Ok.  Any­one who knows me, has read this blog for a while or has seen me speak knows I love to use movies as trans­fer­able con­cepts to try to explain sup­pos­edly com­pli­cated sci­en­tific concepts.

What can the 1985 clas­sic movie “Fletch” teach us about method validation?

Just like when “Gordo” said in the above clip “Awww, come on guys, it’s so sim­ple. Maybe you need a refresher course. [leans arm on hot engine part] Hey! It’s all ball bear­ings nowadays..”,

We say to the lab peo­ple “Awww, come on guys, it’s so sim­ple. Maybe you need a refresher course… Hey! It’s all val­i­da­tion nowadays.”

Method val­i­da­tion has received con­sid­er­able atten­tion in the lit­er­a­ture and from indus­trial com­mit­tees and reg­u­la­tory agencies.

• The U.S. FDA CGMP (1) request in sec­tion 211.165 (e) meth­ods to be val­i­dated: The accu­racy, sen­si­tiv­ity, speci­ficity, and repro­ducibil­ity of test meth­ods employed by the firm shall be estab­lished and doc­u­mented. Such val­i­da­tion and doc­u­men­ta­tion may be accom­plished in accor­dance with Sec. 211.194(a). These require­ments include a state­ment of each method used in test­ing the sam­ple to meet proper stan­dards of accu­racy and reli­a­bil­ity, as applied to the tested prod­uct. The U.S. FDA has also pro­posed an indus­try guid­ance for Ana­lyt­i­cal Pro­ce­dures and Meth­ods Val­i­da­tion (2).
• ISO/IEC 17025 includes a chap­ter on the val­i­da­tion of meth­ods (3) with a list of nine val­i­da­tion para­me­ters. The ICH (4) has devel­oped a con­sen­sus text on the val­i­da­tion of ana­lyt­i­cal pro­ce­dures. The doc­u­ment includes def­i­n­i­tions for eight val­i­da­tion char­ac­ter­is­tics. ICH also devel­oped a guid­ance with detailed method­ol­ogy (5).
• The U.S. EPA pre­pared a guid­ance for method’s devel­op­ment and val­i­da­tion for the Resource Con­ser­va­tion and Recov­ery Act (RCRA) (6). The AOAC, the EPA and other sci­en­tific orga­ni­za­tions pro­vide meth­ods that are val­i­dated through multi-laboratory studies.

Here is the whole point and my logic proof…

1.       I assume that there is a truly sci­en­tif­i­cally val­i­dated and robust method for the analy­sis under­taken in the foren­sic lab.  (To be sure of the use­ful­ness of their analy­sis, we want to exam­ine the lab’s val­i­da­tion stud­ies so as to not make this basic assump­tion.  Oth­er­wise, what’s the point…  With­out this basic prin­ci­ple sat­is­fied that there is a valid method, the lab should not be per­form­ing tests)

2.       Uni­ver­sally, if one fol­lows the val­i­dated method and only if one reli­giously fol­lows the val­i­dated method with no devi­a­tion, we get a valid result that is true (or more cor­rectly put as valid and as true a result as sci­en­tif­i­cally possible)

3.       The con­verse of that is not 100% true.  If we devi­ate from the val­i­dated method in some way or in any way, we MAY get a true result or we MAY not.  We don’t know with­out study­ing the impact of that devi­a­tion.  We do know that the devi­a­tion has caused the result to be not a val­i­dated result.  Not an invalid result, but rather a result that is not val­i­dated accord­ing to the valid method.

4.       As most of the val­i­da­tion stud­ies begin with “Take the filled 10mL grey tube top….”  This would con­note that the method was cre­ated using pre­cisely that a “filled 10mL grey tube top”.  It is the raw mate­r­ial, if you will.  It is like a con­di­tion precedent.

5.       If you do not have the con­di­tion prece­dent (i.e., the 10mL grey tube top), then you are miss­ing a part of the val­i­dated con­di­tions and you have by def­i­n­i­tion a devi­a­tion from the val­i­dated method.

6.       Ergo, you have a result that is not val­i­dated and may or may not be valid and may or may not be true.

To learn more about method val­i­da­tion, I would com­mend to you http://www.labcompliance.com/tutorial/methods/default.aspx

The tuto­r­ial is based upon the book Val­i­da­tion and Qual­i­fi­ca­tion in Ana­lyt­i­cal Lab­o­ra­to­ries, pub­lished by Infoma in 2007 which I highly rec­om­mend as the eas­i­est and best resource that I have come across that explains the impor­tance of method val­i­da­tion, the method of estab­lish­ing a val­i­dated method, the con­se­quences of method val­i­da­tion and the con­se­quences of devi­at­ing from the val­i­dated method.

DEFINITIONS USED IN THIS POST:

Invalid: proven to be not valid, not the same as “not val­i­dated” as “not val­i­dated” means that it may be valid or it may be invalid.

True: the value that char­ac­ter­izes a quan­tity per­fectly in the con­di­tions that exist when that quan­tity is con­sid­ered. It is an ideal value, which could be arrived at only if all causes of uncer­tainty (Type A error and Type B error) are eliminated.

Type A error: a method of eval­u­a­tion by sta­tis­ti­cal analy­sis of a series of observations.

Type B error: any­thing that is not Type A error; a method of eval­u­a­tion by any means other than sta­tis­ti­cal analy­sis of a series of observations.

Valid: doc­u­mented proof that the process under­taken is suit­able for its intended use and achieves the intended reported result cor­rectly and uniquely as free from pos­si­ble from issues of pre­ci­sion and bias as possible.