A P P L I C A T I O N N O T E

Pipet Quality Control: A Microliter of

Prevention…
by Ralph Bertermann

In the day-to-day operation of a chemical, analyt-

ical, quality, or testing laboratory, significant Table 1 Risk of undetected pipet failure for different performance verification intervals
attention is often focused on key analytical instru- Regulation Relevant excerpt
ments, such as spectrophotometers and HPLC 21 CFR Part 58:632 “Equipment used for the generation, measurement, or assessment of
systems. These devices cost many thousands of data shall be adequately tested, calibrated, and/or standardized.”
dollars to purchase, and are also expensive to cal-
ibrate and maintain. There is no question that 21 CFR Part 211:68 “If such equipment is so used, it shall be routinely calibrated,
their proper calibration and maintenance is inspected, or checked according to a written program designed to
essential to the laboratory’s operations, as well as ensure proper performance.”
to ensure compliance with regulations and to pro- 21 CFR Part 211.1643 “The calibration of instruments, apparatus, gauges, and recording
duce valid data. devices at suitable intervals in accordance with an established written
program containing specific directions, schedules, limits for accuracy
There is, however, one critically important and precision, and provisions for remedial action in the event accu-
component of virtually every laboratory’s opera- racy and/or precision are not met. Instruments not meeting estab-
tions that is used in a wide range of methods, lished specifications shall not be used.”
but that is often overlooked when it comes to 21 CFR Part 606.604 “The equipment shall be observed, standardized and calibrated on a
calibration and maintenance: the ubiquitous air regularly scheduled basis as prescribed in the Standard Operating Pro-
displacement pipet. After all, what is there to cedures Manual and shall perform in the manner for which it was
worry about where pipets are concerned? Are designed so as to assure compliance with the official requirements
they not simple, reliable devices that are easy to prescribed in this chapter for blood and blood products.”
use with minimal training? Isn’t annual or semi-
annual pipet calibration enough to ensure data 21 CFR Part 820:725 “Each manufacturer shall ensure that all inspection, measuring, and
validity and compliance with regulations? In a test equipment … is suitable for its intended purpose and is capable
word, no. of producing valid results.”
ANSI/NCSL Z540-1-19946 Calibration Laboratories and Measuring and Test Equipment—
General Requirements (9.1): “All measuring and testing equipment
Taking pipets having an effect on the accuracy or validity of calibrations shall be
calibrated and/or verified before being put into service.”
for granted ISO 9001:20007 Quality Management Systems—Requirements (7.6): “Where neces-
Contrary to widely held beliefs, the air displace- sary to ensure valid results, measuring equipment shall: a) be cali-
ment pipet is a precision instrument, capable of brated or verified at specified intervals, or prior to use, against
consistent volume deliveries in the range below 1 measurement standards traceable to international or national meas-
mL. Even so, pipets can introduce significant data urement standards; where no such standards exist, the basis used for
variability if not used and maintained properly. calibration or verification shall be recorded.”
ISO/IEC 17025:20008 General Requirements for the Competence of Testing and Cali-
Since pipets are precision instruments, they are bration Laboratories, (5.6.1): “All equipment used for tests and/or
subject to the same regulations that apply to the calibrations, including equipment for subsidiary measurements (e.g.
calibration and maintenance of other critical for environmental conditions) having a significant effect on the
laboratory instruments. Pipets therefore require: accuracy or validity of the result of the test, calibration or sampling
shall be calibrated before being put into service.”
• Periodic calibration according to a documented
procedure, along with periodic functional
checks to ensure proper ongoing performance found failure to meet specifications, the labora- quality control at a pharmaceutical API QC lab-
• Periodic maintenance and proper handling tory is required to investigate whether or not oratory are echoed by many laboratory supervisors
• Operation by persons who have demon- the data obtained with that failed pipet could and managers regardless of the class or type of
strated competence in using the instrument. have adversely affected measurements associ- instrument being discussed, including pipets.
ated with research, quality, or product safety.
The GLP and cGMP regulations,1 in line with More frequent verification of pipet perform- Under our formal calibration program, we
recommended practices and various interna- ance, in the form of periodic functional checks, send pipets out for calibration every six
tional standards, are very clear as to what is can help to reduce this risk, and greatly simplify months. Often several pipets are found to
expected in order to ensure the proper opera- the time and expense associated with this be out of tolerance. Our calibration de-
tion of an instrument in a regulated environ- needed remedial action. partment then sends us a form so we can
ment. These sources of guidance all echo the document whether our data results were
same requirements for calibration, a sampling of impacted. By the time we know about the
which is shown in Table 1. As-found failures: A case problem we’ve already shipped product,
and it may have already been consumed by
From a regulatory standpoint, one key reason in point patients. What can we do at this point?
why it is important to ensure a high standard of The concerns expressed below (paraphrased) by a We have no choice but to conclude that
pipet performance is that, in the event of an as- manager whose job responsibilities include pipet there were no significant impacts to the

18 / JUNE 2004 • AMERICAN BIOTECHNOLOGY LABORATORY

 A P P L I C A T I O N N O T E

If that same laboratory tors such as frequency of usage or time since

verifies pipet perform- last maintenance.
ance every four months, • Random failure. Pipet failure commonly re-
following is the risk in sults from random events, such as acci-
working days: three failed dents or mishandling, as opposed to sys-
devices × four months × tematic wear. Such failures cannot be
30 days/month = approx. accurately predicted, and they occur ran-
360 days of data at risk. If domly with respect to the service cycle.
pipets are tested every • Operator technique. Inconsistent or incorrect
three months, the risk pipetting technique is probably the largest
drops to approx. 180 days. single source of liquid delivery variability in
most laboratories. Operators frequently have
not been trained in the correct use of pipets.
Sources • Environmental factors. Pipets perform some-
what differently under different environmen-
Figure 1
of liquid
Risk of undetected pipet failure for different performance verification intervals.
tal conditions, relative to changes in temper-
ature and humidity.
data—that’s the reality. We’d all feel bet-
ter if we could know for sure that our
delivery variability • Device tolerance limits. A small amount of
variability in liquid delivery is due to inac-
As with all precision instruments, it is vitally curacy and imprecision inherent in the pipet
pipets are working properly before we release
important to understand how pipets func- itself. These values are typically known
the product.
tion, and how best to handle and operate based on the manufacturer’s specifications
them to ensure valid and repeatable perform- for the device, which usually reflect best-
The key to peace of mind in such situations is to
ance and results. It is also important to case performance.
verify pipet performance more frequently. As
understand what factors influence the accu-
Figure 1 shows, the number of “at risk” days dur-
racy and precision of pipets, and what opera-
ing which a laboratory is exposed to the conse- For the purposes of this article, discussion is
tor training is required to use them effec-
quences of undetected pipet failure rises, and confined to the first three sources of variability,
tively. Documented schedules and procedures
falls, sharply relative to the frequency of per- since these are by far the most significant.
for calibration (with traceability of measure-
formance checks.
ments to national standards), along with
periodic maintenance, are required to ensure
Take, for example, a laboratory that strives to
that each pipet is functioning within the lab-
Systematic versus random
maintain a target reliability level for pipets of
95%, and which verifies the performance of its
oratory’s established tolerances. failures
population of 100 pipets every six months. If five Systematic pipet failures are those that arise
pipets test outside of established tolerance limits, Pipetting variability arises from these princi- from simple wear, as a function of how often
following is the number of working days of data pal sources: the pipet is used and how frequently it is main-
that have been placed at risk and must be re- tained. Systematic failures can often be pre-
evaluated for mitigation purposes: five devices × • Systematic failure. Some pipet failures occur dicted and prevented by adjusting the service
six months × 30 days/month = 640 days. as a result of predictable wear, based on fac- cycle and calibration interval based on a

20 / JUNE 2004 • AMERICAN BIOTECHNOLOGY LABORATORY

 A P P L I C A T I O N N O T E

review of the as-found performance history of

the device.9,10
quality and safety requirements, all personnel
must be qualified for their particular assign- Compliance
Random pipet failures are those due to acci-
ments. This is spelled out in the regulations,
per 21 CFR Part 211.25 (a), which states, requirements
dents, mishandling, or other unplanned events.
An operator may inadvertently draw liquid into
“Each person engaged in the manufacture, pro-
cessing, packing, or holding of a drug product for pipets
the pipet body, for instance, leading to piston shall have education, training, and experience, A quality control program for liquid delivery
corrosion. The simple act of storing the pipet in or any combination thereof, to enable that per- should start with an understanding of what reg-
a horizontal position may lead to the migration son to perform assigned functions.”3 ulations require in the way of records and oper-
of liquids into the body of the pipet. In the lab- ational practices. As with other precision
oratory environment, random failures cannot be As regards pipets, this means that 1) operators instruments, audit risks for pipets include:
prevented by infrequent, scheduled mainte- who have assigned pipetting duties as part of
nance. This is simply because these failures are laboratory methods must demonstrate compe- • Undetected pipet failures or improper opera-
not dependent on predictable factors such as tency in pipetting technique. 2) An ongoing tion of devices in use
usage frequency or time since last maintenance. investment must be made in operator training. • Inadequate calibration/maintenance data
Random failures can (and do) occur at any In particular, pipet technique training must be • Inadequate training data
point in the maintenance cycle. conducted on a regular basis. 3) Operator com- • Inadequate QC (no established tolerances, no
petence to perform assigned duties must be doc- established calibration frequency, no provision
umented via training records. for on-site performance verification, etc.).
Silent failures
An important aspect of the majority of pipet Based on published studies,11 as well as the In order to ensure compliance with regulations,
malfunctions, whether due to unplanned author’s personal observations, it is clear that the following should be established for pipets:
events or predictable wear, is that they are many experienced laboratory technicians do
silent in nature. Mechanical action pipets have not employ proper or consistent pipetting • Appropriate and realistic tolerance limits for
a number of hidden internal parts (seals, piston, technique. Moreover, even properly trained accuracy and precision based on the labora-
etc.) that can fail, thereby causing the pipet to technicians exhibit person-to-person vari- tory’s process requirements and historical
deliver incorrectly. While some pipet failures ability in technique that can affect results. experience. Manufacturers’ specifications are
are detectable by “look and feel,” the majority Simple steps such as prewetting the pipet tip, based on ideal conditions and are typically
of failures are silent, that is, they are unde- immersing the tip to the proper depth, and too stringent for real-world applications;
tectable by the operator. A typical example of a allowing liquids and equipment to equilibrate hence, reliance on them can result in false
silent pipet failure is a leaking internal seal, to ambient temperature can significantly failures, requiring a costly, documented fol-
which can cause the pipet to underdeliver, improve accuracy and precision. (A concise low-up process. On the other hand, setting
sometimes by a large margin, without the oper- list of “10 Tips” for proper pipetting tech- tolerances too low can negatively affect
ator’s knowledge. nique is available on-line at www.artel- process quality.
usa.com/tips.htm.) • A realistic target reliability goal based on the
laboratory’s quality mandate. Too low a level
The good news is that adequate training is sim- may reduce quality and increase noncompli-
Operator technique ple and inexpensive to provide, and can greatly ance risks. Too high a level may result in
In order to help ensure that products meet improve pipetting technique. higher cost to customers. In a benchmarking

AMERICAN BIOTECHNOLOGY LABORATORY • JUNE 2004 / 21

 A P P L I C A T I O N N O T E

survey conducted by NCSL International quent functional checks can also help to mini- direct control of the laboratory, the labora-
(formerly the National Conference of mize the need for remedial action. tory shall ensure that the function and cali-
Standards Laboratories) in Boulder, CO, the bration status of the equipment are checked
majority of respondents set reliability goals A QC program for pipets should include the fol- and shown to be satisfactory before the
for general-purpose measurement and test lowing best practices: equipment is returned to service.” This point
equipment at 90% or greater. is important because pipets calibrated at an
• An optimal calibration frequency, based on • Verify performance of pipets often enough to environmentally controlled facility often
the established target reliability level and the ensure data validity, based on the laboratory’s deliver incorrectly when returned to the
mean time before failure (MTBF) for the experience with MTBF for the pipet popula- working environment. The ISO 8655:2 stan-
pipet population. tion.12* Performance verification protocols dard, for example, estimates the magnitude of
• A program to train operators in proper pipet- should include provisions for on-site per- imprecision in this context due to typical dif-
ting technique. formance verification immediately preceding ferences in humidity, temperature, and other
• A program for pipet calibration and interim and/or following a critical procedure, in order factors to be on the order of 5%.13
performance checks on the benchtop. to mitigate the potential impact of a mal- • Perform preventive maintenance (cleaning,
Infrequent, off-site calibration does not ade- function. These intermediate checks should seal replacement, relubrication) on a routine
quately address the leading sources of liquid be carried out according to defined proce- basis as determined by the established
delivery error: silent and random malfunc- dures (see ISO 17025, 5.5.10). MTBF. Calibrate all pipets immediately fol-
tion and improper technique. Benchtop cali- • Assign all pipets to specific operators, assay lowing maintenance.
bration not only facilitates more timely methods, or workstations to facilitate the
detection of nonsystematic pipet failures, identification of problematic results.
thereby reducing the need for remedial
action, but also provides a means to evaluate
• Train all operators in proper pipet operation
and pipet storage practices, and verify their
Conclusion
and train operators in proper technique. capability periodically on the benchtop, A program to ensure accurate and precise pipet
under every-day working conditions. performance is not only a good idea; it is a legal
requirement. The consequences of noncompli-
Best practices for • Immediately verify the performance of any
pipet that has just been dropped or otherwise ance, such as remedial follow-up and/or the
invalidation of critical data, can be more time-
pipet quality control mishandled, or that is associated with ques-
tionable data. consuming and expensive than implementing a
proper quality assurance program in the first
Pipets are precision laboratory instruments, and • Verify proper operation of all new pipets
are therefore subject to the same quality control received from a manufacturer, as well as place. In short, taking pipets for granted can be
regulations as other sensitive instruments such pipets that have been sent out for calibration, a painful and costly mistake.
as spectrophotometers and analytical balances. before placing them in service. This quality
This includes regular calibration. concern is specifically addressed by ISO Why aren’t laboratories more diligent about the
17025, 5.5.9, which states: “When, for what- quality of liquid delivery? It usually comes down
The longer a defective pipet remains in service, ever reason, equipment goes outside the to the issue of cost relative to perceived benefit.
the greater the potential for negative conse- Many times, the belief that pipets are beneath
*Ten data points are adequate to verify both accuracy and
quences. The more often calibration is per- precision. Four data points are sufficient as a “quick” concern is rooted in a lack of understanding of
formed, the sooner defective pipets will be check” (per ASTM 1154) or other form of benchtop ver- the nature of pipet failure, coupled with an atti-
detected and taken out of service. More fre- ification process to assess correct and/or accuracy alone. tude of complacency. continued

22 / JUNE 2004 • AMERICAN BIOTECHNOLOGY LABORATORY

 A P P L I C A T I O N N O T E

The reality is that cGMP-compliant pipet QC ity into results up front should not be on 7. Quality Management Systems—Requirements. ANSI/
makes solid competitive sense. It pays for itself the wish list in today’s business and regula- ISO/ASQ Q9001-2000. American National Standards
many times over in cost and time saved tory environment; it should be standard Institute, Washington, DC, Part 7.6, 2000.
through averting the need for remedial action operating procedure. 8. General Requirements for the Competence of Testing and
Calibration Laboratories. ANSI/ISO/IEC 17025:2000.
and assay repeats. Consider the many hours of
NCSL International, 2001, Apr, paragraph 5.4.
paperwork required just to check assay results,
methods, etc., when a pipet is finally found to References 9. Curtis RH. Performance verification of manual action
pipets part I. Am Clin Lab 1994; 13(10):8–9.
be operating incorrectly. Add to that the addi- 1. Guidance for Industry. Center for Drug Evaluation and 10. Curtis RH. Performance verification of manual action
tional resources wasted if the condition was sig- Research, Food and Drug Administration, 1996, para- pipets: part II. Am Clin Lab 1994; 14(1):16–17.
nificant enough to impact results. These graph 58.63. 11. Impact of pipetting technique. Artel Laboratory Report,
expenses, even if they are incurred infre- 2. Code of Federal Regulations (CFR). Title 21, Parts 1–89. issue 2. Artel, Inc., Westbrook, ME, May 1997.
quently, far outweigh the cost of maintaining a Subpart D. U.S. Government Printing Office, 12. Standard Specification for Piston or Plunger Operated
high standard of pipet performance in the first Washington, DC, 1999. Volumetric Apparatus. Specification E 1154-89.
place. A microliter of prevention up front can 3. Code of Federal Regulations (CFR). Title 21, Parts 200– American Society for Testing Materials (ASTM), reap-
result in kilograms of savings later. 299. Subpart D. U.S. Government Printing Office, proved 1997; p. 5, paragraphs 11.2.2 and 11.2.3.
Washington, DC, 1999. 13. ISO 8655-2:2002, Piston-Operated Volumetric Appara-
Robust pipet QC can also reduce rework 4. Code of Federal Regulations (CFR). Title 21, Part 600– tus, Part 2: Piston Pipets. International Organization for
and rejects, compress approval cycles, and 799. U.S. Government Printing Office, Washington, Standardization, Geneva, Switzerland, 2002.
DC, 2002.
eliminate potential compliance problems
5. Code of Federal Regulations (CFR). Title 21, Part 800–
before they happen. Equally important, a 1299. U.S. Government Printing Office, Washington,
standard for liquid delivery that ensures DC, 2002. Mr. Bertermann is a metrology and regulatory compliance con-
compliance with regulations also ensures 6. American National Standard for Calibration: Calibration sultant and Founder of Lighthouse Training Group; 117 S.
that a laboratory has fulfilled an ethical Laboratories and Measuring and Test Equipment— Albert St., Mt. Prospect, IL 60056, U.S.A.; tel./fax: 847-
responsibility to its customers and the com- General Requirements. ANSI/NCSL Z540-1-1994. 392-9796; email: [email protected]; home page: www.light-
munity. For all these reasons, building qual- NCSL International, 1994. housetraining.com.

24 / JUNE 2004 • AMERICAN BIOTECHNOLOGY LABORATORY