Advertisement

Blog Viewer

interoperability woes - usable patient data

By Dennis Tribble posted 01-26-2017 10:06

  

I see a lot of discussion in various locations about interoperability. The vast majority of that discussion appears to presume that our problem is simply the unwillingness of information systems vendors to participate in data sharing. It seems to presume that all healthcare data is of impeccable quality.  I think the problem is more complicated than that. While there can be no doubt that people perceive both business and regulatory issues with sharing health data, let's pretend for a moment that those can be overcome (because they can). Where does that leave us?

I enjoy a good relationship with my personal physician; she is an excellent diagnostician with whom I have had a physician-patient relationship for over ten years, and, I think a good friend. Her practice is owned by a large, multi-state health-system and sports an EMR provided by that health system, along with an on-line patient portal. The data management is atrocious. My medical record is littered with obsolete information: problems that have long since been resolved; drugs I no longer take; allergies I don't have. I had a bout of gout about two years ago; those medications still show up on my medication list. It was once determined that levofloxacin is not effective for my occasional sinus infections; that is dutifully recorded as an allergy.

I have complained about the state of my records; nobody at that practice seems to own the quality of that data. If I complain (and I have), they can't even tell me who I would go to in order to get it fixed. So every time i visit the office, we go through a dance where I inform them that I am still no longer taking the same list of medications I cited during my last visit, and no.. remember... we recorded me as allergic to levofloxacin because that was the only place we could document that it didn't work for my sinus infections. And I have given up even using the portal; it is incomplete and inaccurate.

Some of the problem relates to the structure of the EMR itself; there is no other place to record intolerance to medications, so they get recorded as allergies. Just as we see in hospitals, caregivers who have work to do simply create workarounds on the fly without regard to the downstream effects of their instantaneous data decisions.

But the larger issue seems to be that nobody owns the problem of data quality. There is no process for figuring out on a large scale how, for example, intolerance information should be recorded so that it is not misleading, or problem lists and medication lists stay current.

Further, there seems to be no continuity plan for patient data. During my last physician office visit, I learned that the office was moving to a new system, which meant that I had to completely fill out a stack of new forms demanding information that I know they currently have.

This is not an isolated incident. I had dinner with a good friend recently who described identical problems with his medical record. At the last HIMSS I had a drink at the bar with someone from a company who sells EMR solutions to physicians' offices whom I queried about the extent of the problems I observe. I learned that this is ubiquitous; that offices have nobody whose job it is to manage both the operation and the content of patients' records, and that most policy decisions about the use and maintenance of the electronic health records are left to the one group of people in the practice who haven't a clue about them: the physicians.

By the way, lest we get complacent, let us recognize that our management of patient data within EMR systems in hospitals is often no better. Depending on who collects allergies (it sometimes someone completely unqualified), we likely have codeine allergies (which are probably just codeine-induced belly aches) on patients currently taking acetaminophen with codeine #3, or someone dutifully recording the patient assertion that they are "... allergic to all 'mycins'...". Could information be any less clinically useful? I have yet to read about or hear about anyone in the hospital community that actually manages these garbage cans of information. I have had some tell me that they fear the litigatory consequences of doing so.

So, back to the primary discussion, imagine what happens if my health record now becomes the union of information from three physicians' offices and the two hospitals and three urgent care organizations from which I have received care in the last few decades. All would record a penicillin allergy (highly suspect, by the way) that I reported, but only one of the seven would report an allergy to levofloxacin (which I actually do not have). The medication lists from each of these would likely be different. How is that helpful or useful? How does that improve the quality of care? In this large collection of data, whose job is it to clean this all up so that it contributes useful information to the next healthcare encounter? By the way, self-reporting of allergies is know to be suspect, even when the report is not something as obviously healthcare illiterate as "I am allergic to all mycins." A quick search for frequency of penicillin allergy indicates that it is self-reported in about 10% of the population, while studies of actual prevalence by skin test indicate rates that are much lower. Similarly most reported codeine allergies are actually pseudo-allergies or side effects.

It appears to me that, without both a technical and regulatory infrastructure whose job it is to resolve some of these issues, what technical interoperability gets us is a pile of conflicting data: garbage in = garbage out.

Maybe we need standards we don't currently have around some of this, like "medications don't get perpetuated on a med list if they haven't been refilled in over a year", or "allergies must include a statement of when detected and what the allergic presentation was", or "penicillin allergies must be documented by a skin test." Maybe part of medication reconciliation needs to include capturing allergy information.

I am not certain that I know the answer. I am certain that simply co-mingling patient data from a variety of sources is likely to result in unreliable patient information. How do you think we should solve this?

Dennis A. Tribble, Pharm. D., FASHP

Ormond Beach, FL

DATdoc@aol.com

The opinions in this blog are my own, and not necessarily those of my employer or of ASHP

0 comments
2254 views

Permalink