CSV Testing: What You Need to Know

You can’t have validation without testing. It’s the cornerstone of ensuring reliability and functionality and, for regulators, the testing process demonstrates compliance from life sciences manufacturers and indicates a good faith effort in ensuring patient safety and product quality.

All validation types require testing of some sort. This article explores the testing methods specific to computer system validation (CSV), and how they are changing in response to the FDA’s guidance on Computer Software Assurance (CSA).

CSV Testing Requirements

Before we examine the types of tests we can use for CSV, let’s look at what a test must accomplish to meet the regulatory requirements of the U.S. Food and Drug Administration (FDA) and European Medicines Agency (EMA).

CSV testing is the meticulous process of verifying and documenting that a computer system used in production of an in-scope product (such as medical device or pharmaceutical) fulfills its intended purpose consistently and compliantly. So, tests must show that systems and software effectively achieve their goals, for example a digital interface correctly gathers data from connected readers, or a document control system is correctly sorting documentation.

Tests must be recorded in compliance with data integrity rules and evidence should be collected. This creates a large data burden on CSV testers and makes Computer System Validation one of the most resource intensive forms of validation. This challenge is one reason the FDA announced its guidance on CSA — a risk-based approach to CSV that reduces the burden on teams by using vendor tests or scaled testing on non-high risk products.

Types of CSV Testing

There are three types of CSV tests:

• Scripted Testing
• Unscripted Testing
• Ad-Hoc Testing

Scripted testing is akin to following a meticulously drawn navigation chart. It involves the creation of detailed test scripts or cases before executing them. These scripts outline the expected inputs, actions, and outcomes, providing a structured roadmap for validation.

It’s repeatable, comprehensive, and is backed by documentation to withstand even the heaviest scrutiny. This makes scripted testing the ideal choice for “high risk” systems, which are applications that directly impact product quality or human safety and have the potential to cause or impede a product recall.

Unscripted testing is a more flexible approach, akin to navigating by the stars. Testers rely on their expertise and domain knowledge to explore the system organically, uncovering potential issues that scripted tests might overlook. Where scripted tests record actual results, unscripted tests may use a “Pass/Fail” recording. These tests require less documentation, but still ensure an application is fit for purpose. They are ideal for non-high-risk systems.

These tests mimic real-world usage and are more flexible than scripted tests.

Ad-Hoc Testing is like steering on instinct. It involves spontaneous and unplanned testing as testers explore the system without predefined test cases or formal procedures, relying on their intuition and experience.

These tests are highly flexible and can quickly identify critical issues that may be missed in formal testing. They also may leverage a “Pass/Fail” function and because of their variability, are recommended for use in non-high-risk systems.

Vendor Testing are tests executed by the software or system developer to demonstrate a product’s compliance. This can reduce the burden on life sciences validation teams, but it is recommended manufacturers maintain evidence of a vendor’s tests and execute appropriate scrutiny to ensure such testing was accurate. This is also recommended for non-high-risk systems.

Kneat’s CSA expert, Darren Geaney, explains how to effectively leverage Supplier Activities so you can take advantage of CSA’s benefits without exposing your company to risk in this on-demand webinar. Watch it now.

Digital Validation to Execute CSV Testing

Even these changes to CSV testing can only go so far if you’re performing them with paper-based or hybrid validation processes. Digital validation, enabled with Kneat Gx, can drastically reduce cycle times, improve compliance, and speed time to market. See how Fujirebio Diagnostics, cut their CSV work time in half with Kneat.

About the Author

Tristan Worden – Senior Content Marketing Strategist, Kneat

Tristan has over a decade of experience communicating complex ideas to both general and specific audiences with almost five years in the compliance and validation sector. He is passionate about ensuring validation professionals have the information they need to do their jobs effectively.