Statistical uncertainties in asbestos-laboratory classification through the RICE scheme

Our plain-English summary of Statistical uncertainties in asbestos-laboratory classification through the RICE scheme, by T.L. Ogden (The Annals of Occupational Hygiene, 1984).

The study

• Author: T.L. Ogden
• Journal: The Annals of Occupational Hygiene
• Published: 1984;28(4):449-52
• Cited by: 2 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

In plain English

This short communication from the UK Health and Safety Executive analysed the statistics of the RICE scheme (Regular Interlaboratory Counting Exchanges). The scheme sorted laboratories counting asbestos fibres on membrane filters into three categories using counts on 32 slides. Category 1 required at least 24 results between 0.65 and 2.6 times the 'Magiscan' image analyser count. Category 2 laboratories failed this standard but had 24 or more counts between 0.45 and 3.8 times Magiscan. The author modelled pass chances with the binomial distribution, which numerical simulations closely matched. Assuming RICE slides represented a laboratory's routine samples, a laboratory with more than about 85% of routine slides within category 1 limits had a greater than 95% chance of passing a 32-slide trial. Under the same assumption, one with less than about 60% within the limits had less than a 5% chance. The author concluded that doubling the trial to 64 slides would not give a proportionate improvement in selectivity. Making RICE slides more representative of routine samples was likely to be more productive. No workers or health outcomes were studied.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Short communication: statistical (binomial probability) analysis of a laboratory proficiency-testing scheme, checked by numerical simulations
• Who was studied: Laboratories evaluating asbestos on membrane filters in the UK RICE scheme, each classified on counts from 32 slides (simulated trials only; no actual laboratory results reported)
• Exposure: Not a health exposure study. The subject is asbestos fibre counts on membrane filters compared with the 'Magiscan' image analyser using the Manchester Asbestos Programme. Category 1 required 24 or more results between 0.65 and 2.6 times Magiscan. Category 2 covered laboratories that failed category 1 but had 24 or more counts between 0.45 and 3.8 times Magiscan. No fibre types are named.
• Main results:
  • Simulated RICE trials agreed very well with the binomial model's predictions
  • If RICE slides are representative of a laboratory's routine samples, a laboratory with more than about 85% of its routine slides within category 1 limits would have a greater than 95% chance of passing a 32-slide trial
  • Under the same assumption, a laboratory with less than about 60% of its results within the category limits would have less than a 5% probability of passing
  • Selectivity ratio (the q needed for a 95% pass chance divided by the q needed for a 5% pass chance) was about 1.7 for 16 slides, about 1.42 for 32 slides and about 1.28 for 64 slides
• The authors' conclusion: The author concluded that doubling to 64 slides would not give a proportionate improvement in selectivity. Improving how representative RICE slides are of routine samples was likely to be more productive than enlarging trials.
• Limitations the authors state:
  • The calculated probabilities assume RICE slides are representative of a laboratory's routine samples. The author considered it likely that they often are not, and that a complete match is impossible.

Related studies

• A proposal for harmonising laboratory performance assessment criteria in national asbestos fibre counting schemes — M.C. Arroyo, J.M. Rojo (2001)
• Within-laboratory quality control of asbestos counting — T.L. Ogden, T. Shenton-Taylor, et al. (1986)
• A PC-based quality control (QC) program for an industrial hygiene laboratory — D.K. Verma, K. Ferguson, et al. (2006)
• Reducing environmental risk associated with laboratory decommissioning and property transfer — R. Dufault, E. Abelquist, et al. (2000)
• Evaluation of airborne asbestos exposure from routine handling of asbestos-containing wire gauze pads in the research laboratory — E. Garcia, D. Newfang, et al. (2018)
• A comparison of fibre counting across three European national proficiency testing schemes — A.D. Jones, M.C. Arroyo, et al. (2005)

Browse more research in The Annals of Occupational Hygiene, or return to medical research.