Six Sigma in Clinical Labs: Data Leads to Changes in Quality Control

In May 2024, Attipoe received months’ worth of Six Sigma data about his laboratory’s tests as part of a pilot program developed by IVD manufacturer QuidelOrtho, of which Saint Peter’s is a customer.

The premise of the pilot centered on whether Six Sigma could influence a lab to change its quality control (QC) practices to improve the quality of test results.

“If you think about running a laboratory, you have all these different tests, and all of them have different capabilities, precision requirements, and guidelines. It’s a lot to keep track of,” says Johanna Miller, MS, global product manager and data scientist at QuidelOrtho. With Six Sigma, “you can take your entire menu, and every assay gets a number, which makes it easy to rank them relative to each other.”

Six Sigma seeks to reduce errors in workflows and processes. The approach was pioneered by Motorola in the 1980s as a way to increase quality. Over time, Six Sigma in clinical labs has proved a useful tool for organizations that need to cut costs and improve diagnostic efficiencies.

The goal of Six Sigma is to achieve a level of quality that is nearly perfect when reviewed against a million processes or opportunities. To do so, labs must collect or otherwise receive data on the volume of processes and variances. To calculate the Six Sigma level, divide the number of opportunities by the number of defects and multiplying that figure by 1 million. This analysis can be done using software or manually through sampling.

According to the Six Sigma Institute, the following are broad Sigma quality levels:

• Six Sigma processes produce only 3.4 defects per million opportunities (DPMO).
• Five Sigma gives 233 DPMO.
• Four Sigma results in 6,210 DPMO.
• Three Sigma yields 66,807 DPMO.
• Two Sigma produces 308,537 DPMO.

When dealing with volume-based processes, such as medical lab testing or manufacturing, the numbers add up if there are errors.

Overall, the news was good when Saint Peter’s received its Six Sigma report from QuidelOrtho.

“The Six Sigma metrics were applied to our clinical chemistry assays, and we were already doing remarkably well: We had 58 percent of our assays at Six Sigma based on data gathered in the background,” Attipoe says.

However, a handful of important assays did not perform as well in the data:

Aspartate aminotransferase (AST) tests

This liver enzyme test was high on Saint Peter’s list to improve the metrics for, Attipoe says.

“With AST, we just went through basic laboratory troubleshooting,” he notes. “We did recalibration, fresh reagents, and fresh QC, and after just doing that, we went from a Two Sigma to a Five Sigma for AST.”

The effort highlighted that enforcing traditional quality improvement steps can increase the performance of an assay.

Carbon dioxide (CO₂) tests

Tackling problems with CO₂ tests proved to be more complex because the assay is subject to evaporation, adverse environmental conditions, and variability in sample handling, Attipoe says.

To improve the performance of CO₂ tests, “we implemented something called dropper tips on our QC bottles,” he continues. “This reduced ambient air exposure to our QC material, helped us be precise with our aliquoting volumes, and also prevented contamination from excessive pipettes going in and out of the QC bottles.”

The result? Saint Peter’s rose from a Two Sigma to a Four Sigma for its CO₂ tests, which Attipoe characterizes as great progress.

Iron tests

An investigation into calibration timelines showed an interesting gap for iron test metrics.

“The manufacturer said to calibrate every six months, but when we looked at the data provided by QuidelOrtho, we noticed that at around the five-month mark, we started to see a little bit of a shift in our QC data,” Attipoe recalls.

Based on this revelation, Saint Peter’s decided to go beyond manufacturer recommendations and instead calibrate iron tests every five months, all to avoid QC complications.

After making adjustments for AST, CO₂, and iron tests, Saint Peter’s received an updated Six Sigma data report from QuidelOrtho in June. “We went from 58 percent of assays being Six Sigma to 65 percent,” Attipoe reports—a significant improvement.

As evidenced by QuidelOrtho’s pilot data report to Saint Peter’s, Six Sigma in clinical labs can be a useful benefit in a variety of ways.

Medical laboratory leaders may want to ask their own IVD partners if they can produce Six Sigma metrics as a way to pursue quality improvement. This may be information that is easily available to the lab from a vendor and thus save the laboratory the effort of compiling the data on its own.

“If your processes execute thousands of times, you need that consistency, and you need a way to monitor it,” Miller concludes.

https://www.clinicallab.com/six-sigma-in-clinical-labs-data-leads-to-changes-in-quality-control-28121