Quality control and instrumentation

Quality control and instrumentation are quiet point-earners: the concepts are stable, the math is small, and the same ideas appear on every laboratory exam. If you can compute a coefficient of variation, name the Westgard rule a control failure violates, and state what Beer's law relates to what, you convert a whole section into near-automatic points. This lesson keeps the definitions crisp and the calculations short.

Accuracy and precision are not the same

These two words describe different kinds of correctness, and the board loves to test the distinction.

• Accuracy is how close a result is to the true value. Poor accuracy means a consistent offset, called bias or systematic error.
• Precision is how close repeated results are to each other (reproducibility). Poor precision means scattered results, called random error.

A method can be precise but inaccurate (tight cluster in the wrong place) or accurate but imprecise (centered on target but scattered). You want both.

Mean, standard deviation, and coefficient of variation

Three statistics summarize a set of control results.

• The mean is the average, the center of the data.
• The standard deviation (SD) measures the spread of results around the mean. A larger SD means more scatter, so a smaller SD means better precision.
• The coefficient of variation (CV) expresses the SD as a percentage of the mean, which lets you compare precision across methods and concentration levels:

CV (%) = (SD / mean) × 100

Worked example: computing CV

A control has a mean of 100 mg/dL and an SD of 4 mg/dL. Then CV = (4 / 100) × 100 = 4%. Now compare it with a second method whose mean is 200 and SD is 4: its CV = (4 / 200) × 100 = 2%. Even though both have the same SD, the second method is more precise relative to its concentration. A lower CV is better.

Levey-Jennings charts and Westgard rules

A Levey-Jennings chart plots daily control values over time against the mean and the ±1, ±2, and ±3 SD lines. It lets you see at a glance whether a run is in control and whether error is drifting in.

The Westgard multirules turn the chart into accept or reject decisions. Learn the common ones and what type of error each flags.

Rule	What it means	Type of error
1_2s	One control beyond 2 SD	Warning only; inspect further
1_3s	One control beyond 3 SD	Random error; reject
2_2s	Two consecutive beyond the same 2 SD	Systematic error; reject
R_4s	Range between two controls exceeds 4 SD	Random error; reject
4_1s	Four consecutive beyond the same 1 SD	Systematic error; reject
10x	Ten consecutive on the same side of the mean	Systematic error; reject

The big picture: random error shows up as sudden scatter (1_3s and R_4s), while systematic error shows up as a shift or trend on one side (2_2s, 4_1s, 10x). A shift is a sudden jump to a new level (often a new reagent lot), while a trend is a gradual drift (often a deteriorating reagent or light source). The 1_2s rule is only a warning that tells you to check the other rules, not an automatic rejection.

Diagnostic sensitivity and specificity

These describe how well a test identifies disease and health.

• Sensitivity is the ability to detect people who truly have the disease. High sensitivity means few false negatives: Sensitivity = TP / (TP + FN).
• Specificity is the ability to correctly identify people without the disease. High specificity means few false positives: Specificity = TN / (TN + FP).

Worked example: sensitivity and specificity

A test is run on 100 patients with the disease and 100 without. It correctly flags 90 of the sick (so TP = 90, FN = 10) and correctly clears 80 of the healthy (so TN = 80, FP = 20). Then sensitivity = 90 / (90 + 10) = 90%, and specificity = 80 / (80 + 20) = 80%. A memory hook: a highly sensitive test is good for ruling out disease when negative; a highly specific test is good for ruling in disease when positive.

Instrumentation principles

The board asks what each instrument measures and the principle behind it.

Spectrophotometry and Beer's law

A spectrophotometer passes light through a colored solution and measures how much is absorbed. Beer's law states that absorbance is directly proportional to concentration (and to path length):

A = ε × b × c

where A is absorbance, ε is the molar absorptivity, b is the path length, and c is concentration. So if a solution absorbs twice as much light, it holds roughly twice the concentration. Note that transmittance moves the other way: as concentration rises, more light is absorbed and less is transmitted.

Electrophoresis

Electrophoresis separates charged molecules in an electric field, so molecules move based on their charge and size. It is the basis of serum protein fractionation and hemoglobin separation.

Chromatography

Chromatography separates a mixture by how its components partition between a mobile phase and a stationary phase. High-performance liquid chromatography (HPLC) and gas chromatography (GC) are the workhorse examples for drugs and small molecules.

Ion-selective electrodes

An ion-selective electrode (ISE) measures the activity of a specific ion (such as sodium or potassium) by the voltage it generates, a potentiometric principle. ISE is the standard way analyzers report electrolytes.

Automation ties these principles together in high-throughput, random-access analyzers, but the underlying measurement is always one of these classic methods.

Common traps

• Swapping accuracy and precision. Accuracy is closeness to the true value; precision is agreement among repeats.
• Treating a 1_2s flag as an automatic rejection. It is only a warning to check the other rules.
• Calling a 10x or 4_1s violation a random error. Consecutive one-sided results are systematic error.
• Reversing the sensitivity and specificity formulas. Sensitivity uses false negatives; specificity uses false positives.
• Thinking absorbance falls as concentration rises. Absorbance rises with concentration; transmittance falls.
• Comparing precision by SD alone across different concentrations. Use the CV, which normalizes for the mean.

Recap

Accuracy is closeness to truth (bias), precision is agreement among repeats (random error), and the CV, SD over mean times 100, lets you compare precision fairly. Read control data on a Levey-Jennings chart and apply the Westgard rules, remembering that 1_3s and R_4s flag random error while 2_2s, 4_1s, and 10x flag systematic shifts and trends. Sensitivity uses TP over TP plus FN and rules out disease when negative; specificity uses TN over TN plus FP and rules it in when positive. For instruments, Beer's law ties absorbance to concentration, electrophoresis separates by charge, chromatography by phase partition, and the ion-selective electrode reports electrolytes by potentiometry.