Originally Posted By: Caoimh?n P. Connell
This post was automatically imported from our archived forum.
Hello Mr. Plummer:
Actually, the correct spelling "mould" is becoming preferred even south of the border. For me, it is easy, as I spent time as a youth growing up in Ireland where we learned the King's English.
On to DQOs...
In 17 years as a forensic industrial hygienist, and having performed hundreds of mould assessments, and collected thousands of samples, I would say that none (without exception) of the mould sampling and analysis I have encountered by Home Inspectors were collected, analyzed and/or interpreted correctly. In the overwhelming vast majority of cases, the information the HI provided to the client was simply flat-wrong, and not even supported by the sampling and analysis results. Very frequently, the data presented by the HI actually contradicted the conclusions and recommendations of the HI.
In my experience, in the realm of microbial issues, poor/inappropriate sampling and/or analysis is the primary root of the client?s bad information, and the poor choices as a result of that bad information, and the root of the law suit that follows.
Remembering that Home Inspectors and Industrial Hygienists are different breeds. And that industrial hygienists don?t perform home inspections (even if we are performing inspections in a home). We specialize in understanding sampling theory and the statistical validity and limitations of sampling and analysis. When I purchased my last homes, I relied heavily on the competence of the professional Home Inspector to do what I am not competent to do. However, he knew me and stuck to HIS profession and didn?t get into
mine ? as a result, he didn?t even attempt to perform environmental sampling (radon, moulds, asbestos, pesticides, etc).
Now having said that - The establishment of data quality objectives is a QA/QC part of a larger decision making process; the result of analysis will be the pivotal point upon which those decisions are going to be made. The DQOs ensure, through their prescription, that a sufficient number of samples are collected from statistically representative locations in an acceptable manner by a recognized method. The DQOs will further specify that the samples are submitted to a laboratory that is capable of proficiently analyzing the samples to within a definable uncertainty, using valid methods (not all mould labs can do this). Lastly, the sample results are interpreted according to by ?PARCC? parameters (defined below) to find out if the DQOs were ultimately met. The parameters of the DQOs themselves can be very simple or very complex. DQOs are what make your data meaningful (and tenable in the event you get sued). Without DQOs, you
don?t have data, you have numbers or names on a lab report that possibly CANNOT be interpreted by anyone, since ?data? has no intrinsic meaning outside of
a priori decision criteria.
The DQO process is so entrenched in environmental sampling that it is discussed in great detail in many broad-spectrum governmental sampling protocol. For example, one of the ?bibles? of general environmental sampling is the US EPA SW846. The QA/QC chapters should be ESSENTIAL reading to any Home Inspector who wants to begin collecting samples for mould or other environmental contaminants.
The SW 846 describes DQOs thusly:
?
2.1 DATA QUALITY OBJECTIVES
Data quality objectives (DQOs) for the data collection activity describe the overall level of uncertainty that a decision-maker is willing to accept in results derived from environmental data. This uncertainty is used to specify the quality of the measurement data required, usually in terms of objectives for precision, bias, representativeness, comparability and completeness. The DQOs should be defined prior to the initiation of the field and laboratory work. The field and laboratory organizations performing the work should be aware of the DQOs so that their personnel may make informed decisions during the course of the project to attain those DQOs. More detailed information on DQOs is available from the U.S. EPA Quality Assurance Management Staff (QAMS) (see references 2 and 4).?
Now we are getting to the nub of the issue ? ??usually in terms of objectives for precision, bias, representativeness, comparability and completeness.? These are the ?PARCC? parameters I mentioned above:
Precision:
How reproducible are measurements?
Accuracy:
How close is the value to the true value?
Relevancy:
Do the data speak to the
a priori question being asked?
Comparability (Points of reference):
Can decisions be based by comparing the results against regulatory or nationally accepted guidelines or at least arbitrary guidelines that we established
before we sampled?
Completeness:
Have the DQOs been met?
Without picking on Mr. Goryl, let?s look at the example which started this discussion. A black smudge in the shower. ?Should I sample?? becomes, ?Why would I sample?? Is it to find out if the house has mould? No. We already know that ALL houses have mould, so that question has already been answered. Is it to find out if the house has Bacteria? No. We already know that ALL houses have Bacteria, so that question also has already been answered. Is it to find out if the mould is
Stachybotrys (the notorious, but largely harmless ?toxic black mould? of science fiction)? No. We already know that ALL houses in the U.S. contain
Stachybotrys and that there has not ever been even one confirmed case of
Stachybotrys induced mycotoxicosis from residential indoor air; so THAT question has already been answered. Should we sample to find out which species or genus it is? Well, unless we first answer the question ?Is it a mould or Bacteria? we don?t know if it HAS a genus or species. In the event that it does belong to a particular genus or species, which decisions will we make differently depending on the identification of genus or the species? For example, if it is
Alternaria, will I recommend anything different than if it belongs to the
Cladosporia or
Memnoniella? Am I a competent microbiologist or industrial hygienist whose background permits me to make that differentiation? OK, then let?s sample to see if it?s a mould or Bacteria. OK ? which and why? Well to see if the house has a mould problem. Answer: Since when does a 3 square inch patch of mould in a shower constitute a mould problem in a house? If I want to find out if the house has a mould problem, shouldn?t I design a sampling scheme that will reliably answer that question?
In short, collecting a sample of the smudge in the shower, in the absence of any DQOs, leaves you with a lab result that it ENTIRELY meaningless. Let?s say for example, Mr. Gorly did collect a swab of the smudge, and the lab report came back and stated ?2.6E3 CFUs
Cladosporium.? Jolly good. Now, what does that mean? Does that indicate the house has a mould problem? No. Does it indicate a health hazard exists in the house? No. Does it mean the house contains 2.6E3 CFUs
Cladosporium? No. Does it even mean that the
smudge contains 2.6E3 CFUs
Cladosporium?
NO! Does it even mean that I actually analyzed the smudge?
NO!!! For all we know, the 2.6E3 CFUs
Cladosporium could be the normal background count of the entire shower wall, and the smudge is a biofilm growing on that mould.)
Then, what does it mean? Answer: The laboratory report is completely and entirely meaningless because there were NO
a priori questions being asked and there were no DQOs established to determine if the lab results would properly answer the unasked question. Finally, we have NO information that the lab is even capable of properly analyzing for mould! (Any lab that tries to interpret the data should NEVER be used.)
Result ? If the lab report is used, there will be confusion, bad advice, poor conclusions, possible law suits, and a complete waste of the client?s money.
If you would like to see the practical establishment of DQOs, you can see how I wrote the sampling protocol for the State of Colorado in their new regulations for the assessment of methamphetamine contamination in homes that have been identified as meth-labs. Those sections are
Appendix A and
Attachment to Appendix A of the regulation found at
http://www.cdphe.state.co.us/op/regs/boardofhealth/101403methlabrules.pdf and the complete risk assessment model is at
http://www.cdphe.state.co.us/hm/methlabcleanuplevelsupport.pdf
I hope that sheds some light on mould sampling. By the way, I am on the ASTM International committee that has been tasked with writing the new Standard Sampling and Analysis protocols for moulds. My peers are not unlike me, and these are the thought processes that will go into the standard.
But then, those are just my opinions, and I think rain is wet; so what do I know, eh?
Cheers,
Caoimh?n P. Connell
Forensic Industrial Hygienist
www.forensic-applications.com
(Having just learned that I can edit these posts, I have erased my old website and added our new web address - that is the only edit made - Cheers, Oct. 29, 2005)
