The Evolution Revolution

 

The Evolution Revolution

When I first started working as a laboratorian, now close to 50 years ago, I had multiple responsibilities, to be seeing patients, to be looking at test results and help patient’s docs understand those results, and to be involved in research in the arena of infectious diseases and their diagnoses, and write manuscripts. 

I likely would have stayed with the career path, but in 1983 I got the opportunity to get involved in developing a program in Proficiency Testing (PT), probably better phrased today as External Quality Assessment (EQA). 

Back in those days EQA was seen as the best way to detect when laboratories were making performance and diagnostic errors, which could result in an incorrect test report, and potentially lead to a fatal outcome of a patient.  It was infinitely superior to what Sunderman had created because it was more realistic.

For me this was my career game changer.

Over the years we spent a lot of time evolving EQA beyond being a simple testing program to one that could challenge laboratories and laboratorians knowledge and skills, and contribute to education for the collective varied group of people working in the lab.

EQA  evolved more broadly through creating supplemental challenging programs like monitoring equipment and safety and risks.  It was not only as important as accreditation audits and visits, but some could argue that it was far greater importance because EQA occurred at least 4 times a year or more, while accreditation audits were only once every 4 years. 

But here’s the kicker, as the study of laboratory quality extended, it became clear that at least 70-80 percent of errors happened long before standard EQA worked  as an error detector… these were called pre-examination errors, and another 5-7 percent of errors occurred after the testing phase was completed (post-examination error) 

So while EQA was clearly doing something, was it doing enough? 

If not today, then tomorrow or soon after, the laboratory world is going to go through a MAJOR evolutionary change.  Laboratory computers already capture or have access to tons of information like staff name, age, activity, training status, workload, and shifts, and health breaks, injuries.  Being part of the hospital network it will have patient names, age, room location, medication records, and maybe even their meals.  It will assuredly have access information on who collected the sample, and when, and how it was transported and by who.  It will know when the sample arrived and where it was stored, and who set it up and who tested and reported.  It probably has access to information from the blood culture analyzer, the PCR and the Maldi Tof.   It will assuredly have access to laboratory room temperature and humidity, reagent lot and age, instrument ID calibration date and maintenance history.  All this and tons more will be integrated into an AI Machine Learning System able to detect deviations from what would be an expected result, not 4 time a year, but once or twice a day.

And it likely will have information from the last accreditation, the deviations from ISO 15189 and when were they last corrected and monitored.

This is going to look not like EQA but EQA on steroids.  This will provide people in the laboratory, maybe people in administration, and maybe the people in the accreditation body with data on an “as needed basis”.  The major impact will be that it may inform not as an after the fact, but more as a going forward probability of error pending.

This will not be another step along the evolving evolutionary pathway…  It will be a revolutionary upheaval. 

Perhaps the greatest change will not be that AI detects more errors, but that it detects increasing risk before errors occur. Rather than waiting for EQA, or  quality control failures or patient complaints, laboratories may receive an early warning that combinations of staffing, workload, specimen transport, reagent performance, instrument behaviour or environmental conditions are beginning to resemble situations that previously resulted in errors.

In contrast to where we have come through the development and implementation of EQA, quality management may be more able to shift from investigating yesterday's mistakes to preventing tomorrows.

Standing back, my guess it will be viewed at being a contributor to a successful patient centred care program, loved and cherished by administrators alike.

But from my perspective, scary……