ERMI, Research Data or Commercial Product?

[This petition initiated by Dr. Wei Tang on October 14, 2007 is his personal opinion and it is NOT the position of any organization or business that he is affiliated or working with.]

Standardization of a sampling and analytical method requires years of collaboration from experts in related disciplines and must be based on industry-recognized protocols. It's not a government agency's role to write a “standardized” method based on a few studies and then ask an entire industry to recognize and accept it.

When conducting indoor mold growth investigations, one of the industry-recognized methods is to compare a suspected problem area to a non-problem reference area which represents a normal background level of fungal biomass. ERMI, developed by EPA, uses a reference internal to the sample itself based upon ten fungal species (Group II). It is a new and unproven fungal ecology model developed for indoor mold growth investigation and was developed with little involvement of mycologists. As with any new theory or method, it must be proven or validated with extensive research data over time.

Following are some of the important issues surrounding ERMI that still need to be addressed:

(1) Basic Fundamentals:
The proponents of ERMI ignore many of the established fundamental practices used in performing indoor environmental investigations. For example, it is widely recognized that sampling is only one aspect of a comprehensive building investigation. ERMI uses only one sample and it is being marketed by some labs to less-than-qualified mold inspectors or homeowners as a substitute for a properly performed field investigation. Water intrusion and dampness in a building is the root cause of microbial growth problems which have the potential to impact the health of occupants. These moisture issues need to be corrected. ERMI does not address or locate the cause of microbial growth in buildings. All areas of significant mold growth need to be identified and removed. ERMI presently only places emphasis on 26 Group I species out of hundreds of possible fungal species that can be found in indoor environments. Even more inappropriate is the fact that any significant mold contamination in the building from any of the 10 species in Group II will lower the ERMI score. For this reason, ERMI cannot be used as a Post-Remediation Verification (PRV) tool. In fact, current “Mold-Specific Quantitative PCR” (MSQPCR) cannot responsibly be used for PRV because it does not detect all groups of fungal spores.

(2) Flawed ERMI Calculation Formula and Data Variation:
ERMI is calculated by using a unique unproven formula, which is the “sum of the logs from Group I concentrations” from which the “sum of the logs from Group II concentrations” is subtracted.

ERMI = (sum of the logs from Group I concentrations) - (sum of the logs from Group II concentrations)

since:
Log(A) + Log(B) + Log(C) = Log (A x B x C)
Log(D) - Log(E) = Log (D/E)

therefore:
ERMI = Log(all Group I concentrations multiplied together/all Group II concentrations multiplied together)

As a result, ERMI represents fungal ecology as numeric data that has not been fully studied and it produces questionable results. For example, assume two samples, Sample 1 and 2, have identical Group II background. Sample 1 has 10 species detected in Group I with 10 spores each. Sample 2 has 1 species detected in Group I with 10,000,000,000 spores. Total concentration and species diversity have always been important parts of traditional data interpretation methods. Sample 2 with 10,000,000,000 spores from one single species represents a much more significant mold growth problem than Sample 1 which contains 100 spores from 10 different and diverse species. The sum of Group I spore concentration is 100,000,000 times lower than that of Sample 2, however, ERMI will give the exact same score for both samples.

Also as stated in (1), if any significant mold contamination in the building is from any of the 10 species in Group II, the ERMI score will be lowered instead of increased. It is clear that ERMI does not reflect “total mold burden”. This flaw in the calculation formula also contributes to its huge variation (or random error) of minus 3 to plus 3 (+/- one standard deviation) in a log scale. A wide range of 6 index difference translates into a lot of uncertainty. There is also a good possibility that the actual ERMI will fall outside of the range of +/- 3 of reported ERMI (+/- one standard deviation).

(3) Method Development:
Some of those chosen ten internal reference species (Group II) are considered by many mycologists to be species commonly found in water damaged buildings. These species are not appropriate choices to be use as background contamination or as an internal reference standard. Even when using appropriate internal references, a single carpet dust sample alone is hardly considered by most consultants to be representative of an entire house.

It has never been properly demonstrated that ERMI correlates to an independent set of reliable data representing “mold burden”. ERMI quartiles established based on HUD’s (2006) American Healthy Home Survey are merely a ranked order of ERMI values themselves, and they were not correlated with an independent set of data that establishes or defines “mold burden”. The definition of "mold burden" is based upon the 26 Group I mold species found in 11 carpet or floor dust samples from 6 Cleveland homes in which infants developed pulmonary hemorrhage. They were selected by comparing to 26 samples collected from reference homes in the same area and are presumed to be representative of universal "mold burden" and its associated health effects in occupants of all ages in all residential buildings throughout the country. It also assumes that any of those 10 Group II species cannot be a problematic mold source in a building. These assumptions have never been properly tested or proven and are not realistic.

(4) Method Validation:
MSQPCR developed by EPA can be a good tool for IEQ investigation in some specific applications. However, ERMI, which utilizes this valuable technology but applies an unproven method of data interpretation, has not been properly validated using existing industry-recognized methods for indoor mold investigation in extensive side-by-side comparison studies. It does not even locate the water problems or the mold sources, which are what consultants and homeowners need to know to create a more healthful environment. Its claim of superiority to other investigation methods is unproven and has no basis.