By Victoria Higgins, PhD, FCACB
What if the reference intervals we rely on every day — those trusted “normal” ranges that guide clinical decision-making — are leading us astray? For decades, labs have defaulted to the central 95% interval, which is based on the idea that the same approach will work for every test and every population. Increasingly, that assumption is being questioned.
Today’s session, titled “Approaches and assumptions in laboratory reference interval determination” will challenge outdated norms and explore modern, evidence-based alternatives. ADLM 2025 attendees will examine the history of reference intervals, compare direct and indirect derivation methods, and hear about real-world cases where traditional approaches have failed.
“Reference intervals play a huge role in how test results are interpreted by clinicians — and ultimately, in the decisions they make about patient care,” says Tony Badrick, PhD, MBA, FADLM, who will open the session. “If the intervals are off, even slightly, it can lead to over- or under-diagnosis, which can have serious consequences.”
Badrick will take the audience back in time to explore the origins and limitations of the current paradigm. Then he’ll explain why it’s time to move beyond it. “Looking ahead helps us imagine what’s possible … but looking back is just as important,” Badrick notes. “Understanding the history behind our current reference intervals helps highlight why change is needed in the first place.”
Joining him is Daniel Holmes, MD, FRCPC, who will dive into the technical underpinnings of how reference intervals are determined. He will compare direct methods, which rely on well-characterized healthy cohorts, with indirect approaches, which extract reference intervals from routine patient data using statistical modeling.
The session’s final speaker, Joe El-Khoury, PhD, DABCC, FADLM, will underscore why change is urgently needed using real-world examples, including thyroid stimulating hormone (TSH) and alanine aminotransferase (ALT). “Take TSH, for instance: Current reference intervals have been linked to significant overdiagnosis of subclinical hypothyroidism,” Badrick emphasizes. “In fact, estimates suggest that up to 90% of people taking levothyroxine may not actually need it.”
This session will explore alternatives such as clinical decision limits, personalized reference intervals, and z-scores, highlighting how each option has specific advantages and limitations. These alternatives are practical tools that labs can implement, especially when traditional methods fall short.
The speakers are calling for change — which demands action from all laboratorians. But how should laboratorians take the first step? “Every test is different, and so the approach needs to be tailored,” Badrick says. Implementing such an approach requires considering how each test is used clinically (e.g., what level of sensitivity or specificity is needed?) and incorporating biological variation and pre-analytical factors. “Changing a reference interval isn’t something we do in isolation,” Badrick says. “It’s essential to engage our clinical colleagues, understand their needs, and make decisions together.”
The message is clear: It’s time to evolve. Reference intervals should reflect real biological variability and be grounded in clinical utility — not historical convenience. “The one-size-fits-all approach we currently use for reference intervals is outdated — literally rooted in methods that trace back to 18th-century astronomical observations,” Badrick adds. “Evolving our approach here isn’t just a technical improvement. It’s about delivering better care for both clinicians and the patients they serve.”
