Method for determining reference changes from patients' serial data: example of cardiac enzymes

Area of Science:

• Clinical Chemistry
• Biostatistics
• Laboratory Medicine

Background:

• Serial laboratory test results can show significant changes within reference intervals.
• Establishing reliable reference change limits is crucial for interpreting longitudinal patient data.

Purpose of the Study:

• To develop and validate a statistical method for calculating reference change limits from routine patient data.
• To apply this method to cardiac enzyme analytes: creatine kinase (CK), CK isoenzyme MB (CK-2), lactate dehydrogenase (LD), and LD isoenzyme 1 (LD-1).

Main Methods:

• Developed a statistical method to derive reference change limits from routine patient data.
• Applied the method to cardiac enzyme data from 2029 consecutive patients, excluding those with myocardial infarction or myocarditis.
• Validated the method using conventional statistical approaches on a small group of hospitalized patients without cardiac symptoms.

Main Results:

• Derived clinically applicable reference change limits for CK (-39 to 27 U/L), CK-2 (-8 to 7 U/L), LD (-86 to 85 U/L), and LD-1 (-19 to 15 U/L).
• Obtained similar reference change limits from a smaller group of hospitalized patients without cardiac symptoms.
• Demonstrated the method's applicability on unselected patient data.

Conclusions:

• A statistical method can effectively calculate clinically applicable reference change limits from routine laboratory data.
• This method aids in identifying significant physiological changes in serial laboratory results.
• The approach is valuable for improving the interpretation of laboratory diagnostics in clinical practice.