Browse topics
Biostatistics and epidemiology
24 notes, 75 questions and 201 flashcards.
Tier 1 is core knowledge expected of every trainee. Tier 2 is deeper or less common.
Cardiology
Clinical pharmacology
General medicine
- Absolute, relative and attributable risk, NNT and NNHFree sampleCalculating relative risk, relative and absolute risk reduction, NNT and NNH from trial data, why baseline risk drives the NNT, and attributable and population attributable risk.Tier 1, medical sciences concept, 3 questions, 9 flashcards
- Bias, validity and precision in clinical studiesFree sampleRandom versus systematic error, internal and external validity, the main selection and information biases, and a worked example of how misclassification moves an odds ratio.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Case-control studiesFree sampleSampling on the outcome, choosing controls, calculating unmatched and matched odds ratios, the rare disease assumption, and the biases that case-control studies are prone to.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Choosing a statistical test: parametric and non-parametric testsMatching a test to the outcome type, the number of groups, pairing and the shape of the data, with a worked chi-squared calculation and the parametric and non-parametric pairs.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Cohort studies: prospective and retrospectiveHow cohort studies run forwards from exposure to outcome, the measures they yield (risk, rate, RR, attributable risk), prospective versus retrospective designs, and their strengths and weaknesses.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Confounding and effect modificationWhat makes a confounder, how stratification separates confounding from effect modification, the ways to control confounding, and the hormone therapy story as the classic example.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Data types, summary measures and the normal distributionTypes of data, mean, median, standard deviation and interquartile range, skew, and how the normal distribution sets reference ranges and DXA T-scores.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Diagnostic test accuracySensitivity, specificity, predictive values and likelihood ratios, with a worked pre-test to post-test probability calculation.Tier 1, medical sciences concept, 3 questions, 10 flashcards
- Incidence, prevalence, attack rate and case fatalityDefinitions and worked calculations for incidence (risk and rate), prevalence, attack rate, case fatality and population attributable risk, and how they relate to one another.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Odds ratios and their relationship to relative riskHow to calculate an odds ratio from a 2×2 table, why case-control studies report it, and when it does and doesn't approximate the relative risk.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- P-values, type I and II errors and statistical powerWhat a p-value does and doesn't mean, type I and type II errors, what sets power and sample size (with worked calculations), multiplicity, subgroups and interim analyses.Tier 1, medical sciences concept, 4 questions, 9 flashcards
- Randomisation, allocation concealment, blinding and intention-to-treatWhat randomisation, allocation concealment and blinding each protect against, how intention-to-treat differs from per-protocol analysis, with a worked NNT example and the RoB 2 domains.Tier 1, medical sciences concept, 3 questions, 9 flashcards
- Standard deviation, standard error and confidence intervalsWhy the standard error shrinks with sample size, how to build a 95% confidence interval for a mean, a proportion and a difference, and how to read intervals against no effect and clinical relevance.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Survival analysis, Kaplan-Meier curves and hazard ratiosCensoring, the Kaplan-Meier product-limit method, the log-rank test and Cox hazard ratios, with a worked conversion from a hazard ratio to an absolute benefit and NNT.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Systematic reviews and meta-analysisHow a systematic review is built, how a meta-analysis pools trials by inverse-variance weighting, reading a forest and funnel plot, heterogeneity (I²), publication bias and GRADE.Tier 1, medical sciences concept, 3 questions, 9 flashcards
- Trial designs: parallel, crossover, factorial, cluster and non-inferiorityWhen to use parallel, crossover, factorial, cluster and adaptive designs, the cluster design effect, and how to read a non-inferiority result against its margin.Tier 1, medical sciences concept, 4 questions, 8 flashcards
- Trial endpoints: primary, surrogate and compositePrimary and secondary endpoints, multiplicity, when a surrogate can be trusted (CAST, ILLUMINATE, ACCORD), and how to dissect a composite endpoint with a worked example.Tier 1, medical sciences concept, 3 questions, 8 flashcards
- Cross-sectional and ecological studiesPrevalence surveys and population-level comparisons, the prevalence = incidence × duration relationship, why neither design shows temporality, and the ecological fallacy.Tier 2, medical sciences concept, 3 questions, 8 flashcards
- Crude, specific and standardised ratesWhy crude rates mislead when populations differ in age structure, and worked examples of direct standardisation and the standardised mortality ratio (indirect standardisation).Tier 2, medical sciences concept, 3 questions, 7 flashcards
Geriatric medicine
Medical oncology
- Oncology trials: dose escalation, response criteria and survival endpointsThe 3+3 dose-escalation design, RECIST 1.1 and iRECIST response with a worked tumour measurement, and how to read overall survival, progression-free survival and hazard ratios in cancer trials.Tier 2, medical sciences concept, 4 questions, 9 flashcards
- Screening: principles, lead-time and length bias, overdiagnosis, benefits and harmsWilson-Jungner criteria, the four Australian cancer screening programs (including lung from July 2025), lead-time, length and overdiagnosis bias with worked numbers, and why only a mortality reduction in an RCT proves a screening test works.Tier 2, medical sciences concept, 3 questions, 9 flashcards