year and they are making a big effort to increase their resources for supporting the transition to electronic health records. Their journal will be publishing, over the three months from September, updated practice briefs specifically speaking to EHR issues. Their current practice briefs, like the one linked above, are in their public section and you will find a lot there about implementations and what to do when you have some parts electronic in one system, others in another system, and some on paper (as most of us will for some time yet.).

I have gotten quite interested in this entire topical area, participating in both the HL7 Functional Standards workgroup for the Legal EHR, and AHIMA's work group with a similar assignment. If your organization is large enough to have a part time or full time medical records professional on staff, make sure they participate in your product evaluation process. If you are in a smaller group, you may want to work through your area medical society or hospital medical staff structure to secure some advising on this. It seems truly bizarre to me that anyone would sell an "Electronic Medical Record" that cannot qualify as a bona fide medical record, which is why I refer to them as EHRs.

Most physicians do not realize that the business rules built into clinical systems may actually raise their risk of legal problems. For example, one of the EHRs mentioned several times on this board sends the billing information for an in-house lab test to the PM system when it is ordered, not when it was done. If, as occasionally happens, the test is not done (patient forgets and leaves, cannot urinate, etc.) the PM system dutifully bills for the test anyway. A tough event to explain to an auditor. Another EHR mentioned on this board has different sets of forms provided at installation depending on how attentive the customer is to basic medical records business rules. They have a compliant set of forms, and they have a non-compliant set of forms. If you are going to pay big money for a system, which would you prefer. A third EHR mentioned here has an option to have the system installed with the audit function's on/off "switch" available to the user. In other words, you can retrospectively change the documentation with no evidence that the change was made after the encounter was signed. If this was discovered in your system by an investigator, what then would be the value of your records in that system?

If any of you happen to have COPIC (serves Colorado and, perhaps Nebraska?) as your malpractice carrier, you may have access through them to an excellent resource by the name of Barbara Drury, who gave an excellent presentation at HIMSS 2005 on quantifying the risk impact to your practice of an EHR implementation.

If there is interest in this topic area, I will be happy to write more on this thread. If not, I won't bother you all further. There is growing interest in the se topics in the EHR standards community, especially at HL7 and within the Certification Commission for Health Information Technology (CCHIT), as well as AHIMA. The information base will grow very fast over the next six months as upcoming publications highlight these concerns.

Please, caveat emptor. The first thing your EHR should be is a true medical record.

Reed D. Gelzer, MD, MPH, CHCC
Wallingford, CT


(Also happy to correspond off-thread) (r.gelzer@snet.net)

[Kursk]

Thanks for the information Reed, nicely done, please carry on.

When did you make you career change and are many trying to follow in your footsteps? What training (if any) did you find useful or of interest to the employer market?

[RDGelzer]

Thanks for the information Reed, nicely done, please carry on.

When did you make you career change and are many trying to follow in your footsteps? What training (if any) did you find useful or of interest to the employer market?


One of my more recent consulting clients read my resume and called it "eclectic" which I thought quite polite... I made the shift from full time clinical to full time IT analytic work incrementally over about 5 years, reading and going to trade shows and trying to figure out which products had included basic medical records principles in their design. Once you bang away at a few dozen different systems you can separate the wheat from the chaff pretty quickly. AHIMA has been invaluable, as has participating in the HL7 consensus process for standards development.

Actually it was that part-time job in medical school that most specifically launched me in this direction. There was a particular case involving a falsified medical record that I found fascinating because once I proved it did not meet the basic requirements of a medical record, it was almost as if it did not exist. As you can imagine, the case settled very fast.

At this point, I have not found any employers particularly interested in my work in medical/legal product vetting, but I am finding interest at AHIMA, and in some practices shopping for EHRs, in the testing protocols my partner and I have been developing, one of which will be published in AHIMA's journal sometime early 2006. Most of my current work has to do with assessing and mitigating data quality issues, though it is information integrity that is more my interest as you can tell by my willingness to bend the ear of a willing listener. I will be presenting at the District of Columbia Health Information Management Association annual meeting Sept. 9 on "Data Standards: Electronic Health Records as new sources of variability in Data Quality" and am awaiting a reply from HIMSS on a presentation proposal, with some AHIMA associates, on the Legal EHR at the 06 meeting in San Diego. I think that the response from that will give me a better idea whether to recommend a focus on that area as a professional move. For now, its just a very good idea for all our sakes that we strongly encourage EHR vendors to include in their implementations support an orientation to how to use their systems correctly as medical records systems if, indeed, it is possible to do so with their systems at all.

Reed D. Gelzer, MD, MPH, CHCC
Wallingford, CT

[RDGelzer]

I am working on site visits for EHR evaluations. Anyone know of an ENT in or not far from Macon Georgia using an EHR, that might allow an observation in return for a functional assessment?

Thanks for any recommendations.

RDGelzer

[alborg]

Can Reed's post be placed into its own folder? Very interesting (and important) read...

Al

[RDGelzer]

Thank you for the interest in this topical area. I would be happy to point folks in the direction of identifying how to include risk considerations (and data quality considerations) in your selection or implementation process, or how you can do some basic tests in your own system to test its ability to follow some fundamental medical records principles.

Here is a simple test on Authorship functions:

First, find out if it is possible to set up test patients in your system that will, in some manner, be isolated from your real patient populaton. It would best to not fiddle around with your live system if possible. Your system may have a test environment or means to indicate a patient record is outside your patient populaton (like giving them a name like Fake Test and indicating that they are deceased)

Second, a fundamental requirement of a medical record is that the authorship of documentation is recorded accurately. Each user must be uniquely identified in the system in some manner and their identity linked to the specific documentation that they record. One could possibly do this non-electronically but the most straightforward way, and the best use of the technology is still to have each user have their own unique system ID.

To run a simple test on Authorship:
Create a new encounter for Fake Test. Using a Unique ID #1 (such as a nurse or MA), record typical intake information such presenting complaint, vitals, etc. Then pick some clinical information that can appropriate change in a second interview, such as on intake the patient is recorded as being a smoker and history of ulcer (peptic). Transfer the encounter to Unique ID #2 (Physician, PA, APRN, etc.) and record different findings from those of ID #1 as commonly occurs upon closer questioning, the intake information was not correct, it was a duodenal ulcer, not peptic, etc., or patient actually quit smoking a year ago. Complete the encounter and close it, and look at the final documentation.

1. Is all info recorded by User #1 recorded appropriately to their ID?
2. Is it clear or at least available in some fashion what the original documentation was for #1 as distinct from #2 before changes made by the latter?
3. Is all info recorded by User #2 assigned to #2?

Some EHRs will not allow identification of who documented what, showing that the person who signs the encounter (User #2 in this case) did all the documentation. Similarly, some EHRs will show that all the documentation was done at the time it was signed, and not show when the documentation actually began.

Depending on how a given system behaves, your office may have to establish certain rules of use. For example, some systems will show the correct date on a partially completed note that is saved. Users may be tempted to avoid having incomplete notes by leaving them open and then completing them the next day. The problem is that then the system may date the encounter the later date. This can cause problems such as the date of service on the bill is different than the date documented and signed by the provider. In this case, the system may be set up to close all encounters at some hour of the day (say, 11:59PM) so that the date of service will still be correct, with the encounter showing completion, appropriatey, the next day. (The ideal, though, is clearly to complete notes on the date of service.)

If there is interest, I can also post a number of other things to look for to assure the system behaves, or can be used in a manner that shows appropriate adherence to basic medical records rules. If your office staff includes medical records expertise, make sure to involve them. If you do not have access to such a professional, you might consider arranging assistance through your medical society or hospital medical staff, or contact your state chapter of American Health Information Management Association.

I invite others to suggest topics. If more substantial, quantified information is needed, you can contact me directly. This is a topic of intense interest to me as, among other things, our government (your tax dollars and mine) are going to be subsidizing the implementation of EHRs to the tune of hundreds of millions of dollars. For this to begin before effective standards are in place will inevitably lead to the installation of EHRs that cannot create a bona fide medical record.

RDGelzer

[mel]

Thanks for the very informative posts, and welcome

[RDGelzer]

Thanks Mel. Hope to see this topic generate some traffic. It looks like it will be a while before we see functional standards finalized in a way that can be used by facilities and clinicians to evaluate EHR products.

RDGelzer

[alborg]

How about archiving your patient notes? I've heard of folks go as far as having a 3rd party s.a. a lawyer archive sequential backups of EMR information to demonstrate that the data is original and hasn't been changed after the fact. Is this a bit much? Heck, I don't go that far with my paper charts! What do you recommend, Reed?

Also- how can we prevent a court order from taking the whole EMR including other charts whose information can be used against the defendant by showing patterns of care?

I have my own self-made EMR to which I realize I need to upgrade to making it more lawsuit capable.

Many thanks,

Al

[mel]

Also- how can we prevent a court order from taking the whole EMR including other charts whose information can be used against the defendant by showing patterns of care?
Al

Can they do that?? :confused:

[RDGelzer]

Al, I cannot point to a specific practice as "bullet-proof" but, just to touch on a couple of main points.

First, please read the AHIMA Brief on "Maintaining a Legally Sound Health Record" and pay close attention to the specific guidelines there. Keep notes on areas where you may have questions about your system and address them in a methodical way.

(Regarding "Maintaining...", watch for the updates that will be published in three parts starting either in Sept. or Oct. I may have already mentioned thus so apologize if its a duplicate but you can join AHIMA as an associate member for only $140 and get their journal, well worth that cost IMO. Also, they are very hungry to broaden their membership and especially to gain physician input on how to convey many of these issues.

Back to the AHIMA Brief... Note the short description of the rules for medical records and records on computers. Here is an example of a specific requirement you will need to think thru. A requirement in the Federal Rules of Evidence for business records on computers is that the system be able to demonstrate its mode of operation at the date in question for a particular event, such as an encounter. Even most commercial EHRs do not have sufficient metadata specificity to be able to show how particular prompts, alerts, system behaviors, etc. worked at some prior date. So, if you are periodically improving your program in a way that substantially changes the way it behaves, hopefully you are saving periodic iterations in some form. In order to secure this against suspicion of tampering, a system either has to have very robust security features and access logs, or one can just periodically create an image of the system for off-site storage. (I am stretching the bounds of my technical knowledge here, but I am told this can be done as a regular function of the system's operations). There are companies that do this commercially, like Iron Mountain, or you can consult your attorney on other means to do it.

The most important thing is to:
1. Establish, in writing, what your compliance program is for your medical records, including those on paper and on computer.
2. Where you make discretionary decisions, document why you chose to do so, as a matter of due diligence.


I know it seems like a pain, but think about your expectations for how other businesses protect their records of your life. Would you be satisfied if your bank, your brokerage, etc. kept their electronic records in the same state as
the main thing is to know what you are doing and why, and reference professional guidelines. Again, AHIMA has practice guidelines you can refer to.

Keep in mind that these policies and procedures can also serve to establish your disaster recovery process. That few hundred you spend quarterly will look awfully cheap compared to the cost of a recovery service if you ever have a catastrophic failure. (Plus the cost of refunding to insurers any payments you receive for encounters whose documentation you lose, which they can theoretically demand if there is no evidence that services were provided.)

When in doubt, establish a policy that is referenced and thoughtful, and review it at least annually. If you are depending on your system to be part of your legal health record, make sure it meets the basic standards for same.

The AHIMA Working Group I participated in has applied to present at 2006 HIMSS in San Diego, so watch the agenda there. AHIMA's national convention in October, also in San Diego, will address this topic.

I anticipate that this topical area is going to rapidly assume center stage within the next 6-18 months.

Hope that helps.

RDGelzer

[Kursk]

Thanks, good stuff. I agree that specific identification (ownership) of each piece of the documentation in the record is important but unfortunately it is cumbersome and leads to information clutter. One of our ERs prints the electronic encouter for the PMD - seems like every line has a different sig/time for every nurse and doctor that sees the patient in the ER and EVERY action no matter how meaningless to the overall care process is documented. This 'over documentation' while nice and legal is not conducive to medical communication, nor does it comprise a useful user interface. It is yet another indication of the tail wagging the dog, a metaphor for so many other useless but expensive things we do in medicine to CYA. Unfortunately, such processes seem to trump care and communication which are more important by orders of magnitude. It would be nice to see the ERs supress the parts of the record that aren't immediately pertinent to the medical problem. This should be easy to do yet we are still receiving 5-6 pages of documentation for simple ER problems and even more for more complex cases. This is ridiculous. Naturally the signal to noise ratio is very small.

[alborg]

>>> Can they do that??

Of course! But you can delete the primary table's patient list leaving alone only the sueing patient's information so that the lawyer can't see any other patient data. But you have to have your system prepared...

Al

[RDGelzer]

Thanks, good stuff. I agree that specific identification (ownership) of each piece of the documentation in the record is important but unfortunately it is cumbersome and leads to information clutter. One of our ERs prints the electronic encouter for the PMD - seems like every line has a different sig/time for every nurse and doctor that sees the patient in the ER and EVERY action no matter how meaningless to the overall care process is documented. This 'over documentation' while nice and legal is not conducive to medical communication, nor does it comprise a useful user interface. It is yet another indication of the tail wagging the dog, a metaphor for so many other useless but expensive things we do in medicine to CYA. Unfortunately, such processes seem to trump care and communication which are more important by orders of magnitude. It would be nice to see the ERs supress the parts of the record that aren't immediately pertinent to the medical problem. This should be easy to do yet we are still receiving 5-6 pages of documentation for simple ER problems and even more for more complex cases. This is ridiculous. Naturally the signal to noise ratio is very small.

Kursk, thanks for the great description of a specific (and annoying) way to deal with medical records requirements, etc. That is a very good set of observations and very helpful to have the examples from experience. However, at the end of the day, what you described is, as you say, a poorly designed system.

(Disclaimer-I will mention, from time to time, various products here as examples of particular functions or means to handle certain tasks. Please do not take these as endorsements, nor anything but illustrations. There are many well designed systems out there that I have never seen. There are others I am more familiar with by virtue of requests from clients to test or evaluate in specific settings...)

The problem with what you described is that it is mixing metadata with data. A lot of the older products do this, tho not all. Some do the opposite, they take the metadata and bury it too far to be useful or, worse, do not save any at all.

If you get a chance, look at three different approaches to the problem you describe.

1. MediNotes Chart Power. To some extent the User can set up different primary views with varying degrees of information in that view (errors, corrections, task transfers, etc.) . The supporting information is very readily available if it is needed, no special programming or vendor support is needed for auditing but what the User looks at as the primary view is a very clean note presentation. The info needed by an auditor is just a different set of screen views and you need not bother with them in your clinical use of the system.

2. CyraMed (formerly XocDoc-glad they changed it...) This behaves like an analogue of a form in some respects, (but with the advantage of structured data capture). If changes are made, you get a new view. The changes are preserved in a series of iterations of the same document but that is invisible to the user. What the User sees is a final and clean view, with all the metadata in the background and recoverable by paging back thru the different iterations if needed. Again, most clinicians will never need that but it is there for auditing, security, integrity support, etc.

3. EPIC (and other systems with MEDCIN embedded like PMSI, Allscripts, CHCS II) With successive changes in a document as it passes from user to user, a series of iterations are saved (when the User understands the necessity of doing things in a particular way...) but in the same user view. What you end up with is a long chain of interations of the same documentation but, again when used appropriately, each is a partial version of the cumulative activity thru the encounter. It looks like heck when printed but you quickly learn that what is pertinent is the last presentation which is the entire note final note, more or less "clean". In the user view, it is actually the first view, the the prior iterations in a string below it if you scroll down. If you do not need all the iterations, you simply use the first view you arrive at when looking at the note.

Again, these are three different ways of handling the requirement of the preservation of metadata, either in a system behavior or as a data set. Each has advantages, disadvantages I suppose and each system has its other strengths and weaknesses. I offer them only as examples of designs that do not cause the low signal/noise ratio you describe, while also preserving the fact that, noise in one setting is signal in another (in court...).

Hope that helps. This is exactly the sort of thing that really helps people I think. Point them in the direction of products they can get their hands on to see that similar problems can be handled differently but nonetheless well.

RDGelzer

[tonicdrink]

I am curious, which of those EMRs were you referring to?

"A third EHR mentioned here has an option to have the system installed with the audit function's on/off "switch" available to the user. In other words, you can retrospectively change the documentation with no evidence that the change was made after the encounter was signed. If this was discovered in your system by an investigator, what then would be the value of your records in that system?"

[RDGelzer]

I am curious, which of those EMRs were you referring to?

"A third EHR mentioned here has an option to have the system installed with the audit function's on/off "switch" available to the user. In other words, you can retrospectively change the documentation with no evidence that the change was made after the encounter was signed. If this was discovered in your system by an investigator, what then would be the value of your records in that system?"

I will give you a hint, it is on the CMS DOQ-IT Program's list of participating vendors. Just another illustration that just because an EHR is on a list, does not mean it does not require thorough investigation. If you look at any EHR vetting instrument, make sure you can identify whether it is vendor-reported information or whether someone has independently, objectively evaluated the product.

RDGelzer

[RDGelzer]

Not a lot of traffic on this thread...

To the moderators: HIMSS, AHIMA, and Alliance (NAHIT) are looking for means to convey information to physicians that will be supportive of EHR implementation, such as readiness assessment tools, boilerplate medical records policies and procedures, Compliance guidance for small practices, resources for practice managers for EHR implementations, etc. Would this board be interested in hosting communications for that purpose, as well as for soliciting feedback on those materials?

Please communicate with me with DocsBoard in the subject line

Thanks,

RDGelzer

[RDGelzer]

If you would like a reply, please respond.

RDGelzer


I will give you a hint, it is on the CMS DOQ-IT Program's list of participating vendors. Just another illustration that just because an EHR is on a list, does not mean it does not require thorough investigation. If you look at any EHR vetting instrument, make sure you can identify whether it is vendor-reported information or whether someone has independently, objectively evaluated the product.

RDGelzer

[DaffyDuck]

Whats the story on Soap

[RDGelzer]

Have not had occasion to evaluate Soap, SoapNotes, SoapWare... Would welcome an opportunity to do so.

RDG :)