Calculations and Statistics
Expect calculation questions involving dilution, recovery, clearance, anion gap, osmolal gap, and basic statistics. The dilution equation C1V1 = C2V2 underlies all dilution problems; remember that a 1:10 dilution means one part sample plus nine parts diluent, not one plus ten.
Statistics: mean, standard deviation, coefficient of variation (CV = SD/mean × 100), standard error of the mean, and confidence intervals. Understand sensitivity, specificity, predictive values, and likelihood ratios, and how prevalence changes predictive values but not sensitivity or specificity. Receiver operating characteristic curves compare tests.
Quality statistics: Levey-Jennings charts plot control values against time with ±1, ±2, and ±3 SD limits; Westgard rules detect systematic versus random error. Regression analysis (Deming or Passing-Bablok for method comparison) and Bland-Altman plots assess agreement. Reference interval determination uses nonparametric percentile methods (central 95%).
Clinical calculations: anion gap = Na − (Cl + HCO3); osmolal gap = measured osmolality − calculated (2Na + glucose/18 + BUN/2.8); creatinine clearance = (U × V)/(P × time); corrected calcium = measured Ca + 0.8 × (4.0 − albumin).
Worked example: a sample diluted 1:5 reads 40 mg/dL; original = 40 × 5 = 200 mg/dL. If a control has mean 100 and SD 4, a value of 112 is +3 SD and violates the 1-3s rule, suggesting random error requiring rejection.
Trap: applying a dilution factor in the wrong direction or confusing SD with SEM.