EPA/600/R-10/057 | July 2010 | www.epa.gov/ord
United States
Environmental Protection
Agency
SINGLE-LABORATORY VERIFICATION
OF CULTURE-BASED ANALYTICAL
PROCEDURES FOR ESCHERICHIA COLI
O157:H7iNWATER
STUDY REPORT
National Homeland Security Research Center Office of Research and Development
United States Environmental Protection Agency, Cincinnati, Ohio 45268
-------
-------
Single-Laboratory Verification
of Culture-Based Analytical
Procedures for Escherichia coli
O157:H7 in Water
STUDY REPORT
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
NATIONAL HOMELAND SECURITY RESEARCH
CENTER
CINCINNATI, OH 45268
National Homeland Security Research Center Office of Research and Development
United States Environmental Protection Agency, Cincinnati, Ohio 45268
-------
-------
Acknowledgments
This report presents results of a laboratory study funded by the National Homeland Security
Research Center (NHSRC) within the U.S. Environmental Protection Agency's (EPA's)
Office of Research and Development (ORD) under the direction of Sanjiv R. Shah to
verify culture-based analytical procedures forE. coll O157 in water. Computer Sciences
Corporation (CSC) provided technical support and data evaluation under EPA contract EP-
C-05-045.
The contributions of the following persons and organizations are gratefully acknowledged:
Study Workgroup Participants
• Ayaad Assaad (EPA, Office of Pesticide Programs)
• Michele Burgess (EPA, Office of Emergency Management [OEM])
• Stephanie Harris (EPA, Region 10)
• Marissa Mullins (EPA, OEM)
• Malik Raynor (EPA, Office of Water [OW])
• James Sinclair (EPA, OW, Office of Ground Water and Drinking Water)
Subject Matter Experts
• Cheryl Bopp (Centers for Disease Control and Prevention)
• Diana Cantrell, Victor Cook, and Robert Mandrell (U.S. Department of Agriculture)
• William Cray, Mike Grant, and Steve Weagant (U.S. Food and Drug Administration)
Volunteer Participant Laboratory
• Rick Danielson, Rosie Newton, and Jim Truscott (BioVir Laboratories, Inc.)
-------
Disclaimer
This document has been reviewed in accordance with EPA policy and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or recommendation
for use. Neither the United States Government nor any of its employees, contractors, or their
employees make any warranty, expressed or implied, or assume any legal liability or responsibility
for any third party's use of, or the results of such use of, any information, apparatus, product, or
process discussed in this document, or represent that its use by such party would not infringe on
privately owned rights.
Questions concerning this document or its application should be addressed to:
Sanjiv R. Shah
National Homeland Security Research Center
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
USEPA-8801RR
Washington, DC 20460
(202) 564-9522
shah.saniiv(@.epa.gov
If you have difficulty accessing these PDF documents, please contact Nickel.Kathy @,epa. gov or
McCall.Amelia@,epa.gov for assistance.
-------
Contents
Acknowledgments iii
Disclaimer iv
Table of Contents v
Tables vii
Acronyms viii
Section 1.0-Background 1
Section 2.0-Study Objectives and Design 3
2.1-Study Preparation 3
2.1.1 -Identification of Laboratory 3
2.1.2-Preparation of Spiking Suspensions 3
2.2-Sample Matrices 4
2.3-Sample Analyses 4
2.4-Quality Control Analyses 6
Section 3.0-Study Management 7
3.1-Schedule 7
3.2-Participant Laboratory 7
Section4.0 -DataReporting, Validation, and Censoring 9
4.1-Data Reporting 9
4.2-Data Validation 9
4.3 - Censored Data 9
Section 5.0-Re suits 11
5.1-Initial Analyses 11
5.2-Pilot Analyses 11
5.3-Optimization Analyses 15
5.4-Drinking Water Analyses 19
5.5 - Combined Results of Verification of E. coli O157 Analyses
in Drinking and Surface Water. 19
Section 6.0-Data Analysis and Discussion 23
6.1-Initial Analyses 23
6.2-Pilot Analyses 23
6.3 -Optimization Analyses 23
6.3.1 - Evaluation of Results for Spiked Surface Water and PBS Samples 23
6.3.2-Evaluation of Results for Unspiked Surface Water. 24
6.4-Drinking Water Analyses using Optimized Procedure 24
6.5 - Rainbow Agar. 25
Section7.0-Conclusion 27
Section 8.0-References 29
-------
Appendices
Appendix A- Study-Specific Instructions A-l
Appendix B - Study Plan B-l
Appendix C- Spiking Protocol C-l
Appendix D - Data Reporting Forms D-l
Note: Appendices are attached as distributed to the workgroup and laboratory and may contain
minor differences in language in comparison to the Study Report.
-------
Tables
Table 1. Summary of Sample Analyses..
Table 2. Schedule for Verification of E. coli O157 Culture-based
Procedures in Water Matrices 7
Table 3. Pilot Analyses, Preliminary: Fresh (Non-Stressed) E. coli O157
Culture for Sample Spiking 12
Table 4. Pilot Analyses, Preliminary: Freeze-Stressed E. coli O157 Culture for Sample Spiking 12
Table 5. Pilot Analyses: Fresh (Non-Stressed) E. coli O157 Culture for Sample Spiking 13
Table 6. Pilot Analyses, Repeat: Fresh (Non-Stressed) E. coli O157 Culture for Sample Spiking.. ..14
Table 7. Pilot Analyses, Follow-on: Independent Evaluation of Antibiotics
and Elevated Temperature 15
Table 8. Optimization Analyses, Results Summary: PBS Spiked withE. coli O157 16
Table 9. Optimization Analyses, Results Summary: Surface Water Spiked withE. coli O157 17
Table 10. Optimization Analyses, Results Summary: Unspiked Surface Water. 18
Table 11. Drinking Water Analyses, Results: Verification of the Optimized Procedure
forE. coli O157 19
Table 12. Drinking Water and Surface Water Analyses, Results: Verification of the
"Optimized" Procedure E. coli O157 20
-------
Acronyms
(Included in the Study Report and/or Appendices)
ACV Acriflavin/cefsuludin/vancomycin
ANOVA Analysis of variance
ASTM American Society for Testing and Materials
ATCC® American Type Culture Collection
BHI Brain heart infusion
BSL Biosafety level
°C Degrees Celsius
CPU Colony forming unit
CSC Computer Sciences Corporation
EEB EHEC enrichment broth
EHEC Enterohemorrhagic E. coli
EPA U.S. Environmental Protection Agency
HEPA High efficiency paniculate air
IMS Immunomagnetic separation
LTB Lauryl tryptose broth
MF Membrane filtration
mBPW Modified buffered peptone water
MPN Most probable number
MUG 4-methylumbelliferyl-p-D-glucuronide
NCTC National Collection of Type Cultures
NHSRC National Homeland Security Research Center
OEM Office of Emergency Management
ORD Office of Research and Development
OSWER Office of Solid Waste and Emergency Response
OW Office of Water
PBS Phosphate buffered saline
PCR Polymerase chain reaction
PSI Pounds per square inch
QA Quality assurance
QAPP Quality assurance project plan
QC Quality control
RSD Relative standard deviation
SAP Standard Analytical Protocol
SD Standard deviation
SME Subject matter expert
-------
STEC Shiga toxin-producing E. coli
TC-SMAC Tellurite cefixime sorbitol MacConkey
TNTC Too numerous to count
TSA Tryptic soy agar
TSAYE Tryptic soy agar with 0.6% yeast extract
TB Tryptone broth
-------
-------
Section 1.0
Background
Subsequent to the anthrax attacks in the fall of 2001,
federal and state personnel were tasked with a mission
to provide response, recovery, and remediation for
biological incidents. However, it is recognized that
no standardized set of methods exists for collection,
isolation, and analysis of these biological agents in
environmental samples. This document presents results
of the single-laboratory verification study (Study) of
culture-based procedures for the identification and
quantitation of Escherichia coli serotype O157 (E. coli
O157) in water samples, a potential bioterrorism agent.
E. coli O157:H7 is a virulent member of the pathogenic
enterohemorrhagic E. coli (EHEC) group. Members
of the EHEC group, including E. coli O157:H7, that
produce one or more Shiga toxins are often referred to
as Shiga toxin-producing E. coli (STEC). Infection with
STEC in humans has been associated with a spectrum of
diseases, including gastroenteritis, hemorrhagic colitis,
and hemolytic-uremic syndrome, the latter a potentially
fatal complication in children less than 5 years of age
and the elderly. EHEC/STEC is endemic in cattle and
other domestic animals, and transmission to humans
occurs primarily through contaminated water, vegetables
contaminated by agricultural runoff, unpasteurized milk
and juice, and undercooked meat products (Reference
8.1).
The culture-based procedures used to evaluate water
matrices during the Study were adapted from Standard
Methods for the Examination of Water and Waste
Water, 21st Edition (Reference 8.2), and the journal
article "Evaluation of Techniques for Enrichment and
Isolation of Escherichia coli O157:H7 from Artificially
Contaminated Sprouts" by Weagant and Bound
(Reference 8.3).
The initial analytical procedure evaluated during
the Study was based on EPA's "Draft Standardized
Analytical Procedure for Escherichia coli O157:H7 in
Environmental Samples (March 2008)." The procedure
included the use of EHEC enrichment broth (EEB),
tellurite cefixime sorbitol MacConkey (TC-SMAC) agar,
and biochemical and serological confirmation. Results
indicated that the procedure was not acceptable for
either the reference matrix (phosphate buffered saline
[PBS]), or the two matrices of interest (drinking water,
surface water), and thus, the procedure was modified
based on input from subject matter experts (SMEs)
and workgroup guidance, and was optimized prior to
verification.
In the final optimized procedure, a water sample
(e.g., drinking water, surface water) is inoculated
into modified buffered peptone water (mBPW) and
incubated at 36°C ± 1.0°C for 2 - 2.5 hours followed
by incubation at 42.0°C ± 0.5°C for a total of 20 - 24
hours. Tubes with growth (turbidity) are submitted to
immunomagnetic separation (IMS) and sub-cultured
onto TC-SMAC and Rainbow® agars. After growth at
42.0°C ± 0.5°C for 18 - 24 hours, TC-SMAC plates are
examined for 2 - 3 mm colorless/gray colonies typical of
E. coli O157:H7. On Rainbow® plates, typical colonies
are black/gray. Isolated typical colonies are submitted to
serological typing using E. coli O157 latex agglutination
reagent (e.g., Oxoid DR0620M or equivalent), followed
by biochemical characterization using commercially
available test strips (e.g., API 20E® or equivalent).
Quantification of E. coli O157:H7 is determined using
the most probable number (MPN) technique. Tubes
that confirmed positive forE. coli O157:H7 are used to
determine MPN (Reference 8.4). Based on workgroup
discussion, a nine-tube MPN as opposed to the standard
fifteen-tube MPN was utilized to reduce the burden
(2000 tubes and 4000 plates for a nine-tube MPN versus
3400 tubes and 6800 plates for a fifteen-tube MPN) on
the laboratory once IMS, which was performed on each
individual tube, was added to the procedure.
Note: Serological typing of the H7 antigen or polymerase
chain reaction (PCR) for the H7 gene was not conducted
during the single-laboratory verification Study. In order
to confirm E. coli O157:H7, additional serological or
PCR analyses would be necessary.
-------
-------
Section 2.0
Study Objectives and Design
The primary objective of this Study was to verify
culture-based procedures for identification and
enumeration of E. coli O157. As indicated in the Study
Plan (Appendix B), this Study was originally designed
to verify the procedures for water, solid, and paniculate
matrices. However, during this phase, only water
matrices were evaluated, and only these are included
in this report. A phased approach is being taken for
evaluation of other matrices. Two sets of objectives
were identified for the Study: Study objectives and data
quality objectives.
Study Objectives
• Characterize analytical procedure performance
(recovery and precision) for a reference matrix
(PBS)
• Characterize analytical procedure performance
(recovery and precision) for environmental matrices
of interest (drinking water, surface water)
• Determine whether the analytical procedure requires
revision prior to multi-laboratory validation
To accomplish these objectives, the Study was conducted
in four phases, as described below.
• Phase 1. Identification of a qualified laboratory
to participate in the Study, preparation of spikes,
assessment of analytical procedures by preliminary
analyses of water matrices (PBS [reference matrix],
drinking water, surface water), review of data,
identification of issues, and revision of protocol
after consultation with SMEs
• Phase 2. Pilot analyses: Evaluation of water
matrices using a modified procedure with antibiotics
(acriflavin, cefsulodin, vancomycin), assessment
of freeze-stressed cultures and unstressed (fresh
cultures) laboratory-prepared spikes, review of data,
identification and resolution of issues
• Phase3. Optimization analyses: Evaluation of four
treatment options (antibiotics, elevated incubation
temperature, both, or neither) to optimize recoveries
from PBS and surface water samples and resolve
issues (e.g., inhibition of target) associated with
high ambientfoackground microbial levels observed
in surface water samples, statistical analyses of
treatments, and revision of procedure
• Phase 4. Analyses of drinking water using
the optimized procedure (elevated incubation
temperature) according to the optimization
instructions for elevated temperature
Data Quality Objective
Data produced under this Study were generated
according to the analytical and quality assurance/
quality control (QA/QC) procedures specified in the
Study-specific instructions (Appendix A) and Standard
Analytical Protocol (SAP). This ensured data integrity
and validity for all matrices evaluated and allowed
the Study workgroup to use the results to identify any
necessary revisions of the SAP.
2.1 -Study Preparation
Prior to the Study, the following activities were
completed, including identification of an appropriate
laboratory and development and evaluation of the
Spiking Protocol (Appendix C).
2.1.1 - Identification of Laboratory
A laboratory was identified that was (1) representative
of the general user community, (2) had experience
analyzing environmental samples forE. coli O157:H7,
and (3) had access to representative matrices. To reduce
Study costs, a volunteer laboratory was recruited. To
reduce the burden on the laboratory and encourage
participation, National Homeland Security Research
Center (NHSRC) provided the media, reagents, and
supplies needed for the Study. The requirements and
responsibilities of the laboratory are detailed in Study-
specific instructions (Appendix A) and the draft SAP.
2.7.2 - Preparation of Spiking Suspensions
The Study Plan (Appendix B) included the use of
two spike types, BioBalls and laboratory-prepared.
However, due to production difficulties with BioBalls,
only laboratory-prepared spikes were used during the
Study. During each week of the Study, E. coli O157:H7
(ATCC® 700728™/NCTC 12900) laboratory-prepared
spiking suspensions were propagated in 1% lauryl
tryptose broth (LTB) and incubated at 35.0°C ± 0.5°C
for 20 ± 4 hours. Serially diluted spiking suspensions
were used to spike samples. The laboratory enumerated
spiking suspensions on the same day that samples
were spiked and analyzed. Spiking suspensions were
enumerated using spread plate technique (in triplicate)
on tryptic soy agar (TSA) according to the Spiking
Protocol (Appendix C).
-------
2.2 - Sample Matrices
The laboratory analyzed water matrices to provide a
means for evaluating and optimizing the performance of
the analytical procedure. During the Study, a reference
matrix (PBS) and two matrices of interest (drinking
water and surface water) were evaluated. The reference
matrix was analyzed to provide a means for assessing
performance using a standard matrix that could be
duplicated in the future during routine use of the method.
The following water matrices were evaluated during the
Study:
Reference Matrix
• PBS
Water Matrices
• Chlorinated drinking water (laboratory tap,
dechlorinated with sodium thiosulfate)
• Surface water (reservoir)
2.3 - Sample Analyses
For preliminary analyses, a single unspiked and spiked
sample was evaluated per matrix. For Study analyses
samples, one unspiked PBS or two unspiked drinking or
surface water samples, as appropriate, were evaluated
by the procedure to determine background E. coli O157
concentrations. Results of preliminary analyses were
used to identify and resolve issues. The laboratory
conducted additional analyses to modify/optimize the
protocol to improve recoveries. For each set of analyses,
the data was evaluated by the workgroup prior to
requesting the laboratory to conduct additional analyses.
Table 1 summarizes the number and type of samples that
were evaluated to meet the objectives listed in Section 2.
-------
Table 1. Summary of Sample Analyses
Analysis Phase
Initial Analyses
Pilot Analyses:
Stressed /
Unstressed Spike
Pilot Analyses
Pilot Follow-on
Analyses
Optimization
Analyses
Drinking Water
Analyses
Matrix
Sterile PBS
(Reference
Matrix)
Surface Water
Drinking Water
Sterile PBS
(Reference
Matrix)
Surface Water
Sterile PBS
(Reference
Matrix)
Surface Water
Drinking Water
Sterile PBS
(Reference
Matrix)
Surface Water
Sterile PBS
(Reference
Matrix)
Surface Water
Sterile PBS
(Reference
Matrix)
Drinking Water
Spiking
Description
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Unspiked
Lab-Prepared Spike
Procedure/Treatment
Original procedure: March 2008 draft
document
Modified procedure with antibiotics
(acriflavin, cefsulodin, & vancomycin),
elevated incubation temperature
(42.0°C), and IMS
Modified procedure with antibiotics
(acriflavin, cefsulodin, & vancomycin),
elevated incubation temperature
(42.0°C),andIMS
With IMS and:
Antibiotics (acriflavin, cefsulodin &
vancomycin)
Elevated incubation temperature
(42.0°C)
With IMS and:
Antibiotics (cefsulodin & vancomycin)
Elevated incubation temperature (42.0°C)
Both (antibiotics and elevated incubation
temperature)
Neither (antibiotics nor elevated
incubation temperature)
With IMS and elevated incubation
temperature (42.0°C)
No. of
Replicates
1
1
1
1
1
1
1
2
1
2
1
4
2
4
2
4
1 per
treatment
2 per
treatment
1 per
treatment
2 per
treatment
7 per
treatment
14 per
treatment
14 per
treatment
14 per
treatment
1
1
2
4
-------
2.4 - Quality Control Analyses
The participant laboratory performed the following
QC analyses:
• Method Blank: The laboratory analyzed a
sterile, unspiked PBS method blank during
each week of analyses to verify the sterility
of equipment, materials, and supplies.
• Sterility Checks: To evaluate the sterility
of media and buffer, the laboratory
incubated a representative portion of each
batch (PBS, mBPW, TC-SMAC, Rainbow,
and TSA) for 24 ± 2 hours at 35.0°C -
37.0°C or 42.0°C ± 0.5°C, as appropriate,
and observed for growth. In addition,
sterility checks were conducted each day
samples were analyzed.
• Positive and Negative Controls: For the
purpose of the Study, positive controls
for selective agars and broths are those
organisms that provide the characteristic
growth and/or colony morphology of the
target organism. Negative controls are
those organisms that do not provide the
characteristic target organism growth
or reactions. For biochemical and
serological analyses positive and negative
controls are denned by their reaction
(e.g., Pseudomonas aeruginosa is oxidase
positive and E. coll O157:H7 is oxidase
negative). The following positive and
negative controls were evaluated during
each week of the Study:
• E. coli O157:H7 (ATCC® 700728™/
NCTC 12900): positive control (target
organism)
• E. coli (ATCC® 25922™): negative
control (non-target organism)
• Pseudomonas aeruginosa (ATCC®
27853™): positive or negative control
for biochemical tests, as appropriate
-------
Section 3.0
Study Management
This Study was designed under the direction of NHSRC
within the U.S. Environmental Protection Agency
(EPA's) Office of Research and Development (ORD)
with consultation and input provided by workgroup
members and SMEs. The EPA technical lead was Sanjiv
R. Shah. Coordination of activities for the Study was
conducted by Computer Sciences Corporation (CSC).
3.1 -Schedule
The duration of the Study was April 2008 to March
2009. The Study schedule is provided in Table 2.
Analyses of additional matrices may be conducted at a
later date.
Table 2. Schedule for Verification of E. coll O157 Culture-Based Procedures in Water Matrices
Analysis Phase
Initial Analyses
Pilot Analyses
Optimization Analyses
Drinking Water Analyses
Date
April
June 23, 2008
7 - 12, 2008
-September 13, 2008
November 24, 2008 - January 6, 2009
March
16 - 24, 2009
3.2 - Participant Laboratory
The participant laboratory was BioVir Laboratories,
Inc., (Rick Danielson, Rosie Newton, and Jim Truscott),
located at 685 Stone Road, Unit 6, Benicia, California
94510-1126.
-------
-------
Section 4.0
Data Reporting, Validation and Censoring
4.1 - Data Reporting
The laboratory submitted the following data to CSC for
review and validation:
• Completed cover sheet with sample collection and
QC information
• Completed sample-specific reporting forms
• Completed calculations on data reporting forms
• Documentation of any additional information that
would assist in evaluating the data
4.2 - Data Validation
CSC used data review checklists to ensure that each data
package was complete and that each sample result met
the Study-specific and method-specific requirements.
The review for each sample confirmed the following:
• Original forms were submitted
• Incubation times were met
• Incubation temperatures were met
• Media sterility checks were performed and
acceptable
• Positive and negative controls were performed and
exhibited the appropriate response
• Samples were spiked with the appropriate dilution
• All procedures were performed according to Study-
specific instructions and analytical procedures
• Calculations were correct
This process was performed independently by two
data reviewers, each of whom entered the results into
separate spreadsheets designed for data review and
validation for this Study. The results were compared
to verify consistency and identify potential data entry
errors. Based on data review, the data described below
were noted, and were considered either (1) valid and
acceptable or (2) invalid and unacceptable for inclusion
in subsequent data analysis.
Optimization Analyses
• The extremely high spike level (10,200 colony
forming units [CFU]/100 mL) for the November
24, 2008, analyses resulted in values above the
analytical range (e.g., >1099) of the method and did
not allow for accurate determination of recoveries.
Thus, the data was considered invalid and not
included in subsequent data analyses.
• Unspiked surface water replicates analyzed on
December 8, 2008 resulted in MPN values of
1099 and 3.01. Due to this disparity and based on
discussion with the laboratory, the 1099 value was
considered invalid and not included in subsequent
data analyses.
4.3 - Censored Data
In addition to the numerical sample results generated
during this Study, low censored ("less than") results
also were generated for unspiked samples that had non-
detects.
The easiest approach to dealing with the low censored
values for this Study would have been to eliminate them
from data analysis. However, because the censored
results provide valuable information, the censoring limit
(the "less than" value itself [1.08]) was used in data
analysis for these samples.
In addition to low censored values, 17 high censored (or
"greater than") results (excluding November 24, 2008
results) were observed. The censoring limit, the "greater
than" value itself, (e.g., >109.9 or >1099) obtained
during optimization and drinking water analyses were
used for data analysis.
-------
-------
Section 5.0
Results
This section includes results of unspiked and spiked
samples for Initial Analyses (Section 5.1), Pilot Analyses
(Section 5.2), Optimization Analyses (Section 5.3),
Drinking Water Analyses (Section 5.4), and Combined
Results of Verification of E. coli O157 Analyses in
Drinking and Surface Water (Section 5.5). A summary
of sample analyses is included in Section 2.3, Table
1. Only valid results are included in this section;
a detailed description of invalid data is included in
Section 4.2. Recoveries were based on the E. coli O157
concentrations in spiked samples minus ambient E. coli
O157 concentrations in unspiked samples.
5.1 - Initial Analyses
Recoveries were 6%, 0%, and 13% for PBS, drinking
water, and surface water, respectively using the original
procedure described in the March 2008 document
(enrichment in EEB and plating on TC-SMAC followed
by biochemical and serological confirmation).
5.2 - Pilot Analyses
Results of the pilot analyses were used to determine
if procedural modifications were appropriate. These
modifications included enrichment in mBPW at 36.0°C
± 1.0°C for two hours, addition of antibiotics (acriflavin,
cefsulodin, and vancomycin) and incubation at 42.0°C
± 0.5°C for an additional 20 - 24 hours. Tubes with
growth were submitted to IMS and IMS beads were
plated onto TC-SMAC and Rainbow® agars and
incubated at 42.0°C ± 0.5°C for 18 - 24 hours followed
by serological and biochemical confirmation. During
the pilot analyses the workgroup thought it would be
beneficial to evaluate the use of a freeze-stressed culture
for spiking samples to potentially mimic organisms
that may be stressed in the environment. Both
freeze-stressed and fresh (non-stressed) cultures were
evaluated. All other analyses conducted during the study
utilized fresh cultures. Laboratory-specific results are
provided in Tables 3 (non-stressed), 4 (stressed), and 5
(pilot analyses conducted the week of August 8, 2008).
Due to very low recoveries observed during the analyses
conducted the week of August 8, 2008, pilot analyses
were repeated the week of September 1, 2008 (Table 6).
Since recoveries observed for analyses conducted both
weeks (August 8, 2008 and September 1, 2008) were
low, additional analyses were conducted to evaluate the
use of antibiotics and elevated incubation temperature
independently. Results of this limited evaluation
(follow-on analyses) are provided in Table 7.
-------
Table 3. Pilot Analyses, Preliminary: Fresh (Non-Stressed) E. coli O157 Culture for Sample Spiking
Sample ID
Spike Level '
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent Recovery
PBS Samples
Unspiked
Spiked
NA
3.2
3.2
0-0-0
1-1-0
1-0-0
<1.08
2.6
1.257
48
6
Surface Water Samples
Unspiked
Spiked
NA
3.2
3.2
0-0-0
3-2-0
2-1-0
<1.08
12.68
4.647
363
111
1 Observed spike level was one log lower than target spike level and may account for variability in recovery between replicates
To evaluate whether the analytical protocol would
recover environmental (stressed) organisms, a freeze-
stressed culture was used to spike samples prior to
analyses (Table 4).
Table 4. Pilot Analyses, Preliminary: Freeze-Stressed E. coli O157 Culture for Sample Spiking
Sample ID
Spike Level 1
fCFIT/100mUl
MPN Combo
E. coli O157
nVTPN/100 mUl
Percent
Recovery
PBS Samples
Unspiked
Spiked
NA
26
26
0-0-0
3-3-2
3-3-3
<1.08
109.9
>109.9
418
>418
Surface Water Samples
Unspiked
Spiked
NA
26
26
0-0-0
3-3-3
3-3-3
<1.08
>109.9
>109.9
>418
>418
'Enumeration of the spikes on selective (TC-SMAC) and non-selective media demonstrated an approximate 103 reduction on the
selective media, as anticipated. The spike level noted above was based on TC-SMAC enumeration.
Based on recoveries observed for analyses evaluating
the modified procedure using a fresh culture (Table 3),
the laboratory was asked to proceed with additional
analyses to complete the verification study. The
laboratory analyzed one unspiked PBS sample, two
unspiked drinking water and surface water samples,
and four spiked samples per matrix according to the
modified procedure and Study instructions. Analyses
were conducted the week of August 8, 2008. Results are
provided in Table 5.
-------
Table 5. Pilot Analyses: Fresh (Non-Stressed) E. coli O157 Culture for Sample Spiking
Sample ID
Spike Level
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent Recovery
PBS Samples
Unspiked
Spiked
NA
2840
0-0-0
3-3-3
3-3-0
3-3-0
3-3-0
<1.08
>109.9
23.98
23.98
23.98
4
1
1
1
Drinking Water Samples
Unspiked
Spiked
NA
2840
0-0-0
0-0-0
3-3-0
3-3-0
3-3-0
3-3-0
<1.08
<1.08
23.98
23.98
23.98
23.98
1
1
1
1
Surface Water Samples
Unspiked
Spiked
NA
2840
0-0-0
1-3-0
3-3-1
3-2-1
3-3-0
3-3-1
<1.08
6.207
46.22
16.95
23.98
46.22
1
0
1
1
Due to the low recoveries observed for analyses
conducted the week of August 8, 2008 (Table 5), the
laboratory was asked to repeat the analyses the week of
September 1, 2008 to determine which data set (Table
3 or 5) was indicative of the "true" performance of the
analytical procedure.
-------
Table 6. Pilot Analyses, Repeat: Fresh (Non-Stressed) E. coli O157 Culture for Sample Spiking'
Sample ID
Spike Level
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent Recovery
PBS Samples
Unspiked
Spiked
NA
1490
0-0-0
3-3-1
3-2-0
3-3-2
3-3-2
<1.08
46.22
12.68
109.9
109.9
3
1
7
7
Drinking Water Samples
Unspiked
Spiked
NA
1490
0-0-0
0-0-0
3-3-1
3-3-0
3-3-0
3-3-0
<1.08
<1.08
46.22
23.98
23.98
23.98
3
2
2
2
Surface Water Samples
Unspiked
Spiked
NA
1490
3-3-2
3-3-2
3-3-3
3-3-3
3-3-3
3-3-2
109.9
109.9
>109.9
>109.9
>109.9
109.9
0
0
0
0
'Analyses stopped after mPBW
Recoveries observed for analyses conducted the week
of September 1, 2008 were 7% or less, which was
consistent with results from analyses conducted during
the week of August 8, 2009 (Table 5). These results
indicated that additional optimization of the analytical
protocol would be required prior to conducting
verification analyses. To determine if the use of both the
antibiotic solution and elevated incubation temperature
were inhibiting growth of the target organism, the
laboratory conducted limited analyses to evaluate
use of the antibiotic solution and elevated incubation
temperature independently. Results for those analyses
are provided in Table 7.
-------
Table 7. Pilot Analyses, Follow-on: Independent Evaluation of Antibiotics and Elevated Temperature'
Sample ID
Spike Level
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent
Recovery
PBS Samples - Option 1: Antibiotics (acriflavin, cefsulodin, & vancomycin) Only
Unspiked
Spiked
NA
58
0-0-0
3-3-0
3-3-1
<1.08
23.98
46.22
39
78
Surface Water Samples - Option 1: Antibiotics (acriflavin, cefsulodin, & vancomycin) Only
Unspiked
Spiked
NA
58
0-0-0
1-2-0
3-3-0
<1.08
4.23
23.98
7
39
PBS Samples - Option 2: Elevated Incubation Temperature Only
Unspiked
Spiked
NA
58
0-0-0
3-3-0
3-3-2
<1.08
23.98
109.9
39
188
Surface Water Samples - Option 2: Elevated Incubation Temperature Only
Unspiked
Spiked
NA
58
3-3-0
3-3-0
3-3-3
23.98
46.22
>109.9
38
>148
'Analyses were conducted the week of September 8, 2008
5.3 - Optimization Analyses
During Phase 3 (optimization analyses), the laboratory
evaluated four treatments to determine which treatment
was optimal. The four treatments evaluated were: (1)
neither antibiotics nor elevated incubation temperature
(42.0°C), (2) elevated incubation temperature, (3)
antibiotics (cefsulodin and vancomycin [acriflavin
was not used]), and (4) both antibiotics and elevated
incubation temperature. During the optimization
analyses each treatment was incorporated into the
modified procedure: mBPW, IMS, plating on TC-
SMAC and Rainbow® agars, followed by serological
and biochemical confirmation, as appropriate. Seven
runs were conducted, with 2 samples per treatment,
per run, resulting in 14 spiked samples per treatment
for each matrix (surface water and PBS). In addition,
one unspiked or two unspiked samples were analyzed
during each run for PBS and surface water, respectively.
Summary results are provided in Tables 8 (spiked PBS),
9 (spiked surface water), and 10 (unspiked surface
water). All unspiked PBS samples were negative
(<1.08 MPN/100 mL). Table 10 presents E. coli O157
concentrations (MPN/100 mL) rather than recoveries
because the samples were not spiked prior to analyses.
-------
Table 8. Optimization Analyses, Results Summary: PBS Spiked with E. coli O157 '2 3
Date
Spike
Level
(CFU/
100 mL)
Mean
Recovery
(%)
Min
Recovery
(%)
Max
Recovery
(%)
SD
(%)
RSD
(%)
Overall
Mean
(%)
Overall
SD (%)
Pooled
within-
run SD
(%)4
Overall
RSD
(%)
Pooled
within-
run RSD
(%)5
Neither (Antibiotics nor Elevated Incubation Temperature)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
90
114
44
131
89
109
53
114
44
24
34
65
127
114 6
44
238
144
154
52
0
0
151
78
63
58
0
0
116
87
58
84
66
71
79
63
Elevated Incubation Temperature (42.0°C)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
127
69
74
169
89
65
127
25
44
100
34
65
127
114 6
105
238
144
65
0
63
43
98
78
0
0
91
58
58
87
0
86
63
55
73
57
Antibiotics (cefsulodin and vancomycin)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
90
48
64
145
76
94
53
48
23
52
7
33
127
48
105
238
144
154
52
0
58
132
97
85
58
0
91
91
128
91
75
68
76
91
80
Both (Antibiotics and Elevated Incubation Temperature)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
90
81
105
169
195
33
53
48
105
100
46
33
127
114 6
105
238
343 6
33
52
47
0
98
210
0
58
58
0
58
108
0
97
97
93
93
91
91
96
96
65
65
1 ">1099 MPN/100 mL" was replaced with "1099 MPN/100 mL" for calculation of summary statistics
2 Two replicates were run per sample
3 Samples were spiked with overnight (fresh) cultures
4 Pooled within-run standard deviation (SD) determined by calculating the square root of the mean of the run variances
5 Pooled within-run relative standard deviation (RSD) was determined by calculating the square root of the mean of the squared run
RSDs
6 Recovery calculated based on a high-censored (>) MPN value
-------
Table 9. Optimization Analyses, Results Summary: Surface Water Spiked with E. coli O157 '•2 3
Date
Spike
Level
(CFU/
100 mL)
Mean
Recovery
(%)
Min
Recovery
(%)
Max
Recovery
(%)
SD
(%)
RSD
(%)
Overall
Mean
(%)
Overall
SD (%)
Pooled
within-
run SD
(%V
Overall
RSD
(%)
Pooled
within-
run RSD
(%)5
Neither (Antibiotics nor Elevated Incubation Temperature)
12/1/2008
12/3/2008
12/8/2008 7
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
64
79
33
168
148
109
2
46
23
99
-9
64
127
112 6
44
237
305 6
153
88
47
15
98
222
63
137
59
45
58
150
58
87
94
102
109
87
Elevated Incubation Temperature (42.0°C)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
72
114
103
96
188
140
18
114
103
96
40
140
127
114 6
103 6
96
336 6
140
77
0
0
0
210
0
106
0
0
0
112
0
103
81
84
79
58
Antibiotics (cefsulodin and vancomycin)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
127
26
62
52
92
94
127
5
21
6
43
33
127 6
47
103 6
99
141
154
0
30
58
66
69
85
0
116
93
127
75
91
66
57
54
86
93
Both (Antibiotics and Elevated Incubation Temperature)
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
863
963
1047
462
320
713
64
5
74
124
94
94
1
4
44
9
46
33
127
5
105
238 6
143
154
89
1
43
162
69
85
138
13
58
131
73
91
66
74
83
113
90
1 ">1099 MPN/100 mL" was replaced with "1099 MPN/100 mL" for calculation of summary statistics
2 Two replicates were run per sample
3 Samples were spiked with overnight (fresh) cultures
4 Pooled within-run SD determined by calculating the square root of the mean of the run variances
5 Pooled within-run RSD was determined by calculating the square root of the mean of the squared run RSDs
6 Recovery calculated based on a high-censored (>) MPN value
7 The 1099 MPN/100 mL unspiked result was removed for calculation of summary statistics
-------
Table 10. Optimization Analyses, Results Summary: Unspiked Surface Water
1,2
Date
Mean
(MPN/
100 mL)
Min
(MPN/100
mL)
Max
(MPN/100
mL)
SD (%)
RSD
(%)
Overall
Mean
(MPN/100
mL)
Overall
SD (%)
Pooled
within-
run SD
(%)s
Overall
RSD
(%)
Pooled
within-
run RSD
(%V
Neither (Antibiotics nor Elevated Incubation Temperature)
11/24/2008
12/1/2008
12/3/2008
12/8/2008 5
12/15/2008
12/22/2008
1/6/2009
5.90
5.18
20.75
3.01
3.01
121.42
5.42
1.08
3.01
3.01
3.01
3.01
3.05
3.61
10.72
7.36
38.50
3.01
3.01
239.80
7.23
6.82
3.08
25.10
0.00
0.00
167.41
2.56
115.51
59.34
120.92
0.00
0.00
137.87
47.25
25.11
65.24
64.05
259.86
86.79
Elevated Incubation Temperature (42.0°C)
11/24/2008
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
1.08
3.01
3.01
23.17
17.10
22.87
100.77
1.08
3.01
3.01
3.61
14.69
3.01
42.73
1.08
3.01
3.01
42.73
19.50
42.73
158.80
0.00
0.00
0.00
27.66
3.40
28.09
82.07
0.00
0.00
0.00
119.39
19.90
122.82
81.45
24.43
42.06
34.44
172.18
72.08
Antibiotics (cefsulodin and vancomycin)
Date
11/24/2008
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
Mean
(MPN/
100 mL)
1.17
2.04
6.88
19.31
3.01
10.96
2.18
Min
(MPN/100
mL)
1.08
1.08
1.08
10.99
3.01
7.23
1.08
Max
(MPN/100
mL)
1.26
3.01
12.68
27.63
3.01
14.69
3.28
SD (%)
0.12
1.36
8.20
11.77
0.00
5.27
1.55
RSD
(%)
10.65
66.65
119.20
60.93
0.00
48.10
71.29
Overall
Mean
(MPN/100
mL)
6.51
Overall
SD (%)
7.67
Pooled
within-
run SD
(%)s
5.83
Overall
RSD
(%)
117.82
Pooled
within-
run RSD
(%V
65.33
Both (Antibiotics and Elevated Incubation Temperature)
11/24/2008
12/1/2008
12/3/2008
12/8/2008
12/15/2008
12/22/2008
1/6/2009
1.17
1.08
1.08
1.87
1.08
3.48
1.84
1.08
1.08
1.08
1.14
1.08
3.01
1.08
1.26
1.08
1.08
2.60
1.08
3.96
2.60
0.12
0.00
0.00
1.03
0.00
0.67
1.07
10.65
0.00
0.00
55.31
0.00
19.23
58.36
1.66
0.96
0.62
57.89
31.51
1 The "<" values were replaced with actual value (e.g., <1.08 was replaced with 1.08) for calculation of summary statistics
2 Two replicates were run per sample
3 Pooled within-run SD was determined by calculating the square root of the mean of the run variances
4 Pooled within-run RSD was determined by calculating the square root of the mean of the squared run RSDs
5 The 1099 MPN/100 mL result was removed for calculation of summary statistics
-------
After the results of the optimization analyses were
analyzed and evaluated, the workgroup determined that
treatment two (elevated incubation temperature) was
the optimal treatment option. The final culture-based,
optimized procedure is as follows: (1) enrich samples in
mBPW at 36.0°C ± 1.0°C for 2 - 2.5 hours followed by
incubation at 42.0°C ± 0.5°C for an additional 20 - 24
hours, (2) submit tubes with growth to IMS, (3) plate
IMS beads onto TC-SMAC and Rainbow® agars and
incubate at 42.0°C ± 0.5°C for 18 - 24 hours, and (4)
submit isolates to serological and biochemical test-based
identification/confirmation.
5.4 - Drinking Water Analyses
Based on workgroup acceptance of the procedure, the
laboratory was asked to evaluate two, unspiked and
four, spiked drinking water samples using the optimized
procedure. Results are provided in Table 11.
Table 11. Drinking Water Analyses, Results: Verification of the Optimized Procedure for E. coli O157
Sample ID
Spike Level
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent Recovery
PBS Samples
Unspiked
Spiked
NA
438
0-0-0
3-3-1
3-3-1
<1.081
462.2
462.2
105
105
Drinking Water Samples
Unspiked
Spiked
NA
438
0-0-0
0-0-0
3-3-3
3-3-2
3-3-2
3-3-1
<1.081
<1.081
>1099
1099
1099
462.2
250
250
250
105
5.5 - Combined Results of Verification
of E. co/i O157 Analyses in Drinking
and Surface Water
Results for both drinking water and surface water
samples using the "optimized" procedure (pre-
enrichment in mBPW at 36°C, incubation at 42.0°C,
IMS, plating on TC-SMAC and Rainbow® agars,
and serological and biochemical confirmation) are
summarized in Table 12. Surface water results are from
Tables 9 and 10 and drinking water and PBS results are
from Table 11. Verification results were compiled for
drinking water and surface water to facilitate comparison
of recoveries for both matrices. Please note not all PBS
verification results are provided in Table 12.
-------
Table 12. Drinking Water and Surface Water Analyses, Results: Verification of the "Optimized" Procedure
E. coli O157
Date
Sample ID
Spike Level
(CFU/100 mL)
MPN Combo
E. coli O157
(MPN/100 mL)
Percent Recovery
(corrected for ambient
concentrations ')
PBS Samples
3/16/2009
Unspiked
Spiked
NA
438
0-0-0
3-3-1
3-3-1
<1.08
462.2
462.2
105
105
Drinking Water Samples
3/16/2009
Unspiked
Spiked
NA
438
0-0-0
0-0-0
3-3-3
3-3-2
3-3-2
3-3-1
<1.08
<1.08
>1099
1099
1099
462.2
250
250
250
105
Surface Water Samples
Date
12/1/2008
12/3/2008
12/8/2008
Sample ID
Unspiked
Spiked
Unspiked
Spiked
Unspiked
Spiked
Spike Level
(CFU/100 mL)
NA
863
NA
963
NA
1047
MPN Combo
0-0-0
0-0-0
3-3-2
3-1-3
0-0-1
0-0-1
3-3-2
3-3-3
3-1-0
1-0-0
3-3-2
3-3-2
E. coli O157
(MPN/100 mL)
<3.008
<3.008
1099
158.8
3.008
3.008
1099
>1099
42.73
3.61
1099
1099
Percent Recovery
(corrected for ambient
concentrations ')
127
18
114
114
103
103
-------
Date
12/15/2008
12/22/2008
1/6/2009
Sample ID
Unspiked
Spiked
Unspiked
Spiked
Unspiked
Spiked
Spike Level
(CFU/100 mL)
NA
462
NA
320
NA
713
MPN Combo
2-1-0
1-2-2
3-3-1
3-3-1
3-1-0
0-0-1
3-2-1
3-3-2
3-1-3
3-1-0
3-3-2
3-3-2
E. coli O157
(MPN/100 mL)
14.69
19.5
462.2
462.2
42.73
3.008
149.4
1099
158.8
42.73
1099
1099
Percent Recovery
(corrected for ambient
concentrations ')
96
96
40
336
140
140
1 Background levels of E. coli O157 in unspiked surface water samples (Table 10) were subtracted from spiked surface water levels
to calculate percent recovery
Using the optimized procedure, results indicate that
elevated incubation temperature (42.0°C) combined
with IMS sufficiently reduced background organisms
in surface water samples, resulting in good recoveries.
Recoveries for the reference matrix (PBS) and drinking
water were also good with this procedure. As there was
no evidence of background organisms in drinking water
samples, IMS may not be necessary for this matrix.
-------
-------
Section 6.0
Data Analysis and Discussion
After lengthy evaluation of multiple analytical
procedures for identification and enumeration of E.
coli 0157 in water matrices, a culture-based procedure
was optimized to overcome high microbial background
in surface water samples. While not evaluated during
this study, water quality parameters (e.g., pH, turbidity)
may be collected during the use of this procedure. As
indicated in Section 2, solid and paniculate matrices
were not evaluated during this phase of the effort.
6.1 - Initial Analyses
Preliminary water analyses were conducted using a
laboratory -prepared E. coli O157:H7 spiking suspension
and the procedure in the draft March 2008 document
(enrichment in EEB, plating on TC-SMAC, and
biochemical and serological confirmation). Problems
of overgrowth with background organisms were
encountered with surface water but not with PBS or
drinking water samples. However, recoveries were poor
for all three matrices: PBS, drinking water, and surface
water (See Section 5.1).
SMEs were consulted regarding the high background
in surface water samples. Potential options for method
optimization were discussed including the use of mBPW
as the primary enrichment media, elevated incubation
temperature (42.0°C), antibiotics (acriflavin, cefsulodin,
and vancomycin), IMS, and plating on both TC-
SMAC and Rainbow® agars. The SMEs recommended
mBPW because it has been used successfully to
recover E. coli O157:H7 in environmental matrices.
IMS was recommended as a step to separate E. coli
O157 from the background flora, and antibiotics and
elevated temperature were recommended to reduce
background flora. The use of the two plating media was
recommended to ensure that as many E. coli O157:H7
strains as possible could be detected by the method.
Since E. coli O157:H7 is not as thermotolerant as
other E. coli strains, the SMEs recommended a 42.0°C
incubation as opposed to a 44.5°C incubation utilized for
other E. coli strains.
6.2- Pilot Analyses
Based on workgroup recommendation, analyses were
conducted to evaluate SME suggested modifications,
including the use of mBPW as the pre-enrichment
medium, antibiotics (acriflavin, cefsulodin, and
vancomycin), elevated temperature (42.0°C), and IMS.
Initial results indicated the modifications were needed
(see Tables 3 and 4). However, additional analyses of
all three matrices (PBS, drinking water, surface water)
resulted in recoveries ranging from 0 - 7%, indicating
that further optimization would be required (see Tables
5 and 6). Limited pilot analyses with the modified
method (mBPW and IMS) utilizing either antibiotics
(acriflavin, cefsulodin, and vancomycin) or elevated
temperature (42.0°C) independently (Table 7) provided
improved recoveries, in comparison to samples analyzed
using both (Tables 5 and 6). As a result, the workgroup
recommended that the method should be optimized with
regard to the use of antibiotics and elevated temperature.
6.3 - Optimization Analyses
Based on pilot analyses, it was suspected that the
synergistic effect of both elevated temperature and
antibiotics was detrimental to recovery ofE. coli O157.
SMEs and workgroup also speculated that acriflavin
might be too harsh, especially when used in combination
with additional antibiotics; therefore, optimization
analyses were conducted with only cefsulodin and
vancomycin. As a result, based on the workgroup
recommendation, optimization analyses were conducted
to evaluate four different treatment options: (1) neither
antibiotics nor elevated incubation temperature (42.0°C),
(2) elevated incubation temperature, (3) antibiotics
(cefsulodin and vancomycin [acriflavin was not used]),
and (4) both antibiotics and elevated incubation
temperature. Results of spiked surface water and PBS
samples are described in Section 6.3.1 and results of
unspiked surface water samples are discussed in Section
6.3.2.
6.3.1 - Evaluation of Results for Spiked Surface
Water and PBS Samples
Comparisons of the effect of the different treatment
options on spiked surface water and PBS sample
recovery were performed using two-way Analysis of
Variance (ANOVA) models. Sample run was included
in the models to control for between-run variability. A
run-by-treatment interaction was also assessed for each
matrix. Due to the high frequency of high-censored
results (>1099 MPN/100 mL), analyses were performed
with recoveries for high-censored samples calculated
using (1) the censoring value (1099 MPN/100 mL) as
the value when calculating recoveries for high-censored
data, (2) the next possible MPN value (4622 MPN/100
mL, assuming one additional positive tube, if a 0.01
dilution had also been evaluated) as the value when
calculating recoveries for high-censored data, and (3) all
high-censored data removed. The high spike level
-------
(10,200 CFU/100 mL) for the first run (November
24, 2008) resulted in some values above the analytical
range (e.g., >1099), such that no uncensored result
would produce a percent recovery greater than 11%. As
a result, statistical analyses were also performed with
and without data from this run. Analyses were also run
with and without log-transforming the recoveries as the
data did not give a clear indication of how the results
were distributed within a treatment, matrix, and run.
In all cases, based on the ANOVA models, 1) no run-
by-treatment interaction was observed (i.e., the effect
of treatment on recovery did not differ significantly
between runs) and 2) no significant difference was
observed between optimization treatments for either
matrix.
6.3.2 - Evaluation of Results for Unspiked
Surface Water
Since E. coli O157 was observed in the unspiked
surface water samples for all seven runs, statistical
analyses on unspiked surface water sample results were
also performed, providing an unexpected opportunity
to assess method performance on ambient E. coli
O157 (Table 10). Comparisons of the effect of the
different treatment options were also performed for
the MPN/100 mL results observed in unspiked surface
water samples using two-way ANOVA models. Sample
run was included in the models to control for between-
run variability. Run-by-treatment interaction was
also assessed for each matrix. MPN values were log-
transformed prior to analyses. Based on the ANOVA
models, no run-by-treatment interaction was observed
(i.e., the effect of treatment on MPN/100 mL did not
differ significantly between runs). Interestingly, a
significant difference in log MPN/100 mL was observed
between the four different optimization treatments (p =
0.0032). To identify which specific treatments differed
from which other treatments, comparisons between
treatments were performed using Tukey's pairwise
comparisons. Based on Tukey's:
• mean unspiked surface water results were
significantly higher when "neither" treatment (101.8
mean MPN/100 mL) was applied, compared to
when "both elevated temperatures and antibiotics
were applied" (1.7 mean MPN/100 mL),
• mean unspiked surface water results were
significantly higher when "elevated incubation
temperature" (24.4 mean MPN/100 mL) was
applied, compared to when "both elevated
temperature and antibiotics were applied" (1.7 mean
MPN/100 mL), and
• "use of antibiotics" (6.5 mean MPN/100 mL) was
not significantly different from any of the other
treatments.
It should be noted that the mean MPN/100 mL for
"neither" is being driven by a single 1099 MPN/100
mL result observed on December 8, 2008. The
laboratory could not readily identify any issues, and
they also expressed concern regarding the high value.
In recognition that this single result had a substantial
impact on the mean for that treatment, the statistical
analyses described above were repeated with that data
point removed. Again, a significant difference in log
MPN/100 mL was observed between the four different
treatments (p = 0.0014) and, although the mean result
for "neither" dropped from a mean of 101.8 MPN/100
mL to 25.1 MPN/100 mL, Tukey's comparisons revealed
the same statistical results, as noted above; that is, the
exclusion of the 1099 MPN/100 mL had no effect on the
result of the statistical analyses (Table 10).
Although the "neither" treatment provided the higher
mean recovery (25.1 MPN/100 mLvs. 24.4 MPN/100
mL for elevated temperature), it was determined
that some element of selectivity was required in the
procedure (i.e., elevated incubation temperature) to
address issues with background organisms in some
surface water matrices. Results indicated that elevated
incubation temperature sufficiently reduced background
organisms in the surface water matrix (background
organisms were not observed in PBS and drinking water
matrices). Thus, based on the optimization results and
workgroup recommendations, the procedure was revised
to include the use of elevated incubation temperature
(42.0°C), along with mBPW, and IMS.
6.4 - Drinking Water Analyses using
Optimized Procedure
Since optimization analyses were conducted using PBS
and surface water matrices, follow-on analyses were
conducted using the optimized procedure (mBPW,
elevated incubation temperature, and IMS) to verify
that the procedure performed acceptably with drinking
water. The estimated percent recoveries, ranging from
105% to 250% were considered acceptable. It should
be noted that while 250% recovery may seem high, this
is not surprising given the variability in the 1099 MPN
value (the 95% confidence interval for this value is 225 -
3075). It should also be noted that while using a 15-
tube MPN set-up would increase laboratory burden and
reduce the number of samples that could be processed
per day, it would also result in tighter confidence
intervals and potentially minimize the impact of an
additional positive tube (e.g., a 5-5-0 combination is 240
MPN/100 mL and a 5-5-1 combination is 350 MPN/100
mL). It should also be noted that an MPN value is not
considered absolute quantitation, as a direct plate count
would be, because values are based on the probability of
tube being positive.
-------
6.5 - Rainbow® Agar
The use of Rainbow® agar was problematic throughout
the study for the following reasons. The agar is only
available from a single vendor and was often on
backorder during the study. In addition, the laboratory
indicated that it was very difficult to obtain isolated
colonies on Rainbow® agar and in many instances when
typical colonies were observed on TC-SMAC, there
were no typical colonies on the corresponding Rainbow®
agar plates. In addition, the participant laboratory
commented on the varying quality of the media from lot-
to-lot. It is recommended that a panel of different E. coli
O157:H7 strains should be evaluated as a preliminary
step to the multi-laboratory validation study, to confirm
whether the use of Rainbow® agar is necessary.
-------
-------
Section 7.0
Conclusion
Based on the results of the Study, the optimized culture-
based procedure has acceptable method performance
for identification and enumeration of E. coli 0157 in
PBS, drinking water, and surface water samples. The
optimized procedure includes: (1) cultures are enriched
in mBPW at 36.0°C ± 1.0°C for 2 - 2.5 hours followed
by incubation at 42.0°C ± 0.5°C for an additional 20 -
24 hours, (2) tubes with growth are submitted to IMS,
(3) IMS beads are plated onto TC-SMAC and Rainbow®
agars and incubated at 42.0°C ± 0.5°C for 18 - 24
hours, and (4) isolates are then submitted to serological
and biochemical confirmation. Time to results for the
procedure as verified is approximately 96 hours from
receipt of samples. However, time to results would be
approximately 48 hours if only serological confirmation
were conducted (e.g., no biochemical confirmation) or
72 hours if alternate biochemical analyses were used
(e.g., Vitek®).
As a result, the optimized procedure merits multi-
laboratory validation to assess method performance
and set quantitative QC criteria, so that the procedure
can be considered for potential implementation on a
national scale. The draft SAP will be revised to include
the optimized procedure prior to multi-laboratory
validation. Although the method was evaluated using
IMS for all three water matrices, consideration may be
given to evaluating drinking water samples both with
and without the IMS step during preliminary phases of
the multi-laboratory evaluation to confirm its necessity
in this matrix. Due to issues associated with Rainbow®
agar, we recommend, as a preliminary step to the
multi-laboratory validation study, evaluating a panel of
different E. coli O157:H7 strains to determine whether
the use of Rainbow® agar is necessary.
-------
-------
8.1 Feng, P. and Weagant, S.D. 2009. "Diarrheagenic
Escherichia coliT In Bacteriological Analytical
Manual Online. 8th Edition. G.J.Jackson,
R.I. Marker, and R. Handler (eds.). Center for
Food Safety and Applied Nutrition, U.S. Food
and Drug Administration, http://www.fda.gov/
Food/ScienceResearch/LaboratoryMethods/
Bacteriological AnalyticalManualBAM/ucni070080.
htm
8.2 Hunt, M.E. and Rice, E.W. 2005. "Section
9020 - Quality Assurance/Quality Control," and
"Section 9260F - Pathogenic Escherichia coliT In
Standard Methods for the Examination of Water and
Wastewater. 21st Edition. A.D. Eaton, L.S. Clesceri,
E. W. Rice, A.E. Greenberg, and M.A.H. Franson
(eds.). Washington D.C.: American Public Health
Association, American Water Works Association,
and Water Environment Federation.
8.3 Weagant, S.D. and Bound, A.J. 2001. "Evaluation
of Techniques for Enrichment and Isolation
of Escherichia coli O157:H7 from Artificially
Contaminated Sprouts." International Journal of
Food Microbiology. 71(1): 87-92.
8.4 Klee,A. J. 1993. "A Computer Program for the
Determination of the Most Probable Number and
its Confidence Limits." Journal of Microbiological
Methods. 18(2): 91-98.
Section 8.0
References
-------
-------
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |