﻿ Data Reduction Method Explained

# Data Reduction Method Explained

LGC Maine Standards uses a data reduction process in which target values are calculated, limits are calculated around these target values and then the actual recovered values are compared to the targets.  The method used is simple linear regression, which means a line that fits the equation y = mx + b, where m is the ‘slope’ and b is the ‘y-intercept’, is used.  The statistical method used to calculate this line is called ‘least squares fit’.

## Target Value Calculation

Many methods can be used to calculate target values and two examples are provided in each VALIDATE® product insert.  This explanation outlines the method used when data is processed at LGC Maine Standards or when a laboratory uses our free of charge MSDRx® software.

VALIDATE® products are manufactured to provide an equal distance between Levels 1 to 5.  Therefore, actual recovered values can be used to calculate target values.

First, simple linear regression, ‘least squares fit’, is performed on two or three consecutive levels of recovered values to calculate the target values.  Appropriate equal distance values are input as the ‘x’ values and recovered values for the levels are used as ‘y’ values to calculate a regression equation.

Next, using the equation calculated above, equal distance values of 1 to 5 are entered into the equation as ‘x’ values.  The resulting ‘y’ values are the target values.

## Limits

After calculating the target values, the next step is to calculate the ‘Limits’.  The Limits are based on the Clinical Laboratory Improvement Act of 1988 (CLIA ’88) analytical limits, if defined, and/or other published industry-accepted limits for each specific analyte.  Fifty percent of the analytical limits is used as the allowable error for linearity.

The allowable error for linearity is subtracted from the target value to establish the lower limit.  The upper limit is calculated by adding the allowable error for linearity to the target value.

The CLIA ’88 ‘criteria for acceptable performance’ can be found on the internet in two locations:

For routine chemistry:

http://www.gpo.gov/fdsys/pkg/CFR-2003-title42-vol3/xml/CFR-2003-title42-vol3-part493.xml#seqnum493.931

For toxicology:

http://www.gpo.gov/fdsys/pkg/CFR-2003-title42-vol3/xml/CFR-2003-title42-vol3-part493.xml#seqnum493.937

In cases where CLIA ’88 has not defined limits of performance for a test procedure, limits are suggested that coincide with performance characteristics of a similar chemistry.

For example, GGT does not have CLIA limits defined.  The GGT chemistry has performance characteristics similar to that of AST.  Therefore, the performance limits for AST are applied to GGT.  