Our Quality Assurance (QA) program at CompuChem is one of the best in the industry; our QA procedures have held up under the scrutiny of some of the most stringent state, federal, and commercial approval programs in the country. Our QA department staff consists of senior technical personnel, all with undergraduate degrees, and each with more than ten years of practical laboratory experience.

CompuChem's QA staff currently consists of two individuals including the Vice President of Quality Assurance and Technology (VPQAT) and the Quality Assurance Manager.

CompuChem's QA staff routinely conduct system audits. These audits apply specific audit methods, evaluates audit findings, and report these findings to laboratory management.

Senior management assesses the effectiveness of the quality management controls and the adequacy of resources available to achieve and assure them. The QA department auditing activities are considered in assessing the overall quality management systems. QA auditing activities are summarized in a report to the VPQAT who reports to executive staff on a monthly basis. The VPQAT also assesses the integrated quality assurance program and its performance. The assessment focuses on how well the QA program is working and identifies management problems that hinder the organization in achieving its objective quality requirements. Follow-up by the VPQAT includes an evaluation of the effectiveness of management's corrective actions.

QA Program - Method Validation

CompuChem validates all analytical methods before they become analytical services to our clients. To validate a method, one of our senior chemists reviews the published method description, assessing the instrumentation needed to perform the method, the QA/QC requirements, and any potential safety hazards. Based on his/her review, the chemist formulates a plan of testing and submits this plan to laboratory management for approval. If the chemist's proposal is approved, a full-scale method validation study is performed. All method validation studies must include:

• a method detection limit (MDL) determination (repeated annually)
• a precision and
accuracy determination

Once the method validation study has been approved by our Quality Assurance department, the laboratory in which the method is performed must write a standard operating procedure (SOP) describing the laboratory procedures involved. All SOPs must comply with CompuChem's internal SOP writing guidelines, as well as the NELAC Standards and are subject to our document control program. Our QA department approves all SOPs before they are distributed within the laboratories.

CompuChem performs annual MDL determinations for all methods currently in use. Included in CompuChem's MDL studies are verification of Practical Quantitation Limits (PQLs).

Method Validation Studies

Prior to analysis of any samples for approved methods not previously offered for sale by the laboratory, a method validation study must first be performed. A chemist familiar with the extraction/preparation procedures and the instrumental detection systems reviews the as-written method. The chemist looks for safety hazards, applicability of available instrument systems, new equipment requirements, any discrepancies in the written method, and the QA/QC requirements. A plan of testing approach is discussed with the laboratory director, technical directors, and executive management. Analyses of laboratory spiked matrices are performed to determine overall method recoverability. A draft SOP is written after the method validation study has been performed.

Documentation includes an MDL study, a precision and accuracy determination, and a draft SOP. This and any other pertinent information is then forwarded to the QA department for final approval. Any deviations from the published method must be noted in the SOP. Once approval by QA and SOP document control is completed, product codes can be developed and the new method can be offered to clients.

Data Quality Objectives

Data quality objectives (DQOs) are used in planning environmental data collection activities. They establish the level of data that are needed to support decisions regarding the site. They establish the level of uncertainty in results that a decision-maker is willing to accept. They can be used to define QA/QC programs specific to a project or data collection activity. DQOs have been established for programs under which the laboratory provides analytical services. Internal project support staff work closely with the client and regulatory agencies to ensure that DQOs will be met by the analytical results provided. The project management teams at CompuChem convey project- and client- specific requirements to the laboratory by using and distributing a unique project profile sheet (PPS) and meeting on a regular basis.

When out-of-control events occur, they are documented and corrective action is implemented. Corrective action is the process of identifying, recommending, approving, and implementing measures to counter unacceptable procedures, deviations from laboratory standard practices and procedures, or out of control quality control (QC) performance that can affect data quality. To the extent possible, samples are reported only if all QC measures are acceptable. If a QC measure is found to be out of control, and the data are to be reported, all samples associated with the failed QC measure are reported with the appropriate data qualifiers. How such events are documented depends on the type of event and area where the excursion occurred. Most out-of-control events identified during routine sample processing are restricted to single samples, batches, or data reports. Some examples of these types of events include:

• incorrect client identifier on chain-of-custody (COC) documents
• improper or broken sample containers
• unpreserved or improperly preserved samples
• lab accident during sample processing
• accuracy failures for surrogate and/or spike standards
• precision failures between duplicates
• variations in internal standard responses
• contamination of method blank
• missed holding time
• errors or omissions in data reports

Corrective action may occur prior to, during and after initial analysis. Corrective actions are performed prior to release of the data from the laboratory. The corrective action is addressed in the SDG Narrative by the case reviewer and on the analytical worksheets, signed by the analyst and reviewed by the supervisor. The types of excursions that occur and the corrective action documentation that results during different stages of sample processing are discussed in the following sections.

Corrective Action Types
CompuChem has two separate corrective action policies that are in effect and are initiated, depending on the item requiring corrective action. As a result of external and internal system audits or external and internal data audits, identified defects, findings, and even observations will result in the generation of a Corrective Action Report (CAR) by a representative of the QA department. The CAR identifies the deficiency and is generally sent to the supervisor of the appropriate area. It also may specify the action step needing to be taken. A turnaround time requirement is specified and the CAR is returned to the QA department for approval. CAR documentation is maintained in the QA department and is available to external auditors.

Sample-specific corrective actions are addressed in the laboratory Standard Operating Procedure (SOP) collection. Sample-specific corrective actions include repeats due to unacceptable surrogate recoveries, internal standards not meeting acceptance criteria, dilution re-injection requirements, etc. These types of corrective actions do not require CAR documentation. For certain occurrences, laboratory notices are prepared and are included in the SDG Narrative section of the deliverable package, as well as discussed in the narrative itself. Most of these notices are prepared after discussions with a QA department representative. They do identify the event and address the corrective action. The Case Reviewer is responsible for assessing them and addressing all sample-specific corrective actions in the report narrative. Case Reviewers are the most senior members of the technical staff and are given the authority to approve laboratory notices as the laboratory managers' designee.

When project issues are brought to our attention, the appropriate Project Manager is usually first contacted. They complete a Client Inquiry form and take it to the personnel responsible for addressing the issue. Sometimes, a CAR will be initiated based on the concern.  The policies are further discussed below.

Sample Receipt Discrepancies

Any discrepancies noted during the sample receiving process are documented on the COC form. Information about discrepancies is provided to the appropriate Project Manager, who is responsible for contacting the client about the problem and documenting the resolution in a phone log. If the client instructs the laboratory to proceed with sample analysis, a Quality Assurance (QA) Notice explaining the problem is included with the data package and the discrepancy is addressed in the SDG Narrative by the Case Reviewer. This process is commonly used when samples are received at higher than required temperatures or have been improperly preserved. If a sample is broken during transport to the laboratory, the Project Manager notifies the client, who, in turn, notifies the field crew to resample if possible. Notification of potentially high concentration samples should be conveyed from the client and entered into the LIMS laboratory instructions field.

Out-of-Control Events at the Bench

Out-of-control events noted and corrected at the bench by technicians or analysts are recorded on sample preparation worksheets, sample analysis worksheets, instrument runlogs, calibration or temperature logbooks, or instrument maintenance logbooks, as applicable. This includes condition codes on analysis worksheets and in the LIMS. Condition codes are used to describe the status of a sample analysis and whether reanalysis is required. The codes can be indicative of trends. A QC failure normally generates a request for re-extraction and reanalysis, but depends on the occurrence or method. For instance, an MS/MSD does not have to be repeated for a surrogate failure when the original unspiked field sample also demonstrates the same failure, confirming sample matrix interference. Other QC failures may prompt the reanalysis of samples and qualifying of results. The analyst makes the determination that corrective action is needed. The type of corrective action needed has been specified in SOPs through approval of the Vice President of QA and Technology. The case reviewer, authorized by the Laboratory management for release of the data, further documents these incidences in the SDG narrative.

Data Report Errors or Omissions

If errors or omissions in data reports are found during any of the review processes, the organics data are corrected online in the ThruPut data management system. The type of correction made and the person who made the correction are documented in the audit trail within ThruPut. A QA notice or laboratory notice may be included with any reported data that does not meet all quality control criteria or which requires explanation. This notice, or the SDG narrative, describes any out-of-control events and corrective actions taken to remedy them. The Case Reviewer, who authorizes release of the data report, writes the SDG narrative.

Out-of-Control Events during QA, Management, or External Audits

During various internal or external audits, an out-of-control event might be observed that can adversely affect laboratory quality on a more global basis. These events might affect, for example, entire analytical systems; projects; data usability or integrity; or sample integrity and security, or personnel safety. An unacceptable result on a performance evaluation (PE) study or internal blind PE study can also be an indication of a systematic problem. Other examples include:

• >temperature excursions on successive days in a sample storage refrigerator
• improperly calibrated analytical equipment
• poor spike recoveries on multiple extractions of a laboratory control sample
• contaminated glassware or sample/standard storage
• inaccurate/incomplete SOP or personnel not following SOP
• deviations in COC documentation or procedure
• obliterations, writeovers, or other improper data corrections
• expired standards on the "active" shelf of a refrigerator
• open or unlabeled waste containers
• inadequate hood velocity

These out-of-control events and the corrective action taken to resolve them are documented in several ways, such as:
An interoffice memorandum is written to the responsible party describing the out-of-control event and requiring a written response including corrective action by a specified date.
After internal or external audits, a formal report is written to all involved laboratory areas that describes any non-conformances or deficiencies found.
Written responses must detail corrective actions taken. For external audits, these responses are included in the formal written response to the auditor.
A Corrective Action Report form is completed when an external or internal client identifies a deficiency. This form documents what the out-of-control situation is, who is responsible for correcting the problem, what action is to be taken, and the target date for implementation. The form is used to enter information into a tracking database.
The nonconformance is closed once QA or lab management has determined that implementation is complete and that system control is restored. Reports are kept on file.
Corrective action will be initiated as a result of performance or system audits, split samples, or data audits.
Acceptance limits for CLP methods are published in the Statement-of-Work (SOW). Limits for non-CLP methods or limits that are not specified in the method are subject to change based on updated intra-laboratory statistical performance data. Generally, acceptance criteria are advisory unless the client requires specific corrective action steps for failures.

Client-Identified Discrepancies

Clients who have a question or problem with a data report will call their Project Manager (PM). The PM completes a Client Inquiry form and directs the issue to the appropriate lab supervisor for resolution. Most often these issues are related to poor copy quality, a missing page from the report, or a technical question. The PM handles client inquiries and provides a copy of the resolved inquiry to QA. Depending on the circumstances, QA may elect to issue a CAR based on the inquiry. The PM provides a response and provides supplemental documentation within an agreed upon time frame.
Where a complaint, or any other circumstance, raises doubt concerning the laboratory's compliance with our own policies or procedures, or with the quality of our tests, the laboratory ensures that those areas of activity and responsibility involved are promptly addressed and audited. Records of the complaint and subsequent actions are maintained.