Printer Friendly

Quality management: a must for laboratories too.

What is true about quality management in corporations is also true in an analytical laboratory. An analytical lab is a microcosm of its parent corporation and must satisfy two types of customers: internal (researchers and plant process engineers) and external (purchasers of products).

Since the adoption of Total Quality Management (TQM) by American industries in the eighties, quality management is no longer a passing fancy but has become a way of life. There should no longer be a debate about whether an organization should follow quality management practices; the only question is how best to implement these practices, for the very survival of a modern organization depends heavily on its ability to do so. According to Edward Deming [1], "Charles Darwin's law of survival of the fittest holds in free enterprise as well as natural selection. The only survivors will be the companies with constancy of purpose for quality, productivity, and service."

Lab and Customer

What is true about quality management in corporations is also true in an analytical laboratory. An analytical lab is a microcosm of its parent corporation and must satisfy two types of customers: internal (researchers and plant process engineers) and external (purchasers of products). The crucial role an analytical laboratory plays in customer satisfaction cannot be overemphasized. Think about the function of a testing laboratory in the following cases:

The plant produces an on-specifications product, but the testing lab defines it as off-specification. The plant must rework or reblend the product.

The laboratory certifies a truly off-specification product as on-specification. The situation becomes worse because of the resulting customer dissatisfaction.

Customers expect useful, consistent, accurate, and timely data. A laboratory must show its customers that it cares about their success. Customers should be customers by choice. Violations of customers' reasonable expectations, especially if substantial or repeated, will be perceived as lack of caring.

A manufacturing plant makes products for its customers; a laboratory produces data for its clients. There should be no difference in quality, whether the material is tangible (from a factory) or intangible (from a lab). The laboratory must have a vision and plan for excellence, together with goals and objectives. Outlined below are descriptions of practices that will help a laboratory achieve customer satisfaction.


* Be the best in quality.

* Achieve the best quality through continuous improvement.

* Deliver, on time, a consistent product that meets customers' expectations and that instills their confidence in the laboratory's reliability and dedication to quality.


* Make the right analysis the first time; eliminate repeated analyses and wrong information.

* Maintain and operate the laboratory in an ergonomically and environmentally safe fashion.

* Use well-trained, qualified analysts.

* Manage analytical procedures and operations to ensure consistency and accuracy of data.

* Provide excellent analytical capabilities using appropriate instrumentation and the most efficient and cost-effective means.

* Use statistical techniques such as analytical reference standards, industry cross-checks, and round-robins to understand and reduce sources of variation in the analytical process.

* Maintain statistical quality assurance (SQA) training and skills in laboratory staff, and improve quality continuously by forming laboratory quality improvement and quality action teams (QIT/QAT).

* Document the laboratory's practices, test methods, calibration, QA, and sample data.

* Obtain independently assessed QA approvals such as ISO 9000, if appropriate.


The laboratory must ensure that it operates in a manner that is safe for its staff as well as for the outside environment. Standard safety precautions for handling reagents, chemicals, solvents, gases, and radiation sources must be followed. Protective equipment must be worn as needed. The laboratory should have an inventory of all chemicals, their quantity, and their shelf life, as well as the material safety data sheets for each. A quick check of the quality of incoming chemicals should be made before adding them to the inventory. Wastes generated in the lab must be disposed of in a safe and environmentally acceptable manner. Documents must be maintained for the period of time required by regulatory agencies.


Laboratory staff must be properly trained and qualified. Although employees in most labs are divided among managerial, professional, and technician categories, the overall staff is responsible for safe operation of the laboratory equipment; maintenance of security of data and product information; performance of analyses using proper methods, techniques, and standards, quality and statistical integrity of data; documentation of data, test methods, calibration, and quality control (QC); updates of control charts for all testing; corrective actions; identification of the long-term analytical needs and development of plans to meet those needs; participation in the lab's QIT/QAT activities to improve laboratory performance; and industry cross-checks.

Laboratory staff must be familiar with test methods, instrumentation, and quality management practices used there. The performance of newly trained staff must be monitored before they carry out analyses for customers, and each lab should have a continuous retraining program so that staff can improve performance.


For any tests a laboratory may be expected to conduct, proper instrumentation and a program for replacing outdated equipment are essential. There must be a plan for routine preventive maintenance, and records of repair should be kept. It's a good idea to establish a separate logbook for each piece of equipment and include its maintenance history.


A laboratory must document fully its test methods and duplicate exactly those requested by customers or spelled out in product specifications. Without customer approval and proof of equal or better results and accuracy, no substitution may be made. Required test methods must conform to standard methodology such as that established by the American Society for Testing and Materials (ASTM), the U.S. Environmental Protection Agency, the Association of Official Analytical Chemists, and the U.S. Food and Drug Administration. A main source of error is deviation, whether intentional or inadvertent, from standards. The laboratory should have a system that periodically checks conformance.


Different test methods require different calibrations, and decisions about appropriate calibration and frequency are made on a case-by-case basis. However, all analytical testing requires calibration, and calibration practices must be thoroughly documented. If an instrument is out of calibration, under no circumstances can data be sent to customers.


It is surprising how many otherwise good laboratories do not practice SQA. This process should be viewed as an integral part of the complete analytical sequence rather than as an unnecessary added burden. A written schedule should be available for the SQA frequency, and all lab personnel must be familiar with it.

Frequency of SQA analysis depends on the sample type and the instrument. Certainly at least one SQA standard should be analyzed with each lot of customer samples. Preferably, one SQA standard should be analyzed before and after a series of analyses; even better, a few SQA samples should be interspersed with several real samples to continuously monitor the analytical data quality. A general rule of thumb is one SQA sample per five to ten samples. If the SQA sample indicates flawed performance, no data can be sent to the customer. Immediate remedial action must be taken.

Control Charts

All QA data generated in the laboratory must be plotted promptly on a control chart. Analysts who produce the data - not the supervisor - should plot the charts to foster a sense of ownership and encourage problem solving. Charts should be displayed near the analysis station for easy reference. Charts must not be considered an end; statistical behaviour indicated must be acted upon, following run rules, and if the run rules show out-of-control data, no customer sample analyses can be sent out. Immediate remedial action must be taken and documented. In manufacturing plants it's a good idea to have two control charts for the same analysis, one using a reference sample to measure the lab variance and a second using day-to-day plant products to measure the process variance. In a well-controlled laboratory, the analytical precision should be much better than the process precision. Eventually, the lab or process control charts can become part of a certificate of analysis, which is sent to customers to build their confidence in the laboratory.

Using the control charts or SQA data, a laboratory should calculate monthly or quarterly the precision for individual tests. These sigmas, compared with those given in standard methods (e.g., ASTM), can provide a clear picture of performance against industry standards and pinpoint a need for corrective action. Periodic review should be a key feature of any continuous improvement program.

Reference Materials

Whether for calibration of QC, some form of reference (RM) is necessary - more may be required for calibration. Commercial sources include the National Institute of Standards and Technology. The calibration of RMS should be traceable to national standards, and the traceability must be preserved in the lab's document system.

Often, no suitable commercial standards, particularly for metal analysis, are available. In such cases, a large enough quantity of a plant product should be obtained. After ensuring that the material is homogeneous, one can analyze it for species of interest using classical primary methods of analyses such as titrimetry or gravimetry. However, methods such as atomic absorption spectroscopy, inductively coupled plasma atomic emission spectroscopy, and X-ray fluorescence spectrometry cannot be used for certifying these primary calibration standards because these methods are calibration-dependent. They may, however, be useful for confirming data obtained by the primary methods. If possible, two different primary methods should be used, and these analyses must be replicated in sufficient numbers to represent the aliquots of the parent material.

It is not a good idea to use the same material as a calibration standard and for QC. The latter should be similar in matrix to the type of samples being analyzed, although this is not always feasible. When a commercially available RM cannot be used for QC, it's easy enough to segregate a reasonably large quantity of plant product; analyze it for the tests of interest twenty to thirty times; and, on the basis of these data, calculate the average value and sigma limits for the control chart.


Interlaboratory cross-checks help measure a laboratory's capability against its competitors. When one lab is less precise than others, the cause must be sought and eliminated. In cases of quality complaints, cross-checks provide better confidence in a customer's or supplier's data quality. Participation in cross-checks as part of continuous improvement means studying data results and correcting deficiencies.

Cross-checks should be conducted on homogeneous and stable materials. Precautions must ensure that the material does not degrade or decompose between dispatch and receipt. A sufficient number of laboratories and samples should be included in a cross-check to obtain statistically meaningful conclusions. ASTM recommends using a minimum of five laboratories and more than ten different samples. Data obtained should be analyzed for statistical outliers, which are deleted before calculating the intralab repeatability and interlab reproducibility [2]. In addition to the data tables, graphical data representation is highly useful for a quick synopsis of a lab's performance in the cross-check.


Sometimes it becomes necessary for one laboratory to use another - for example, when the work requires a particular instrument. If subcontracting is not a frequent occurrence, it may not be realistic to check the TQM concepts of other laboratories. But it is good practice to check another laboratory's adherence to test methods, calibration, and QC. It's also a good idea to send a blind QC sample along with the real sample; if the QC sample data are equivalent to those normally obtained, it is safe to accept data on the real samples. The subcontractor should provide a certificate of analysis.


Most modern laboratories use some kind of computerized laboratory information management system (LIMS), a number of which are available commercially. Minimum capabilities of such systems include the following:

* tracking samples from log-in to log-out;

* producing data on backlogged samples;

* calculating sample turnaround time;

* tracking customer samples for accounting;

* filing historical data on calibrations and QC;

* reminding when calibrations and QCs are due;

* charting QC data and flagging statistical run-rule violations when they occur;

* calculating periodic lab precision for various tests;

* providing certificates of analysis;

* comparing analyses against the product specifications;

* if relevant;

* providing historical data bases by tests and products;

* and providing, where possible, an automatic data link with laboratory instruments.

Management should not view a LIMS as a panacea to solve all problems; the system is a clerical tool to streamline and store laboratory data and records. Ways to continuously improve methods, precision, and accuracy must come from laboratory personnel.


Every laboratory should have a quality manual that contains information about its activities. It should have an organization chart and provide details such as the responsibilities of personnel, training policy and programs, a list of equipment, an instrumentation maintenance program and schedules, purchase policies, a chemicals inventory, the SQA program, a calibration program and schedule, QC frequency, standard RM use, and copies of test methods.

Certificates of Analysis

The laboratory needs certificates of analysis to transmit test data to customers. The primary components are the name and address of the testing lab; the customer's name and address; the sample receipt date; a reference number for the report; the customer sample identification and lab identification numbers; test method designation; test result with measurement units; measurement precision, if relevant; details about analyses; an authorized signature; and the data transmission date.

Internally, a certificate of analysis should be fully traceable to calibrations, QC, and any other appropriate information.

ISO 9000 Certification

Over the past decade, registration to ISO 9000 quality systems has become a way of life for companies and laboratories throughout the world. It is certainly a means to systemically document all laboratory activities. However, it should be looked upon as a basic foundation to build a true TQM system. There are other general laboratory accreditation agencies, such as the American Association for Laboratory Accreditation (A2LA) and individual industrial groups that certify such laboratories. If appropriate and useful, a laboratory should obtain such accreditations. General industry guidelines for labs are given by the American Society for Quality Control and ISO [3,4].

Internal Audits

Every laboratory should have a system for periodic checks of its own practices to confirm conformance to documented systems. Even if the lab is subjected to a biannual ISO 9000 audit, it is important to have internal audits because internal reviewers are much more familiar with their lab's requirements than the outside ISO auditors. In these reviews, the laboratory should check for whether it complies with its own quality system and whether the tests are being performed correctly (see insert on p. 20) through such thorough reviews can a laboratory build on its strengths, eliminate deficiencies, and become a high-quality business that can satisfy customers. Findings and recommendations from internal audits should be reviewed by lab management and acted upon to correct any faults.

Other Practices

In addition to the protocols suggested above, industry groups advocate many other general analytical decision-making statistical protocols. These include protocols for situations such as representative sampling, acceptability of replicate testing, data quality disputes among laboratories, data acceptance against product specifications, and use of significant figures (see box this page). In such situations, a laboratory must use these sound protocols rather than make arbitrary decisions.
Industry Standards for Analysis and Data Handling


Standard Subject

D3244 Data acceptance against product specification

D4057 Representative sampling

E29 Use of significant figures

E177 Use of terms "precision" and "bias" in ASTM
 test methods

E548 General criteria used for evaluating laboratory

E882 Accountability and QC in chemical analysis

E994 Laboratory accreditation systems

E1301 Development and operation of laboratory
 proficiency testing programs

E1323 Evaluation of laboratory measurement practices

RR-D02-1007 Data processing from cross-checks


Operation of an analytical laboratory, whether part of a manufacturing plant, a research and development division, or an independent commercial lab, should follow the same sound quality management practices used in other industries. The laboratory's product is data, and the qualities that will satisfy its customers include leadership, training, empowerment, SQA, benchmarking, long-range efforts, teamwork, continuous improvement, and open communications. These must be used in operating successful customer-oriented analytical laboratories.

Essential Elements of Laboratory Total Quality Management

* Goals and objectives.

* Strategies for excellence.

* Customer satisfaction policy.

* Safe operations.

* Personnel policies.

* Training policies.

* Instrumentation.

* Test methods of documentation.

* Use of reference materials.

* Calibration practices.

* Statistical quality assurance.

* Use of control charts.

* Corrective actions.

* Participation in cross-checks.

* Laboratory information management systems.

* Use of subcontractor labs.

* ISO 9000 registration.

* Internal quality and technical audits.

* Continuous improvement programs.

* Use of quality improvement teams.

* Use of standard protocols for data quality.

* Customer responses.

* Certificates of analysis.

Suggested Laboratory Audit Checklist

Laboratory systems

Goals and objectives Organization chart Description of responsibilities Manuals Documentation

Laboratory personnel

Requirements Qualification records Training policies and program SQA training

Safety practices

Safety and emergency procedures Storage of chemicals Scheduled safety inspections Availability of material safety data sheets for all laboratory reagents

Test methods

Documented test methods Review of actual practice against standard method Documented system for updating test methods Review of test performance


Equipment inventory Adequate instrumentation for all tests Maintenance program Logs of service repairs Backup system in case of instrument failure Upgrade program for older instruments

Calibration practices

Procedures Schedules Log of calibration records

QC practices

Schedules Log of QC records Use of control charts Use of statistical run rules Corrective actions taken in case of out-of-control data Calculation of laboratory test precision Review of laboratory test precision versus results given in the standard methods Program to improve test precision

Reference materials

List of reference materials used for calibration and QC Traceability System to prepare local secondary standards Adequate storage


Participation in industry or customer cross-checks Performance rating Actions taken to improve cross-check performance


Sample and data documentation and traceability System capabilities Generation of certificates of analysis Generation of control charts


System to evaluate subcontractor labs Practice in using subcontractor data


ISO 9000 registration Actions taken on ISO audit findings Internal audit system Actions taken on internal audit findings

Other issues

Continuous improvement activities Use of quality action teams TQM training of personnel Practice for representative sampling Practice for replicate testing Practice for rounding off data Sample retention system


1. Deming, W.E. Out of the Crises, MIT Press, Cambridge, MA, 1982, p. 155.

2. Annual Book of ASTM Standards, American Society for Testing and Materials, West Conoshohoken, PA, 1996, Vol 5.03.

3. Quality Assurance for the Chemical and Process Industries, American Society for Quality Control, Milwaukee, WI, 1987.

4. ISO Guide 25 - General Requirements for the Technical Competence of Testing Laboratories, ISO, Geneva, Switzerland, 1990.

R.A. Nadkarni received his PhD from the University of Bombay. Before joining Exxon, he was a Research Associate at the University of Kentucky and a Research Manager of the Materials Science Center analytical facility at Cornell University, Ithaca, NY. He coordinates the analytical QC activities and long-range analytical directions of 20 laboratories worldwide and has been actively involved in the ISO 9000 registrations of all of them.
COPYRIGHT 1997 Chemical Institute of Canada
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Nadkarni, R.A.
Publication:Canadian Chemical News
Date:Mar 1, 1997
Previous Article:Accreditation.
Next Article:The behaviour of viscoplastic materials around spherical particles.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |