Global Strategy for Data Management in the QC Laboratory

The quality control (QC) unit of any global supply chain for a large corporate entity is responsible for the testing of products and materials across all manufacturing sites within the supply chain, as well as various contract testing sites around the world. Each of these sites generates large amounts of laboratory analytical data, including information originating from automated data systems in manufacturing and QC, processed raw data, reports, data archival, sample management, and current Good Manufacturing Practices (cGMP) compliance data.

This paper describes a worldwide strategy for automating and managing analytical data in QC laboratories at the sites that make up the global supply chain for a large corporate entity. This strategy assumes that the combined expertise of staff from each site will be utilized to promote collaboration by leveraging successes and resources currently employed at each site. The concepts contained in this strategy have been successfully demonstrated through projects piloted at each site. Most importantly, the strategy provides a structure and ensures, by its very implementation, that global regulatory compliance issues are avoided completely by the building, generation, and execution of validation protocols into the strategy.

Objectives

Following is a list of the overall goals of the strategy:

  • Share and exchange analytical information among sites by creating a global repository for analytical data, based on the NuGenesis® Scientific Data Management System (SDMS) from Waters Technologies Corp. (Milford, MA)
  • Provide global access to any QC database across the organization through a validated, third-party reporting utility, e.g., Hyperion Brio Enterprise™ reporting (Hyperion Solutions Corp., Sunnyvale, CA) or Crystal Reports (Business Objects Solutions, San Jose, CA)
  • Provide senior management with the capability to trend and statistically analyze laboratory results, based on the NuGenesis SDMS and Hyperion Brio Enterprise reporting
  • Deploy a global laboratory information management system (LIMS) with future global QC laboratory for the organization
  • Align and leverage automated processes across all sites in an effort to minimize costs
  • Utilize successful system processes and best practices from each site
  • Provide long-term support for the applications and technology.

Benefits

The key benefits of the strategy are as follows:

  • Facilitate the creation of a distributed QC laboratory for the organization
  • Provide management with a high-level interpretation of the results to make sound business decisions
  • Link results in analytical reports to the raw data source (e.g., at the instrument) through the global repository
  • Ensure a consistent automation approach across all sites
  • Ensure the long-term integrity of the raw source data.

In order to implement this strategy, resources are needed at each site to make the transition from a manual, paper-based system to an electronic, paperless process. In addition, an experienced core team with knowledge of laboratory operations and the processes at each site must be utilized to assist each site in fully implementing the strategy.

Overview

Vast amounts of analytical data are generated in microbiological, chemical, and biochemical laboratories across the supply chain. These data are in either paper-based or electronic-based form. Sources include:

  • Paper records
  • Notebooks
  • Benchtop data systems
    • Standalone equipment
    • PC-based equipment
  • Processing systems
    • Hand calculations
    • Assay spreadsheets
    • Data files
    • LIMS.

Figure 1 - Data landscape without automation.

A data landscape for analytical data that is not fully automated is shown in Figure 1. In this scheme, data generally are manually transferred from the many analytical data sources to the desired processing mechanism. This approach is very inefficient and costly, and often requires data to be duplicated at multiple targets.1

As individual sites develop their approach to handling analytical data, systems and processes are utilized that facilitate site culture and optimize the manual process without automated systems to track and analyze data. This site-by-site approach works well when laboratories handle data independently of other sites in a global organization. However, integration into a global business community, where information and resources are shared with a parent entity, makes alignment of laboratory practices, terminology, and systems across sites very critical to the leveraging of costs and resources across site cultures and communities.2

Alignment means that some business practices and terminology will have to be changed at each site. Best practices will have to be evaluated and determined for fit at each site. In determining best practices and alignment, each site must keep an open mind and understand how other sites arrived at their current business processes. Understanding the terminology used at each site is extremely important in the alignment process. For example, aligning laboratory metrics might be one goal. Two sites are compared. One site is found to generate and process more samples per month than the other site. A closer examination of the metrics finds that the “less efficient” site logs multiple analytical tests or assays per sample into its LIMS whereas the “more efficient” site logs one test or assay per sample into its LIMS. Final analysis shows the two sites to be equally efficient after the definitions of test and sample are clarified.

Strategy

As discussed above, access to analytical data in the nonautomated laboratory is difficult and chaotic. Since data generation systems and processes are isolated and not interconnected, information must be transferred manually and rekeyed into other systems. Reports are largely paper based and archived to off-site storage. Searching for information is even more difficult. The strategy described in this paper leverages current automated system approaches from all sites within the supply chain. Implementing this strategy will enable a global, seamless integration of analytical information across the organization. Numerous examples of deploying this approach are found in the research and development environment.3–8

Figure 2 - Data landscape with automation.

The strategy outlined in Figure 2 demonstrates how a centralized data repository can be utilized to organize and catalog the chaotic analytical data landscape described previously and shown in Figure 1. This strategy effectively utilizes technology currently deployed and validated at various sites and ties the information together in a central repository. However, the strategy proposed in this paper takes the data management task to a new level of integration. Third-party reporting tools from either Hyperion Brio Enterprise reporting or Crystal Reports can enable two-way access to the analytical data between the repository and the various analytical databases and/or processing applications.

The following sections will describe five overall areas of a comprehensive data management strategy to organize the volume of analytical data generated in the QC laboratory. An overview of each of these areas will be presented. Although it might be the best approach to implement all five areas of the strategy at the same time, it is not required. A modular approach can be taken by evaluating the priorities and current system state at each site, then implementing plans based on critical needs and/or weaker areas. For example, a site may already have a LIMS and want to focus on analytical data management3,7,8 or reporting as a first step.

Analytical data management

Analytical data from the laboratory are generated in two forms: paper or electronic. Paper-based data are entered into notebooks and preprinted forms. Even automated systems, such as the benchtop equipment and processing mechanisms mentioned earlier, produce paper reports that are filed with other paper records. Paper records make sharing of analytical information in a global organization challenging. Electronic systems such as PC-based instruments, assay spreadsheets, and LIMS create data files and database records that can be shared if handled appropriately. Based on actual data from one site, a single chromatography data system generates approx. 2000 files per month while another nonchromatography system generates 10,000 files per month. These data are processed by other systems to generate more analytical results for management and business decisions. The strategy in this document will outline an approach to facilitate the management and sharing of this analytical information with other sites in the global organization.

Figure 3 - Raw data management strategy.

The focus of raw data management is at the raw data and processing level of analytical data. Good data management practices are important to both business operations and cGMP compliance. This area concentrates on keeping data available to the user, minimizing the loss of critical data and downtime on the instrument PC. “Off-the-shelf” systems can readily be deployed at various sites to manage these types of data. For example, the ChemStore™ database (Agilent Technologies, Palo Alto, CA) may be used to manage data from the ChemStore chromatography data systems (CDS), and the NuGenesis SDMS may be utilized to manage raw data files and reports from any non-CDS instrument systems on the company network.3,7,8 This analytical data management approach is shown in Figure 3. As depicted in the diagram, the NuGenesis SDMS can also be used to manage files from the ChemStore database.

The latter system is also used to manage forms and analytical calculations generated by Microsoft® (Redmond, WA) Excel™ spreadsheets. Further regulatory compliance can be applied to Excel and Microsoft Word templates using the Application Control module from the NuGenesis SDMS. This module utilizes the NuGenesis SDMS to capture audit trails and ensure that all versions of the spreadsheet are captured regardless of which workstation is used to run the controlled template.