INDUSTRIAL HYGIENE QUALITY CONTROL FOR SAMPLING AND LAB

INDUSTRIAL HYGIENE QUALITY CONTROL FOR SAMPLING AND LAB ANALYSIS UNIVERSITY OF HOUSTON - CLEAR LAKE 2015

QUALITY Quality product (or service) as one that is free of defects and performs those functions for which it was designed and constructed and produces Client satisfaction. (Juran) Quality Control: System of activities whose purpose is to control the quality of a product or service so that it meets the needs of users. (Taylor)

QUALITY ASSURANCE (QA) An additional QC system implemented to assess the efficacy of the QC system monitoring the product. This control system is referred to as a QA program. Thus, quality assurance could be defined as quality control on quality control.

MANAGEMENT SYSTEM Defines mission, goals, and values of organization. Also provides: financial systems, HR, and administrative functions. The quality management system provides policies, procedures, and organization that defines the quality assurance programs and how they interact with and are supported by the overall management system.

FIELD AND LABORATORY QUALITY MANAGEMENT PLAN Elements: - Organization - Management System - Document Control - Review of Requests/Tenders/Contracts - Subcontracting - Service to Customer - Purchasing

FIELD AND LABORATORY QUALITY MANAGEMENT PLAN Elements: - Control of Nonconformance - Complaints - Improvement - Corrective Action - Preventive Action - Control of Records - Internal Audits - Management Reviews and Reports

TECHNICAL QUALITY REQUIREMENTS Topics: - Selection and Training of Personnel - Selection of Methods - Estimation of Uncertainty - Control of Data - Equipment and Instrumentation - Traceability

TECHNICAL QUALITY REQUIREMENTS Topics: - Sampling - Handling of Test Items - Quality Assurance of Results - Reporting of Results Issues scaled and adjusted to organization size and scope of processes involved.

QUALITY CONTROL (QC) Purpose of quality is to provide a level of assurance that the result of a process will meet specifications. The terms: accuracy, bias, and precision are terms often used to describe how close a result is to the true or expected value.

ACCURACY “Accuracy is qualitative term referring to whethere is agreement between a measurement made on an object and its true (target or reference) value”. [NIST]

BIAS “Quantitative term describing the difference between the average of measurements made on the same object and its true value”. Bias is the difference between the average of observed results and the true value, and is determined over a period of time.

PRECISION Quantitative measurement of normal distribution of results due to random error in the system. measurements. The smaller the standard error, the more precise are the measurements. Precision is a measurement of the variability or standard error observed between the average value and the individual readings. Measures of variability include statistics like the range, variance, standard deviation, coefficient of variation, and the standard error.

PRECISION Measures of variability that are often used to evaluate precision are: range – maximum minus minimum; sample variance – differences between average of a series of measurement and the individual measurements; sample standard deviation – square root of variance; coefficient of variation – standard deviation divided by the mean; and, standard error – estimate of expected error in sample estimate of population mean or the sample SD divided by the square root of size.

QA FOR SAMPLING In order to draw conclusions about airborne contaminant concentration, the extent of current or future worker exposure, efficacy of control measures, samples must be properly collected analyzed. Sample collection and analysis are interrelated, and both are critical components of accurate data production. There must be goals and objectives for each operation.

SAMPLING VARIABILITY Caused by several factors: - Training, attitude, and attention - Representative !!! samples - Environmental factors – T/%RH/BP; sampling handling and transport; contaminant concentrations assessed - Sample collection factors – flow rate, time, and collection efficiency

VARIABILITY Rigid adherence to written sampling methods can reduce inherent variability. Materials must be consistent in quality and use. Equipment and instruments used must be appropriate for the procedures employed. Metrics for monitoring the sampling plan is through use of samples that produce results that provide comparisons: duplicate, split, spiked, and blank samples.

EVALUATION OF SAMPLING PERFORMANCE Control samples: - Duplicate samples – evaluate method - Split samples – e. g. bulk samples to labs - Spiked samples – most common; apply known mass of contaminant on media - Blank samples – field blanks; transport blanks; and, media blanks.

WRITTEN SAMPLING METHOD Use of reference method – NIOSH/OSHA Documentation of modifications, etc. Validated methods. Identify variables that cannot be controlled. Written sampling method/protocol – equipment; sampling time intervals; personal/area; handling and transport; “blanks”; recordkeeping; decontamination process; data check sequences; and personnel training.

ACCEPTABLE MATERIALS Testing of supplies and materials – QA programs Statistical sampling protocols Labels – lot-specific Certificates of Analysis Material QA/QC issues – sampling media Lab/field blanks

SAMPLER CALIBRATION � Calibration – “set of operations used to determine the accuracy of the reading of a test device to a stated uncertainty” [AIHA] � Equipment calibration and recordkeeping � Description of environmental conditions � Realistic pre- and post-calibration intervals � Written methodology for calibration � Mechanisms used for establishing traceability of calibration standards (i. e. NIST) or other recognized organizations.

PORTABLE INSTRUMENTS �Portable instruments = laboratory function. Purpose to provide immediate results useful to help make decisions. �Subject to many of same QA as laboratory. �Users trained on equipment. �Calibration before and after use; standard and routine maintenance. �QC samples for accuracy and precision on a regular basis with appropriate data analysis.

ACCREDITATION Formal recognition by a national or international authority of capability of a lab to perform testing and measurements. Purpose to provide information that will help make informed decisions regarding laboratory selection. Demonstrates lab competence and capabilities. (e. g. AIHA) AIHA – voluntary program; ISO/IEC Standard 17025; inter-laboratory proficiency programs, and other technical requirements.

CONTROL CHART Normal distribution properties: Symmetrical distribution in which the mean, median, and the mode all have the same value. See: Figure 13. 3 +/- 1 SD = 68% +/- 2 SD = 95% +/- 3 SD = 99. 7%

CENTRAL LIMIT THEOREM For random samples of size n drawn from a population with mean and SD, as n increases: mean of samples approaches population mean; SD of samples approaches SE of mean; shape of the distribution will approach the normal.

CONTROL CHART Extend lines that segment the distribution curves by standard deviation, then rotate by 90 degrees to form a control chart. See: Figure 13. 4 mean +/- 3 sigma of average is UCL/LCL mean +/- 2 sigma of average is UWL/LWL

VARIABILITY Two general types in data-producing systems: - assignable (or determinate) causes is systematic error (i. e. control chart data) - unassignable (indeterminate) causes is random error Need two types of control charts – one to deal with bias and another for precision.

CONTROL CHARTS Since bias is related to central tendency, a common type of control chart for bias plots MEANS (xbar). Precision is a measure of variability, and is commonly monitored by the use of RANGES. Combination of charts is referred to an xbar and r chart.

OUTLIER Defined as a data point that “appears to be markedly different from other members of the sample in which it occurs”. Not discarded or deleted, but indicated in set. Data could be: - an extreme value in the distribution; results from some gross deviation from analytical method or math error; so, investigate process and calculations first.

EVALUATE LAB METHODS Most IH methods used address both sampling and lab analysis. Validation. Sampling part of methods is often accepted as published and then evaluated further based on field studies and comparison with other methods. Lab portion of methods should be validated for the analytes, instrumentation, and the procedures involved (i. e. spiked samples).

SPIKED SAMPLES Sample to which has been added a known amount of analyte. The analysis of spike samples can be used to determine the bias and precision of a test method, the accuracy of a lab measurement process, and/or to detect changes in the analytical process. Need to know ranges of concentrations of interest and the relationship between recovery and concentration(s).

REPORTING LIMITS AIHA definition: “the lowest concentration of an analyte in a sample that can be reported with a defined, reproducible level of certainty”. Environmental chemistry limits: - Critical Limit – analyte detection - Detection Limit – distinguish from zero - Quantitation Limit – relatively close to the true value.

SIGNIFICANT FIGURES Labs report results to reflect the “true” value. Number of significant figures implies the precision that can be attributed to the result. General rules to apply: - The least precise measurement determines the number of significant figures. - All digits are retained during the calculation and the final result is rounded to significant digits. - Other rules for significant figures on page 324 of third edition.

UNCERTAINTY Two types of error that contribute to uncertainty: random errors and biases. - Biases – contributors that can be corrected or minimized (e. g. calibration of standards or references by labs, material prep, environ conditions). Overall average deviation. - Random errors – results of contributors that cannot be corrected (e. g. instruments, inability to repeat a process, variability, etc. ). Predominant contributor to the precision control chart. It can be measured but cannot be corrected.

INTERLABORATORY TESTS Proficiency Testing Programs by: American Industrial Hygiene Association (AIHA) Proficiency Analytical Testing (PAT) – evaluate labs analyzing workplace samples by use of reference samples (i. e. metals, silica, organics, asbestos, lead, microbial). Statistical data analysis to assess proficiency according to defined criteria. Round-robin approach.

CALCULATIONS/ PROBLEM EXAMPLES Statistics Normal distributions QA/QC Control charts