BEC485 Control Charts and Process Capability

Percentage step yield equation

100 * (amount of product at end of step/ amount of product at beginning of step)

Define process capability analysis

Comparing the performance of a process against its specifications

Cp equation for normally distributed data with process mean centered on the midpoint of specs

Cp = (USL - LSL) / (6*standard deviation)

Cp equation for when the mean is not centered between spec

Cp = min [ (USL-mean)/(3*SDV) , (mean-LSL)/(3*SDV)

How do we tell if a process is capable based on the Cp value?

If Cp > 1, process is capable

21 CFR requires which yields to be recorded/calculated?

actual yield and percentage of theoretical yield

why do we use 6*standard deviation?

all data falls within +/- 3 SDV of the mean, so 6 standard deviations covers the entire range

When s a process considered capable

if virtually all of the possible variable values fall within the specification limits

Components of a control chart

points: measurement of the parameter at diff times Mean (average): center line drawn to represent mean Upper and Lower control limits: typically drawn +/- 3 SDV from center line Specifications: indicate value(s) that must be met for the parameter. Horizontal lines are drawn to indicate specs

Cp = 1

process is just meeting specifications

Cp<1 means

process variation exceeds specification, so the process is not capable

Cp > 1

process variation is less than the specification, so the process is capable

Actual yield defn

quantity actually produced at any appropriate phase of manufacture, processing, or packaging of a particular drug product

Percentage of theoretical yield defn

ratio of actual yield to the theoretical yield

If our data distribution is much tighter than our specs, what might the FDA say?

that we need to tighten the spec range unless there is good scientific evidence for the specification (like a tolerance limit in a patient for an impurity)

Theoretical yield defn

the quantity that would be produced, based on quantity of components used, in the absence of any loss or error in production

What is a control chart used for?

to determine if a process is in a state of statistical control