High Yield Overview

STUDY DESIGN TYPES

Research Methodologies | Study Hierarchy | Evidence Quality

Level IRCTs and Systematic Reviews

Level IIProspective Cohort Studies

Level IIICase-Control Studies

Level IVCase Series and Expert Opinion

Study Design Hierarchy

Experimental

PatternRCT - Randomization controls confounding

TreatmentHighest quality evidence

Observational Analytical

PatternCohort/Case-Control - No randomization

TreatmentModerate quality evidence

Observational Descriptive

PatternCase Series/Cross-sectional

TreatmentLower quality evidence

Critical Must-Knows

RCT: Random allocation eliminates selection bias and balances known/unknown confounders
Cohort Study: Follows exposed and unexposed groups forward in time to measure outcomes
Case-Control Study: Starts with disease (cases) and no disease (controls), looks backward for exposures
Cross-Sectional Study: Snapshot in time - measures exposure and outcome simultaneously
Case Series: Descriptive study of patients with similar condition - no comparison group

Examiner's Pearls

"
RCT is gold standard for therapeutic interventions but not always ethical or feasible
"
Cohort studies are best for rare exposures; Case-control studies are best for rare outcomes
"
Observational studies are prone to confounding and bias - must use statistical adjustment
"
Registry studies provide real-world effectiveness data but lack randomization

Critical Study Design Concepts

Experimental vs Observational

Experimental: Investigator assigns intervention (RCT). Observational: Investigator observes without intervention (Cohort, Case-Control).

Prospective vs Retrospective

Prospective: Data collected going forward from study start. Retrospective: Uses existing data from past records.

Randomization Importance

Randomization balances: Known confounders, Unknown confounders, Selection bias. Creates comparable groups at baseline.

Internal vs External Validity

Internal: Are results valid within study? External: Can results be generalized to other populations?

At a Glance

Research study designs form an evidence hierarchy with randomized controlled trials (RCTs) at the apex (Level I) because randomization eliminates selection bias and balances both known and unknown confounders. Cohort studies (Level II) follow exposed vs unexposed groups forward in time—best for rare exposures. Case-control studies (Level III) compare cases with disease to controls without, looking backward for exposures—best for rare outcomes. Observational designs are prone to confounding and bias requiring statistical adjustment. The key distinction is experimental (investigator assigns intervention) vs observational (investigator only observes), and internal validity (are results valid within the study?) vs external validity (can results be generalized?).

Mnemonic

RCCCCEStudy Design Hierarchy (Therapeutic Questions)

Randomized Controlled Trials

Level I - Gold standard for treatment

Cohort Studies (Prospective)

Level II - Follow groups forward

Case-Control Studies

Level III - Compare cases to controls

Case Series

Level IV - Descriptive series

Cross-sectional Studies

Prevalence surveys

Expert Opinion

Level V - Lowest evidence

Memory Hook:Research Creates Clear Clinical Conclusions Effectively - from highest to lowest quality evidence!

Mnemonic

FINERChoosing the Right Study Design

Feasible

Can you complete the study with available resources?

Interesting

Does it address an important clinical question?

Novel

Does it fill a gap in current knowledge?

Ethical

Can it be done without harm to participants?

Relevant

Will results impact clinical practice?

Memory Hook:FINER criteria help you choose the right research question and design!

Overview/Introduction

Randomized Controlled Trial (RCT)

Definition: Participants are randomly allocated to intervention or control groups, then followed prospectively to measure outcomes.

Key Features:

Randomization: Eliminates selection bias and balances confounders
Prospective: Follows participants forward in time
Control Group: Provides comparison to measure treatment effect
Blinding: Can be single-blind, double-blind, or triple-blind

RCT Variations

Design	Description	Advantage	Disadvantage
Parallel Group	Two separate groups compared	Simple analysis, most common	Requires large sample size
Crossover	Each participant receives both treatments	Smaller sample needed, controls for individual variation	Requires washout period, carryover effects
Factorial	Tests 2 or more interventions simultaneously	Efficient, can assess interactions	Complex analysis, increased sample size
Cluster	Groups (hospitals, clinics) randomized, not individuals	Prevents contamination, practical	Larger sample needed, complex statistics

Strengths of RCTs:

Highest level of evidence for therapeutic questions
Minimizes bias and confounding
Establishes causality

Limitations of RCTs:

Expensive and time-consuming
May not reflect real-world practice (narrow inclusion criteria)
Not ethical for harmful exposures
Not feasible for rare outcomes

Understanding these experimental designs is essential for critically appraising treatment studies.

Concepts and Principles

Evidence Hierarchy Principles

The evidence hierarchy is fundamental to understanding study quality:

Level I Evidence: Systematic reviews/meta-analyses of RCTs, or individual high-quality RCTs

Provides strongest evidence for causation
Randomization controls for known and unknown confounders
Gold standard for therapeutic questions

Level II Evidence: Prospective cohort studies, lesser-quality RCTs

Cannot prove causation (association only)
Prone to confounding and selection bias
Appropriate when RCTs are not ethical/feasible

Level III Evidence: Case-control studies, retrospective cohort studies

High risk of recall bias and selection bias
Best for rare diseases or outcomes
Establishes temporal relationship for case-control

Level IV Evidence: Case series, cross-sectional studies

No comparison group (case series)
Cannot establish temporal relationship
Useful for describing disease characteristics

Level V Evidence: Expert opinion, case reports

Lowest level of evidence
Subject to individual bias and experience
May generate hypotheses for future research

Observational Analytical Study Designs

Cohort Studies

Definition: Follows groups with and without exposure forward in time to compare incidence of outcomes.

Types:

Prospective Cohort Study

Process:

Identify exposed and unexposed groups at baseline
Follow both groups forward in time
Measure incidence of outcomes
Calculate relative risk (RR)

Example: Follow surgeons who operate (exposed) vs those who do not (unexposed) to measure radiation exposure and cancer risk.

Strengths:

Can calculate incidence and relative risk
Multiple outcomes can be studied
Temporal relationship clear (exposure precedes outcome)
Less prone to recall bias

Limitations:

Time-consuming and expensive
Loss to follow-up
Not efficient for rare outcomes
Confounding possible

Prospective cohort studies provide Level II evidence.

Case-Control Studies

Definition: Starts with cases (disease present) and controls (disease absent), then looks backward to compare exposure history.

Process:

Identify cases with the disease/outcome
Select controls without the disease (matched or unmatched)
Measure past exposure in both groups
Calculate odds ratio (OR)

Example: Compare patients with AVN (cases) to those without AVN (controls) to assess whether steroid use (exposure) was more common in cases.

Strengths:

Efficient for rare diseases
Faster and cheaper than cohort studies
Can study multiple exposures
Small sample size needed

Limitations:

Cannot calculate incidence or relative risk (only OR)
Prone to recall bias and selection bias
Temporal relationship unclear
Confounding common

Key Point: Case-control studies are Level III evidence - useful for rare outcomes but inferior to cohort studies for establishing causality.

Observational Descriptive Study Designs

Cross-Sectional Studies

Definition: Measures exposure and outcome at a single point in time (snapshot).

Uses:

Prevalence surveys
Screening studies
Hypothesis generation

Example: Survey orthopaedic surgeons to measure prevalence of burnout and correlate with work hours.

Strengths:

Quick and inexpensive
Good for prevalence data
Generates hypotheses

Limitations:

Cannot establish causality
Cannot measure incidence
Temporal relationship unclear (which came first?)
Survival bias

Case Series and Case Reports

Definition: Descriptive study of patients with similar condition - no comparison group.

Uses:

Describe new diseases or rare conditions
Report novel surgical techniques
Generate hypotheses

Strengths:

Simple to conduct
Useful for rare conditions
Hypothesis-generating

Limitations:

No comparison group (no control)
Cannot establish causality
Selection bias
Level IV evidence only

Understanding descriptive studies helps identify when stronger evidence is needed.

Study Design Components

Essential Components of Any Study

Population and Sampling:

Target population: The group about whom conclusions will be drawn
Study sample: Subset of population actually studied
Sampling method: How participants are selected (random, consecutive, convenience)

Exposure and Outcome:

Exposure/Intervention: What is being studied (treatment, risk factor)
Outcome: What is being measured (disease, recovery, complication)
Primary vs Secondary: Main outcome vs additional outcomes

Time Frame:

Prospective: Follow participants forward in time
Retrospective: Look back at existing data
Cross-sectional: Single point in time

Classification

Study Design Classification

Primary Classification of Study Designs

Category	Type	Investigator Role	Examples
Experimental	Randomized Controlled Trial	Assigns intervention	Drug trial, surgical technique comparison
Observational Analytical	Cohort Study	Observes only	Smoking and nonunion, registry studies
Observational Analytical	Case-Control Study	Observes only	Rare disease risk factors
Observational Descriptive	Cross-Sectional	Observes only	Prevalence surveys
Observational Descriptive	Case Series	Observes only	Novel technique reports

Clinical Application

Choosing Design for Therapeutic Questions

Question: Does treatment A work better than treatment B? Best Design: RCT (if ethical and feasible) Alternative: Prospective cohort study

Choosing Design for Rare Outcomes

Question: Does exposure increase risk of rare disease? Best Design: Case-control study Alternative: Large registry cohort

Choosing Design for Prevalence

Question: How common is condition X in population Y? Best Design: Cross-sectional survey Alternative: Registry analysis

Choosing Design for Prognosis

Question: What is the natural history of disease X? Best Design: Prospective cohort study Alternative: Retrospective cohort from registry

Bias and Confounding

Types of Bias

Selection Bias:

Systematic error in how participants are selected
Example: Only including patients who survived long enough to be studied
Prevention: Random sampling, consecutive enrollment

Information/Measurement Bias:

Systematic error in how data is collected
Recall bias: Cases remember exposures better than controls
Observer bias: Assessor influenced by knowledge of group allocation
Prevention: Blinding, standardized measurement

Confounding:

Third variable associated with both exposure and outcome
Creates spurious association or masks true association
Prevention: Randomization, matching, stratification, multivariable analysis

Systematic Reviews and Meta-Analysis

Systematic Review

Definition: Comprehensive, reproducible synthesis of all available evidence on a specific question.

Key Features:

Explicit, pre-specified methods
Comprehensive literature search
Critical appraisal of included studies
Qualitative or quantitative synthesis

PRISMA Guidelines:

Preferred Reporting Items for Systematic Reviews and Meta-Analyses
27-item checklist for transparent reporting
Flow diagram showing study selection process

Meta-Analysis

Definition: Statistical combination of results from multiple studies.

When Appropriate:

Studies are clinically and methodologically similar
Heterogeneity is acceptable (I² less than 75%)
Provides pooled effect estimate with confidence interval

Registry Studies in Orthopaedics

Registry-Based Research

Definition: Large-scale observational studies using data from national or regional registries.

Major Orthopaedic Registries:

AOANJRR (Australian): Largest national registry, over 500,000 THAs/TKAs
Swedish Hip Arthroplasty Register: Established 1979, longest follow-up
National Joint Registry (UK): Over 3 million procedures recorded
American Joint Replacement Registry (AJRR): Growing database

Strengths:

Large sample sizes (100,000s of patients)
Real-world effectiveness data
Long follow-up periods
Detect rare outcomes and complications
Track implant performance

Limitations:

Observational only (no randomization)
Confounding by indication
Variable data quality
Limited clinical detail

Limitations and Pitfalls

Common Pitfalls by Design

RCT Pitfalls:

Underpowered studies (Type II error)
Poor allocation concealment
Unblinded outcome assessors
Per-protocol analysis instead of ITT
Narrow inclusion criteria limiting generalizability

Cohort Study Pitfalls:

Loss to follow-up (over 20% is concerning)
Confounding by indication
Immortal time bias
Selection of exposed/unexposed groups

Case-Control Pitfalls:

Inappropriate control selection
Recall bias (cases remember better)
Selection bias
Cannot calculate incidence or RR

Statistical Measures by Design

Measures of Association

Relative Risk (RR):

Used in: Cohort studies, RCTs
Incidence in exposed / Incidence in unexposed
RR greater than 1 = increased risk with exposure
Can calculate from prospective studies only

Odds Ratio (OR):

Used in: Case-control studies (also cohort, RCT)
Odds of exposure in cases / Odds of exposure in controls
Approximates RR when outcome is rare (less than 10%)
Only measure available from case-control design

Hazard Ratio (HR):

Used in: Survival analysis (time-to-event)
Instantaneous risk of event at any time point
Accounts for censoring and time-varying exposure

Outcomes and Endpoints

Types of Outcomes

Primary Outcome:

Main outcome the study is powered to detect
Should be clinically meaningful
Used to calculate sample size
Only ONE primary outcome (multiple = type I error inflation)

Secondary Outcomes:

Additional outcomes of interest
Exploratory - not powered to detect
Generate hypotheses for future studies

Surrogate vs Patient-Centered:

Surrogate: Lab value, radiograph (e.g., radiographic union)
Patient-centered: Function, pain, quality of life (e.g., PROMIS scores)
Surrogate outcomes may not correlate with patient-centered outcomes

Evidence Base

CONSORT Statement for Reporting RCTs

Schulz KF, Altman DG, Moher D • BMJ (2010)

Key Findings:

CONSORT provides 25-item checklist for transparent RCT reporting
Includes flow diagram showing participant flow through trial
Improves quality and transparency of RCT publications
Endorsed by over 600 medical journals globally

Clinical Implication: CONSORT guidelines ensure RCTs are reported completely, allowing critical appraisal of study quality.

Limitation: Adherence to CONSORT varies - some journals enforce more strictly than others.

STROBE Statement for Observational Studies

von Elm E, Altman DG, Egger M, et al • Lancet (2007)

Key Findings:

STROBE provides reporting guidelines for cohort, case-control, and cross-sectional studies
22-item checklist ensures transparency of methods and results
Separate checklists for each study design type
Improves reproducibility and critical appraisal

Clinical Implication: STROBE guidelines improve quality of observational study reporting in orthopaedics.

Limitation: Does not assess study quality - only reporting completeness.

Hierarchy of Evidence in Orthopaedic Research

Wright JG, Swiontkowski MF, Heckman JD • JBJS Am (2003)

Key Findings:

Level I: High-quality RCTs or systematic review of Level I studies
Level II: Lesser-quality RCTs or prospective cohort studies
Level III: Case-control studies or retrospective cohort
Level IV: Case series
Level V: Expert opinion

Clinical Implication: Understanding evidence hierarchy helps surgeons critically appraise literature and make evidence-based decisions.

Limitation: Not all clinical questions can be answered with Level I evidence due to ethical and practical constraints.

Exam Viva Scenarios

Practice these scenarios to excel in your viva examination

VIVA SCENARIOStandard

Scenario 1: Study Design Selection

EXAMINER

"You want to study whether smoking increases the risk of nonunion after tibial fracture. What study design would you choose and why?"

EXCEPTIONAL ANSWER

For this question, I would choose a prospective cohort study. I would identify a cohort of patients with tibial fractures at baseline and classify them as smokers (exposed) or non-smokers (unexposed). I would then follow both groups forward in time and measure the incidence of nonunion in each group. This allows me to calculate relative risk and establish temporal relationship. A cohort design is preferred over case-control because smoking is not a rare exposure, and I can measure incidence directly. An RCT would not be ethical because I cannot randomize patients to smoke. The main limitation would be confounding - smokers may differ from non-smokers in other ways (age, diabetes, open fractures), so I would need to adjust for these confounders in my analysis.

KEY POINTS TO SCORE

Cohort study is appropriate for common exposures like smoking

Prospective design establishes temporal relationship

RCT not ethical for harmful exposures

Need to address confounding through statistical adjustment

COMMON TRAPS

✗Choosing case-control study - less efficient for common exposure

✗Suggesting RCT - unethical to randomize to smoking

✗Not mentioning confounding and how to address it

LIKELY FOLLOW-UPS

"What are the main sources of bias in a cohort study?"

"How would you minimize loss to follow-up?"

"Could you use a retrospective cohort design instead?"

VIVA SCENARIOChallenging

Scenario 2: Critically Appraising an RCT

EXAMINER

"You are reviewing an RCT comparing operative vs non-operative treatment for displaced ankle fractures. What key features would you look for to assess the quality of this trial?"

EXCEPTIONAL ANSWER

When critically appraising an RCT, I would systematically assess several key features. First, **randomization** - was the allocation sequence truly random and concealed? This prevents selection bias. Second, **baseline characteristics** - are the groups similar at baseline for age, fracture type, comorbidities? If not, randomization may have failed. Third, **blinding** - ideally double-blind, but for surgical vs non-operative this is impossible, so outcome assessors should at minimum be blinded. Fourth, **intention-to-treat analysis** - were all patients analyzed in the group they were allocated to, regardless of crossover? Fifth, **loss to follow-up** - was it under 20 percent and balanced between groups? High attrition threatens validity. Sixth, **sample size and power** - was the study powered to detect a clinically meaningful difference? Finally, I would check if the trial follows CONSORT reporting guidelines. The main risk in this surgical trial would be lack of blinding leading to performance bias and detection bias.

KEY POINTS TO SCORE

Randomization and allocation concealment prevent selection bias

Baseline balance confirms effective randomization

Blinding prevents performance and detection bias (challenging in surgical trials)

Intention-to-treat preserves randomization benefits

Adequate power ensures meaningful results

COMMON TRAPS

✗Not mentioning blinding is often impossible in surgical RCTs

✗Forgetting intention-to-treat analysis importance

✗Not discussing loss to follow-up impact

LIKELY FOLLOW-UPS

"What is allocation concealment and why does it matter?"

"What is the difference between per-protocol and intention-to-treat analysis?"

"How would you handle crossover in analysis?"

MCQ Practice Points

Study Design Question

Q: A researcher wants to study the association between high BMI and knee osteoarthritis. She measures BMI and presence of knee OA in 500 patients at a single clinic visit. What type of study is this? A: Cross-sectional study. Exposure (BMI) and outcome (OA) are measured at the same point in time. This design can measure prevalence but cannot establish causality or temporal relationship.

RCT Advantage Question

Q: What is the main advantage of randomization in an RCT? A: Balances both known and unknown confounders between groups. Randomization creates groups that are comparable at baseline, eliminating selection bias and confounding, allowing isolation of treatment effect.

Case-Control Study Question

Q: When is a case-control study the preferred design? A: For rare diseases or outcomes. Case-control studies are efficient because you start with cases (already have the rare disease) and look backward for exposures. Much faster than waiting for rare outcome to occur in a cohort.

Australian Context

AOANJRR (Australian Registry)

Australian Orthopaedic Association National Joint Replacement Registry:

Established 1999, world's largest national registry
Over 500,000 hip and knee arthroplasties recorded
Mandatory reporting for all Australian hospitals
Annual report provides survival data by implant type
Gold standard for observational arthroplasty outcomes

NHMRC Evidence Levels

National Health and Medical Research Council:

Australian guidelines for evidence hierarchy
Level I to IV similar to international standards
Grades of recommendation (A-D) based on evidence
Used in Australian clinical practice guidelines

Exam Relevance

For the Australian exam, you must understand study design principles and be able to:

Critically appraise a published study
Identify the appropriate design for a clinical question
Explain the limitations of observational vs experimental studies
Interpret registry data from AOANJRR
Distinguish between statistical and clinical significance

Management Algorithm