Data Integrity 101 for Water Labs: What You Need to Know

The implications are steep if a lab technician makes a simple transcription error resulting in the E. coli counts reading lower than they actually are. Moreover, the consequences for the lab that provided the bad data can lead to additional costs, loss of time, and damage to its reputation.

To help water labs prevent these types of mistakes, this article looks at the basics of data integrity, including core principles and best practices to follow, as well as mistakes to avoid.

What Is Data Integrity?

Data integrity focuses on the accuracy, completeness, and traceability of stored data. It is adhered to through the use of standard rules and procedures and is preserved through error-checking and validation routines.

The result for water labs is data that is legally defensible and helps clients make the best decisions. This means that the quality of the evidence is documented and instills confidence in your data and lab operations.

Core Principles for Water Labs

Accreditation to management standards like ISO 17025 and National Environmental Laboratory Accreditation Program doesn't require specific methods for ensuring the overall accuracy, completeness, and consistency of data. However, labs can ensure confidence in their data by following key principles, including:

• Documentation of the sample, calibration, methods, and quality control (QC) results enables a reconstruction of the data.
• Reference materials, standards, and reagents are traceable.
• Training records and proficiency testing results demonstrate analyst competency.
• Analysts handle samples properly, generate data according to the prescribed method, and report it correctly.
• Data quality is documented via QC results.

Your Quality Assurance Plan

Your quality assurance plan is the foundation of your lab's data quality. The program, at a minimum, should document:

• Certifications and accreditations
• Organizational chart, responsibilities, and training requirements
• Supplier quality management procedures, including evaluation of suppliers and verification of materials and equipment
• Control procedures
• Standard operating procedures, including processes for reviewing data
• Internal audit plan
• Quality management system review and QC reporting procedures
• Ethics and fraud prevention policies

Best Practices

Internal audits should include a review of data and lab practices to assess for issues. It's important to note that data integrity principles apply to both paper records and computer systems.

For paper records:

• All data should be kept in bound notebooks with numbered pages to help identify if there are missing pages.
• Pages should not be torn out or have any white-out omissions; if needed, technicians can strike through notes and initial them.
• Notebooks should be reviewed regularly to make sure data is complete and legible.
• Computer printouts can be taped into notebooks and should include a narrative description (equipment without a printout or retrievable file can raise questions).

For computer systems:

• Each user should have their own login rather than using a single shared login.
• Permissions should be set to prevent unauthorized editing or destruction of data.
• Original data should be saved with any data correction or changes recorded and justification provided according to your established data handling processes.

Preventing Lab Fraud

The EPA recognizes lab fraud as a serious issue. Employee training should address lab fraud, defined by the EPA as "the deliberate falsification of analytical and quality assurance results." Lab fraud includes:

• Fabricating data or misrepresenting facts in a narrative
• Modifying samples, analytical results, or instrument records
• Deleting non-compliant data, such as peaks on a chromatograph
• Falsifying equipment records
• Changing computer clocks so it looks like samples were analyzed within the holding time
• Inadequate handling of equipment issues
• Not reporting fraudulent behavior

Some labs have been fined millions and leaders sent to prison over data falsification. Your lab's training program should review ethics and data handling policies, including an open-door policy where technicians feel comfortable reporting fraudulent behavior.

Mistakes to Avoid

Of course, not all data quality mistakes are deliberate. Some of the most common unintentional mistakes include:

• Sharing logins on computer systems
• Recording results on scrap pieces of paper and then later transferring them to lab notebooks, a common cause of transcription errors
• Pressuring employees to process too many samples in a short amount of time, which often comes at the expense of quality
• Attributing anomalous data to equipment malfunction without further investigation

Data integrity is particularly important in water labs where the results impact public health decisions. The risks are high for labs that do not implement formal practices to protect data quality and can result in both regulatory penalties and loss of business. By instituting basic data quality procedures—and verifying employees follow them—lab managers can ensure the reliability of their data and their lab as a whole.