A probability model for evaluating building contamination from an environmental event

Our plain-English summary of A probability model for evaluating building contamination from an environmental event, by R.C. Spicer, H.J. Gangloff (Journal of the Air & Waste Management Association, 2000).

The study

• Authors: R.C. Spicer, H.J. Gangloff
• Journal: Journal of the Air & Waste Management Association
• Published: 2000;50(9):1637-46
• Cited by: 3 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Asbestos dust and bioaerosol sampling data from suspected contaminated zones in buildings allowed development of an environmental data evaluation protocol based on the differences in frequency of detection of a target contaminant between zones of comparison. Under the assumption that the two test zones of comparison are similar, application of population proportion probability calculates the significance of observed differences in contaminant levels. This was used to determine whether levels of asbestos dust contamination detected after a fire were likely the result of smoke-borne contamination, or were caused by pre-existing/background conditions. Bioaerosol sampling from several sites was also used to develop the population proportion probability protocol. In this case, significant differences in indoor air contamination relative to the ambient conditions were identified that were consistent with the visual observations of contamination. Implicit in this type of probability analysis is a definition of "contamination" based on significant differences in contaminant levels relative to a control zone. Detection of a suspect contaminant can be assessed as to possible sources(s) as well as the contribution made by pre-existing (i.e., background) conditions, provided the test and control zones are subjected to the same sampling and analytical methods.

In plain English

This technical paper described a statistical method for judging whether a building zone is contaminated. It compares how often a contaminant exceeds a chosen level in a suspect zone and in a comparison zone. The authors applied it to more than 400 settled-dust samples, analyzed by transmission electron microscopy, from a multistory office building after a fire. The fire was on a floor that still had amosite spray-applied fireproofing, and chrysotile was present on some pipefittings and floor coverings. In this paper, p is the probability that zone 2 levels are greater than zone 1 levels, and 0.90 or higher was treated as significant. Floors considered fire-contaminated (278 samples) showed no significant difference in total asbestos from floors considered clean (108 samples). They had significantly less chrysotile (p = 0.028) and significantly more amosite (p = 1.00). Dust above ceilings and behind wall units had more of every asbestos type than surfaces below the ceiling. The authors concluded that the asbestos pattern was inconsistent with the fire and supported residual contamination that predated it. They also applied the method to airborne fungi at eight water-damaged sites.

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: Methodological (statistical) paper presenting a population proportion probability model, applied to existing asbestos dust sampling data from one building after a fire and to airborne fungal sampling data from eight sites
• Who was studied: Settled dust from a multistory office building (location and date not stated) after a fire on a mid-level floor: more than 400 dust samples grouped into zones for several comparisons. Also more than 700 culturable fungal spore air samples, split between indoors and outdoors, from eight water-damaged sites, July 1998 through August 1999
• Exposure: Asbestos in settled dust (total asbestos, chrysotile and amosite) after a fire in a building partly built with amosite spray-applied fireproofing, with chrysotile on some pipefittings in shafts and in floor coverings. Several floors had undergone asbestos abatement in the years before the fire.
• Main results:
  • Comparison of two adjacent zones in the allegedly contaminated area: 2 of 12 samples in zone 1 and 15 of 22 samples in zone 2 exceeded 330,000 asbestos structures per square centimeter (the combined median). The probability that total asbestos was greater in zone 2 than in zone 1 was 0.998, and the same was true for chrysotile.
  • In the same two-zone comparison, the probability that amosite levels were greater in zone 2 than in zone 1 was 0.001
  • Floors considered contaminated by the fire (zone 2, 278 samples) versus floors considered uncontaminated (zone 1, 108 samples): the text reports no significant difference in total asbestos (p = 0.209; Table 4 prints 0.167)
  • In that same comparison, chrysotile was significantly less in the floors considered contaminated (p = 0.028) and amosite was significantly higher there (p = 1.00)
  • Within the allegedly contaminated floors, dust above the acoustical tile ceiling and behind perimeter window ventilation units had significantly higher total asbestos, chrysotile and amosite than floors, desktops and shelves (p = 1.00, p = 1.00 and p = 0.998, respectively)
  • Ventilation ductwork on a roof setback (24 samples) had significantly higher total asbestos and chrysotile than the allegedly contaminated floors (278 samples), p = 1.00 for both, but amosite was not significantly elevated
  • For airborne fungi, at least one fungal type was significantly elevated indoors compared with outdoor air (p greater than 0.90) at each site
• The authors' conclusion: The authors concluded that the distribution of asbestos and its mineral types in the settled dust was highly erratic and inconsistent with the origin of the fire and the alleged transport mechanisms. They concluded that the analysis instead supported residual asbestos contamination from sources within the building that predated the fire.
• Limitations the authors state:
  • The method assumes that the two zones being compared are the same and were sampled and analyzed with identical methods
  • The method will not show a significant difference between two zones unless enough samples are taken in each zone
  • For fungi, the culture media are selective, so comparisons can only be made for species able to grow on the media used

Related studies

• Relative congener scaling of Polychlorinated dibenzo-p-dioxins and dibenzofurans to estimate building fire contributions in air, surface wipes, and dust samples — J.D. Pleil, M.N. Lorber (2007)
• Contamination of the air with mineral fibers following the explosive destruction of buildings and fire — J.A. Hoskins, R.C. Brown (1994)
• Physical characteristics and health effects of aerosols from collapsed buildings — S.H. Gavett (2006)
• Probability of cancer in lung nodules using sequential volumetric screening up to 12 months: the UKLS trial — M.W. Marcus, S.W. Duffy, et al. (2019)
• Composition of heavy metals and airborne fibers in the indoor environment of a building during renovation — M.T. Latif, N.H. Baharudin, et al. (2011)
• External validation and recalibration of the Brock model to predict probability of cancer in pulmonary nodules using NLST data — A. Winter, D.R. Aberle, et al. (2019)

Browse more research in Journal of the Air & Waste Management Association, or return to medical research.