"These are the times of one standard"
How important is it for a laboratory to go for accreditation?
Accreditation is extremely essential as it is a question of standardisation. Standards for labs were set as late as December 1999. Before, there were only guidelines. The present iso 17025 addresses technical competency; training requirements; equipment maintenance; calibration traceability and uncertainty in measurement. This is a global standard for laboratory performance.
Politicians like to claim there's no money for state-of-the-art laboratories because the poor must be fed, first
It is a question of two paradigms. On the one hand, we are addressing the export market. On the other, we are not able to address local requirements. The local economy is unable to afford the quality cost. The day our per capita income goes up, we will also be able to afford quality in each and every sector. Today, whatever expertise we have in select pockets serves the exports market. But things have to change in future. Globalisation is happening. These are the times of one standard, one test report: global acceptance.
Laboratories test food products, but industrialised countries reject exports found laced with contaminants. Is there a way to ensure labs test products properly, and exports not suffer?
Thanks to the accreditation system, testing activities are being standardised in the country. But there are problems with sampling. There is an iso standard for sampling but how many follow it?
What is the general sampling procedure? Who collects the sample -- industry or the laboratory?
Sampling is a contractual requirement. A lab does what it is told to do. If the lab is told to collect samples, it does. In such cases it mentions the day the sample was drawn, and by which method. In most cases, sampling is done by the consumer (industry) because of affordability. Testing costs a fortune -- testing one water sample for about 32 pesticides costs about Rs 10,500, but can come down to Rs 2000 if only one pesticide has to be tested -- and if you add sampling to it the cost doubles. And this enhances the cost of the product.
What practice do you follow in Vimta labs?
Our certificate is backed by a standard clause: "samples not drawn by us unless otherwise stated". And this is also a requirement under the National Accreditation Board for Testing and Calibration Laboratories (nabl) guidelines.
Can you differentiate between a sample of raw water and treated/distilled water? Are there chances of industry sending manipulated samples?
For us it is a water sample. The customer labels the sample. We test the sample for specified parameters and present a report. Manipulation does not generally happen, but there can be exceptions to anything. As a contract lab, we are neutral; we cannot doubt. Trust is important. Just as the customer has to trust the lab, so the lab must trust the customer.
Recently Pepsi put out results of its product tested, as claimed, by Vimta. As per the results, in some cases raw water does not have pesticide residue, but the treated water does. Why this discrepancy?
A brief comment. The matter is sub judice, and let's respect our judicial system. Raw water testing is done seasonally and there is no batch number. Treated water sample is usually drawn on the day of production and has a batch number. The two samples cannot be correlated, because the raw water sample need not be the same as the treated water sample. They are often different, so no point comparing them.
Post-soft drink controversy, many state governments are testing soft drinks without following any standardised procedure. How do we ensure standardisation happens?
Standardisation of procedures has happened at global level as late as July 2003 for residue analysis, when the Codex Alimentarius Commission released its guidelines. So talking about it at the national level is premature.
Standardisation is very complex; it involves a lot of science, technology and -- now -- a little politics. It requires a proficiency testing, a test conducted by a capable organisation to evaluate the technical competency of a lab. A test sample having homogeneity, stability and integrity is sent to different labs. Then a statistical analysis is done by which inter-lab and intra-lab variations are quantified. This test will tell you what is a median and which tests are possibly incorrect.
Unless a proficiency testing is carried out, things will not be clear. Confusions will remain. We are on the right track, but with a little less speed. We need to pick up now.