Procedures for Detection and Quantitation
Technical Workgroup
Federal Advisory Committee on Detection and Quantitation Approaches and Uses in Clean Water Act (CWA) Programs
Summary of Technical Work Group Conference Call #4
August 10, 2005
1:00 – 3:00 p.m. EDT
The next Technical Work Group call is scheduled for Wednesday, August 24, from 1:00 – 3:00 PM EDT. Please see below for the schedule of subsequent calls.
Action Items
Richard Reding will:
- Work with Triangle to further refine the glossary of terms for discussion at the next Technical Work Group meeting.
- Distribute public comments from Ken Miller via email to Technical Work Group members.
For discussion at the next meeting, the subgroup on definitions of detection and quantitation (Steve Bonde, Cliff Kirchmer, Ken Osborn, John Phillips, David Rocke, and Brad Venner) will:
- Continue refining definitions for detection (L c , L d and MDL).
- Further define terms for quantitation (draft circulated 18 July 2005).
- Complete definitions for remaining terms identified by the federal advisory committee (i.e. false positive, false negative, uncertainty).
For discussion at the next meeting, the subgroup on matrix of characteristics and procedures (Richard Reding, Timothy Fitzpatrick, Larry LaFleur, Jim Pletl, Richard Burrows, Richard Rediske, and Bill Ingersoll) will:
- Continue modifying the document describing procedure characteristics and begin filling in the matrix.
A third subgroup (Larry LaFleur, Brad Venner, Cliff Kirchmer, and Richard Burrows) was formed to examine the draft document describing limitations of the current MDL and ML procedures (“MDL_ML-Issues.doc”, 30 June 2005). The ideas and comments from the discussion among Technical Work Group members at the 10 August 2005 meeting of the MDL procedure will be added to the document.
- Larry LaFleur will lead the subgroup in a discussion to refine the document in preparation for the September 29-30 federal advisory committee meeting.
Triangle will:
- Work with Richard Reding to further refine the glossary of terms for discussion at the next Technical Work Group meeting.
- Redistribute the MDL and ML limitations document (see above) to all Technical Work Group members.
- Send protocols for email communication to all Technical Work Group members.
- Distribute to Technical Work Group members the composition of each of the subgroups.
Welcome and Introductions
Robert Wheeler, facilitator, welcomed participants to the call at 1:00 PM EDT . He conducted a roll call of Technical Work Group members and observers, and briefly reviewed the agreed-upon protocols for the teleconference established on the June 24 teleconference call. (The protocols are listed in the June 24 meeting summary.) He reminded Technical Work Group members of their assignment from the federal advisory committee. He said Technical Work Group members were tasked with developing:
- Definitions for terms (detection, quantitation, false positive/negative, uncertainty, L c, L d, and L q) and a glossary of additional terms associated with detection and quantitation.
- A matrix of existing detection and quantitation procedures.
- Characteristics for existing detection and quantitation procedures.
- A narrative description of concerns about the existing MDL and ML procedures.
Mr. Wheeler reiterated a message sent via email by Richard Reding clarifying that the Technical Work Group has several “subgroups” rather than “subcommittees” that are serving the full Technical Work Group. Due to administrative constraints, it is important to use the subgroup nomenclature.
Mr. Wheeler asked for comments on the draft summaries from the July 13 and July 25 Technical Work Group meetings. There were no comments, and the summaries were approved as drafted. Responding to a question from Mr. Wheeler, Technical Work Group members said they supported the level of detail in the summaries.
Email Protocol for Technical Work Group and Subgroup Communication
In response to suggestions for document management from Technical Work Group members, Mr. Wheeler suggested and Technical Work Group members approved the following protocol for titling the subject line of emails.
- For emails related to the full Technical Work Group, the subject line of the email will read: “TWG – [specific subject]”.
- For emails related to the subgroup for definitions, the subject line of the email will read: “TWG Def Sub – [specific subject]”.
- For emails related to the subgroup for characteristics and the matrix, the subject line of the email will read: “TWG Matrix Sub – [specific subject]”.
- For emails related to the subgroup for issues and concerns of the existing MDL, the subject line of the email will read: “TWG MDL Sub – [specific subject]”.
Report of the Subgroup for Definitions and Discussion
Mr. Wheeler turned attention to the substance of the meeting. He asked John Phillips to describe the draft product from the subgroup for definitions distributed via email on 4 August 2005 (“Detection Definitions Proposal.doc”).
Mr. Phillips reported that the subgroup had substantial discussions of the detection definitions. To date, the subgroup has agreed to keep the detection definitions from IUPAC (L c and L d ) and EPA (MDL). He said the subgroup also planned to add additional “layperson’s” language to the definitions to facilitate understanding among federal advisory committee members of each of the definitions. He added that in the next iteration of the subgroup product, it will have an expanded definition of the MDL procedure.
The Technical Work Group discussed the practical differences between the MDL and the critical level (L c ). Brad Venner of EPA explained that the two can be similar under certain circumstances. In simplistic terms, the MDL procedure emphasizes an estimate-testing framework, whereas the IUPAC procedure emphasizes a hypothesis-testing framework. Mr. Venner explained the advantage of an estimate-testing framework is that it can avoid the confusion associated with statistical testing. Cliff Kirchmer responded by saying that the current verbal definition of the MDL is inconsistent with the associated equation and graphical representation of the procedure. In his opinion, the MDL is a detection limit and not a critical level, and represents a “special case” identified by Currie (1968) and Wilson ( Chemical Analysis of Water). The Technical Work Group felt this discussion was important and needed to be documented and further discussed.
During the course of discussion of definitions for detection, Technical Work Group members discussed the need to complete the glossary of additional terms, and the utility of such a document for discussions among federal advisory committee members. Mr. Wheeler reminded Technical Work Group members that there are currently three glossaries that have been distributed via email. Richard Reding suggested he would work with Triangle Associates to compile into one document all of the useful terms for purposes of discussing detection and quantitation at the federal advisory committee. The compiled glossary will be presented prior to and discussed at the next Technical Work Group meeting.
Discussion on Issues/Concerns of MDL Procedure
Larry LaFluer suggested the federal advisory committee would benefit from understanding the shortcomings of the current MDL and ML. Mr. Wheeler reminded Technical Work Group members that Richard Reding had distributed a document in June describing the current issues and concerns of the MDL. Mr. Wheeler asked Technical Work Group members to review, comment and expand on that document.
After further discussion, Technical Work Group members agreed to task a third subgroup, led by Larry LaFleur, with reviewing the current draft document of MDL shortcomings and evaluate whether issues from the two other subgroups (definitions and matrix) are included. The MDL subgroup will focus on describing the shortcomings in layperson terms and on encapsulating procedural and definitional issues. Mr. Wheeler reminded MDL subgroup members to incorporate information from the discussion earlier in the meeting.
Report of the Subgroup for the Characteristics-Procedure Matrix and Discussion
Mr. Wheeler thanked Tim Fitzpatrick, Jim Pletl, and John Phillips for their efforts to define characteristics in the matrix, and asked Mr. Fitzpatrick to describe the product from the subgroup. Mr. Fitzpatrick explained that the subgroup had limited time to draft and discuss the document that describes each of the characteristics. The main issue with which they struggled was whether the intent was to capture what each of the procedures currently does or what each could do. Technical Work Group members agreed to focus the product on what each of the procedures currently does, and if time allows, add what the procedures might address. After discussing a few of the characteristics, Mr. Wheeler suggested that the subgroup would benefit from having additional discussions on the characteristics and associated revisions to the matrix, and bringing a refined product to the Technical Work Group for discussion at the next meeting.
Public Comment
Mr. Wheeler invited observers to offer comments.
Ken Miller, a contractor to EPA’s Engineering and Analysis Division, offered two comments.
- Regarding the definition for L c, in terms of a blank, especially in the context of contamination, the term “blank” implies that the concentration assumed under the null hypothesis could be greater than 0, and that the decision being made is not necessarily whether there is any amount of the analyte in the sample, but whether there is a greater amount of analyte in the sample than what could be caused by some contamination observed in the blank. In Currie’s work, he seems to use a null hypothesis of mu=0, though he may be assuming that any contamination level is subtracted prior to testing.
Defining L c in terms of a blank also seems to affect how a false positive is defined. This definition would seem to assume that any blank result exceeding L c would be a false positive whether the analyte is present in the blank or not, while if you defined L c in terms of concentration zero rather than a blank, a contaminated blank exceeding L c would be a true positive rather than a false positive, because the true concentration is not 0.
- For the criterion about false negatives, there could be some confusion from asking “Does the procedure account for false negatives?” Among the different detection procedures, some include an L c and L d, while others include only an L c. For procedures with an L d, one is estimating the concentration for which the false negative rate is some given rate (e.g. 5%). (While concentrations lower than the L d will have a higher false negative rate and concentrations higher than the L d will have a lower false negative rate.) If a procedure does not have an L d, then there is some concentration for which the rate of false negatives is 5%. However, the procedure does not give you any way to know what it is. Would a better criterion be: “Does the procedure specify a concentration for which the false negative rate is some given value?”
Wrap-up and Closing
Mr. Wheeler reviewed the schedule of remaining calls and suggested that, in order to approve products for distribution to the federal advisory committee, the Technical Work Group would likely need one additional call prior to the September federal advisory committee meeting. Members agreed to add one meeting: Tuesday, September 13 from 1:00 – 3:00 PM. Therefore, with the additional meeting, the dates and times (all are EDT) of future Technical Work Group meetings are as follows:
- Wednesday, August 24 from 1:00 – 3:00 p.m.
- Wednesday, September 7 from 1:00 – 3:00 p.m.
- Tuesday, September 13 from 1:00 – 3:00 p.m.
- Wednesday, September 21 from 1:00 – 3:00 p.m.
- Wednesday, October 5 from 1:00 – 3:00 p.m.
- Wednesday, October 19 from 1:00 – 3:00 p.m.
Mr. Wheeler briefly noted the action items from the call, thanked Technical Work Group members for their participation, and closed the meeting at 3:00 p.m. (EDT).
Attendance
Technical Work Group Members
- States: Timothy Fitzpatrick and Cliff Kirchmer
- Environmental Laboratories: Steve Bonde and Richard Burrows
- Environmental Community: Richard Rediske
- POTWs: Jim Pletl
- Industry: Larry LaFleur and John Phillips
- US EPA: Richard Reding (Office of Water) and Brad Venner (Office of Enforcement and Compliance Assurance)
- Other Federal:
- Triangle Associates - Derek Van Marter and Robert Wheeler
Observers
- Joanne Dea (US EPA Conflict Prevention and Resolution Center)
- Marion Kelly (US EPA Engineering and Analysis Division)
- Ken Miller (Consultant for EPA Office of Water)
![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif)