1. Introducing Stats-Free Quality January 16, 2014

#1: The Journey Begins 
View this email in your browser


Happy New year! This will to be a pivotal year for laboratory quality control!

In 2014, labs will begin to implement CLSI EP23 guidelines that "The laboratory should manage risk by implementing QCPs (Quality Control Processes) that serve to ensure test result quality is appropriate for clinical use."  and  "At the least, the ability of the QC procedures to detect medically allowable error should be evaluated."

In 2014, labs will be introduced to Stats-Free QualityTM!  

Quality OptimiZer Reports , AWEsome Numbers Inc.Stats-Free Quality (SFQ): "Stats-Free QC assesses and verifies that patient results meet  defined clinical standards.  Results are conveyed in  standardized grades and graphics with interpretive comments and action flags."    (Z Brooks, 2014) 
Image: Report components from Quality OptimiZer (AWEsome Numbers Inc.) 

 (Of course the software uses stats, but the people do not.)   
Zoe Brooks, September 2013  I have reached the conclusion (and finally found the nerve to say out loud) that the only way to improve quality control is to dramatically change how it is taught and implemented.   After decades of teaching "this is what the statistics mean to patient care,"  it makes sense to me to let the computer calculate and interpret statistics, and teach the people to understand patient risk and respond  to action flags.

Forty years ago, in 1974, I became Chief Technologist in the 50 bed hospital in Port Perry, ON. We boiled blood sugars; did tilt-tube pro-times; washed glass pipettes and the flame photometer blew up quite regularly.  Forty years ago, we used Levey-Jennings charts with a 1-2s rule.  We had peer comparison reports with Youden plots and SDIs.  Allowable error limits had been recommended for years(1); Critical Systematic Error was introduced to assess accuracy & precision (2).  Ontario launched the Laboratory Proficiency Testing Program.   

Twenty-five years ago, in 1989, CLIA 88 was the hot topic. Westgard multi-rules were the recommended standard (3).  Westgard was publicizing Q.C. selection grids (4).  I was Technical Coordinator for six Northern Ontario labs.  We had computerized QC with Westgard Rules.  We used the internet to transmit and receive peer reports and an FTP site to send data from remote labs to my office.  Twenty-five years ago, the experts recommended setting allowable error limits(1),  evaluating method quality(2) and selecting Q.C. rules to match critical systematic error(3)I have been using that process since then.  I seem to be in the minority.

Today's analytical processes bear no resemblance to those of 25 to 40 years go. 
Today's Q.C. practices are virtually the sameDespite a small mountain of literature and innumerable meetings and committees ... today's labs have not implemented the processes and practices that are required to create clinically-effective quality control. 

Let's literally look at laboratory quality differently.   I believe it is entirely possible for everyone who cares about  laboratory quality to assess medically valididity of patient results - without statistics.  This top-down, patient-focused process requires a change in today's philosophy and process, not in the software or controls.

Your Invitation to
The Journey from Stats to Stats-Free

25 Years in 25 Weeks

It took me 25 yThe Journey to Stats-Free Qualityears to reach the conclusion that the only way to improve quality control is to dramatically change how it is taught and implemented.   I traveled a unique path to reach that conclusion.   It would  make no sense to ask you to blindly accept my conclusions without seeing the evidence I have seen.  Fortunately, AWEsome Numbers Inc. has provided a platform to allow you to follow my 25 year journey and decide for yourself.   Please get involved!  Ask, test, question, simulate, challenge, suggest, improve, lead! 

With your support, this 25 week series will be an open dialogue on clinical quality!   Open doors.  No politics or commercialism. Everyone welcome.

I believe there is a Stats-Free way for everyone to understand clinical quality.  I hope you will be part of this.  Although I call it Stats-Free, it is also free to use the most powerful statistics available - as long as the input and output is verified clinically effective .   Let's create an evolving QC processes that continuously improves clinical quality. 

Enroll in Webinar Series
Enroll in FREE 25 week 'Webinar+' series.
Live webinars:  Every Thursday, 2:30 PM EST, beginning January 30rd
30 minute presentation of each week's topic, plus Q&A. 

Jan 30  2:30  EST:  Stats-Free QC.  The Journey Begins

Feb 6  2:30  EST:  
 "If we only knew then what we know now
        ... oh wait, we did!"

Link to  the 25 week outline, references, discussion forums and more.
This week's online discussion: 
     "Do you think it is OK for 1 result in 20 to be medically invalid? That is 50 misleading results in every 1000 you report.   Can the error rate ever be guaranteed at zero?"
This  25 week journey concludes in Chicago,  July 27th, with an AACC  Afternoon Workshop Risk Based Quality Control – No Stats Required

Planned stops include a full day workshop June 20th, at the CSMLS LabCon 2014 in Saskatoon 
Q_ality needs U. What do U need? 

Where the Stats-Free journey goes from there is up to you!

This change requires YOU.  Please join the discussions, spread the word, participate in and follow the studies, be part of the process!      Zoe

This is where social media comes in.  Spread the word!

Why Stats-Free?  This fails the DIMS test

.... because today's Q.C. fails the DIMS test.

1.  It is NOT OK for one patient result in 20 to be medically invalid. 

The definition of allowable error has drifted in the past 25 years.

  • The traditional definition: The amount of error that can be tolerated without invalidating the medical usefulness of the analytical result (Carey and Garber - 1989)
  • Biological Variation Definition:
    The amount of error calculated from intra- and inter- individual biological variation
  • Regulatory Definition: The amount of error that can be tolerated without failing requirements established by a regulatory body
  • Statistical Definition: A statement such as “This result is expected to be within X% of the true result Y% of the time. X is the performance standard. Typical values for Y are 95% (2 SD), 99.7% (3 SD) and 99.9997% (Six Sigma)."
  • Dictionary definition: the amount of error that can be tolerated without invalidating the medical usefulness of the analytic result. Allowable error has a 95% limit of analytic error; only 1 sample in 20 can have an error greater than this limit. (Mosby's Medical Dictionary, 8th edition. © 2009, Elsevier.)
2.    There are NOT 192 right ways to do Q.C.  In a recent survey, at least one person selected each of 6 sources for the assigned mean, 8 sources for the assigned SD and 4 combinations of Q.C. rules.  With these choices, there are potentially 192 different Q.C. practices in use. 
Insanity: Doing something over and over the same way and expecting a different outcome
Absurdity: Doing something 192 different ways and expecting the same outcome!

3.     Despite everyone's best efforts, current Q.C. practices are  pitifully ineffective to prevent patient errors.  (References) 

4.     We know today's Q.C. is broken.  It's time to stop doing what we have always done ...

Todya's QC makes no Sense

5.  Most lab staff have zero to ten hours education in stats & QC.  You cannot possibly teach method validation, statistical QC, Six Sigma, QC design and troubleshooting in that time frame to today's high school grads.  In recent surveys, that's what supervisors and techs want new grads to know.  Why are we teaching 250,000 people how to interpret a 1-2s flag, when the computer could just tell them what it means and what they should do?

A Different Perspective

Visit  AWEsome Numbers on Facebook!
Our goal -  clear communication between all the stakeholders in laboratory quality.
AWEsome Facebook
Thanks for caring for quality!             
Please forward this to friends and colleagues.

Zoe Brooks     January 16, 2014
  1. Tonks, DA. A study of the accuracy and imprecision of clinical chemistry determinations in 170 Canadian laboratories. Clin. Chem. 1963;9:217-33.
  2. Westgard JO, Carey NR, Wold S. Criteria for judging precision and accuracy in method development and evaluation. Clin. Chem. 1974;20:825-33.
  3. Westgard JO, Barry PL, Hunt MR, Groth T.  A multi-rule Shewhart chart for quality control in clinical chemistry.  Clin Chem 1981;27:493-501.
  4. Cavenaugh, E.  A Method for Determining Costs Associated with Laboratory Error. Am J Public Health. 1981 August; 71(8): 831–834.
  5. Westgard JO, Qualm EF, Barry PL. Selection grids for planning quality control procedures. Clin Lab Science 1990;3:271-278.
  6. Visit www.awesome-numbers.org for more references,studies, publications, links and more
Please note.  I am not being critical of any individual person, institution or organization.  I want to express my most sincere respect, admiration and appreciation to  Dr. Westgard and so many others.  This process does not contradict  today's  published  recommended, principles and processes; it just implements them differently. 
Last modified: Tuesday, 4 February 2014, 11:40 AM