Why Biostatistics Matters

The Clinician’s Statistical Toolkit

Learning Objectives

  1. Explain why statistical literacy is essential for clinical practice
  2. Distinguish between qualitative (categorical) and quantitative (numerical) data
  3. Classify variables using NOIR: Nominal, Ordinal, Interval, Ratio
  4. Choose appropriate statistics based on measurement scale
  5. Distinguish descriptive from inferential statistics
  6. Identify independent and dependent variables
  7. Recognise how statistical errors cause real patient harm

Clinical Hook: The CAST Trial

The Deadly Assumption (1980s)

  • Post-MI patients have PVCs
  • PVCs predict sudden cardiac death
  • Assumption: suppress PVCs → reduce mortality

The CAST Result (1989)

  • Antiarrhythmics doubled mortality (RR = 2.5)
  • Trial stopped early
  • Estimated tens of thousands of excess deaths per year in the US alone

Surrogate Endpoint Fallacy

Suppressing a risk marker is not the same as preventing the disease. The marker (PVCs) was treated instead of the patient.

More Examples of Statistics Gone Wrong

Case What happened Module
HRT Observational studies said it prevented heart disease. RCT (WHI) showed it caused it. 6, 14
Hydroxychloroquine Preliminary data drove global prescribing. RCTs showed no benefit. 8
Vioxx Cardiovascular risk hidden in subgroup analyses. Withdrawn after ~88,000 excess MI. 14

As a clinician, you will read research every week. This course gives you the tools to read it critically.

What Is Biostatistics?

Three practical functions for clinicians:

Summarise

When you look at lab reports over time or compare ward outcomes — that’s descriptive statistics.

Conclude

A BP of 142/90 in one reading doesn’t confirm HTN. A p = 0.04 doesn’t prove a drug works. Statistics handles uncertainty.

Evaluate

Understanding study design, bias, and analysis lets you separate reliable evidence from noise.

Data Types in Medicine

Figure 1

The NOIR Classification

Stevens’ Four Levels of Measurement

Scale Order? Equal intervals? True zero? Example
Nominal No No No Blood group, Sex
Ordinal Yes No No NYHA class, GCS
Interval Yes Yes No Temperature (°C)
Ratio Yes Yes Yes Height, Weight, BP

The Ratio Test

Can you meaningfully say “twice as much”? If yes → Ratio. If no → Interval or lower.

  • “14 g/dL Hb is twice 7 g/dL” ✓ → Ratio
  • “40°C is twice as hot as 20°C” ✗ → Interval

NOIR: Permissible Operations

Figure 2

Choosing the Right Statistics

Scale Central Tendency Spread Tests
Nominal Mode Frequency Chi-square, Fisher
Ordinal Median IQR Mann-Whitney, Spearman
Interval Mean, Median SD, IQR t-test, ANOVA, Pearson
Ratio Mean, Geo. mean SD, CV All parametric tests

Common Mistake

Computing the mean of NYHA class or Likert scale data. These are ordinal — use the median and non-parametric tests.

Descriptive vs Inferential Statistics

Figure 3

The Research-to-Bedside Pipeline

Figure 4

Variables and Their Roles

Figure 5

Confounders

A confounder is associated with both exposure and outcome, distorting the apparent relationship. Coffee → Lung cancer? No — smoking is the confounder.

Key Takeaways

  1. Statistics is not optional — flawed reasoning has killed patients (CAST, HRT, HCQ)

  2. NOIR classification determines which statistics and tests are valid

  3. Interval ≠ Ratio — the “true zero” test (Temperature °C vs Height cm)

  4. Descriptive → Inferential — from “what I see” to “what is true”

  5. Independent → Dependent — beware of confounders

Further Learning

Videos:

Books:

  • Bland M. An Introduction to Medical Statistics. Ch 1–2.
  • Indrayan A. Medical Biostatistics. 4th ed. (Indian examples)