Data Quality in the Modern Analytical Lab

1
(No Transcript)
2
Data Quality in the Modern Analytical Lab

• Dr. Roy-Keith Smith
• Apichemical Consultants
• East Lansing, MI
• 1 May, 2007

3
General Considerations

• Legally defensible (Enforcement value of data)
• Analytically valid (Analytical value of data)

4
Common Foundation

• Approved methods by program
• Necessary calibrated equipment and supplies
• Trained personnel
• Quality Assurance Plan
• Certifications/Approvals

5
Public Labs

• Allocated funds from legislated budgets
• Slow response to changing regulations and
  technology
• Ponderous personnel systems and salary
  restrictions
• Ample training opportunities and funds
• Time to do things right and repeat work if needed

6
Commercial Labs

• Rapid response to changing regulations and
  technology (few decision makers)
• Immediate hire and fire ability generally higher
  salaries
• Little money available for training
• Things have to be done correctly the first time
• Over-riding consideration

7
(No Transcript)
8
Implications of

• Analysis of environmental samples is tedious and
  expensive
• Commercial labs concentrate on high dollar/high
  volume techniques
• Minimize services with high labor needs and low
  profit margin
• always generates the possibility of abuse

9
Fiscal Common Sense

• Minimize overhead and expenses
• Maximize sample through-put
• Increase sales force
• Maximize profit margin
• This is difficult and the first area to be
  critically examined is Quality Assurance

10
QA Issues

• QA is EXPENSIVE!
• On the surface its non-revenue generating
• Trained QA personnel are expensive
• Certifications are expensive and time-consuming
  to obtain and maintain
• Performance Evaluation samples are expensive

11
Other expenses

• Initial and continuing calibration is expensive
• Matrix spikes and Laboratory Control samples are
  expensive
• Method Detection Limit studies are time-consuming
  and expensive
• Initial Demonstrations of Competence are
  time-consuming and expensive

12
And when the QC results are not acceptable

• Time and supplies used are wasted while the
  situation is corrected and the samples
  re-analyzed.

13
Two Commercial Attitudes

• The high road
• The low road
• Most commercial labs exist somewhere between the
  two

14
High Road

• Require a culture of QA in the laboratory and
  develop a reputation for high standards and
  impeccable ethics.
• Accept a lower profit margin with the recognition
  that sometimes a loss is going to be incurred.
• Spend the money to hire and train professional
  analysts.

15
High Road (continued)

• Hire a high quality, expert QA supervisor.
• Give them the power and flexibility to oversee a
  rigorous QA program.
• Continually examine and re-examine QC procedures
  and results for each and every test process
• Continually train lab personnel in QA
• Provide copies of QA Manual

16
High Road (continued)

• Stand behind analytical results and answer all
  questions asked to the best of your ability
• Offer complete results and QC data packages to
  clients even when not requested and at no extra
  cost
• Offer training to customers to help them
  understand analysis and QA, especially those
  without chemistry backgrounds

17
High Road (continued)

• Seek to obtain and then maintain the highest
  quality laboratory certifications possible.
  NELAP is the beginning.

18
Low Road

• And then there are the practices and symptoms of
  the low road, where the objective is to maximize
  profits and trim all avoidable expenses.

19
Badges, we dont need no stinkin badges

• Concentrate on business that does not require any
  sort of laboratory certification/approval.
• Avoid possible situations where outsiders conduct
  a detailed examination of laboratory procedures
  and practices.
• Refuse to produce any data related to possible QC
  results.

20
Badges (continued)

• QA Manual consists of price list and a notarized
  statement stating We do good work

21
(No Transcript)
22
Slash and Burn

• The business plan depends upon a steady flow of
  new customers
• Repeat business is acceptable, however if a
  customer asks questions of any sort, drop them
  like a hot rock.
• For the client who develops regulatory problems
  the lab pulls a vanishing act.

23
Ill show you mine, if you show me yours

• Performance Evaluation results can make or break
  a lab.
• Labs share preliminary PE sample answers.
  Examination of a variety of answers for the
  sample(s) allows one to chose the best values to
  submit to the sponsor.

24
Water, wastewater, solid waste, its all the same

• Particularly prevalent in large laboratories -
  GC, GCMS, HPLC, ICP, and ICPMS are only set up
  for one method, most commonly a SW-846 method,
  and then all samples are processed using the
  method.
• The most common rationalization is that the
  SW-846 methods are the most modern and thus the
  best, and the other methods are not as good.

25
All the same (continued)

• This ignores differences in sample prep,
  surrogates, calibration, target analyte list, etc.

26
QA Manual? Its around here somewhere.

• A comprehensive and annually revised QA Manual is
  a major component of a QA Program. It is also
  tedious and expensive to write and update.
  Inability to produce a QA Manual is a good sign
  the lab lacks a QA Program.

27
Bubba here just got promoted to QA Manager

• In smaller labs, randomly assigning the position
  of QA Manager to anybody, is a common practice.
  Often they are a relative or good friend of the
  owner. This is in addition to other duties such
  as sales rep, or field services manager.

28
A little bit doesnt mean anything

• Laboratory contamination in blanks and samples
• Methylene chloride in the volatiles lab
• Phthalates in the semivolatiles lab
• Falsified blanks and blank correction
• Any results must be rejected where blank
  contamination is suspected

29
Sale TODAY ONLY, hurry down for the best prices

• Sales people exist to sell. They will do most
  anything to complete the sale. They are
  generally incapable of answering technical and QA
  questions. Avoid labs where you can only talk to
  the sales people, rather than the people actually
  doing the analysis.

30
Bottom Line

• Buyer Beware.
• You the client bear the regulatory burden when
  results/methods are not acceptable.
• The environmental services industry is
  essentially immune to regulatory action.
• Do not award analytical services contracts based
  solely on lowest price

31
Bottom Line

• Write bid specifications to include a rigorous QA
  Program, PE sample result submission, and an
  on-site inspection.
• You know what is required for legally defensible
  and analytically valid data within your program.
  Insist that your Contract Laboratory have the
  same knowledge and dedication.

32
References

• Smith, R-K, 1995, Water and Wastewater Laboratory
  Techniques, WEF, Alexandria VA 1999, Lectures on
  Wastewater Analysis Interpretation, Handbook of
  Environmental Analysis (4th Ed.) 2000,
  Interpretation of Organic Data 2001,
  Interpretation of Inorganic Data 2003, Guide to
  Environmental Analytical Methods, 5th Edition,
  Genium Publishing, New Amsterdam NY

33
(No Transcript)
34
(No Transcript)