AIHA Laboratory Quality Assurance Manual Ten Eric L. Botnick, CIH Ram $, Suga, SC, CIH, CHMM Kenneth 7, White, CIH, CSP A Publication by American Industrial Hygiene AssociationLaboratory Quality Assurance Manual Third Edition Edited by Eric L. Botnck, CH, Ram S, Suga, SC, CI, CHMM, fd Kenneth T Whit, CNT, CSP‘This manval was developed by experts with backgrounds, taining, and experience in laboratory quality assurance, storking with information and conditions existing atthe time of publication, The American Industrial Hygiene Association (AIHA), as publisher, and the authors and editors have been diligent in ‘ensuring tht the material and methods adessed in this book reflec prevailing oseupational health and safty and industrial hygiene practices. It is possible, however, that eemain procedures discussed will requis modification because of changing federal, state, and local regulations, oF heretofore unknown developments in research, AIHA and the authors and editors disclaim any Habiliy, loss o risk resulting directly or ininectly from use of the practices andiortheries discussed in this book. Moreover it isthe users responsibility to say informed of any changing federal, state, of lea regulations that might affect the material eantained herein, and the polices adopted specially in the user's workplace Specific mention of manufacturers and products in this book doesnot represent an endorsement by AINA Please send all comments regarding this Lab Quality Assurance Manual to ‘rie botnick@monsanio.com ‘Copyright 2005 by the American Industrial iygiene Association. All rights reserved. No part ofthis publication may be reproduced in any form, by photostat, mero, retrieval system, or any other means, without pri written permission ofthe publisher ISBN 1-931504.63.6 2700 Prospsity Avemue, Suite 250 Fairfax, VA 22031 Stock No, LOAMOS-699Laboratory Quality Assurance Manual ‘Third Edition Contents Page Figures, Tables, and Forms ¥ Preface si 1, Introduction 1 LL Introduction, 1 12 Purpose. 1 13 Paliies. 14 Definitions 1 1.5. Elements ofa Quality Assuranoe Program, ei a 2 16 List of Abbreviation oun eats i a 2 17. References 4 2._Personnel and Training 5 ion, and Control of Errors eee ees 34 Sources and Types of Ero. 9 3.5 Sunifcan! Figures ond Rounding 9 Sle rao ace a + Refrenoe 38 Furthor Reading 12 Sampling Procedures... zs a sas 13 42 Quality Control (QC) Samples 14 “43. Field Fquipment Calibration Considerations iS 44 Diflision Monitor Considerations Sl 45 Sampling ldetifieation, 11 {46_ Sampling Data Sheets. vo a 17 ‘42 Shipping, Field Preservation, and Procauion assum sa Aa tra TE 5. Sample Receipt and Handling. ar 5.1 Introduction, en 52. Chain of Custody 2 53 Information Reporting 22 6. Intradaboratory and Intertaboratory Quality Testing. 2 6.1 Introduction 29 62. Ints-lboratory Considerations 29 63 Inter aborwry Considerations boos cn bos 30 7. Analytical Methods at 7.1 Intraduotion, 3 72. Method Planing. 3 73 Method Validation Docu 33 74. Method Format 34 American Indusval Hygiene Association iitLaboratory Quality Assurance Manual Third Edition Figures, Tables, and Forms Figures Chaprer 3 Figure 3.1 Additive Eeror, 4 Figure 3.2 Proportional Eror 5 Chapter 4 Figure 41 System Flow Chart 13 Chapter § Figure $.1 Sample Seal. 23 Chapter 9 Figure 9.1 Duplicates-Range Variation. en na en “4 Figure 9.2Youden Tow — Sample Plt 45 Figure 9.3 Reference Contol Chast 46 Figure 9.4 Duplicate Control Chart Wrong Method a Figure 9-8 Duplicate Control Chart — Britton’s Method, a7 Figure 0.6 Duplicate Contol Chart — Taylor's Method a8 Figure 97 Duplicate Conttol Chart — NYSDOH Method. a8 Tables Chapter 3 ‘Table 3.1 Sources of Potential Ero rw se bow 10 Chapter 7 Table 7.1 QC Table 39 Chapter 9 ‘Table 9.1 Example Data 46 Forms Chapter 4 ‘tachment A: Industrial Hygiene Sampling Data Sheet 20 Chapter § Adachment Bi: Air Sampling Worksheet 24 Atachment C: Air Sampling Report. 36 Aachment D: Chain-of Custody Record 27 American Indusval Hygiene Association vLaboratory Quality Assurance Manual “Third Edition 75 Analytcal Teems and Limits. 7.6 Quality Contol Sample Aceeptnce Crit 17 References Equipment Calibration and Maintenance. 8.1 Introduction, 82 Instruments 83 Glassware and Volumetric Apparatus S4 Lleetncel Power Sourve Data Validation and Interpretation 9.1 Introduction 9.2. Data Review Procedure, 93. Quality Control Chart 9.4 Occasional Samples 95. Statistical Analysis for Quality Contol of Precision 9.6. Rejoeton of Results. 9.7 Comctive Action Pan 98 System Aus 9.9) Referenees. 10, Reporting and Record Keeping LT Introduction, 02 Elements of the Analsial Report 103 Recond Retention Program 14 Sample Retention Program 11, Analysis of Fiber Containing Samples. HL Introduction, 11.2 Sample Shipment 13 Contamination Checks. 11.4 Fiber Counting Using Phase Contrast Microscopy (PCN) 1.5 Aishorne Asbestos Fibers by Transmission Electron Miroscopy 11.6 Asbestos Bulk Identification Using Polarized Light Microscopy 1137 References 12, Quality Assurance Considerations for Blologleal Monitoring 121 Introduction, 122 Definition . 123 Regulations and Exposure Indie 124 Project Initiation 125 Sample Collection 12.6 Sample Analysis 122 Quality Assurance Quality Control 128 Reporting Results 129 References, 36 39 30 41 al al 2 2 4 8 a3 m3 45 46 49 SI si 52 33 53 53 55 "39 a s 65 aS “6S 66 70 7 2 75 76 MOB a al pearance memmeancamaatt American Indusval Hygiene AssociationLaboratory Quality Assurance Manual Third Edition Chapter 2 Personnel and Training 21 Introduc Qualified, properly trained personnel are a key aspect of satisfactory laboratory performance. Although good fciities and equipment are essential, they are of litle value without well-trained petsonnel. A laboratory, therefore, must establish adequate qualifications for personnel and Provide an ongoing training program to develop and maintain the necessary abilities. Ako, Petsonnel should be knowledgeable of the sampling and analysis processes, inorder to facilitate {ener communication with field personne! and clients. Effective eoordinston is essential to provide laboratory clients with saisictory analytical services, 22 Personnel Qualifications ‘The minimum qualifications for laboratory personnel should be those curently sot Forth in AIA acerectation and proficieney testing programs, of other relevant accreditation qualification Programs, Some laboratories may want to establish supplementary qualifications, such as adkitional experience in specifi areas or on specifi topics, oF requiring that some personnel have ‘advanced scientific degrees and/or additional camistry education, 23 Personnel Trai ig 2.3.1 General Considerations ‘QA program must include and provide forthe ongoing taining of beratory persomn. This training program must be in print, adaptable, and include demonstrations of proficiency. A rscord ‘of qualifications and training must be kept foreach person, This the responsibilty ofmanagem otto establish and implement a traning program that ensures that methods and procedures ste carried out propery. The program should include, a minim: sampling techniques, analytes! methods, specifiy. accuracy and precision monitoring, limits oF detection, potential interferences, and regulatory’authortaive His 2.3.2 Specific Training It's the responsibility of laboratory management (however tiled) to ensure that all aboratory petsonne! are introduced tothe QAM and ttsined peopetly in laboratory techniques, methods oF analysis, and QA QC procedures. The laboratory should have a writen training program or ‘checklist to ensure that ll Iboratory personnel rseive adequate training in policies, procedures, methods, instrumentation, caleulations, reporting, quality coatrl, and safety, with special ‘emphasis on th safe handling of toxie and carcinogenic materials. Training of new personne! should be done on the job under the guidance of qualified, experienced analysts, This taining should be supplemented with attendance at seminars and courses, such as those presented by instrament manufactrers. so, eoutses in the fundamentals of industrial hygiene, industrial hygiene chemistry, ahoratory accreditation, quality conto, ssbestos analysis, lead sampling ond analysis, ete, are presented by various college and university educational resource centers, in Addition to the National Institute for Occupational Safety and Heath (NIOSH), the US. Environmental Protcction Agency (EPA), and AIHA. Finally there should be opportunities for Petsonnel to ksep curent in their fick of interest through attendance at appropriate technical meetings, such asthe annual American Industral Hygiene Conferenee & Exposition (Hee). An analyst must be required to demonstrate profcisney in the performance of new methods or techniques as the last stp in the training process before beginning to analyze actual fd samples. ‘This ean be done with the sttisfactory analysis of Known samples, such as certified reference American Indusval Hygiene Association 5Laboratory Quality Assurance Manual Third Edition Chapter 3 Error and Sources of Error 31 Introduction An essential tsk in a QA program is the detection of error ara its elimination of contol, This ‘chapter discusses eror in sampling and analysis. Error and is various types are defined, and ways to detcot and contral error aze presented briefly, Also, specific sources of error are given for ‘consideration during sampling and analysis, As the first step inthe detsction of err, the analyst must be familiar with the definitions ofthe lypes of cror that might Be encountered, an the relationship of error Io aecuraey and precision. An understanding of these definitions increases awareness ofthe potential for eror aa allows sources and types of eror to be descted nd classified, Once the sures and types of error are idomitid, th elimination or contol of err should follow: 3.2 Definitions Error cannot be defined without frst defining accuracy. Therefore, accuracy is defined as the degree of agreement between the measured value and he tre value. rors defined as the degree ‘of disagreement ofa measured value with an accepted rferenee value (ie the “tue” value) Error may have both systematic and random compononts!"" Systematic error is synonymous with dotsminate error, assignable crpor, oF “bias and is @ ‘comsstont deviation in the results of @ measurement fom an aecepted reference value, The ‘of bias may or may not be known, but is considered to be assignable. Systmatiectrors ean be classified as “additive” o¢ “proportional” The errors additive if it is # constant vale, regard ‘of the amount or concentration of the analyte (See Figure 3.1). It is possible to sce the diference between additive and proportional error by plotting the observed values asa function of theoretical values. Additive err would produce a linear plot with slope of one (just like ta that had no ‘tematic ero) but the y-interoept would not be zero it would be the same a the eer value ANALYTICAL VALUE THEORETICAL VALUE Figure 3.1 — Additive error. American Indusval Hygiene Association 7Laboratory Quality Assurance Manual “Third Edition 24 materials (CRM), extra sts of industrial hygiene profiieney analytical testing (PAT) program samples from AIHA, extra sets of environmental lad proficiency analytical testing (ELPAT) samples ftom AIA, or blind samples prepared by somone else ithe laboratory. The use of Sondard reference materials (SRM) for Such purposes is discouraged, as it depletes the supply of the materials, and other sources for samples are readily available. The laboratory’s routine QC program show serve a8 a continuing check on the analysts performance cori must he kept to document the qualifications and taining ofall laboratory personnel. The ‘minimum information in these records should ince formal education, prior experienc, taining receivod on the job, et. Those records should also ineluds an outins, a Sasription of the course an overview of the techaigue oF procedure, the dates trining took place, and the person of btganization providing it. There also needs Yo bs a record indicating the qualifications of the course instructor. A designated management representative should sign the record, indicating satisfactory completion ofeach phase of traning and authorization to operate specific equipment ‘or perform specific analytical methods! Accreditation recognition programs may have specie ‘and more extensive recordkeeping requirements, References Anna 1. American Industrial Hyglene Association (ATHA): “AIA Laboratory Programs Foye aa one/LaboratoryServieesthmblgap htm, 2. US. Environmental Protection Agency (EPA): Principles. In Quality Assurance Handbook for sr Pollution Measurement Systoms, vol. 1 (EPA Pub. Ne. 600-9-76-008}. Cincinnati EPA, 1976, 6 American Indusval Hygiene AssociationLaboratory Quality Assurance Manual “Third Edition A systematic errr is sid to be proportional ifthe magnitude ofthe ero is proportional tothe concentration of the analyte. 1f here isa systematic erro, thon a plot of observed values as a function of theoretical values would demonstrate a slope diferent fiom unity. Systematic ror ray be a more complex funetion of concentration and could result i a euviinear slope for sel a plot (se Figure 3.2). Although specific soursos of determinate (systematic) ctror ate listed for various aspoets of sampling and analysis in a fate postion of this chapter the following ae luseative examples! + Poe recovery ofthe analyte from the sample media + Incorrect preparation of standards + Improper instrument calibration + Coataminated or outdated reagents + Human visual acuity deficiencies + man carelessness or personal bias + Improper selection of stmpling or analytic + Improper shipment and storage of samples + Interferences fiom other elements or ompourds method indeterminate e ments, Ra Random error is synonymous wit or o¢ non-assignable error an often is Aofined in toms of the pcesion of meas ndom err i characterized by ieguat ‘aration in ropated observations of acasurements, Because the variations ae random in natu they are unpredictable individually, The eause of this type of error generally cannot be asigne Random error statistics usually are represented by & normal dstebution—a distribution having a Frequency vs. size curve tat is belFshapedl and symmetrical about the arithmetic mean size However, measurements of small entities with possibly large exors in the upward but not dovsnward dieetion (Zero-fimited, non-negative measures) give skewed distribution curve that is offen represented by a lognormal distribution (ic, the logarithms of the measurements ane ‘normally distributed) Random ctor canbe estimated statistically fom the proesion of replicate ‘determinations, Sources of random erro ae identified typically a8 potential soures af variation inherent withthe spestie steps of an established sampling and analytical method, Examples ince variations in ANALYTICAL VALUE THEORETICAL VALUE Figure 3.2 — Proportional error. 8 American Indusval Hygiene AssociationLaboratory Quality Assurance Manual Third Edition 33 34 35 physical measurements of volume and mass, variations in instrament response, electrical line Volage transients and sample heterogeneity For a quantity or concentration that is derived from the measurement of several variables, an estimate of th overall uncertainty in the devived quantity caused by random error can be made using the law ofthe propagation of erors, This is dane by combining the contnbution of each ‘variable. Implicit im the law is the assumption that cach contnbuting variable is independent. For linear relationships, the overall variance is the weighted average of the individual variances, with degre of frecdom used a8 weighing factors Por prodicts and ratios, he square ofthe relative ‘rors (ic. the souares of the coeffcints of variation) are adkitive. The overall variance isthe Square root ofthe sum ofthe squares of the coe cients of variation for each measured variable ‘Nore thorough presenation ofthis mathematical approach i given by Mandel snd Nanni! An example of the use of the law of the propagation of errs is provided by NIOSH in the validation of personal air sampling and analytical methods!) ‘The concentiation of an air contaminant is the ratio of @ measured mass of the contaminant to a measured air volume ‘Consequently, NIOSH defines the overall coefficient of variation asthe square root ofthe sum of the squares ofthe cocficicnts of variation for sampling (CVs), analysis and desorption efficieney (CV gp and flow (CV Detection, Elimination, and Control of Errors ‘The QA measures discussed in other chapters of this manual are useful in the detection Laboratory Quality Assurance Manual Third Edition 3.6 INOTE: Sce vendor's troubleshooting guide for typical instrumentel problems or contact vendor's application specialist for unusual mechanical problems | Matrix effects Methodology Incorrect recoveries of QC samples Excessively high blanks Deviations from documented procedures Improper calculations of mass concentration Inadequate extraction efficiency data over the range of analysis Data Interpretation Improper evaluation of reported results Discounting large breakthrough results on solid sorbent samples Erroneous — Spa = Beer Te eR 8 Co BShoerne aad 83 | sw scm — = e Cane PP roe ExeSETEM [anes Tao zt 1 Loa ne 3 | pasbscines trolserarorhes i § TWEE i SRT S| pump "ERE ge a Gime T poromeren S| iiss | ee ST oie 5 | Séilng orale sal i » ow uel SB ‘asanecTER dtl SERRE eo, VOLUME SAUPLED CONCENTRATION. ‘Oita STANDARE “ a Hs aty sey 3 5 - - 5h. ee : gl _