---
title: "Lab Repeat Passing QAQC when it looks like it should fail"
canonical: "https://servicedesk.maxgeo.com/space/DS/1828290570/Lab%20Repeat%20Passing%20QAQC%20when%20it%20looks%20like%20it%20should%20fail"
format: markdown
---
## **Troubleshooting Guide: Lab Repeat QC Rule**

**Issue Description**

In some databases, a Quality Control (QC) rule requires that **100% of lab repeats must be within a 10% difference** from the original result.

**Example:**

- A repeat with an **18.18% difference** is marked as **Failed**.
- However, repeats with **100%** and **22.22% differences** are marked as **Passed**.

This raises the question: **Why do some repeats outside the 10% threshold pass while others fail?**

---

**Explanation**

This behavior is **not a bug**—it reflects how the **duplicate QAQC logic** is designed to work.

**Key Factors in Duplicate Evaluation**

1. **Percentage Difference Threshold**
  - The system checks the **difference between the original result and the repeat**.
  - Typically, a threshold of **20%** is used for this comparison.
2. **Proximity to Detection Limits**
  - If the results are **very close to the detection limit**, even a **small absolute change** can result in a **large percentage difference**.
  - In such cases, the system considers the **context of the values** and may **pass the repeat** despite a high percentage difference.

**Why Some Repeats Pass Despite High Differences**

- When a result is near the detection limit, the system **adjusts its evaluation** to account for the sensitivity of the measurement.
- This means a repeat with a large percentage difference may still **pass** if the absolute values are low and close to the detection threshold.

---

**Summary**

- **Not all repeats with >10% difference will fail.**
- The system uses a combination of **percentage difference** and **proximity to detection limits** to determine pass/fail status.
- This ensures that results near the detection limit are not unfairly penalized due to naturally higher variability.