Work by some labs ‘sloppy’

Press, 30 May 1989, Page 5

Work by some labs ‘sloppy’

By

NIGEL MALTHUS

Sloppy work by commercial chemistry laboratories in New Zealand has led to bad business, environmental and legal decisions, according to a Department of Scientific and Industrial Research chemist, Dr Max Robertson.

Dr Robertson, a Government analyst with the Chemistry Division in Christchurch, said he knew of some errors which had cost thousands of dollars.

Sloppy work was found in laboratories throughout the country, even some run by local authorities on which the public depended for proper testing of drinking water quality. “Some are very good value; some I wouldn’t give you tuppence for their results,” he said. Dr Robertson has been an assessor, and is now on the Registration Advisory Committee of the Testing Laboratory Registration Council (Telarc), a body set up by the Government in 1975 to monitor laboratory standards.

While there could never be absolute guarantees of accuracy, people needing important analyses done should always go to a laboratory registered as having met Telarc standards, he said. Dr Robertson said two common problem areas were “confidence levels,” a measure of the accuracy of testing, and “level of detection,” defining the sensitivity of the methods used.

Even some Telarc laboratories had been found not to calculate confidence levels correctly, while others did not even try, he said.

Dr Robertson would not name those involved, but said that in one recent case a particular material was required to have a chloride level of no more than 10 parts per million, or it would be rejected as prone to corrosion. A laboratory found 11 p.p.m. in a New

Zealand-made sample, so the material was bought overseas, at an extra cost (to the taxpayer) of more than $1 million, plus the loss of work locally.

Dr Robertson said it was then discovered that the testing method was accurate to plus-or-minus five p.p.m., at a confidence level of 95 per cent — meaning that it was as likely to have contained only six p.p.m., or as much as 16 p.p.m. “We (D.5.1.R.) believe that material actually did comply,” he said. The laboratory involved later applied for Telarc registration but voluntarily closed down rather than meet the cost of necessary improvement.

Dr Robertson said that another instance involved a trade union which asked a laboratory to measure the level of a particular toxin in a workplace. The laboratory, believing its method could detect one p.p.m., found nothing and declared it safe.

The real level of detection, however, was 10 p.p.m. — the same as the theoretical "L.D.50” dose.- The laboratory involved might not have found any toxin, even if there had been enough to kill half the workers, said Dr Robertson.

“A lot of people are fobbing off cheap analyses around the country,” he said. A proper analysis might cost twice as much as “a quick cheapie, often done by an unqualified person,” but was still cheap compared with the cost of having an export shipment rejected by the customer, he said.