Learn the differences between absolute and relative uncertainty, when to use each method, and best practices for accurate reporting. Enhance your lab’s precision and reliability with our comprehensive guide to navigating uncertainty in clinical measurements.
Understanding Measurement Uncertainty: Absolute vs. Relative
Last week, I received an email from a former colleague about measurement uncertainty. To my horror, I realised I had a half-written post on this very topic that I hadn’t finished. So, I am finally completing it now, hoping to provide some useful insights on measurement uncertainty for others in the medical laboratory sciences field.
What is Measurement Uncertainty?
Measurement uncertainty is a critical concept in laboratory science, particularly in the medical laboratory sciences field, where accuracy and precision are paramount. When reporting our final expanded uncertainty, we can choose between two methods: absolute uncertainty and relative uncertainty. The choice depends on how we calculate the uncertainty and how we want the result to be interpreted.
Absolute Uncertainty in Medical Laboratory Sciences
Absolute uncertainty represents the range of values around a reported result within which the expected “true” value is believed to lie, with a predetermined level of confidence. For example, a result of 10 ± 1 indicates that the true result is expected to fall between 9 and 11 with, say, 95% confidence.
Importantly, absolute uncertainty is always reported in the same units as the result. This is crucial in the medical laboratory sciences because using different units can lead to misinterpretation. Absolute uncertainty is an absolute number, determined specifically at the points where measurements were made, influenced by the contributors used in the uncertainty budget.
Best Practices for Reporting Absolute Uncertainty
- Ensure Matching Units: Always report absolute uncertainty in the same units as the result to avoid misinterpretation, especially in clinical settings.
- Validate Normal Distribution: Confirm that IQC results follow a normal distribution before reporting, as this affects the standard deviation’s validity as an uncertainty measure.
- Report Specific Points: Use absolute uncertainty when reporting a specific result, particularly near clinical decision limits where precision is critical.
Relative Uncertainty: A Useful Tool
Relative uncertainty is the ratio of the uncertainty (absolute) to the reported result. In our earlier example, the relative uncertainty would be 1/10 or 0.1, which is unitless. This method can also be extended using the Coefficient of Variation (CV%) as a measure of imprecision. When calculated this way, the result is expressed as a percentage, providing a relative standard deviation.
Pros and Cons of Relative Uncertainty
Pros:
- Ease of Understanding: Relative uncertainty is straightforward and allows for easy extrapolation of absolute uncertainty at any given result by converting the percentage to an absolute value. This is especially useful when IQC levels conform to basic requirements like “Normal/Abnormal” but do not span clinical decision limits, which are often of primary interest in medical laboratory sciences.
Cons:
- Risks of Extrapolation: Extrapolating uncertainty to results in ranges not specifically assessed for imprecision can lead to inaccuracies.
Comparing Absolute and Relative Uncertainty
Reporting absolute uncertainty across an entire measurement range can be risky. It assumes the uncertainty is constant across all potential results, which might not be true. If this assumption is incorrect, relative uncertainty may appear significantly higher for “low” results compared to “high” ones. In reality, imprecision often varies across an assay range, making relative uncertainty a more adaptable method. It keeps relative uncertainty constant, while absolute uncertainty varies according to the reported result.
Best Practices for Relative Uncertainty
- Use for Broad Assessments: Relative uncertainty is ideal for summarising performance across an assay range and monitoring against performance specifications.
- Consider Variability: Recognise that imprecision may vary across an assay range; relative uncertainty helps account for this variability.
- Avoid Unwarranted Extrapolation: Be cautious when applying relative uncertainty to ranges without direct measurements.
Combining Absolute and Relative Uncertainty Approaches
Assigning an absolute uncertainty to an entire assay can be challenging if we are unsure that all contributors (not just the imprecision component) are constant across the assay range. However, absolute uncertainty is what users typically want for a specific result within an individual or at a clinical decision point.
In practice, absolute and relative uncertainties are not mutually exclusive; both provide valuable insights. Relative uncertainty is helpful for summarising performance across an assay range and simplifying monitoring against performance specifications. In specific situations where a result is reported with its uncertainty, absolute uncertainty is preferable, as it uses the same units as the result, making interpretation easier.
Additional Considerations for Measurement Uncertainty
In addition to imprecision, other uncertainty contributors in the measurement process can influence how we report uncertainties. Experimentally derived Type A uncertainties can be expressed absolutely or relatively. However, Type B uncertainties often have less information, requiring careful consideration of how uncertainty propagates through functional relationships in the measurement process.
Conclusion
The choice between absolute and relative uncertainty depends on the context and specific needs of the analysis or clinical decision. Both have their places and can be used effectively when their strengths and limitations are understood. I am grateful to my former colleague (who is still my colleague!) for reminding me of this topic and prompting these reflections. If only I had finished this when I started it a year ago, my answers might have been clearer—lesson learnt on my part!
For further reading on measurement uncertainty and its applications in the medical laboratory sciences, explore other resources on our site. Remember, understanding the intricacies of uncertainty measurement can significantly impact your lab’s accuracy and reliability, ultimately improving patient care.
