Research Methodology, Statistics and Critical Appraisal

Quick Answer

Hierarchy of Evidence:

1. Systematic reviews/Meta-analyses (highest)
2. Randomised Controlled Trials (RCTs)
3. Cohort studies
4. Case-control studies
5. Case series/Case reports
6. Expert opinion (lowest)

Key Statistical Concepts:

P-Value:

• P <0.05: Statistically significant (5% chance of Type I error)
• P >0.05: Not statistically significant
• Limitation: Does NOT indicate clinical importance or magnitude of effect

Confidence Interval (CI):

• 95% CI: Range within which true effect lies with 95% certainty
• CI includes null value (1.0 for RR/OR, 0 for difference): Not statistically significant
• Narrow CI: More precise estimate
• Wide CI: Less precise, smaller sample

Number Needed to Treat (NNT):

• NNT = 1 / Absolute Risk Reduction (ARR)
• Interpretation: Number of patients treated to prevent one adverse outcome
• Lower NNT: More effective intervention
• Example: NNT of 20 means treat 20 patients to prevent one event

Critical Appraisal Checklists:

For RCTs (CONSORT):

• Randomisation method appropriate?
• Allocation concealment?
• Blinding (participants, investigators, outcome assessors)?
• Intention-to-treat analysis?
• Complete follow-up (>80%)?
• Sample size adequate (power >80%)?

For Systematic Reviews (PRISMA):

• Clear PICO question?
• Comprehensive search strategy?
• Quality assessment of included studies?
• Assessment of heterogeneity (I² statistic)?
• Funnel plot for publication bias?

Common Biases:

• Selection bias: Non-representative sampling
• Confounding: Third variable affecting exposure and outcome
• Measurement bias: Inaccurate outcome assessment
• Publication bias: Positive results more likely published
• Recall bias: Cases remember exposure better than controls

---

Indigenous Health Considerations

Research methodology and evidence-based practice present unique considerations for Aboriginal, Torres Strait Islander, and Māori populations, particularly regarding culturally safe research practices, community engagement, and addressing health disparities through rigorous methodology. For Aboriginal and Torres Strait Islander peoples, historical research exploitation (including the infamous Harry Bailey chelation therapy experiments and non-consensual genetic research) has created profound mistrust of Western research methodologies. Contemporary research with Indigenous communities must follow National Health and Medical Research Council (NHMRC) guidelines for ethical conduct, requiring genuine partnership, community control, and benefit-sharing.

Community-based participatory research (CBPR) is the preferred methodology for Indigenous health research, recognising that Aboriginal and Torres Strait Islander communities possess sophisticated knowledge systems and research capabilities. This approach positions community members as co-researchers with equal authority in study design, implementation, and knowledge translation. The AIATSIS (Australian Institute of Aboriginal and Torres Strait Islander Studies) Code of Ethics mandates: (1) Indigenous self-determination, (2) Indigenous leadership, (3) impact and value, (4) sustainability and accountability. Research that does not demonstrate direct community benefit should not proceed.

Māori health research in Aotearoa New Zealand operates under Te Ara Tika: Guidelines for Māori Research Ethics, which emphasises tika (research design), pono (conduct), and aroha (outcomes). Māori Data Sovereignty principles assert that Māori communities own data about themselves, with strict protocols for access, use, and interpretation. The Te Mana Raraunga (Māori Data Sovereignty Network) provides governance frameworks ensuring research serves Māori interests and prevents data colonisation.

Statistical considerations in Indigenous health research must account for small population sizes, clustering in remote communities, and complex survey design effects. Standard statistical methods assuming simple random sampling often underestimate variance and overestimate precision. Cluster sampling adjustments, finite population corrections, and stratified analyses are frequently required. Power calculations must account for higher non-response rates and attrition in longitudinal studies with Indigenous cohorts.

Critical appraisal of research involving Indigenous populations requires additional domains beyond standard checklists: Was the research Indigenous-led or co-designed? Did the community define the research question? Were cultural safety protocols established? Was Indigenous knowledge valued equally with Western scientific knowledge? Has the community approved publication and knowledge translation? Healthcare providers applying evidence-based medicine must recognise that much published research excludes Indigenous populations or analyses them as aggregated "ethnic minorities," limiting generalisability. ANZCA trainees must be trained to critically appraise whether research findings apply to their Indigenous patients or whether culturally adapted approaches are required. The Lowitja Institute (Australia's National Institute for Aboriginal and Torres Strait Islander Health Research) provides resources for culturally safe evidence-based practice.

---

Study Designs

Hierarchy of Evidence

Levels of Evidence: [1,2,3]

Randomised Controlled Trials (RCTs)

Definition: [4,5] Participants randomly allocated to intervention or control group, minimising confounding and selection bias.

Key Features:

• Randomisation: Computer-generated, stratified by key variables
• Allocation concealment: Preventing foreknowledge of assignment (centralised, sealed envelopes)
• Blinding:
  • Single-blind: Participants unaware of allocation
  • Double-blind: Participants and investigators unaware
  • Triple-blind: Including outcome assessors
• Control group: Placebo, standard care, or no intervention
• Intention-to-treat (ITT): All randomised participants analysed in original group

CONSORT Statement: [6,7] Standard for reporting RCTs (25-item checklist):

• Title/abstract: "Randomised" in title
• Introduction: Scientific background
• Methods: Eligibility criteria, settings, interventions, outcomes, sample size, randomisation, blinding, statistical methods
• Results: Participant flow, recruitment, baseline data, numbers analysed, outcomes, ancillary analyses, harms
• Discussion: Limitations, generalisability, interpretation

Strengths:

• Minimises confounding through randomisation
• Causality most robust
• Allocation concealment prevents selection bias

Limitations:

• Expensive and time-consuming
• Strict inclusion/exclusion limits generalisability
• Ethical constraints (cannot withhold effective treatment)
• Hawthorne effect (behaviour change from being observed)

Systematic Reviews and Meta-Analyses

Definition: [8,9] Systematic, transparent review of all relevant studies on a specific question, often with statistical pooling (meta-analysis).

Key Features:

• PICO question: Population, Intervention, Comparison, Outcome
• Comprehensive search: Multiple databases, grey literature, hand-searching
• Inclusion/exclusion criteria: Predetermined, explicit
• Quality assessment: Risk of bias (Cochrane Risk of Bias tool, Newcastle-Ottawa Scale)
• Data extraction: Standardised forms, independent dual extraction
• Meta-analysis: Statistical pooling if appropriate (homogeneity)

PRISMA Statement: [10,11] Preferred Reporting Items for Systematic Reviews and Meta-Analyses (27-item checklist):

• Title: "Systematic review" identified
• Abstract: Structured summary
• Introduction: Rationale, objectives
• Methods: Eligibility criteria, information sources, search strategy, study selection, data collection, items/summary measures, synthesis, risk of bias, additional analyses
• Results: Study selection, study characteristics, risk of bias, results of individual studies, synthesis, risk of bias across studies, additional analyses
• Discussion: Summary of evidence, limitations, conclusions
• Funding: Source, role

Heterogeneity Assessment: [12,13]

• Clinical heterogeneity: Different populations, interventions, outcomes
• Methodological heterogeneity: Different study designs, quality
• Statistical heterogeneity: Variability in effect estimates
  • I² statistic: 0-40% low, 30-60% moderate, 50-90% substantial, 75-100% considerable
  • Chi-squared test: P<0.10 suggests heterogeneity
• Fixed effects model: Assumes one true effect
• Random effects model: Assumes distribution of true effects (more conservative)

Publication Bias: [14,15]

• Funnel plot: Asymmetry suggests missing small negative studies
• Egger's test: Statistical test for funnel plot asymmetry
• Trim and fill: Adjusts for missing studies
• Sensitivity analyses: Excluding low-quality studies

Strengths:

• Highest level of evidence
• Increases precision (narrower CI)
• Resolves uncertainty from conflicting studies

Limitations:

• Quality limited by included studies
• Publication bias
• Heterogeneity may preclude pooling
• Outdated quickly

Cohort Studies

Definition: [16,17] Groups (cohorts) classified by exposure status and followed forward in time to observe outcomes.

Types:

• Prospective: Identified now, followed into future
• Retrospective: Historical cohorts, outcomes already occurred
• Ambidirectional: Both directions

Key Features:

• Exposure: Measured at baseline
• Follow-up: Regular assessment for outcome occurrence
• Comparison: Exposed vs unexposed cohorts
• Outcomes: Incidence rates, relative risk (RR)

Analysis:

• Relative Risk (RR): Risk in exposed / Risk in unexposed
  • RR=1: No association
  • RR>1: Increased risk (harm)
  • RR<1: Decreased risk (protection)
• Attributable Risk: Risk in exposed - Risk in unexposed (excess risk due to exposure)
• Population Attributable Risk: Proportion of disease in population due to exposure

Strengths:

• Can study multiple outcomes from one exposure
• Incidence data available
• Temporal sequence clear (exposure before outcome)

Limitations:

• Expensive and time-consuming
• Attrition/loss to follow-up
• Confounding variables
• Not suitable for rare diseases (few cases)

Case-Control Studies

Definition: [18,19] Cases (with disease) compared to controls (without disease) regarding past exposure.

Key Features:

• Case selection: Diagnostic criteria clearly defined
• Control selection: Same population as cases, same exclusion criteria
• Matching: Controls matched to cases on confounders (age, sex)
• Exposure assessment: Retrospective (records, interviews)
• Analysis: Odds ratio (OR)

Analysis:

• Odds Ratio (OR): (Cases exposed / Cases unexposed) / (Controls exposed / Controls unexposed)
  • OR=1: No association
  • OR>1: Increased odds (harm)
  • OR<1: Decreased odds (protection)
• Rare disease assumption: OR approximates RR when disease rare (<10%)

Strengths:

• Efficient for rare diseases
• Quick and inexpensive
• Can study multiple exposures
• Small sample sizes adequate

Limitations:

• Recall bias: Cases remember exposure better
• Selection bias: Difficult to select appropriate controls
• Temporal ambiguity: Exposure and disease timing unclear
• **Cannot calculate incidence or attributable risk

Cross-Sectional Studies

Definition: [20,21] Exposure and outcome measured simultaneously at one time point ("snapshot").

Uses:

• Prevalence estimation
• Hypothesis generation
• Planning health services

Limitations:

• Cannot establish temporal relationship
• Survivors only (prevalence-incidence bias)
• Not suitable for studying causality

Case Series and Case Reports

Definition: [22,23] Description of characteristics and outcomes in series of patients with specific condition/treatment.

Uses:

• Hypothesis generation
• Rare disease description
• Adverse event signal
• Teaching

Limitations:

• No control group
• Selection bias
• Cannot infer causality

---

Statistical Concepts

Measures of Central Tendency and Dispersion

Central Tendency: [24,25]

• Mean: Sum of values / number of values (affected by outliers)
• Median: Middle value (robust to outliers)
• Mode: Most frequent value

Dispersion:

• Range: Maximum - minimum (sensitive to outliers)
• Interquartile Range (IQR): 25th to 75th percentile (robust)
• Standard Deviation (SD): Square root of variance (for normal distributions)
• Standard Error (SE): SD / √n (precision of mean estimate)

Normal Distribution:

• Mean = Median = Mode
• 68% within 1 SD, 95% within 2 SD, 99.7% within 3 SD
• Parametric tests applicable

Skewed Distribution:

• Positive skew: Tail to right (mean > median)
• Negative skew: Tail to left (mean < median)
• Non-parametric tests required

Hypothesis Testing and P-Values

Null and Alternative Hypotheses: [26,27]

• H₀ (Null): No difference/no association
• H₁ (Alternative): Difference/association exists

P-Value:

• Probability of observing results (or more extreme) if H₀ true
• P <0.05: Reject H₀ (statistically significant)
• P >0.05: Fail to reject H₀ (not statistically significant)

Type I and Type II Errors: [28,29]

• Type I Error (α): False positive (saying difference when none)
  • Typically set at 0.05 (5%)
  • P-value is probability of Type I error
• Type II Error (β): False negative (missing real difference)
  • Typically set at 0.10-0.20 (10-20%)
  • Power (1-β): Probability of detecting true difference
  • Typically 80% or 90%

Clinical vs Statistical Significance:

• Statistically significant: P<0.05, unlikely due to chance
• Clinically significant: Meaningful impact on patient outcomes
• May be one without the other

Confidence Intervals

Definition: [30,31] Range of values within which true population parameter lies with specified probability (usually 95%).

Interpretation:

• 95% CI: If study repeated 100 times, 95 of CIs would contain true effect
• Narrow CI: More precise estimate (larger sample, less variability)
• Wide CI: Less precise (smaller sample, more variability)

For Different Measures:

• Mean difference: CI includes 0 → Not statistically significant
• Risk Ratio (RR): CI includes 1 → Not statistically significant
• Odds Ratio (OR): CI includes 1 → Not statistically significant
• Hazard Ratio (HR): CI includes 1 → Not statistically significant

Advantages over P-values:

• Shows magnitude of effect
• Shows precision
• Allows clinical interpretation

Sample Size and Power Calculations

Factors Affecting Sample Size: [32,33]

• Effect size: Larger effect requires smaller sample
• Variability (SD): More variability requires larger sample
• α (Type I error): Smaller α requires larger sample
• β (Type II error) / Power: Higher power requires larger sample
• Dropout rate: Account for attrition

Common Formulas:

Comparing Two Means:

n = 2 × [(Zα/2 + Zβ) × SD / (Mean1 - Mean2)]²

Comparing Two Proportions:

n = [(Zα/2 × √(2p(1-p)) + Zβ × √(p1(1-p1) + p2(1-p2))) / (p1 - p2)]²

Practical Example:

• Expected difference: 10 mmHg BP reduction
• SD: 15 mmHg
• α=0.05 (Z=1.96), Power=80% (Z=0.84)
• n ≈ 36 per group (72 total)

Power Calculation Post-Hoc:

• What was the probability of detecting the observed effect?
• Underpowered studies may miss clinically important effects

Measures of Association and Effect

Absolute Risk Reduction (ARR): [34,35]

• ARR = Risk in control - Risk in intervention
• Example: Control risk 10%, intervention risk 6%, ARR = 4%

Relative Risk Reduction (RRR):

• RRR = (Risk control - Risk intervention) / Risk control
• Example: (10-6)/10 = 40% relative risk reduction

Number Needed to Treat (NNT):

• NNT = 1 / ARR
• Example: ARR = 0.04, NNT = 25
• Interpretation: Treat 25 patients to prevent one adverse outcome
• NNT <10: Very effective
• NNT 10-20: Moderately effective
• NNT >50: Marginally effective

Number Needed to Harm (NNH):

• NNH = 1 / Absolute Risk Increase
• For adverse effects of intervention

Odds Ratio (OR):

• Used in case-control studies
• OR approximates RR when disease rare (<10%)

Hazard Ratio (HR):

• Used in survival/time-to-event analyses
• HR=0.5 means 50% reduction in hazard (risk) at any time point

Survival Analysis

Kaplan-Meier Curves: [36,37]

• Plot probability of survival over time
• Accounts for censored data (lost to follow-up, still alive at study end)
• Visual comparison of groups
• Log-rank test: Compares survival curves between groups

Cox Proportional Hazards Model:

• Multivariable survival analysis
• Adjusts for confounders
• Proportional hazards assumption: Hazard ratio constant over time

Regression Analysis

Linear Regression: [38,39]

• Continuous outcome, continuous predictor
• Y = a + bX
• R²: Proportion of variance in Y explained by X (0-1)

Logistic Regression:

• Binary outcome (yes/no), any predictor
• Odds = e^(a + bX)
• Odds Ratio: Change in odds per unit change in predictor

Cox Regression:

• Time-to-event outcome, any predictor
• Hazard Ratio: Change in hazard per unit change in predictor

Confounding Adjustment:

• Include potential confounders in multivariable model
• Change-in-estimate method (>10% change indicates confounding)

---

Critical Appraisal

Systematic Approach to Appraising Evidence

Validity: [40,41]

• Are the results credible? (internal validity)
• Can we trust the findings?

Importance:

• Are the results clinically meaningful?
• What is the magnitude of effect?

Applicability:

• Can we apply findings to our patients?
• External validity, generalisability

Critical Appraisal of RCTs (RAMMbo)

R - Recruitment: [42,43]

• Was randomisation truly random?
• Was allocation concealed?
• Were groups similar at baseline?

A - Allocation:

• Was the process unbiased?
• Centralised telephone/computer preferred

M - Maintenance:

• Was follow-up complete (>80%)?
• Was intention-to-treat analysis used?
• Were patients analysed in groups originally randomised?

M - Measurement:

• Were outcome assessors blinded?
• Were outcomes measured objectively?
• Was there differential ascertainment?

b - Balance:

• Were prognostic factors balanced?
• Were co-interventions similar?
• Was compliance adequate?

o - Objective outcomes:

• Were outcomes clinically relevant?
• Were all important outcomes reported?

Additional Checks:

• Sample size: Was power calculation performed? Was it adequate?
• Blinding: Were participants, care providers, outcome assessors blinded?
• Side effects: Were adverse events reported?

Critical Appraisal of Systematic Reviews

AMSTAR 2 Checklist: [44,45]

Critical Domains:

1. PICO components described?
2. Comprehensive literature search?
3. Duplicated study selection and data extraction?
4. Included studies assessed for risk of bias?
5. Funding sources of included studies reported?
6. Risk of bias in individual studies considered in synthesis?
7. Risk of bias across studies (publication bias) discussed?

Assessment of Heterogeneity:

• Was heterogeneity assessed? (I², Q statistic)
• Was it explored? (subgroup analyses, meta-regression)
• Was random effects model used when appropriate?

Sensitivity Analyses:

• Did authors test robustness? (exclude low-quality studies)
• Did they assess publication bias? (funnel plot, Egger's test)

Common Biases in Research

Selection Bias: [46,47]

• Non-random selection of participants
• Healthy user bias (healthier people choose intervention)
• Prevention: Random sampling, consecutive recruitment

Confounding: [48,49]

• Third variable associated with both exposure and outcome
• Control: Randomisation (RCT), restriction, matching, stratification, multivariable analysis
• Residual confounding: Unmeasured confounders

Information/Measurement Bias:

• Recall bias: Cases remember exposure better (case-control)
• Observer bias: Outcome assessors not blinded
• Measurement error: Inaccurate instruments

Publication Bias: [50,51]

• Positive results more likely published
• Detection: Funnel plots, statistical tests
• Mitigation: Trial registration, publication of negative results

Attrition Bias:

• Differential loss to follow-up between groups
• Prevention: ITT analysis, minimise dropout

Performance Bias:

• Differential co-interventions between groups
• Prevention: Standardised protocols, blinding

Statistical vs Clinical Significance

P-values and Clinical Importance: [52,53]

Small Effect, Large Sample:

• May be statistically significant (P<0.05)
• But clinically trivial
• Example: BP reduction of 2 mmHg, P=0.01 (n=10,000)

Large Effect, Small Sample:

• May not reach statistical significance (P>0.05)
• But clinically important
• Example: Mortality reduction 20% vs 10%, P=0.08 (n=100)

Interpretation:

• Look at confidence intervals for magnitude
• Consider NNT for clinical utility
• Evaluate patient-important outcomes (mortality, quality of life, not surrogate markers)

---

Research Ethics

Ethical Principles

The Belmont Report (1979): [54,55]

Respect for Persons:

• Autonomy: Informed consent, right to withdraw
• Protection of vulnerable populations

Beneficence:

• Maximize benefits
• Minimize harms
• Risk-benefit assessment

Justice:

• Fair distribution of research burdens and benefits
• Equitable selection of participants

Informed Consent

Requirements: [56,57]

• Disclosure: All relevant information (purpose, procedures, risks, benefits, alternatives)
• Comprehension: Participant understands (appropriate language, health literacy)
• Voluntariness: Free from coercion or undue influence
• Competence: Capacity to consent
• Documentation: Written consent (usually)

Vulnerable Populations:

• Children (assent + parental consent)
• Cognitively impaired (surrogate decision-maker)
• Prisoners, pregnant women, foetuses (special protections)
• Indigenous communities (community consent)

Research Ethics Committees (REC/IRB)

Role: [58,59]

• Review research protocols
• Ensure ethical standards met
• Protect participant welfare
• Monitor ongoing research

Review Levels:

• Full committee review: High-risk research
• Expedited review: Minimal risk
• Exempt: Educational research, public data

Indigenous Research Ethics

Specific Considerations: [60,61,62]

NHMRC Values and Ethics (Australia):

1. Spirit and Integrity: Research conducted with spirit of respect
2. Reciprocity: Mutual benefit, sharing knowledge
3. Respect: For Indigenous knowledge, culture, identity
4. Equality: Partnership, not exploitation
5. Survival and Protection: Strengthen Indigenous culture
6. Responsibility: Accountability to community

Māori Research Ethics (Te Ara Tika, NZ):

• Tika (Correct): Appropriate research design
• Pono (Honest): Transparent conduct
• Aroha (Respectful): Beneficial outcomes

Data Sovereignty:

• Indigenous communities own data about themselves
• Control over collection, analysis, interpretation, storage
• Right to withdraw consent

---

Special Considerations in Anaesthesia Research

Anaesthesia-Specific Challenges

Blinding Difficulties: [63,64]

• Cannot blind anaesthetist to intervention
• May blind patient, surgeon, outcome assessor
• Sham procedures raise ethical concerns

Standardised Care:

• Anaesthesia involves multiple variables (drugs, doses, techniques)
• Standardisation vs clinical flexibility
• Protocol adherence monitoring

Outcome Measurement:

• Many anaesthesia outcomes subjective (pain, nausea)
• Need objective measures (recovery time, complications)
• Patient-reported outcomes (PROMs) increasingly important

ANZCA Research Framework

ANZCA Clinical Trials Network (CTN): [65,66]

• Coordinates multi-centre anaesthesia research
• Provides infrastructure, methodology support
• Mentorship for trainee researchers

Research Requirements for Fellowship:

• Understanding of research methodology
• Critical appraisal skills
• Evidence-based practice
• May participate in research project

Indigenous Health Considerations in Research

Note: This section expands on the Quick Answer section above for comprehensive cultural safety training.

Aboriginal and Torres Strait Islander Research: [67,68,69]

Historical Context:

• Legacy of exploitation and unethical research
• Harry Bailey experiments (1960s chelation therapy)
• Non-consensual genetic research
• Medical experimentation without informed consent
• Stolen Generations research

Current Framework:

• NHMRC Guidelines: Ethical conduct with Aboriginal and Torres Strait Islander Peoples
• AIATSIS Code of Ethics: Indigenous self-determination, leadership, impact, sustainability
• Community-based participatory research (CBPR): Community as co-researchers

Key Principles:

1. Indigenous leadership: Research led by Indigenous researchers
2. Community control: Community decides research questions, methods, dissemination
3. Benefit-sharing: Direct benefit to community (not just academic publications)
4. Cultural safety: Protocols for respectful engagement
5. Data sovereignty: Community owns and controls data
6. Capacity building: Training Indigenous researchers

Research Ethics:

• Beyond individual consent: Community consent required
• Elders and community leaders: Must approve research
• Aboriginal Community Controlled Health Services (ACCHS): Often gatekeepers
• Transparent agreements: Intellectual property, publication rights

Statistical Considerations:

• Small population sizes: Many communities <500 people
• Cluster sampling: Communities clustered geographically
• Complex survey design: Requires adjusted analyses
• Non-response: Higher in Indigenous populations
• Attrition: Significant in longitudinal studies

Dissemination:

• Community feedback: Results returned to community first
• Plain language: Non-technical summaries
• Cultural protocols: Permission for publication
• Indigenous authorship: Community members as co-authors

Māori Research Ethics (Aotearoa New Zealand): [70,71,72]

Te Ara Tika Principles:

• Tika (Appropriate): Culturally appropriate research design
• Pono (Truthful): Honest, transparent conduct
• Aroha (Respectful): Research that shows care and respect

Māori Data Sovereignty:

• Te Mana Raraunga: Māori Data Sovereignty Network
• Principles:
  • Rangatiratanga: Māori control over Māori data
  • Whakapapa: Data identifies relationships and connections
  • Whanaungatanga: Data builds relationships
  • Kotahitanga: Data unified for collective benefit
  • Manaakitanga: Data shared with care and respect
  • Kaitiakitanga: Stewardship and guardianship of data

Health Research Council (NZ):

• Māori health research funding: Dedicated funding streams
• Kaupapa Māori research: Māori-led, Māori-focused
• Ethics approval: Māori ethics committees available

Research Partnerships:

• Whānau involvement: Family-centred research
• Iwi (tribal) engagement: Tribal authorities involved
• Māori Health Workers: Essential research team members
• Kaumatua (elders): Guidance and oversight

Applying Evidence to Indigenous Patients: [73,74,75]

Critical Appraisal Questions:

• Did the study include Indigenous participants?
• Were Indigenous participants analysed separately or aggregated?
• Are findings generalisable to Indigenous populations?
• Were cultural factors considered in the intervention?
• Is there Indigenous-specific evidence?

Evidence Gaps:

• Most RCTs exclude Indigenous populations
• Genetic studies rarely include Indigenous peoples
• Pharmacokinetic studies predominantly in Caucasian populations
• Different disease profiles may alter risk-benefit

Clinical Decision-Making:

• Evidence + Clinical Judgment + Patient Values
• Cultural adaptation of evidence-based interventions
• Shared decision-making with Indigenous patients
• Recognition of Indigenous knowledge systems
• Consultation with Aboriginal Health Workers/Māori Health Workers

ANZCA Training Implications:

• Cultural safety training: Mandatory for all trainees
• Indigenous health curriculum: Understanding disparities
• Critical appraisal skills: Evaluating applicability to Indigenous patients
• Research ethics: Understanding community-based research
• Partnership approach: Working with Indigenous communities

---

ANZCA Exam Focus

Written Examination (SAQ)

High-Yield Topics:

1. Study design selection: Which design for which question?
2. Hierarchy of evidence: Levels and why
3. P-values vs CI: Interpretation and limitations
4. NNT calculation: Clinical utility
5. Critical appraisal: RAMMbo for RCTs, AMSTAR for reviews
6. Sample size factors: Power, effect size, alpha, beta

Common SAQ Scenarios:

Scenario 1: "A study randomised 200 patients to receive either drug A or placebo for postoperative pain. At 24 hours, pain scores were lower in the drug A group (mean difference 10 mm on VAS, 95% CI 2-18 mm, P=0.02). Calculate the NNT if baseline risk of moderate-severe pain in placebo group was 40%. Critically appraise this study. (20 marks)"

Scenario 2: "Explain the difference between relative risk and absolute risk reduction, and when each would be used. (10 marks)"

Viva Voce Examinations

Expected Viva Themes:

Theme 1: Study Design

• "What study design would you use to investigate whether a new antiemetic reduces PONV?"
  • Key points: RCT, PONV outcomes, power calculation

Theme 2: Statistics

• "What does a P-value of 0.03 mean? What are its limitations?"
  • Key points: 3% chance of Type I error, doesn't show magnitude or importance

Theme 3: Critical Appraisal

• "How would you critically appraise a systematic review?"
  • Key points: PICO, search strategy, quality assessment, heterogeneity, publication bias

Theme 4: Ethics

• "What are the key ethical principles in research?"
  • Key points: Respect for persons, beneficence, justice

Viva Scenario Example

Examiner: "You are reading a paper comparing two anaesthetic techniques. The P-value for the primary outcome is 0.04. What does this tell you?"

Candidate Response Framework:

1. Interpretation:

   • "A P-value of 0.04 means there is a 4% probability of observing this result (or more extreme) if there were truly no difference between the techniques"
   • "Since this is less than 0.05, it is considered statistically significant"
   • "We would reject the null hypothesis of no difference"

2. Limitations:

   • "However, the P-value does not tell us the magnitude of the difference"
   • "It doesn't tell us if the difference is clinically important"
   • "With a large enough sample size, trivial differences can be statistically significant"

3. Better Approach:

   • "I would want to see the confidence interval to understand the precision and magnitude"
   • "I would look at the NNT to understand clinical utility"
   • "I would assess whether the outcome is patient-important"

Examiner Follow-up: "The mean difference in recovery time was 5 minutes with a 95% CI of 1-9 minutes. How would you interpret this?"

Candidate: "The 95% confidence interval tells us that we can be 95% confident that the true difference in recovery time lies between 1 and 9 minutes. Since the entire interval is above zero and excludes the null value, this supports the statistically significant finding. However, a 5-minute difference in recovery time, while statistically significant, may not be clinically important to patients, so I would consider this a modest effect of questionable clinical significance."

---

Short Answer Questions

SAQ 1: Study Design and Statistics

Question: (20 marks) You are planning a study to investigate whether a new regional anaesthesia technique reduces chronic post-surgical pain at 6 months after total knee replacement.

a) What study design would be most appropriate? Justify your answer. (5 marks)

b) What outcomes would you measure? Distinguish between primary and secondary outcomes. (5 marks)

c) List three sources of bias that could affect this study and how you would minimise each. (6 marks)

d) What factors would you consider in your sample size calculation? (4 marks)

---

Model Answer:

a) Study Design (5 marks):

Design: Randomised Controlled Trial (RCT) [1 mark]

Justification: [4 marks]

1. Randomisation distributes confounders evenly between groups [1]
2. Causality - strongest evidence for effectiveness of intervention [1]
3. Chronic pain is objective outcome that can be blinded [0.5]
4. Comparison to standard care (control group) essential [0.5]
5. Temporal sequence clear (intervention before outcome) [0.5]
6. Prospective design allows complete follow-up [0.5]

Alternative acceptable: Prospective cohort study (if RCT not feasible, but loses causal inference)

b) Outcomes (5 marks):

Primary Outcome: [2 marks]

1. Incidence of chronic post-surgical pain at 6 months [1]
   • Definition: Pain >3/10 NRS persisting >3 months after surgery
   • Measured by validated pain scale (NRS 0-10)
2. Or proportion with moderate-severe chronic pain [0.5]
3. Need for analgesics at 6 months [0.5]

Secondary Outcomes: [3 marks - any 3]

• Pain intensity at rest and movement [0.5]
• Quality of life (SF-36, EQ-5D) [0.5]
• Functional outcomes (knee function scores) [0.5]
• Analgesic consumption (opioid-sparing) [0.5]
• Patient satisfaction [0.5]
• Adverse effects (nerve injury, local anaesthetic toxicity) [0.5]
• Cost-effectiveness [0.5]
• Time to return to work/normal activities [0.5]

c) Bias and Minimisation (6 marks):

Selection Bias: [2 marks]

• Issue: Non-representative sampling, differential recruitment [0.5]
• Minimisation: Consecutive recruitment, clear eligibility criteria, computer randomisation, allocation concealment [1.5]

Performance Bias: [2 marks]

• Issue: Differential co-interventions, non-blinding [0.5]
• Minimisation: Standardised anaesthetic protocol, blinding of outcome assessors, patient blinded to group allocation [1.5]

Attrition Bias: [2 marks]

• Issue: Differential loss to follow-up (6 months), missing data [0.5]
• Minimisation: Multiple contact attempts, intention-to-treat analysis, sensitivity analyses for missing data [1.5]

Other acceptable biases:

• Detection/measurement bias (blinding of assessors)
• Recall bias (standardised assessment intervals)
• Confounding (randomisation, stratification)

d) Sample Size Factors (4 marks):

1. Effect size: Expected difference in chronic pain incidence between groups [1]
2. Baseline risk: Incidence of chronic pain with standard care (typically 20-30%) [0.5]
3. Alpha (Type I error): Usually 0.05 (two-tailed) [0.5]
4. Power (1 - Type II error): Usually 80% or 90% [0.5]
5. Attrition/dropout rate: Expect 10-20% loss at 6 months [0.5]
6. Variance: Standard deviation if continuous outcome [0.5]
7. Minimum clinically important difference: What difference matters to patients? [0.5]

---

SAQ 2: Critical Appraisal

Question: (15 marks) A published RCT compared sugammadex versus neostigmine for reversal of neuromuscular blockade in 500 patients. The results showed:

• Primary outcome (residual paralysis at PACU): 5% vs 15%, P=0.001
• Mean difference in reversal time: 3 minutes (95% CI 1-5 minutes)

a) Calculate the Absolute Risk Reduction (ARR), Relative Risk Reduction (RRR), and Number Needed to Treat (NNT). (6 marks)

b) Interpret the confidence interval for the mean difference in reversal time. (3 marks)

c) List three limitations of the P-value in this context. (3 marks)

d) What additional information would you need to critically appraise this study? (3 marks)

---

Model Answer:

a) Calculations (6 marks):

Absolute Risk Reduction (ARR): [2 marks]

• ARR = Risk control - Risk intervention
• ARR = 15% - 5% = 10% or 0.10 [1]
• Interpretation: Sugammadex reduces residual paralysis by 10 percentage points [1]

Relative Risk Reduction (RRR): [2 marks]

• RRR = (Risk control - Risk intervention) / Risk control
• RRR = (0.15 - 0.05) / 0.15 = 66.7% [1]
• Interpretation: Sugammadex reduces relative risk of residual paralysis by 67% [1]

Number Needed to Treat (NNT): [2 marks]

• NNT = 1 / ARR
• NNT = 1 / 0.10 = 10 [1]
• Interpretation: Need to treat 10 patients with sugammadex instead of neostigmine to prevent one case of residual paralysis [1]

b) Confidence Interval Interpretation (3 marks):

1. Range: We are 95% confident that the true mean difference in reversal time lies between 1 and 5 minutes [1]
2. Significance: The entire confidence interval is above zero (excludes null value), supporting the statistically significant finding [1]
3. Precision: The relatively narrow interval (4-minute range) suggests reasonable precision given the sample size [0.5]
4. Clinical relevance: While statistically significant, a 3-minute difference may be of limited clinical importance in many settings [0.5]

c) P-value Limitations (3 marks):

1. Magnitude not shown: P-value doesn't indicate the size of the effect (10% ARR vs 1% ARR could both have P=0.001 with different sample sizes) [1]
2. Clinical importance: Statistical significance doesn't mean clinical importance (may be statistically significant but trivial clinically) [1]
3. Sample size dependence: With very large samples, trivial differences become statistically significant [0.5]
4. Binary thinking: P=0.051 vs P=0.049 doesn't represent fundamentally different findings [0.5]

d) Critical Appraisal Information (3 marks):

Methodological details: [2 marks]

• Randomisation method and allocation concealment [0.5]
• Blinding (patients, assessors, analysts) [0.5]
• Intention-to-treat analysis performed? [0.5]
• Complete follow-up rate (>80%)? [0.5]
• Sample size calculation and power [0.5]

Clinical relevance: [1 mark]

• Were groups similar at baseline? [0.5]
• Side effects/adverse events reported? [0.5]
• Generalisability to my patient population? [0.5]
• Funding source and conflicts of interest? [0.5]

(Any 3 points acceptable)

---

SAQ 3: Research Ethics

Question: (15 marks) You are planning a research study involving Aboriginal patients in a regional hospital.

a) What are the three core ethical principles outlined in the Belmont Report? (3 marks)

b) List four specific ethical considerations when conducting research with Aboriginal and Torres Strait Islander peoples. (4 marks)

c) What is "community-based participatory research" and why is it important in Indigenous health research? (4 marks)

d) What is meant by "data sovereignty" in the context of Māori health research? (4 marks)

---

Model Answer:

a) Belmont Report Principles (3 marks):

1. Respect for Persons: [1]

   • Recognition of autonomy (informed consent)
   • Protection of those with diminished autonomy (vulnerable populations)

2. Beneficence: [1]

   • Obligation to maximize benefits and minimize harms
   • Risk-benefit assessment
   • "Do no harm" (non-maleficence)

3. Justice: [1]

   • Fair distribution of research burdens and benefits
   • Equitable selection of participants
   • No exploitation of vulnerable groups

b) Aboriginal Research Ethics (4 marks):

1. Community consent: Beyond individual consent, requires community approval through elders/leadership [1]
2. Aboriginal leadership: Research should be led by or in genuine partnership with Aboriginal researchers [1]
3. Benefit-sharing: Research must provide direct benefit to Aboriginal communities, not just academic publications [0.5]
4. Cultural safety: Research protocols must ensure respectful, safe engagement [0.5]
5. Data sovereignty: Communities own and control data about themselves [0.5]
6. Capacity building: Training Aboriginal researchers and building community research capacity [0.5]

(Any 4 points acceptable)

c) Community-Based Participatory Research (4 marks):

Definition: [1.5 marks]

• Collaborative approach where community members are full partners in all phases of research (design, conduct, analysis, dissemination) [1]
• Recognises community expertise and priorities research questions relevant to community needs [0.5]

Importance in Indigenous Research: [2.5 marks]

1. Addresses power imbalance: Shifts from "research on" to "research with" Indigenous peoples [0.5]
2. Cultural appropriateness: Ensures research methods respect cultural protocols [0.5]
3. Community priorities: Research questions address community-defined health needs [0.5]
4. Trust building: Addresses historical exploitation and builds research partnerships [0.5]
5. Knowledge translation: Results more likely to be used and benefit community [0.5]

d) Data Sovereignty (4 marks):

Definition: [1.5 marks]

• Right of Indigenous peoples and communities to own, control, access, and possess data about themselves [1]
• Includes determining what data is collected, how it's used, who has access, and how it's stored [0.5]

Māori Context (Te Mana Raraunga): [2.5 marks]

1. Rangatiratanga: Māori control over Māori data [0.5]
2. Governance: Māori-led governance of data repositories [0.5]
3. Access: Māori decide who can access and use data [0.5]
4. Benefit: Data must benefit Māori communities [0.5]
5. Protection: Safeguards against misuse or unauthorised access [0.5]

---

References

1. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924-926. PMID: 18436948.

2. Murad MH, Asi N, Alsawas M, et al. New evidence pyramid. Evid Based Med. 2016;21(4):125-127. PMID: 27378393.

3. Harbour R, Miller J. A new system for grading recommendations in evidence based guidelines. BMJ. 2001;323(7308):334-336. PMID: 11498496.

4. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c332. PMID: 20332509.

5. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c869. PMID: 20332511.

6. Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med. 2001;134(8):663-694. PMID: 11304107.

7. Devereaux PJ, Manns BJ, Ghali WA, et al. The reporting of methodological factors in randomized controlled trials and the association with a journal policy to promote adherence to the Consolidated Standards of Reporting Trials (CONSORT) checklist. Control Clin Trials. 2002;23(4):380-388. PMID: 12161084.

8. Oxman AD, Guyatt GH. Validation of an index of the quality of review articles. J Clin Epidemiol. 1991;44(11):1271-1278. PMID: 1941023.

9. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ. 2009;339:b2535. PMID: 19622551.

10. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. PMID: 19631507.

11. Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1. PMID: 25554246.

12. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-560. PMID: 12958120.

13. Ioannidis JP, Patsopoulos NA, Evangelou E. Uncertainty in heterogeneity estimates in meta-analyses. BMJ. 2007;335(7626):914-916. PMID: 17974687.

14. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. PMID: 9310563.

15. Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002. PMID: 21784880.

16. Grimes DA, Schulz KF. Cohort studies: marching towards outcomes. Lancet. 2002;359(9303):341-345. PMID: 11843010.

17. Mann CJ. Observational research methods. Research design II: cohort, cross sectional, and case-control studies. Emerg Med J. 2003;20(1):54-60. PMID: 12533370.

18. Grimes DA, Schulz KF. Compared to what? Finding controls for case-control studies. Lancet. 2005;365(9468):1429-1433. PMID: 15836893.

19. Schulz KF, Grimes DA. Case-control studies: research in reverse. Lancet. 2002;359(9304):431-434. PMID: 11844534.

20. Mann CJ. Observational research methods. Research design II: cohort, cross sectional, and case-control studies. Emerg Med J. 2003;20(1):54-60. PMID: 12533370.

21. Setia MS. Methodology Series Module 3: Cross-sectional Studies. Indian J Dermatol. 2016;61(3):261-264. PMID: 27293243.

22. Black N. Why we need observational studies to evaluate the effectiveness of health care. BMJ. 1996;312(7040):1215-1218. PMID: 8634569.

23. Vandenbroucke JP. Case reports in an evidence-based world. J R Soc Med. 1999;92(4):159-163. PMID: 10450268.

24. Altman DG, Bland JM. Standard deviations and standard errors. BMJ. 2005;331(7521):903. PMID: 16223828.

25. Bland JM, Altman DG. Statistics notes: Measurement error. BMJ. 1996;313(7059):744. PMID: 8813014.

26. Goodman SN. Toward evidence-based medical statistics. 1: The P value fallacy. Ann Intern Med. 1999;130(12):995-1004. PMID: 10383350.

27. Goodman SN. Toward evidence-based medical statistics. 2: The Bayes factor. Ann Intern Med. 1999;130(12):1005-1013. PMID: 10383351.

28. Colquhoun D. An investigation of the false discovery rate and the misinterpretation of p-values. R Soc Open Sci. 2014;1(3):140216. PMID: 26064558.

29. Ioannidis JP. Why most published research findings are false. PLoS Med. 2005;2(8):e124. PMID: 16060722.

30. Greenland S, Senn SJ, Rothman KJ, et al. Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. Eur J Epidemiol. 2016;31(4):337-350. PMID: 27209009.

31. Cumming G, Finch S. Inference by eye: confidence intervals and how to read pictures of data. Am Psychol. 2005;60(2):170-180. PMID: 15740449.

32. Bland JM. The tyranny of power: is there a better way to calculate sample size? BMJ. 2009;339:b3985. PMID: 19861363.

33. Schulz KF, Grimes DA. Sample size calculations in randomised trials: mandatory and mystical. Lancet. 2005;365(9467):1348-1353. PMID: 15823387.

34. Cook RJ, Sackett DL. The number needed to treat: a clinically useful measure of treatment effect. BMJ. 1995;310(6977):452-455. PMID: 7873954.

35. McQuay HJ, Moore RA. Using numerical results from systematic reviews in clinical practice. Ann Intern Med. 1997;126(9):712-720. PMID: 9139558.

36. Pocock SJ, Clayton TC, Altman DG. Survival plots of time-to-event outcomes in clinical trials: good practice and pitfalls. Lancet. 2002;359(9318):1686-1689. PMID: 12020548.

37. Spruance SL, Reid JE, Grace M, et al. Hazard ratio in clinical trials. Antimicrob Agents Chemother. 2004;48(8):2787-2792. PMID: 15273082.

38. Bland JM, Altman DG. Statistics notes: The odds ratio. BMJ. 2000;320(7247):1468. PMID: 10827089.

39. Bland JM, Altman DG. Statistics notes: Logarithms. BMJ. 1996;312(7038):1169. PMID: 8620150.

40. Guyatt GH, Sackett DL, Sinclair JC, et al. Users' guides to the medical literature. IX. A method for grading health care recommendations. Evidence-Based Medicine Working Group. JAMA. 1995;274(22):1800-1804. PMID: 7500513.

41. Guyatt GH, Oxman AD, Ali M, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383-394. PMID: 21195583.

42. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17(1):1-12. PMID: 8721797.

43. Schulz KF, Chalmers I, Hayes RJ, et al. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA. 1995;273(5):408-412. PMID: 7823387.

44. Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. PMID: 28935782.

45. Shea BJ, Hamel C, Wells GA, et al. AMSTAR is a reliable and valid measurement tool to assess the methodological quality of systematic reviews. J Clin Epidemiol. 2009;62(10):1013-1020. PMID: 19230606.

46. Rothman KJ, Greenland S. Precision and validity in epidemiologic studies. In: Rothman KJ, Greenland S, Lash TL, eds. Modern Epidemiology. 3rd ed. Philadelphia: Lippincott Williams & Wilkins; 2008:128-147.

47. Groves RM, Fowler FJ Jr, Couper MP, et al. Survey Methodology. 2nd ed. Hoboken, NJ: John Wiley & Sons; 2009.

48. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Philadelphia: Lippincott Williams & Wilkins; 2008.

49. Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Epidemiology. 2007;18(6):805-835. PMID: 18049195.

50. Dickersin K. The existence of publication bias and risk factors for its occurrence. JAMA. 1990;263(10):1385-1389. PMID: 2304243.

51. Easterbrook PJ, Berlin JA, Gopalan R, et al. Publication bias in clinical research. Lancet. 1991;337(8746):867-872. PMID: 1672966.

52. Greenland S, Senn SJ, Rothman KJ, et al. Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations. Eur J Epidemiol. 2016;31(4):337-350. PMID: 27209009.

53. Bland JM, Altman DG. Best (but oft forgotten) practices: the p-value, deconstructed. Am J Clin Nutr. 2015;102(4):759-764. PMID: 26315499.

54. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington, DC: Department of Health, Education, and Welfare; 1979.

55. Faden RR, Beauchamp TL. A History and Theory of Informed Consent. New York: Oxford University Press; 1986.

56. Flory J, Emanuel E. Interventions to improve research participants' understanding in informed consent for research: a systematic review. JAMA. 2004;292(13):1593-1601. PMID: 15467062.

57. Appelbaum PS, Lidz CW, Grisso T. Therapeutic misconception in clinical research: frequency and risk factors. IRB. 2004;26(2):1-8. PMID: 15206360.

58. Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA. 2000;283(20):2701-2711. PMID: 10819955.

59. Levine RJ. The nature and definition of informed consent in various research settings. In: National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington, DC: Department of Health, Education, and Welfare; 1979: Appendix C.

60. National Health and Medical Research Council. Values and Ethics: Guidelines on Ethical Matters in Aboriginal and Torres Strait Islander Health Research. Canberra: Commonwealth of Australia; 2003.

61. Australian Institute of Aboriginal and Torres Strait Islander Studies. Guidelines for Ethical Research in Australian Indigenous Studies. Canberra: AIATSIS; 2012.

62. Fredericks B. Which way that health? Indigenous health and the neoliberal agenda. In: Broom A, Dixit S, eds. Health, Culture and Society: Conceptual Legacies and Contemporary Applications. New York: Routledge; 2018:81-96.

63. Myles PS, Gin T. Statistical methods for anaesthesia and intensive care. 1st ed. Oxford: Butterworth-Heinemann; 2001.

64. Miller RD, Cohen NH, Eriksson LI, et al. Miller's Anesthesia. 8th ed. Philadelphia: Elsevier Saunders; 2015.

65. ANZCA Clinical Trials Network. Available at: https://www.anzca.edu.au/research/anzca-clinical-trials-network

66. Myles PS. The Australian and New Zealand College of Anaesthetists Clinical Trials Network. Anaesth Intensive Care. 2008;36(5):629. PMID: 18714621.

67. Sherwood J. Colonisation – It's bad for your health: the context of Aboriginal health. Contemp Nurse. 2009;33(2):228-240. PMID: 20166921.

68. Paradies Y, Ben J, Denson N, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One. 2015;10(9):e0138511. PMID: 26398658.

69. Anderson K, Devitt J, Cunningham J, et al. 'All they said was my kidneys were dead': Indigenous Australian patients' understanding of their chronic kidney disease. Med J Aust. 2008;189(9):499-503. PMID: 19012551.

70. New Zealand Ministry of Health. Te Ara Tika: Guidelines for Māori Research Ethics. Wellington: Ministry of Health; 2010.

71. Te Mana Raraunga. Māori Data Sovereignty Principles. Available at: https://www.temanararaunga.maori.nz/

72. Hudson M, Milne M, Reynolds P, et al. Treaty of Waitangi and Māori Health. N Z Med J. 2010;123(1317):6-8. PMID: 20321602.

73. Anderson I, Robson B, Connolly M, et al. Indigenous and tribal peoples' health (The Lancet-Lowitja Institute Global Collaboration): a population study. Lancet. 2016;388(10040):131-157. PMID: 27108232.

74. National Health and Medical Research Council. Ethical conduct in research with Aboriginal and Torres Strait Islander Peoples and communities: Guidelines for researchers and stakeholders. Canberra: Commonwealth of Australia; 2018.

75. Gracey M, King M. Indigenous health part 1: determinants and disease patterns. Lancet. 2009;374(9683):65-75. PMID: 19577646.