J. Neville, PK-Jarvis Microbiological LLC, Novi, MI; P. Epstein, Clayton Group Services, Novi, MI; S. Rupkey, Clayton Group Services, Chicago, IL; M. Hodgson, Clayton Group Services, Edison, NJ.
Bioaerosol sampling is performed to estimate microbial air concentrations in suspected building areas. Since safe bioaerosol exposure levels have not been established yet, indoor levels are compared with outdoor levels as the background control.
Industrial hygienists (IH) typically want to test the null hypothesis “does the concentration and/or biodiversity of indoor bioaerosols equal that found outdoors.” Two types of errors can be made when testing the null hypothesis. The Type I error (false positive–α), will occur when the null hypothesis is rejected when in reality it should not. This may cause the IH to make a recommendation to remediate when it is not necessary. The Type II error (false negative–β), which is more destructive, will occur when the null hypothesis was not rejected when in reality it should have been. This may cause the IH to make a recommendation not to remediate when it is necessary.
Rates of Type II errors can be lowered if we increase the statistical power of sampling by increasing the number of samples. Statistical power (1-β) is the probability of rejecting the null hypothesis when in fact it is false and should be rejected. Like other areas of biological research, bioaerosol sampling strategies typically neglect use of statistical power analysis. Quite often, very few samples are taken because of client budget constraints. This increases the probability of making Type II errors.
In this preliminary study, we employ statistical power analysis on bioaerosol samples that were taken from various U.S. geographic locations. We will show the number of samples required to obtain an 80% probability that we are not making Type II errors. Data from spore trap and culture samples will be presented. Statistical power concepts as they apply to different bioaerosol sampling techniques will be discussed.
F. Wu, S. Huang, Aemtek Inc., Fremont, CA.
One of the challenges that mold investigators face today is the selection of primary culture medium for recovery of fungi. The current commercially available media are based on receipts developed for traditional applications in either mycology or plant pathology. Each medium favors certain fungi while it inevitably limits the growth of other fungi. For example, in order to capture both high water activity and osmophilic fungi, both malt extract agar (MEA) and Dichloran 18% (DG18) are used in fungal sampling protocols, but neither medium is ideal. In addition, several other media are often used for growth and identification of fungi of special interest. Recently, alternatives to MEA and DG18, gypsum amended MEA (GMEA), and salt MEA (SMEA), respectively, were suggested as having added benefit of supporting Stachybotrys growth. This study evaluated performances of MEA, GMEA, SMEA, and DG18 in fungal recovery. Six common indoor fungi were employed, including species of Aspergillus, Cladosporium, Scopulariopsis, Penicillium, Stachybotrys, and Ulocladium. Conidium suspensions were used as inoculates to measure germination and growth rate. The results showed that Stachybotrys and Ulocladium grew faster on GMEA than they did on MEA, while the other tested fungi showed no significant difference on MEA and GMEA. Performances of SMEA and DG18 were generally similar, except that Ulocladium grew faster on SMEA than on DG18. However, all tested fungi seemed to have smaller biomass on GMEA and SMEA than on their respective alternatives.
C. Meyer, Aerotech Laboratories Inc., Phoenix, AZ.
The absence of a standard method for the enumeration of fungal spores in indoor air has led to a diverse range of analytical approaches and statistical analysis of data. During the analysis of spore trap cassettes, laboratories typically analyze anywhere from 15 to 100% of the sample. Depending on the nature of the sample, the percentage of the sample analyzed can have a significant effect on the final results. In outdoor air samples where the fungal spores are relatively homogeneous and free-floating, our data shows relatively little influence. In sharp contrast, air samples from indoor environments often contain aggregate fungal conidia and other multi-spore structures. Real world experience with over 1 million samples analyzed indicates that these fungal structures are deposited on spore traps in a heterogeneous pattern. Fifty-five Air-O-Cell cassette samples were analyzed via both the 15% and the 100% counting method and the results compared. When compared to reading 100% of the sample, the 15% technique failed to detect Stachybotrys in 9% of the samples. Eighteen percent of the samples were skewed due to heterogeneous deposition of spores. Twenty-five percent of the samples missed 5 or more genera and 9% of the samples failed to detect spores in concentrations less than or equal to 227 m3.
R. Fahrion, Huntsman Chemical, Port Neches, TX.
Fungal proliferation indoors was a very expensive and inconvenient problem in Texas during the summer of 2001. During that time about 1800 Air-O-Cell samples were taken from problematic building constructions along the Texas coast. The objectives of this study were to characterize fungal bioaerosol concentrations among several different populations and answer questions regarding the experienced symptoms. From the entire population, 47.70% reported a health related symptom in conjunction with indoor mold growth. Sinus congestion and headache were the most frequently reported symptoms. Aerosolized indoor fungal loads ranged from 40 s/m3 to 267,213 s/m3 (mean = 4974 s/m3). Total load, Aspergillus/Penicillium load, and Stachybotrys loads were analyzed using odds ratios, ANOVA, and t-test methods. The statistics obtained from the symptom reporting population indicated mean concentration values for experienced symptoms were 9146.63 s/m3 for total fungal load, 8994.60 s/m3 for Aspergillus/Penicillium load, and 218.95 s/m3 for Stachybotrys bioaerosol load. The threshold limit value for sample size N = 239 was 7245 s/m3.
D. Robertson, R. Billups, Air Quality Sciences Inc., Marietta, GA.
Stachybotrys chartarum has been identified as an important indicator organism in water damaged buildings. Although the role of this fungus in indoor air health complaints in indoor environments remains controversial, it is an excellent indicator of high water activity. Recovery of this organism on cellulose agar has been standard practice in the industry for several years. Our studies indicate that Stachybotrys chartarum is difficult to recover on either 2% malt extract agar (MEA) or cellulose agar, as it does not compete well with other, more aggressive fungi that are frequently recovered from water damaged buildings. Further, when S. chartarum is recovered, the actual colony forming units (CFU)/gram or CFU/m3 are difficult to determine accurately. Stachybotrys selective agar (SSA) is a selective media with antifungal and antibacterial agents that inhibit the growth of other competing fungi and bacteria. In this study, dust samples submitted for routine analysis were dilution plated on 2% MEA, dichloran glycerol agar (DG-18), cellulose agar, and SSA. S. chartarum colonies were rarely recovered on the 2% MEA or DG-18 agars. Minimal recovery was noted on cellulose agar but significant levels of S. chartarum colonies were recovered on SSA from the same samples. Characteristic colonies were visible at 3 days of growth and identification could be verified on days 5–7 of incubation at 25ºC. Recovery of S. chartarum on SSA is more frequent than on cellulose agar, as it is selective for this organism, and provides counts that are more accurate due to discrete colony morphology and inhibited competition. Further, SSA has a more rapid turnaround time relative to general isolation agars and is cost effective relative to non-culture analyses specific for S. chartarum.
D. Daugherty, T. Bowie, G. Hoch, E. Lu, ENVIRON, Emeryville, CA.
This paper discusses the difficulties in analyzing fungal sampling data from limited samples during an initial fungal investigation. According to the American Conference of Governmental Industrial Hygienists’ (ACGIH’s) guidance Bioaerosols: Assessment and Control, sample size determines the probability of detecting a significant difference in results from fungal sampling studies. To increase the probability that an investigator can determine a significant difference and limit the possibility of a false-negative or false-positive result, field blanks and duplicate samples would be required. However, in the practical application of fungal sampling at commercial and residential sites, an investigator is rarely provided with the time or resources to conduct a sample-intensive study that would adequately address all the areas of sampling uncertainty. This paper describes the results of field studies where, by necessity, only a limited number of samples could be obtained. The objective of these limited studies was to provide initial information on whether further, more in-depth studies should be performed at each site. This paper discusses variations in sampling results for three methods commonly employed in initial fungal investigations: nonviable fungal air sampling, viable fungal carpet dust sampling, and nonviable fungal wall space sampling. The evaluation of nonviable fungal air sampling focuses on results from duplicate samples and outdoor air samples, which indicate potential sampling variations up to a factor of two to three. For viable fungal carpet dust sampling, results from field studies indicate that fungi counts for low weight samples are more variable and may appear elevated due to the small sample size. Results from nonviable fungal wall space sampling at control and potentially impacted areas are compared to results from visual inspections during remediation activities of the sampled areas. Implications of the observed sampling variations for limited fungal sampling investigations and the lessons learned from these observations are discussed.
R. Rottersman, G. Crawford, M. Simmons, B. Caddick, Boelter & Yates Inc., Park Ridge, IL.
Mold growth in carpet is often not visually apparent. Dust extract samples are routinely collected to evaluate fungal burden. Do the test results suggest fungal growth or simply background levels? Interpreting the test results can be challenging. Results are usually reported as colony forming units per gram (CFU/g). CFU/ft2 and grams of dust per square foot (g/ft2) are also important in quantifying fungal burden and soiling. There are some guidelines available to help interpret samples expressed in CFU/g, however, there is little information for interpretation in CFU/ft2 or g/ft2. A database of values is needed to establish ranges encountered in various settings. Variables such as the type and history of the building can influence results and interpretations. Forty-four dust samples were collected from carpet in 11 school buildings. Samples were collected with a pump and MCE filter cassettes with a collection tube. The tube was passed over the surface three times. Samples were taken from areas where heavy foot traffic would not be expected. The median concentrations in schools on malt extract agar were 203,486 CFU/g, 11,970 CFU/ft2, and 0.04 g/ft2. The same methodology was applied to commercial office buildings where median concentrations were 66,176 CFU/g, 4,060 CFU/ft2, and 0.05 g/ft2. The average fungal burden in school carpet is higher than commercial office buildings even though the dust loading was similar. Review of the data also indicated that results expressed as CFU/ft2 were almost always less than CFU/g except in cases where there was a history of water damage or mold growth in the carpet. A number of data sets from types of buildings are included in the study. Development and review of this type of data would be useful to develop guidelines to aid in data interpretation and more accurately represent the fungal burden of carpet.
Posted May 30, 2004