USMLE Biostatistics and Epidemiology Rapid Mastery: Find the denominator, read the study, and know what the number is allowed to mean (MyBigStep High Yield Series)

Biostatistics is the strangest subject on the USMLE. The mathematics stops at division. There are maybe forty ideas in the whole of it. And it is still where competent candidates lose marks, on every Step, year after year.

Almost every mark lost in this subject is one of two things: the right calculation performed on the wrong group of people, or the right calculation followed by the wrong conclusion. Sensitivity and positive predictive value use the same numerator and different denominators. A five year survival improvement after screening is compatible with a program that has not postponed a single death. A significant p value cannot tell you whether to prescribe anything.

This book is built around that specific failure. It teaches the subject as a set of decisions rather than a set of formulas, and it names, in each case, the error the decision prevents.

What is inside The denominator framework. Four moves you run on every item before touching a number. Once you can name who is in the denominator, the formula is a formality. 22 original figures , every one generated from the numbers printed beside it: the two by two grid with its four different denominators, the effect of prevalence on predictive values, an ROC curve, confidence intervals against the null, skew, a Kaplan-Meier curve with censoring, a forest plot, a funnel plot, and the two screening biases drawn on one timeline. Nine decision algorithms. Naming the design. Naming the bias. Choosing the test. Deciding whether a result is real. Reading a trial abstract in sixty seconds. Judging a screening program. Getting from association to causation. 17 comparison tables , each with a final column naming the one feature that separates the things that get confused: confounding against effect modification, lead time against length time bias, the three ratios, superiority against non-inferiority. 105 original questions , five options each, with an explanation for the correct answer and for every distractor. The answers sit in a separate section, so you commit before you see the key. 72 calculation drills in six sets, with answers, to make the arithmetic automatic. A Final Review Command Center. Every formula in the subject on two pages, and forty rules, each naming the specific error it prevents. A seven day plan and a three day plan , each naming what to skip as well as what to review. A weakness tracker that records why you missed an item, with six error codes, because the reason repeats and the topic does not. Step 1, Step 2 CK and Step 3 emphasis set out explicitly, so you read the parts your examination rewards. How it was checked Every worked number in this book is recomputed from its own inputs by the software that builds the file, and a disagreement between a printed value and its formula fails the build rather than becoming a note for a proofreader. Every figure is generated from the numbers printed beside it, so a figure cannot disagree with the text that describes it. All text, tables, algorithms, figures, questions and drills were written for this book.

Please note 192 pages. Second edition, verified 24 August 2026. Produced with the assistance of artificial intelligence tools, under human direction. MyBigStep is not affiliated with, endorsed by or sanctioned by the NBME, the FSMB or the USMLE program. This book does not state how much of any examination is biostatistics, because nobody outside the NBME can know that. It is a preparation guide and sits alongside a question bank rather than replacing one.