Introduction to Laboratory Medicine and Pathology

1. Request: A doctor prepares a pathology request form specifying the required tests.
2. Collection: The patient provides a sample (blood, stool, urine, saliva, or tissue) at a collection center, which is then transported to the lab.
3. Analysis: Laboratory staff perform the actual testing on the samples.
4. Report: Results are compiled into a report containing reference ranges and clinically relevant information.

Exam Tip: While the laboratory provides the data, the doctor always makes the final assessment and interpretation.

Sources of Variation and Error

In an ideal world, a "perfect" test would exactly measure a substance and accurately predict health status—where a positive result always indicates disease and a negative result always indicates health. In reality, several factors affect reliability.

Biological vs. Analytical Variation

Variation in test results stems from two primary sources:

• Biological Variation: This is the natural fluctuation in a human body. It can be intraindividual (changes within one person due to age, diurnal/time-of-day cycles, or seasonal changes) or interindividual (differences between different people's "setting points").
• Analytical Variation: This refers to the imprecision of the laboratory assay itself.

The precision of an assay is measured by the Coefficient of Variation (CV), which represents the variation as a percentage of the mean. A lower CV indicates higher precision.

Analytical Error: Systematic vs. Random

When performing analysis, labs must account for two types of error:

• Random Error: Unpredictable fluctuations caused by mechanical variation, electrical interference, or specimen issues like clots.
• Systematic Error: Consistent, predictable errors often caused by method bias, instrument bias, or calibration issues.

The Three Stages of Error

Errors can occur at any point in the request-test-report cycle:

1. Pre-analytical Errors: These are the most frequent and occur before the sample is tested. Examples include wrong patient identification, incorrect collection containers, improper transport temperatures, or failure to follow fasting instructions.
2. Analytical Errors: Occur during the testing phase, involving issues with precision, bias, or interference.
3. Post-analytical Errors: Occur after the test is complete, such as transcription errors, incorrect data download, or failure to interpret reference intervals correctly.

Reference Intervals

A reference interval (or reference range) is the range of values considered normal for healthy individuals in a specific population. These are typically calculated as $\pm 2$ standard deviations from the mean, meaning the range covers approximately $95%$ of a healthy population.

Because biological variation is so significant, reference ranges are not "one size fits all." They must be adjusted based on:

• Age: For example, Alkaline Phosphatase (ALP) levels are naturally higher in children due to bone growth.
• Sex: Differences in muscle mass or hormones can alter ranges for analytes like Creatine Kinase (CK).
• Fasting Status: Some tests require a fasting state to be accurate.
• Time of Day: Some analytes follow diurnal rhythms.

Quality Management: QC, QA, and QS

To mitigate errors and ensure patient safety, laboratories employ a tiered system of quality management:

• Quality Control (QC): The immediate, day-to-day process of using "control" materials (not patient samples) to check if an assay is performing within limits. Laboratories often use Westgard Rules—a set of statistical rules (like the $1_{3s}$ or $2_{2s}$ rules)—to decide whether a test run should be accepted or rejected.
• Quality Assurance (QA): A broader approach that includes the policies, protocols, and documentation used to deliver quality across the entire laboratory process.
• Quality Systems (QS): The overarching mechanism that captures incidents, manages staff accountability, and uses audits to ensure the entire system complies with documentation.

Key Distinction: A calibrator is used to adjust a measurement procedure to a known value, whereas a control is used to monitor the ongoing performance of that procedure.