Quality Assurance and Laboratory Safety
Core concepts
Separate the three quality domains: quality control (statistical monitoring of a method), quality assurance (the broader programme ensuring quality), and quality improvement (systematic change). Know the rules for accepting or rejecting a run: Westgard multirules (1-2s warning, 1-3s, 2-2s, R-4s, 4-1s, 10x), Levey-Jennings chart interpretation, and the meaning of bias versus imprecision. Understand calibration, traceability, uncertainty of measurement, and the role of reference intervals.
Safety requires knowledge of standard precautions, transmission-based precautions, and the hierarchy of controls (elimination, substitution, engineering, administrative, PPE). Know biological safety levels, sharps handling, spill management, chemical hazard communication, fire safety, and waste segregation. Understand the post-exposure pathway for needlestick injury, including source testing and post-exposure prophylaxis timing.
Traps in questions
A 1-2s rule is a warning, not a rejection. A shift (systematic bias) differs from a trend (progressive drift). Do not confuse precision with accuracy - precision is reproducibility, accuracy is closeness to truth. PPE is the last line of defence, not the first. A spill of a biological agent requires defined absorbent and disinfectant procedures, not simple wiping.
Worked rule of thumb
When a QC result falls outside limits, ask in order: is it random or systematic? Random errors produce single outliers; systematic errors produce shifts, trends, or multiple consecutive points on one side of the mean. Investigate the reagent, calibrator, and instrument before repeating the patient samples.
Laboratory Instrumentation
Core concepts
Understand the operating principles of the major analyser classes: spectrophotometry (Beer-Lambert law), nephelometry and turbidimetry, ion-selective electrodes, flame photometry, chromatography, mass spectrometry, and immunoassay formats (competitive versus sandwich). Know the components of an automated chemistry analyser (sample and reagent probes, reaction cuvettes, photometer, wash station) and the sources of carryover. For haematology, understand impedance and flow-cytometric differential counting, and the principles behind coagulation analysers (optical, mechanical, chromogenic).
For molecular and microbiology instrumentation, know thermal cyclers, real-time fluorescence detection, capillary electrophoresis, and MALDI-TOF mass spectrometry for organism identification. Understand the difference between continuous-flow and discrete analysers, random access versus batch operation, and the concept of open versus closed systems.
Traps in questions
Beer-Lambert law linearity fails at high absorbance - dilution is required. Turbidimetry measures light transmitted; nephelometry measures scattered light at an angle. Ion-selective electrodes are susceptible to pseudohyponatraemia in hyperlipidaemia and hyperproteinaemia. Carryover between samples causes falsely elevated results in the following specimen. A two-point calibration does not establish linearity across the analytical range.
Rule of thumb
For any instrumentation question, identify the physical principle first (absorbance, scatter, electrochemistry, fluorescence, mass-to-charge), then the analyte property it exploits.
Methodology and Statistics
Core concepts
Statistical literacy underpins every other domain. Know descriptive statistics (mean, median, standard deviation, coefficient of variation, percentiles) and the distinction between parametric and non-parametric tests. Understand diagnostic test statistics: sensitivity, specificity, predictive values, likelihood ratios, and receiver operating characteristic curves with area under the curve. Know the meaning of correlation (Pearson versus Spearman) and regression, and the difference between agreement and correlation (Bland-Altman, kappa).
Method comparison and evaluation require knowledge of precision studies (repeatability and within-laboratory), linearity, limit of detection and quantification, interference, and reference interval establishment. Understand the hierarchy of evidence and the interpretation of confidence intervals, p-values, type I and type II errors, and power.
Traps in questions
A p-value does not measure effect size or clinical importance. Correlation does not imply agreement. Sensitivity and specificity are properties of the test; predictive values depend on prevalence. A likelihood ratio combines both sensitivity and specificity and is prevalence-independent. Do not confuse standard deviation with standard error of the mean. Coefficient of variation is SD divided by mean, expressed as a percentage.
Worked example
For a test with sensitivity 90% and specificity 90% in a population with 10% prevalence: positive likelihood ratio = 0.90 / 0.10 = 9; negative likelihood ratio = 0.10 / 0.90 = 0.11. A positive result raises the post-test probability substantially; a negative result nearly excludes disease.
Rule of thumb
For any diagnostic study, ask: what was the reference standard, what was the spectrum of patients, and was the test applied blindly?