Research and Evidence-Based Medicine

Research methodology and evidence-based medicine (EBM) form the intellectual backbone of modern general practice. Australian GPs encounter an enormous breadth of clinical presentations — from acute minor illness to complex multimorbidity — and must navigate an ever-expanding body of literature to deliver safe, effective, and cost-conscious care. This article provides a foundational guide to the statistical concepts, study designs, evidence hierarchies, and critical appraisal skills that underpin evidence-based decision-making in Australian primary care.

The concept of EBM was formalised in the early 1990s by a group at McMaster University in Canada, led by David Sackett and Gordon Guyatt. Since then, it has become embedded in Australian medical education, specialist training, and continuing professional development (CPD). The Royal Australian College of General Practitioners (RACGP) mandates that GPs maintain skills in literature appraisal as part of the Fellowship (FRACGP) curriculum and the CPD programme. The Australian Commission on Safety and Quality in Health Care (ACSQHC) similarly emphasises evidence-based clinical governance through the National Safety and Quality Health Service (NSQHS) Standards.

In Australia, the National Health and Medical Research Council (NHMRC) has developed a nationally recognised hierarchy of evidence and grading system for clinical recommendations. This framework is used by guideline developers including the RACGP, Cancer Council Australia, the Australasian Society of Clinical Immunology and Allergy (ASCIA), and Kidney Health Australia to produce clinical practice guidelines relevant to primary care.

The Australian general practice research landscape is supported by organisations including the Primary Health Care Research and Information Service (PHCRIS), the Australian Primary Health Care Research Institute (APHCRI), and the NHMRC Centre for Research Excellence in Primary Health Care. Each year, Australian GPs contribute to and consume a substantial body of research, with the Australian Journal of General Practice (AJGP, formerly AFP) serving as the principal peer-reviewed journal for the profession.

Understanding the performance characteristics of diagnostic tests is fundamental to clinical reasoning in primary care. GPs order hundreds of pathology and imaging tests each year, and the interpretation of results depends critically on four inter-related concepts: sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

The 2 × 2 Contingency Table

All diagnostic test performance measures derive from the classic 2 × 2 table, which cross-tabulates test results against the true disease status (gold standard):

Sensitivity and Specificity

Positive and Negative Predictive Values

Worked Example — Prostate-Specific Antigen (PSA) Screening

Consider a PSA test with 80% sensitivity and 60% specificity applied to two populations:

Likelihood Ratios

Likelihood ratios (LRs) combine sensitivity and specificity into a single measure and are more clinically useful for individual patient assessment:

• Positive likelihood ratio (LR+): Sensitivity ÷ (1 − Specificity). An LR+ > 10 provides strong evidence to rule IN a diagnosis.
• Negative likelihood ratio (LR−): (1 − Sensitivity) ÷ Specificity. An LR− < 0.1 provides strong evidence to rule OUT a diagnosis.
• Using LRs in practice: Post-test odds = Pre-test odds × LR. This allows GPs to update the probability of disease after a test result, moving from pre-test probability to post-test probability using the Fagan nomogram or Bayesian calculations.

Receiver Operating Characteristic (ROC) Curves

An ROC curve plots sensitivity (y-axis) against (1 − specificity) (x-axis) across all possible cut-off values. The area under the ROC curve (AUC or C-statistic) summarises overall test performance: AUC = 0.5 is no better than chance; AUC = 1.0 is a perfect test. In clinical practice, an AUC > 0.8 is generally considered good and an AUC > 0.9 excellent.

Research methods in general practice encompass a broad spectrum of quantitative and qualitative approaches. Understanding the strengths, limitations, and appropriate applications of each design is essential for Australian GPs who both consume and generate primary care research.

Quantitative Research Designs

Quantitative research uses numerical data and statistical analysis to test hypotheses, estimate effect sizes, and establish associations or causation. The major designs are described below in descending order of evidentiary strength:

Qualitative Research

Qualitative research explores the why and how of health, illness, and healthcare — areas that quantitative methods cannot adequately address. It is increasingly valued in Australian primary care research for understanding patient experience, clinician decision-making, and the social determinants of health.

Mixed-Methods Research

Mixed-methods research combines quantitative and qualitative approaches within a single study or programme of research. In Australian primary care, this is increasingly the preferred approach for complex interventions, health services evaluation, and implementation science. For example, a study evaluating a new chronic disease management programme might use a cluster RCT to measure clinical outcomes (quantitative) alongside semi-structured interviews with patients and GPs to understand barriers and enablers of implementation (qualitative).

Grey Literature and Real-World Evidence

• Grey literature: Government reports (AIHW, ABS), clinical registries, health department policy documents, conference abstracts, and theses. These sources are important in Australian primary care but are not peer-reviewed in the traditional sense.
• Real-world evidence (RWE): Data from electronic health records, PBS dispensing data, Medicare Benefits Schedule (MBS) claims, and disease registries (e.g., the Australian Orthopaedic Association National Joint Replacement Registry). RWE complements RCT data by capturing outcomes in diverse, real-world populations.
• Practice-based research networks (PBRNs): Australian examples include the NPS MedicineWise General Practice Analysis Team and the BEACH successor studies, which generate data directly from GP consultations.

The Five Steps of EBM

The practice of EBM involves five interconnected steps, originally described by Sackett and colleagues:

The PICO Framework

NHMRC Levels of Evidence

The Australian NHMRC has developed a nationally standardised hierarchy of evidence used by guideline developers across Australia. This framework classifies evidence by study design:

NHMRC Grades of Recommendation

The Evidence Pyramid

The "evidence pyramid" is a widely used visual representation of the hierarchy of evidence, with study types arranged from the broadest base (weakest) to the narrowest apex (strongest):

Key Statistical Concepts for GPs

Critical appraisal is the systematic process of evaluating research evidence for its trustworthiness, value, and relevance. For Australian GPs, who are expected to manage uncertainty and apply the best available evidence across a vast clinical scope, critical appraisal is not an academic exercise — it is a core clinical skill.

The Three Fundamental Questions

Every critical appraisal should address three domains, as articulated in the JAMA Users' Guides series and the CASP (Critical Appraisal Skills Programme) checklists:

Types of Bias

Bias is a systematic error that distorts study findings. Recognising bias is perhaps the single most important critical appraisal skill. The three broad categories are:

Selection Bias

Occurs when the way participants are selected or allocated leads to systematic differences between groups. Examples include:

• Allocation bias: Non-random assignment to treatment groups (e.g., sicker patients getting the intervention).
• Volunteer bias: Healthier, more motivated individuals self-selecting into studies.
• Berkson's bias: Hospital-based case-control studies over-representing conditions associated with hospitalisation.
• Attrition bias: Differential loss to follow-up between groups (e.g., patients with side effects dropping out of the treatment arm).

Minimised by: Randomisation, allocation concealment, intention-to-treat analysis, and high follow-up rates (>80%).

Information Bias (Measurement Bias)

Occurs when data on exposure or outcome are measured inaccurately. Examples include:

• Recall bias: Cases (e.g., mothers of children with birth defects) recall exposures more completely than controls.
• Interviewer bias: The researcher's expectations influence data collection.
• Measurement bias: Inconsistent or inaccurate measurement instruments.

Minimised by: Blinding (double-blind preferred), validated measurement instruments, standardised data collection protocols, and objective outcome measures.

Confounding

A confounder is a variable that is associated with both the exposure and the outcome, but is not on the causal pathway. Classic examples include:

• Smoking confounds the relationship between coffee drinking and lung cancer.
• Socioeconomic status confounds the relationship between diet and cardiovascular disease.
• Age confounds almost everything in medicine.

Controlled by: Randomisation (RCTs), restriction, matching, stratified analysis, and multivariable regression (observational studies).

CASP Checklists

The Critical Appraisal Skills Programme (CASP) provides free, structured checklists for appraising different study types. These are widely used in Australian GP training:

Appraising Specific Study Types

Appraising an RCT

• Randomisation: Was the allocation sequence truly random (computer-generated, random number tables)? Was allocation concealment maintained (sealed opaque envelopes, central randomisation)?
• Blinding: Were patients, care providers, outcome assessors, and data analysts blinded? Double-blinding (patient + clinician) is the minimum standard for drug trials.
• Intention-to-treat (ITT) analysis: Were all randomised patients analysed in the group to which they were assigned, regardless of compliance? Per-protocol analysis alone introduces bias.
• Follow-up: Was loss to follow-up < 20%? Was it balanced between groups?
• Baseline comparability: Were groups similar at baseline? Any imbalances suggest failure of randomisation.
• External validity: Were inclusion/exclusion criteria so strict that the study population doesn't resemble your patients?

Appraising a Diagnostic Accuracy Study

• Was there an independent, blinded comparison with a valid reference ("gold") standard?
• Did the study population include a clinically relevant spectrum of disease (not just severe cases vs. healthy controls)?
• Were sensitivity, specificity, likelihood ratios, and predictive values reported with confidence intervals?
• Was the test evaluated in a population similar to your own (pre-test probability)?

Appraising a Systematic Review

• Was the research question clearly defined using PICO?
• Was the search strategy comprehensive (multiple databases, grey literature, hand-searching reference lists)?
• Was the quality of included studies assessed independently by two reviewers?
• Was there significant heterogeneity between studies? If so, was a random-effects model used, and were subgroup analyses or meta-regression performed?
• Was publication bias assessed (funnel plots, Egger's test)?

Clinical vs. Statistical Significance

Practical Tips for Time-Poor GPs

• Start with the abstract and conclusions, then go to methods: If the methods are flawed, the conclusions are untrustworthy regardless of how positive they sound.
• Read the "Limitations" section: Authors are required to disclose study limitations. If they don't, be suspicious.
• Check the funding source and conflicts of interest: Industry-funded studies are more likely to report favourable outcomes for the sponsor's product.
• Use pre-appraised resources when available: Cochrane Clinical Answers, BMJ Best Practice, NPS MedicineWise RADAR, and the RACGP's AJGP "Clinical Practice" summaries.
• Apply the "back of the envelope" test: Can you explain the key finding to a patient in one sentence? If not, the evidence may not be mature enough to change practice.
• Discuss with colleagues: Journal clubs, peer review groups, and RACGP local groups provide invaluable critical appraisal support.

The final — and arguably most important — step in EBM is the translation of research findings into clinical decisions that benefit individual patients. In Australian general practice, this involves navigating several contextual factors:

Patient-Centred Care & Shared Decision-Making

EBM explicitly includes patient values and preferences as one of its three pillars. Shared decision-making involves presenting evidence in an accessible format, exploring the patient's goals and concerns, and arriving at a management plan that respects both the evidence and the patient's autonomy. Decision aids (available from organisations such as the Wiser Healthcare collaboration and the Consumers Health Forum of Australia) can support this process.

Australian Regulatory & Funding Context

• Pharmaceutical Benefits Scheme (PBS): Not all evidence-based treatments are PBS-listed. GPs must consider cost to the patient and whether authority or streamlined authority is required.
• Medicare Benefits Schedule (MBS): Evidence-based investigations and procedures must align with MBS item descriptors for Medicare rebates to apply.
• Therapeutic Goods Administration (TGA): Medications must be TGA-registered for their indication; off-label use (supported by evidence but not TGA-approved for that indication) requires careful documentation and informed consent.
• National Immunisation Program (NIP): Immunisation recommendations are informed by ATAGI (Australian Technical Advisory Group on Immunisation) and reflect the best available evidence adapted to Australian epidemiology.

Implementing Evidence-Based Guidelines

Clinical practice guidelines (CPGs) synthesise evidence into actionable recommendations. In Australia, major guideline developers include the RACGP (Red Book, Green Book, Clinical Guidelines), Cancer Council Australia, ASCIA, Kidney Health Australia, Lung Foundation Australia, and the National Heart Foundation of Australia. GPs should critically evaluate guidelines using the AGREE II (Appraisal of Guidelines for Research and Evaluation) instrument, considering:

• Rigour of development — Was systematic evidence review conducted? Was the strength of evidence graded?
• Applicability to Australian primary care — Were local epidemiology, healthcare system factors, and resource constraints considered?
• Conflicts of interest — Were guideline panel members free from pharmaceutical industry conflicts?
• Currency — When was the guideline last updated? Evidence changes rapidly.

When Evidence Is Insufficient

Audit & Quality Improvement

The NSQHS Standards (ACSQHC) require health service organisations to use evidence to drive continuous quality improvement. GPs can apply evidence-practice gap analysis through clinical audit, PDSA (Plan-Do-Study-Act) cycles, and practice-based quality improvement activities. The RACGP's Practice Accreditation standards (Standards for General Practices, 5th edition) mandate systematic approaches to evidence-based care delivery.

Certain populations are systematically underrepresented in clinical research, yet they constitute a significant proportion of Australian GP consultations. GPs must be aware of how evidence limitations affect care for these groups.

Aboriginal and Torres Strait Islander Australians experience a disproportionate burden of chronic disease, lower life expectancy (approximately 8 years less than non-Indigenous Australians), and significant barriers to healthcare access. Evidence-based practice in this context requires particular attention to research equity, cultural safety, and the application of evidence within a framework of self-determination and community control.

Applying EBM with Aboriginal and Torres Strait Islander Patients

Key Australian Resources for Indigenous Health Evidence

• NHMRC guidelines on ethical conduct in Aboriginal and Torres Strait Islander health research (2018)
• RHDAustralia — Clinical guidelines for rheumatic fever and rheumatic heart disease
• NACCHO — National policy, workforce development, and clinical resources
• Lowitja Institute — Australia's National Institute for Aboriginal and Torres Strait Islander Health Research
• AIHW Indigenous health reports — Epidemiological data and performance monitoring
• RACGP Specific Interests — Aboriginal and Torres Strait Islander Health

1. 1. Sackett DL, Rosenberg WMC, Gray JAM, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn't. BMJ. 1996;312(7023):71–72.
2. 2. Straus SE, Glasziou P, Richardson WS, Haynes RB. Evidence-Based Medicine: How to Practice and Teach EBM. 5th ed. Edinburgh: Elsevier; 2018.
3. 3. National Health and Medical Research Council (NHMRC). NHMRC additional levels of evidence and grades for recommendations for developers of guidelines. Canberra: NHMRC; 2009.
4. 4. Greenhalgh T. How to Read a Paper: The Basics of Evidence-Based Medicine. 6th ed. Chichester: Wiley-Blackwell; 2019.
5. 5. Grimes DA, Schulz KF. An overview of clinical research: the lay of the land. Lancet. 2002;359(9300):57–61.
6. 6. Critical Appraisal Skills Programme (CASP). CASP Checklists. Oxford: CASP; 2018. Available at: casp-uk.net.
7. 7. Guyatt GH, Rennie D, Meade MO, Cook DJ. Users' Guides to the Medical Literature: A Manual for Evidence-Based Clinical Practice. 3rd ed. New York: McGraw-Hill Education; 2015.
8. 8. Royal Australian College of General Practitioners (RACGP). Standards for General Practices. 5th ed. Melbourne: RACGP; 2020.
9. 9. Australian Institute of Health and Welfare (AIHW). Aboriginal and Torres Strait Islander Health Performance Framework. Canberra: AIHW; 2023.
10. 10. National Health and Medical Research Council (NHMRC). Keeping Research on Track II: A companion document to the NHMRC ethical guidelines for research involving Aboriginal and Torres Strait Islander peoples. Canberra: NHMRC; 2018.
11. 11. Higgins JPT, Thomas J, Chandler J, et al. Cochrane Handbook for Systematic Reviews of Interventions. Version 6.4. London: Cochrane; 2023.
12. 12. Brouwers MC, Kho ME, Browman GP, et al. AGREE II: advancing guideline development, reporting and evaluation in health care. CMAJ. 2010;182(18):E839–E842.
13. 13. O'Connell NE, Wand BM, McAuley J, Marston L, Moseley GL. Intensive strength and balance training for older adults with chronic low back pain (RESOLVE trial): protocol for a randomised controlled trial. BMC Musculoskelet Disord. 2023;24:518.
14. 14. Australian Commission on Safety and Quality in Health Care (ACSQHC). National Safety and Quality Health Service Standards. 2nd ed. Sydney: ACSQHC; 2021.
15. 15. Burchett H, Umoquit M, Dobrow M. How do we know when research from one setting can be useful in another? A review of external validity, applicability and transferability frameworks. J Health Serv Res Policy. 2011;16(4):238–244.