Tuesday, November 20, 2012

Quality and Error and Standardization of Laboratory Reports

Over the last while I have been reading about and hearing presentations on Electronic Health Records.  The concept is that EHRs while expensive, and difficult to establish, once established will revolutionize care.  Dependence on having access to every clinician’s records will become a non-issue.  Test results will be readily and rapidly available side-by-side by date, rather than being in a separate record for each laboratory.  Repeat testing will become a thing of the past.  

Well it turns out that maybe by taking it very slow, we have done the right thing.  It may be that EHRs are going to be our Ford Edsel.  A whole slew of challenges have appeared.  Getting agreement on common EHR language is slow and painful.  And even if we have common language, many laboratory results are highly analyzer specific and not comparable across platforms.  And now some economic cracks are appearing [see: http://www.darkdaily.com/healthcare-observers-disagree-on-cost-effectiveness-of-electronic-health-record-systems-1116#axzz2CnmDcvnZ ]. 
So time will tell us soon enough whether EHRs are actually going to happen.  

But in the meantime, we still have a big problem that needs to be solved, one way or another.   

Several years back, I was in conversation with a group of pathologists all with particular interest in one particular anatomic target.  The group saw great opportunity to band together to create an on-call network so that there would always be pathology support for surgeons but at the same time they would also be free to work on their own research activities on a set schedule.  Because they were working as a team, they collectively agreed that if a doc called, he would get the on-call pathologist rather than a specific one.  Since they all basically had similar knowledge in the specific target, they could say that everyone was receiving a standardized level of care.  All that made infinite sense.  

Unfortunately the system started to break down after barely a few months.  The problem was that when it came to writing reports, each had their own style.  Some would construct separate sections within the report, while others preferred a narrative style.  Some were brief and precise, others were more verbose.  Some were heavily jargon enriched while others preferred to use both the accepted jargon, but also explanatory comments.  When we tried to establish a code to standardize reporting, the push-back was huge.  “No one is going to tell me how I can or will practice my craft.”  End stop.

It’s interesting that we also see this same mind-set in medical microbiology.  As much as we have much commonality to microbiology test methods, there is no requirement that these “standard” test methods actually get used.  There is no requirement that all microbiology laboratories use uniform microbial nomenclature.  There is no requirement that all microbiology laboratories use a common set of interpretive or cautionary notes; where common concepts get reported, they are as often as not crafted with personalized verbiage.  

We microbiologists are a highly individualized and ornery group.  “My laboratory … My way”.  It’s not that we are resistant to change [for a read on Luddites see: http://www.medicallaboratoryquality.com/2011/11/innovation-and-medical-laboratory-part.html ], it’s more that we sometimes get more locked into our own subtleties than in adopting general principles.   

Example 1:
At CMPT, our proficiency testing program, we promote common concepts, such as requiring the reporting of “Methicillin Resistant Staphylococcus aureus (MRSA)” rather than “ Staphylococcus aureus ; R-Oxacillin”.  

Clinicians understand that MRSA is a signal that Infection Control needs to be contacted.  Laboratories that choose to not use that terminology put patients at risk.

Example 2:
I understand that the problem is not necessarily an ego issue.  Sometimes it is our technology.  A number of years back at CMPT we sent out a simulated urine culture with a heavy and pure coliform, suggestive and supportive of a urinary tract infection.  [see: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3327405/ ]  Unfortunately because of the limitations of the LIS systems available at the time the results came back in 14 different styles of units, and rarely with a confirmation comment (consistent with urinary tract infection).  When we send out reports as either 1X10^5 bacteria/mL or 1000X10^5 bacteria/L or 1X10^8 bacteria/L, we can not be particularly surprised when confusion sets in (Note:  all these values are the SAME!). 

What concerns me is that for us, microbiology is what we do and understand.  That is not the case for clinicians.  When we refuse to adopt even simple rules for standardizing the text of reports, then we increase the risk of people misinterpreting our reports, and open the gates to error.    And now that many patients can receive reports directly, without any health care intermediary, our level of hazard and risk grows substantially.  

So here’s the bottom line.  Waiting for EHRs to solve our reporting standardization problems is a time killer and potentially a patient killer.  Simple low tech solutions are available, and we avoid them at our peril.

No comments:

Post a Comment

Comments, thoughts...