Back To Top

RHIOs without Patient ID is like playing with fire


By Beth Just 
As seen in Healthcare IT News
January 1, 2006

If you watch "ER," you know there are creative writers who can dream up just about any ailment sure to keep the viewers checking into the fictitious Chicago hospital. We tune in to keep up with the team of doctors and nurses trying to do their best in an understaffed emergency room.

The medical situations likely baffle even the actors on the show. After all, they're not doctors - they just play them on T.V. In their world, the gravest mistakes require multiple takes. In real medical facilities, mistakes hold far more serious consequences. The wrong medical procedures, drug interactions, failure to act... these are the subjects of real-life dramas being played out in medical facilities all over the world. 

All too often, unfortunately, these mistakes and so many others are the direct result of duplicate and overlayed medical records. This isn't brain science, but to healthcare facilities battling patient information challenges, it's just as crucial to the health and wellbeing of the "patient" - in this case, the Master Patient Index (MPI.) 

As multiple patient records from various medical facilities are electronically linked, it comes down to dirty data in, dirty data out. In the United States, it's not uncommon to find 10 to 40 percent duplicate rates in hospital MPIs. Provisions must be put into place to find errors and correct them before the data gets even dirtier. 

While the push is on to accurately electronically link multiple patient records, now that many healthcare facilities understand and recognize that their MPI is only as good as the integrity of the data it contains, there is a real problem with taking the easy way out - electronically linking on the match of a small number of patient identifying fields. This is a form of the federated model of record linking which matches various patient identifying data elements instead of using a unique identifier. 

Some in the industry suggest linking records electronically on very few data elements gets the job done. This, in my opinion, is asking for trouble. Compromised MPI data - for whatever reason - easily causes duplicate or overlayed records. 

It is incumbent upon medical facilities to do whatever it takes to make sure that MPIs - once electronically linked - are as accurate as they can possibly be. 

"Almost," which unfortunately is the best you can hope for using a limited data set, is just not good enough. Without either a unique identifier or a more robust set of patient identifying fields for linking records, overlays will be created. Overlays cause inaccuracies. Inaccuracies cause patient harm. 

Is it worth the risk? Clearly not.

OK, so you're convinced. You understand the problem. You even know about the solution. What you may not get is that sometimes there's no time to do it right, but always time to do it over. MPI clean up is not one of those times. This is an area of black and white, life and death. It's critical to the health and welfare of patients that electronic linking of medical records is done with the utmost accuracy. 

The federated model assumes that computers are as smart as humans. Remember, dirty data in, dirty data out. The only truly accurate - and yes, cost effective - solution to electronic MPI data linking must include advanced methodology - perhaps adding more unique identifiers such as a middle name or social security number - and human intervention. 

Any automated linking process that does not include a technology-based mechanism to look at the records and see an overlay and subsequently alert a person who can correct the mistake will never assure the kind of accuracy needed. 

Of course, as with many business decisions, it often comes down to cost. Many medical facilities are quick to say they can't afford a more accurate system and therefore choose to "do the best we can." Can you imagine a patient saying that to a doctor? "Just do the best you can in the operating room." 

No patient would tolerate that kind of lackadaisical attitude on the part of a medical professional and they shouldn't have to. Yet, in a very real way, if physicians can't be sure of the information upon which medical decisions are based, they really are only able to make an attempt at doing their best. 

No one should have to deal with those kinds of odds. The ER team has an hour each week to save their patients; real life doctors often have seconds. Only the most accurate, cleanest medical records will allow them to make the most of every moment.