Article 12 part 2 ... Continued ....

Read Time 2 Minutes

Deming: Improving Quality through Testing

Typically Statistical Quality Control talks about sampling plans for AQL – Acceptable Quality Limit. Such plans include Dodge-Romig Sampling Plans and Mil Standard Sampling Plans. These plans provide guidance for sampling to “Accept” or “Reject” a shipment of parts.

In his book, Out of the Crisis [ref. 6], Dr. Deming, talks about a different approach for product testing using economic optimization. Simply stated,

If p = average fraction defective in incoming lot of parts

k1 = cost to inspect one part

k2 = cost to dismantle, repair, reassemble and test an assembly that fails because a damaged part was put into production

Further suppose that the process is in statistical control with defect rate p,

Then the rules for the average minimum total cost are [ref. Deming, page 411]

Case 1: p < k1/k2  No Inspection/Testing

Case 2: p > k1/k2  100% Inspection/Testing

An Example: Suppose a customer buys 100 drums of toluene at Rs 10,000 per drum from a refinery that uses statistical process control and certifies that their current process produces an average of 1% defective drums. Should the customer test each drum?

(a) What if the cost of testing is Rs 200/test?

(b) What if the cost is Rs 50/test?

Answer:

(a) No Testing.

(b) 100% Testing.

Deming’s economic optimization of the testing costs provides an intelligent answer to this problem.

A key assumption is “…refinery that uses statistical process control and certifies that their current process produces an average of 1% defectives.”

Through an ISO process audit and documentation this statement can be verified. Both the Supplier and the Customer can reduce their testing costs. It is a win-win for both the parties.

Application to Pharma Industry:

Instead of using this economic testing optimization at product test level, we can extend the concept to industry level.

Earlier I provided some Bloomberg data. US FDA uses 12 inspectors; they plan to expand to 19.

We know that the current total cost of the “Repair” far exceeds US $500 MM.

So if k1 = 20 inspectors x US $200,000 per inspector per year = US $4 Million

k2 = US $500 Million

A limiting case is p=k1/k2 = 4/500 = 0.8%

This means Indian Pharma Industry could pay for 125 years of testing by 20 FDA inspectors!

Also note that the FDA will not be able to expand to 19 inspectors easily because the US Congress, backed by the majority Republican Party is putting enormous pressure to reduce the cost of the US Government.

So Indian Pharma Industry will have to help its customer, the US FDA.

This implies that it is better for the Indian Pharma Industry (the supplier) to unite, and offer its customer (the US FDA and drug industry) a win-win solution. “We will fund the FDA inspectors; you teach us the methods. We will implement them in the true spirit of the ISO and FDA standards.”

So crank up your Product Development engines... Let us speedup new product development and growth rates. And let the fun begin!

References: