Development of the Regular Interlaboratory Counting Exchanges (RICE) scheme to include visual reference counts and samples from asbestos clearance operations

Our plain-English summary of Development of the Regular Interlaboratory Counting Exchanges (RICE) scheme to include visual reference counts and samples from asbestos clearance operations, by P.W. Brown, N.P. Crawford, A.D. Jones, et al (The Annals of Occupational Hygiene, 1994).

The study

• Authors: P.W. Brown, N.P. Crawford, A.D. Jones, B.G. Miller, W.M. Maclaren
• Journal: The Annals of Occupational Hygiene
• Published: 1994;38(5):687-703
• Cited by: 4 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

The RICE proficiency testing scheme for asbestos fibre counting operated until 1992 with industrial samples and with reference counts produced by an image analyser (Magiscan). The visual counts produced in RICE by the participating laboratories have now been used to develop a reliable reference count for these industrial samples. Concurrently with the routine participation in RICE, laboratories also counted samples from asbestos clearance operations; these counts were used to develop suitable reference counts and performance limits for this type of sample. The width of this satisfactory performance interval is proportional to the natural variation in the data on the samples and is therefore a fixed interval on the log scale (of sample density) for high density samples and a fixed interval on the square-root scale for low density samples (where counting normally stops after a fixed number of fields). These developments led to the introduction of visual reference counts and clearance samples into RICE during 1992.

In plain English

This paper described changes to RICE, the UK proficiency testing scheme for laboratories counting airborne asbestos fibres. Until 1992, reference counts came from an image analyser (Magiscan), and slides came almost exclusively from asbestos-industry workplace monitoring. The authors reanalysed almost 20,000 laboratory counts on 501 industrial slides from RICE Rounds 10 to 16. They also ran four supplementary exchanges (1988 to 1990) in which 295 laboratories counted 621 low-density slides: 503 clearance samples and 118 blanks. They chose the median of at least 15 visual counts as each slide's reference count. 461 industrial and 438 low-density slides met this requirement. For densities above 63.7 fibres per square millimetre, the new limits for satisfactory counting were 0.50 to 2.00 times the reference. Below that density, the limits were fixed distances on a square-root scale, except that below 5.5 fibres per square millimetre the lower limit was set to zero. Recalculated ratings mostly stayed the same as under Magiscan and appeared slightly better overall. The authors concluded that clearance samples and visual reference counts could be introduced; both were introduced in April 1992.

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: Methods development study for a laboratory proficiency testing scheme: reanalysis of existing interlaboratory asbestos fibre count data plus supplementary slide exchanges of low-density samples
• Who was studied: UK laboratories taking part in RICE. Industrial data: almost 20,000 counts on 501 slides from RICE Rounds 10 to 16. Supplementary exchanges in 1988 to 1990: 10,537 counts on 621 slides (503 clearance samples and 118 blanks) by 295 RICE laboratories. Counts by laboratories judged less than satisfactory in routine RICE were omitted.
• Exposure: Airborne asbestos fibre samples on membrane filters, counted by phase-contrast optical microscopy. Before 1992, industrial samples came almost exclusively from workplace monitoring in asbestos textiles, asbestos cement and brake lining manufacture, generally 100 to 600 fibres per square millimetre. Clearance samples and blanks were of lower density, mostly 0 to 90 fibres per square millimetre.
• Main results:
  • For some industrial slides, laboratories' counts lay systematically away from the Magiscan reference count. The authors suggested this meant the reference count for those slides was biased.
  • The median of laboratory counts was chosen as the reference count, trusted only when based on at least 15 counts. Of the 501 industrial slides, 461 had at least 15 counts.
  • For industrial samples, the new outer performance limits were 0.50 to 2.00 times the visual reference count and the new inner limits were 0.65 to 1.55 times. These replaced 0.55 to 2.20 times the Magiscan count.
  • Ratings for Rounds 10 to 16 were recalculated, 2191 ratings in total. Under the Magiscan reference, 1536 (70%) were rating 1, 516 (24%) rating 2 and 139 (6%) rating 3. Under the median visual reference, 1741 (79%) were rating 1, 355 (16%) rating 2 and 95 (4%) rating 3. Most ratings were the same under both systems.
  • Laboratories' counts on the low-density supplementary samples were generally of a high standard, comparable with routine RICE results, even though no feedback was given until all four exchanges were finished.
  • Of the 621 low-density slides, 438 (326 clearance samples and 112 blanks) had been counted at least 15 times and could be used as reference samples.
  • For clearance samples, the spread of counts was roughly constant on a square-root scale. For industrial samples, it was roughly constant on a logarithmic scale.
  • The cross-over density between the two kinds of limit was 63.7 fibres per square millimetre. Below it, the outer limits lay 2.34 below and 3.30 above the reference on the square-root scale.
  • For reference densities below 5.5 fibres per square millimetre, the formula gave a negative lower limit, so the lower limit was set to zero.
  • In the supplementary exchanges, 957 ratings were calculated: 876 (92%) were rating 1, 62 (6%) rating 2 and 19 (2%) rating 3.
• The authors' conclusion: The authors concluded that low-density (clearance) samples and the visual reference system could be introduced into RICE and used alongside industrial samples. They reported that both changes were introduced in April 1992, in Round 23.
• Limitations the authors state:
  • The theory used to decide how many counts a reliable median needs strictly applies only to normally distributed data.
  • The lower share of rating 3 in the supplementary exchanges should be interpreted with care, because the participating laboratories were not exactly the same as those in RICE.
  • Industrial samples showed more outlying counts, but this may reflect their larger number of counts rather than the type of sample.
  • It was not yet known whether having to count samples across a much wider range of densities would affect laboratories' results.

Related studies

• Further observations on the short-term retention and clearance of asbestos by rats, using UICC reference samples — A.P. Middleton, S.T. Beckett, et al. (1979)
• A proposal for harmonising laboratory performance assessment criteria in national asbestos fibre counting schemes — M.C. Arroyo, J.M. Rojo (2001)
• A comparison of fibre counting across three European national proficiency testing schemes — A.D. Jones, M.C. Arroyo, et al. (2005)
• MesoTRAP: a feasibility study that includes a pilot clinical trial comparing video-assisted thoracoscopic partial pleurectomy decortication with indwelling pleural catheter in patients with trapped lung due to malignant pleural mesothelioma designed to address recruitment and randomisation uncertainties and sample size requirements for a phase III trial — C. Matthews, C. Freeman, et al. (2019)
• A simple dispenser for generating dust clouds from standard reference samples of asbestos — V. Timbrell, A.W. Hyett, et al. (1968)
• Statistical uncertainties in asbestos-laboratory classification through the RICE scheme — T.L. Ogden (1984)

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