SECTION 1: Foundations of Paper Critique

1.1 Purpose and Format of the FRCS Part 2 Viva

What the Examiners Are Looking For

The Academic Viva tests three core competencies:

1. Critical Appraisal Skills - Can you identify strengths and weaknesses in research methodology?
2. Statistical Understanding - Can you interpret results and determine clinical significance?
3. Clinical Application - Can you apply research findings to patient care?

Typical Question Progression

The viva typically follows this pattern:

1. Opening (2 min): "Tell me about this paper" - broad overview
2. Study Design (3 min): Questions about methodology, randomization, blinding
3. Results (3 min): Interpreting tables, graphs, p-values, confidence intervals
4. Critique (3 min): Strengths, limitations, bias
5. Application (2-4 min): How would this change your practice?

1.2 How to Read a Surgical Paper (The 3-Pass Method)

With only 10 minutes, you need a systematic approach. Here's the proven 3-pass method:

Pass 1: The Quick Scan (2 minutes)

Pass 2: Methods & Results Deep Dive (5 minutes)

Focus on these key areas:

In Methods:

• Inclusion/exclusion criteria - who was included and why?
• Randomization method - was it adequate?
• Blinding - who was blinded (patient, surgeon, assessor)?
• Primary outcome - what's the main thing they measured?
• Sample size calculation - did they plan adequately?

In Results:

• Patient flow diagram - how many dropped out and why?
• Baseline characteristics table - were groups comparable?
• Main results table - what are the p-values and confidence intervals?
• Any subgroup analyses - treat these with suspicion

Pass 3: Critical Analysis (3 minutes)

Mentally prepare answers to these questions:

1. What are 2-3 strengths of this study?
2. What are 2-3 limitations or sources of bias?
3. Do the results apply to my patients?
4. What would I tell a patient about these findings?

1.3 Understanding Study Designs: Quick Overview

Research studies fall into two broad categories:

1. Observational Studies (No Intervention)

• Cohort studies: Follow groups forward in time (e.g., smokers vs non-smokers → lung cancer)
• Case-control studies: Look backward from outcome (e.g., lung cancer patients vs controls → smoking history)
• Cross-sectional: Snapshot at one time point (prevalence studies)

2. Experimental Studies (Intervention)

• Randomized Controlled Trials (RCTs): Gold standard - random allocation reduces bias
• Quasi-experimental: Intervention without randomization

For detailed study design features, strengths, weaknesses, and when each is appropriate, see Section 2: Research Study Designs Explained

1.4 Basic Statistics

P-Values: What They Really Mean

The p-value is the probability of seeing results at least as extreme as observed, assuming the null hypothesis is true.

Common Misconceptions:

• ❌ "p=0.05 means there's a 5% chance the null hypothesis is true" (WRONG)
• ❌ "p<0.05 means the result is clinically important" (WRONG)
• ✅ "p<0.05 means we can reject the null hypothesis at the 5% significance level" (CORRECT)

Confidence Intervals (CI): More Informative Than P-Values

A 95% CI means: "If we repeated this study 100 times, we'd expect the true value to fall within this range in 95 of those studies."

How to Interpret Confidence Intervals:

Example:

Hazard Ratio = 0.75 (95% CI: 0.60-0.94)

This means:

• Point estimate: 25% reduction in hazard
• We're 95% confident the true reduction is between 6% and 40%
• Since CI doesn't cross 1, this is statistically significant
• The intervention appears beneficial

Statistical vs Clinical Significance

1.5 Bias and Confounding

Selection Bias

Occurs when the study sample doesn't represent the target population.

Examples:

• Volunteer bias: Patients who agree to join trials may be healthier/more motivated
• Survival bias: Only including patients who survived to enrollment excludes early deaths
• Exclusion criteria too strict: Trial only enrolling fit patients but results applied to frail elderly

Performance Bias

Differences in care between groups other than the intervention being studied.

Examples:

• Intervention group gets more follow-up visits
• Different surgeons operate on different groups
• One group receives more supportive care

Prevention:

Blinding of participants and care providers, standardized protocols for both groups.

Detection Bias

Systematic differences in how outcomes are measured between groups.

Examples:

• Assessors knowing which treatment was given look harder for complications
• Patients in intervention arm report symptoms more enthusiastically
• Different tests or thresholds used to detect outcomes

Prevention:

Blinding of outcome assessors, objective outcome measures, standardized assessment protocols.

Attrition Bias

Systematic differences in withdrawals from the study.

Red Flags:

• Different dropout rates between groups
• Missing data not random (e.g., sicker patients drop out)
• Per-protocol analysis instead of intention-to-treat

Confounding

A confounding variable is associated with both the exposure and the outcome, making it appear there's a relationship when there might not be (or masking a true relationship).

Classic Example:

Observational study shows coffee drinkers have higher lung cancer rates. Confounded by smoking (smokers drink more coffee AND get lung cancer from smoking, not coffee).

In Surgery:

• Comparing open vs laparoscopic surgery, but surgeon experience differs
• Comparing two hospitals, but case-mix differs (one takes sicker patients)
• Comparing two time periods, but management protocols changed

How to Control Confounding: