Sunday, April 23, 2023

Human Performance Errors (HPEs) and EQA


I am in the midst of an interesting career that started simply enough by being a laboratory physician, but over time has taken me to into areas that I did not anticipate which included working in laboratory quality from multiple directions. 

In the early 1980s I got involved in developing an External Quality Assessment or EQA program (also known as proficiency testing) which was designed to help laboratorians and accreditation bodies be better aware of human performance errors (HPEs).  That led me towards an active career in all aspects of laboratory quality from standards development to quality management training and then to teaching.  Some would say that sounds like a backwards direction.

 In any event, HPEs are a big problem, not only in laboratories, and usually all work sites.  They either result from slips or distractions, or from mistakes associated with knowledge or procedures failure.  The former are frequent, but often intermittent and inconsistent.  The latter are more likely to be repeated or at least repeatable.  These repeatable HPEs are often associated with poorly written Standard Operating Procedures, or unclear unmonitored training which can result in gaps in knowledge and understanding.  These are Human Performance Error, but while the error may be detected in the work of the technologist/performer, but the REAL PROBLEM is the person who did the insufficient training or wrote the insufficient SOP, i.e. these are management caused systemic errors.  This is not a new finding.  Almost all people who look at error (Deming, Juran, Crosby, Dekker) all say the same thing, management errors cause more harm and more significant harm, but rarely get included in error reports.  (The Pen is More Dangerous than the Wrench)

HPEs differ from equipment or reagent errors.  Equipment and reagent errors can be detected and prevented by a solid Quality Control program.  HPEs can be detected by EQA, especially when they are repeated or repeatable, and silent, meaning that they didn’t show up during an bad event investigation. 

When people are presented with samples that closely simulate typical samples whose content is unknown to the tester, but known by the provider, and error can be captured by the external results evaluator.  This is clearly better than having the error detected after a patient is harmed.

The external evaluation is an essential part of the exercise, because it shines a light on the problem in a way that internal evaluation by itself may not.   When the reviewer is external (as with EQA) there is no motivation to not report the problem… there is no reason to hide missed performance.  It makes EQA the most objective form of testing to detect HPEs in laboratories.

Not everyone likes the idea of being monitored for error and may find ways to escape scrutiny.  During a recent presentation that I gave, a person in the audience said that laboratories would do better if instead of developing programs like EQA, they would develop better techniques so that that errors would not happen.  I disagreed.  I don’t think there is any plan that will prevent all errors.

I noticed with interest that the International Organization for  Standardization (ISO) does not agree either.  The newest version of ISO15189:2022 on Quality and Competence in the medical laboratory expects laboratories to enroll in EQA programs that not only look for human performance during examination testing, it also specifies looking at performance in the pre-examination (between the collection of the sample and it being received in the laboratory) as well as the post-examination  (starting to creating the record or report and ending with contacting the physician).  At the present time I would guess that only about 20 percent of EQA organizations can provide today, so good on the ISO for writing standards that should push program development.