The following is a guest blog post by Lisa Pike, CEO of Versio.
With over a third of healthcare organizations switching to a new EHR in 2014, there is a lot of data movement going on. With the vast amount of effort it took to create that data, it’s a valuable asset to the organization. It can mean life or death; it can keep a hospital out of the courtroom; and it can mean the difference between a smooth-running organization and an operational nightmare.
But when that important data needs to be converted and moved to a new EHR, you realize just how complex it really is.
During a recent conversion of legacy data over to a new EHR, we came across this entry in the Medication List: Raised toilet seat, daily.
Uh, come again??
How about this one? “Dignity Plus XXL [adult diapers]; take one by mouth daily.” What does the patient have, potty mouth?
Now, while we may snicker at the visual, it’s really no joke. These are actual entries encountered in source systems during clinical data migration projects. Some entries are comical; some are just odd; and some are downright frightening. But all of them are a conversion nightmare when you are migrating data.
Patient clinical data is unlike any other kind of data, for many reasons. It’s massive. It requires near-perfect accuracy. It’s also extremely complex, especially when you are not just migrating, but also converting from one system “language” to another.
Automated conversion is a common choice for healthcare organizations when moving data from legacy systems to newly adopted EHRs. It can be a great choice for some of the data, but not all. If your source says “hypertension, uncontrolled,” but your target system only has “uncontrolled hypertension,” that’s a simple enough inconsistency to overcome, but how would you predict every non-standard or incorrect entry you will encounter?
Here are some more actual examples. If you’re considering automated conversion, consider how your software would tangle up over these:
|SOURCE SYSTEM SAYS
|346.71D Chm gr wo ara w nt wo st
|levothyroxine 100 mg
||Should be mcg. Yikes!
||Target system has 20 choices
||NKDA= no known drug allergies.
Having no allergies causes vomiting?
|Massage Therapy, take one by mouth twice weekly
|Tylenol suppositories; take 1 by mouth daily
||Maybe not life-threatening, but certainly unpleasant
|Should have been PMDD
(premenstrual dysphoric disorder)
|Allergy: Reglan 5 mg
||Is patient allergic only to that dosage, or should this have been in the med list?
Confusing allergies and meds can be deadly.
||Centimeters or inches? Convert carefully!
These just scratch the surface of the myriad complexities, entry errors, and inconsistencies that exist in medical records across the industry. No matter how diligent your staff is, I guarantee your charts contain entries like these!
When an automated conversion program encounters data it can’t convert, it falls out as an “exception.” If the exception can’t be resolved, the data is simply left behind. Even with admirable effort, almost no one in the industry can capture more than 80% of the data. Some report as low as 50%.
How safe would you feel if your doctor didn’t know about 20% of your allergies? What if one of those left behind was the one that could kill you? What if a medication left behind was one you absolutely shouldn’t take with a new medication your doctor prescribed? Consider the woman whose aneurysm history was omitted during a conversion to a new EHR, so her specialist was unaware of it. She later died during a procedure when her aneurysm burst. I would say her family considered that data left behind pretty important, as did the treating physician, who could be found liable.
Liable, you say?
That’s right. The specialist could be found liable for the information in the legacy record because it was available….even if it was archived in an old EHR or paper chart.
You can begin to see the enormity of the problem and the potentially dangerous ramifications. Certainly every patient deserves an accurate record, and healthcare providers’ effectiveness, if not their very livelihood, depends on it. But maintaining the integrity of the data, especially during an EHR conversion, is no trivial task. Unfortunately, too many healthcare organizations underestimate it, and clearly it deserves more attention.
There is good news, however. With a well-planned conversion, using a system that combines robust technology with human expertise, it is possible to achieve 100% data capture with 99.8% accuracy. We’ve done it with well over a million patient charts. It isn’t easy, but the results are worth it. Patients and doctors deserve no less.
Lisa Pike is the CEO of Versio, a healthcare technology company specializing in legacy data migration, with a proven track record of 100% data capture and 99.8% quality. We call it “No Data Left Behind.” For more information on Versio’s services or to schedule an introductory conversation, please visit us at www.MyVersio.com or email firstname.lastname@example.org.