PICO Framework

PICO Framework

🟢 Lite — Quick Review (1h–1d)

Rapid summary for last-minute revision before your exam.

Evidence-Based Medicine (EBM) integrates best current research evidence with clinical expertise and patient values to guide primary care decisions. The PICO framework structures clinical questions into Patient/problem, Intervention, Comparison, and Outcome components. Apply the 5A steps: Ask, Acquire, Appraise, Apply, Audit. Hierarchy of evidence ranks systematic reviews of RCTs at the top, followed by individual RCTs, then cohort, case-control, case series, and finally expert opinion. Key biostatistics to memorize: ARR = Risk_control − Risk_treatment, NNT = 1/ARR, RR = Risk_exposed/Risk_unexposed, and OR = (a×d)/(b×c). For diagnostic tests, Sensitivity = TP/(TP+FN), Specificity = TN/(TN+FP), PPV = TP/(TP+FP), NPV = TN/(TN+FP). The Saudi GP Board tests calculation of NNT and interpretation of likelihood ratios using Fagan’s nomogram frequently.

---

🟡 Standard — Regular Study (2d–2mo)

Standard content for students with a few days to months.

Definitions and the EBM Cycle

Evidence-Based Medicine (EBM) is the conscientious use of current best evidence in making decisions about the care of individual patients, integrating clinical expertise, patient preferences, and rigorously published research. The 5A model operationalizes EBM: (1) Ask a focused clinical question, (2) Acquire relevant literature from PubMed/Cochrane, (3) Appraise validity and importance, (4) Apply to the patient, (5) Audit outcomes.

PICO Framework

Element	Question	Example
Patient/Problem	Who?	Adult with type 2 diabetes
Intervention	What action?	Add empagliflozin
Comparison	Versus what?	Placebo + metformin
Outcome	Desired effect?	Reduction in CV mortality

Hierarchy of Evidence

From strongest to weakest: systematic reviews/meta-analyses of RCTs → individual RCTs → cohort studies → case-control studies → case series → expert opinion. Case reports and mechanistic reasoning sit below cohort studies for therapeutic decisions.

Therapy Study Appraisal

Valid RCTs require randomization with allocation concealment, blinding of patients/clinicians/outcome assessors where feasible, complete follow-up (>80%), and intention-to-treat (ITT) analysis to preserve randomization and avoid attrition bias.

Diagnostic Study Appraisal

The index test must be compared against an appropriate reference (gold) standard in a clinically relevant spectrum of disease, with blinded independent interpretation to prevent verification and review bias.

Worked Risk Measures

Suppose a trial reports CVD events in 8/100 on treatment vs 15/100 on placebo. Then ARR = 0.15 − 0.08 = 0.07, RR = 0.08/0.15 ≈ 0.53, NNT = 1/0.07 ≈ 15 (treat 15 patients for ~5 years to prevent one CV event).

Test Performance Measures

Sensitivity rules out disease when negative (SnOut); Specificity rules in disease when positive (SpIn). PPV/NPV depend on prevalence, while likelihood ratios (LR+ = Sensitivity/(1−Specificity); LR− = (1−Sensitivity)/Specificity) remain stable across populations and combine with pre-test odds via Post-test odds = Pre-test odds × LR.

---

🔴 Extended — Deep Study (3mo+)

Comprehensive coverage for students on a longer study timeline.

Applying Evidence to Individual Patients

Translate group-level estimates using the patient’s baseline risk: a low-risk patient may gain marginal absolute benefit, whereas a high-risk patient (e.g., elevated ASCVD score) achieves a much smaller NNT. GRADE rates evidence as High/Moderate/Low/Very Low based on risk of bias, inconsistency, indirectness, imprecision, and publication bias, and grades recommendations as Strong or Conditional/Weak.

Common Biases in Primary Care Research

Selection bias distorts exposure-outcome associations when sampling is non-random; information bias (e.g., recall bias) arises from differential measurement; confounding is addressed through randomization, restriction, matching, stratification, or multivariable regression; lead-time bias falsely prolongs apparent survival in screening cohorts without shifting time of death.

Pre-test to Post-test Probability

LR+ > 10 or LR− < 0.1 generate large and often conclusive shifts in probability. Use Fagan’s nomogram to convert pre-test probability → post-test probability, anchoring management decisions.

Screening Criteria (Wilson & Jungner)

A screening program is justified when the condition is important, the natural history is understood, an acceptable test exists with good sensitivity/acceptable specificity, treatment is effective, and cost/equity are acceptable.

Common Mistakes

Confusing RR with ARR (relative measures exaggerate benefit), using PPV/NPV in populations with different prevalence, ignoring confidence intervals (a CI crossing 1.0 for OR/RR = non-significant), and accepting surrogate endpoints (HbA1c, LDL) without hard outcomes.

Practice Prompts

1. A cohort study finds new T2DM in 30/1000 obese adults over 5 years. Calculate incidence rate per 1000 person-years and interpret.
2. A diagnostic test has sensitivity 90%, specificity 80%, and disease prevalence 10%. Compute PPV and explain why it differs in a clinic with 1% prevalence.

---

Content adapted based on your selected roadmap duration. Switch tiers using the selector above.