image.jpg
image.jpg
Answer these questions for dnb anaesthesia exam
I can see from your notes and the attached PDF that you need comprehensive answers for DNB Anaesthesia stats questions. Let me read the PDF first to understand the full scope, then provide detailed answers.
Reading File
Running Command
cd /home/daytona/workspace/attachments/9b490236-1ad2-4c79-af32-ac252018542a/ && python3 -c " import pdfplumber with pdfplumber.open('STATS QUESTIONS BY DR ANSHUL DIWAKAR.pdf') as pdf: print(f'Total pages: {len(pdf.pages)}') for i, page in enumerate(pdf.pages): text = page.extract_text() print(f'--- PAGE {i+1} ---') print(text) print() "
I now have the full content of the PDF. Let me compile comprehensive exam-ready answers for all the topics from your notes and the PDF.Here are comprehensive, exam-ready answers for all the DNB Anaesthesia stats topics from your notes and the PDF:
DNB ANAESTHESIA - STATISTICS: COMPLETE EXAM ANSWERS
MUST DO TOPICS (2023-2025)
1. P-VALUE + INTERPRETATION
Definition: The p-value is the probability of obtaining results as extreme as (or more extreme than) the observed data, assuming the null hypothesis is true.
Key formula concept:
- p < 0.05 → Result is statistically significant → Reject null hypothesis
- p > 0.05 → Result is NOT statistically significant → Fail to reject null hypothesis
Interpretation table (exam-ready):
| p-value | Interpretation |
|---|---|
| p < 0.001 | Highly significant |
| p < 0.01 | Very significant |
| p < 0.05 | Significant (threshold) |
| p > 0.05 | Not significant |
Common Misinterpretations (DNB favourite):
- p-value does NOT tell you the probability that the null hypothesis is true
- A small p-value does NOT mean a large/clinically important effect
- p-value is affected by sample size - a huge sample can make a trivial difference "significant"
- A non-significant result does NOT prove the null hypothesis
Clinical example: A new opioid reduces post-op pain. p = 0.03 means: if the drug had no effect, there is only a 3% chance of seeing this result by chance. We reject H₀ and accept the drug is effective.
2. NULL HYPOTHESIS
Null Hypothesis (H₀): States there is NO difference, NO effect, or NO association between the variables being studied. It is the "default" position.
Alternative Hypothesis (H₁ or Ha): States there IS a difference/effect/association.
Examples:
- H₀: Drug A and Drug B have equal analgesic effect (μ₁ = μ₂)
- H₁: Drug A and Drug B have different analgesic effects (μ₁ ≠ μ₂)
Process of hypothesis testing:
- State H₀ and H₁
- Choose significance level (α = 0.05 usually)
- Collect data and compute test statistic
- Calculate p-value
- If p < α → reject H₀ → accept H₁
Important: Failing to reject H₀ does NOT prove it is true. It only means there is insufficient evidence to reject it.
One-tailed vs Two-tailed tests:
- Two-tailed: Tests for difference in either direction (μ₁ ≠ μ₂)
- One-tailed: Tests for directional difference only (μ₁ > μ₂ or μ₁ < μ₂)
3. TYPE I AND TYPE II ERRORS
Memory aid: ALFA = fALse Alarm; BETA = Better miss it
| Null Hypothesis TRUE | Null Hypothesis FALSE | |
|---|---|---|
| Reject H₀ | Type I Error (α) False Positive | Correct (Power = 1-β) |
| Fail to Reject H₀ | Correct (1-α) | Type II Error (β) False Negative |
Type I Error (α - False Positive):
- Rejecting a TRUE null hypothesis
- Concluding a drug works when it actually doesn't
- Probability = α (significance level, usually 0.05)
- "Crying wolf"
Type II Error (β - False Negative):
- Failing to reject a FALSE null hypothesis
- Concluding a drug doesn't work when it actually does
- Probability = β
- Power of study = 1 - β (usually set at 80% → β = 0.20)
- "Missing a real signal"
Relationship between errors:
- Decreasing α (e.g., 0.05 → 0.01) reduces Type I error but increases Type II error
- To reduce BOTH: increase sample size
Clinical relevance in anaesthesia:
- Type I Error: Adopting an ineffective drug (waste, potential harm)
- Type II Error: Rejecting an effective analgesic/anaesthetic agent (patient loses benefit)
4. ODDS RATIO (OR) + RISK RATIO - CLINICAL REFERENCE
2x2 Contingency Table (MUST memorize):
Disease + Disease -
Exposed | a | b |
Not Exposed| c | d |
Odds Ratio (OR) = (a×d) / (b×c)
- Used in: Case-control studies (you CANNOT calculate incidence/RR)
- OR = 1 → No association
- OR > 1 → Exposure increases odds of disease
- OR < 1 → Exposure is protective
Risk Ratio / Relative Risk (RR) = [a/(a+b)] / [c/(c+d)]
- Used in: Cohort studies and RCTs
- RR = 1 → No association
- RR > 1 → Exposure increases risk
- RR < 1 → Exposure protective
When OR ≈ RR: When the disease is rare (< 10% prevalence) - the rare disease assumption
Attributable Risk (AR) = Incidence in exposed - Incidence in unexposed
- Tells you how much disease is due to the exposure
Number Needed to Treat (NNT) = 1 / Absolute Risk Reduction
- Lower NNT = more effective treatment
Clinical example: OR = 2.5 for smoking and laryngeal cancer means: smokers have 2.5 times the ODDS of laryngeal cancer compared to non-smokers.
5. SAMPLE SIZE CALCULATION
Why it matters: Too small → miss real effects (Type II error). Too large → wasteful, unethical.
Formula (observational study):
N = Z²pq / d²
Where:
- Z = Z-value for desired confidence level (Z = 1.96 for 95% CI)
- p = expected prevalence/proportion
- q = 1 - p
- d = allowable error (precision)
Factors that INCREASE sample size required:
- Smaller effect size (harder to detect)
- Lower α (more stringent significance level)
- Higher desired power (less β)
- Greater variability in the outcome
- Rare disease/outcome
- Two-tailed test vs one-tailed
Power analysis: Usually conducted to achieve 80% power (β = 0.20) at α = 0.05.
DNB exam tip: The question often asks "what factors affect sample size" - remember: effect size, α level, power (1-β), and variability.
6. RANDOMIZATION + BLINDING - CONCEPTS IN RCT
Randomization
Definition: Process of assigning participants to groups by chance, so each participant has an equal probability of being in any group.
Purpose: Eliminates selection bias and distributes confounders (known and unknown) equally.
Types of Randomization:
| Type | Description |
|---|---|
| Simple | Coin flip equivalent; may give unequal groups |
| Block | Ensures equal group sizes (e.g., blocks of 4 or 6) |
| Stratified | Randomizes within strata (e.g., age groups, ASA grade) first, then randomizes |
| Cluster | Entire groups (wards, hospitals) randomized together |
| Minimization | Dynamic allocation to maintain balance on key variables |
Blinding (Masking)
Definition: Keeping participants, investigators, or both unaware of treatment allocation to prevent bias.
| Type | Who is blinded? |
|---|---|
| Single-blind | Patient only |
| Double-blind | Patient + clinician/researcher |
| Triple-blind | Patient + clinician + statistician |
Why blinding matters:
- Prevents performance bias (patients behaving differently)
- Prevents detection/assessment bias (observer unconsciously rating differently)
- Minimizes the placebo effect
- Reduces reporting bias
Allocation concealment vs blinding:
- Allocation concealment = preventing knowledge of UPCOMING assignment (before randomization) - prevents selection bias
- Blinding = preventing knowledge of CURRENT assignment (after randomization) - prevents performance/detection bias
7. SENSITIVITY AND SPECIFICITY
The 2x2 Diagnostic Table:
Disease + Disease -
Test + | TP | FP |
Test - | FN | TN |
Sensitivity = TP / (TP + FN)
- "How good is the test at detecting disease when it IS present"
- A HIGH sensitivity test, when NEGATIVE, RULES OUT disease
- Mnemonic: SnNout (Sensitive test, Negative result = rules OUT)
Specificity = TN / (TN + FP)
- "How good is the test at correctly identifying those WITHOUT disease"
- A HIGH specificity test, when POSITIVE, RULES IN disease
- Mnemonic: SpPin (Specific test, Positive result = rules IN)
Positive Predictive Value (PPV) = TP / (TP + FP)
- Probability of disease given a positive test
- Affected by disease prevalence
Negative Predictive Value (NPV) = TN / (TN + FN)
- Probability of no disease given a negative test
Likelihood Ratios:
- LR+ = Sensitivity / (1 - Specificity)
- LR- = (1 - Sensitivity) / Specificity
Trade-off: Increasing sensitivity decreases specificity (and vice versa). Visualized by the ROC curve.
Clinical context in anaesthesia:
- Pre-op tests: High sensitivity preferred (don't miss sick patients)
- Confirmatory tests: High specificity preferred (avoid unnecessary cancellations)
VERY LIKELY TOPICS
8. IMPACT FACTOR OF A JOURNAL
Definition: A metric indicating the average number of citations received per paper published in a journal over the past 2 years.
Formula:
Impact Factor (year X) = Citations in year X to articles published in years (X-1) and (X-2) / Total articles published in years (X-1) and (X-2)
Example: IF for 2024 = Citations received in 2024 for papers published in 2022-23 ÷ Total papers published in 2022-23
Significance:
- Higher IF = more influential/widely-read journal
- Important for: journal selection, career advancement, grant applications
- Common high-IF anaesthesia journals: Anesthesiology, British Journal of Anaesthesia, Anaesthesia
Limitations (frequently asked):
- Varies by specialty - cannot compare across fields
- Encourages publication bias toward positive/exciting results
- Review articles inflate IF (naturally get more citations)
- Self-citation manipulation possible
- Only reflects journal-level impact, not individual article quality
- Time lag - doesn't capture long-term influential papers
Alternatives:
- h-index (individual researcher impact)
- Eigenfactor score (accounts for quality of citing journals)
- Altmetric score (social media/news mentions)
- CiteScore
9. STANDARD DEVIATION (SD) AND STANDARD ERROR (SE)
Standard Deviation (SD):
- Measures variability/spread of individual data points around the mean
- Describes the sample itself
- Does NOT change substantially with larger samples
- Used when describing data
Formula: SD = √[Σ(x - x̄)² / (n-1)]
Standard Error (SE):
- Measures how precisely the sample mean estimates the population mean
- SE = SD / √n
- Gets smaller as n increases (more precise estimate with larger samples)
- Used when making inferences about population
Key difference for exams:
| SD | SE |
|---|---|
| Describes spread of data | Describes precision of mean estimate |
| Does not depend on n | Gets smaller as n increases |
| Use in descriptive statistics | Use in inferential statistics |
| Used to draw error bars showing variability | Used for confidence intervals |
Normal Distribution (68-95-99.7 rule):
- Mean ± 1 SD = 68% of data
- Mean ± 2 SD = 95% of data
- Mean ± 3 SD = 99.7% of data
10. CONFIDENCE INTERVAL (CI) - Linked to P-value
Definition: A range of values within which the true population parameter is expected to lie, with a specified level of confidence (usually 95%).
Interpretation of 95% CI: If we repeated the study 100 times, 95 of those CIs would contain the true population parameter.
Key rules for interpreting CI in exams:
| CI for difference in means | Interpretation |
|---|---|
| Does NOT include 0 | Statistically significant (p < 0.05) |
| Includes 0 | NOT statistically significant (p > 0.05) |
| CI for ratio (OR, RR) | Interpretation |
|---|---|
| Does NOT include 1 | Statistically significant |
| Includes 1 | NOT statistically significant |
CI width affected by:
- Sample size (larger n → narrower CI → more precise)
- Confidence level (99% CI is wider than 95% CI)
- Variability of data
Relationship with p-value: If 95% CI for OR does not include 1 → p < 0.05 (significant). These are two ways of expressing the same thing.
11. BIAS IN CLINICAL TRIALS
Definition: Systematic error that leads to incorrect estimation of the true effect.
Types of bias (exam favourite):
| Bias | Definition | How to reduce |
|---|---|---|
| Selection bias | Non-random selection into study groups | Randomization |
| Allocation bias | Researcher influences group assignment | Allocation concealment |
| Performance bias | Different care given to groups apart from intervention | Blinding |
| Detection/Observer bias | Outcome assessed differently between groups | Blinding of assessors |
| Recall bias | Cases remember exposure better than controls | Prospective design; case-control problem |
| Reporting bias | Selective reporting of outcomes | Pre-registration of trials |
| Publication bias | Positive results more likely to be published | Funnel plot, grey literature search |
| Attrition bias | Differential loss to follow-up | Intention-to-treat analysis |
| Confounding | Third variable affects both exposure and outcome | Randomization, multivariable analysis |
Intention-to-Treat (ITT) analysis: Analyze participants in the group they were randomized to, regardless of whether they completed the intervention. Preserves randomization benefits and gives real-world efficacy.
12. RCT (RANDOMIZED CONTROLLED TRIAL)
Definition: An experimental study where participants are randomly allocated to receive an intervention or control, with outcomes compared prospectively.
Hierarchy of evidence (highest to lowest):
- Systematic review / Meta-analysis
- RCT ← Gold standard for therapeutic efficacy
- Cohort study
- Case-control study
- Cross-sectional study
- Case reports/series
- Expert opinion
Key features of a good RCT:
- Randomization (eliminates selection bias)
- Allocation concealment
- Blinding (double-blind ideal)
- Adequate sample size (power calculation done)
- Intention-to-treat analysis
- Pre-registered protocol
- Reported per CONSORT guidelines
CONSORT checklist (2010): 25-item checklist for reporting RCTs. Includes: title, abstract, introduction, methods (eligibility, outcomes, randomization, blinding, sample size), results (flow diagram, baseline, analysis), discussion, and registration details. Accompanied by a CONSORT flow diagram showing participant progress through each stage.
13. ROC CURVE (Receiver Operating Characteristic)
Definition: A graphical plot showing the diagnostic ability of a test across all possible cut-off thresholds, plotting sensitivity (y-axis) vs 1-specificity (x-axis).
Area Under the Curve (AUC):
| AUC value | Interpretation |
|---|---|
| 1.0 | Perfect test |
| 0.9-1.0 | Excellent |
| 0.8-0.9 | Good |
| 0.7-0.8 | Fair |
| 0.6-0.7 | Poor |
| 0.5 | Useless (no better than chance) |
Uses of ROC curve:
- Compare diagnostic tests (the test with higher AUC is better)
- Select optimal cut-off point (point closest to top-left corner)
- At the optimal cut-off, sensitivity and specificity are maximized together
Relationship: Moving the cut-off point changes sensitivity and specificity inversely - every point on the ROC curve represents a different threshold.
RECENT TOPICS
14. CONSORT GUIDELINES
Full form: Consolidated Standards of Reporting Trials
Purpose: A standardized checklist to ensure complete and transparent reporting of RCTs so readers can assess validity and applicability.
Key components:
- 25-item checklist covering: title, abstract, introduction, methods, results, discussion, and other information (trial registration, protocol access, funding)
- CONSORT flow diagram (must-know): Shows number of participants at each stage:
- Assessed for eligibility
- Excluded (with reasons)
- Randomized
- Allocated to each arm
- Lost to follow-up (with reasons)
- Discontinued (with reasons)
- Analyzed (including ITT)
Why CONSORT matters: Poor reporting leads to inability to replicate studies, assess bias, or apply results. CONSORT compliance is required by most major anaesthesia journals.
15. PICO FRAMEWORK (Research Question Framing)
Full form: Population - Intervention - Comparison - Outcome
| Element | Description | Example |
|---|---|---|
| Population | Who is the study about? | Adult patients undergoing laparoscopic cholecystectomy |
| Intervention | What intervention/exposure? | Dexmedetomidine infusion |
| Comparison | Compared to what? | Propofol infusion |
| Outcome | What outcome measured? | Hemodynamic stability and recovery time |
PICO example (anaesthesia):
"In adult patients undergoing general anaesthesia (P), does use of BIS monitoring (I) compared to standard clinical monitoring (C) reduce the incidence of intraoperative awareness (O)?"
Why PICO matters:
- Structures the research question clearly
- Guides database search strategy in systematic reviews
- Helps define inclusion/exclusion criteria
- Part of CONSORT and systematic review protocols
EXTRA TOPICS (From Second Image)
16. TYPES OF RESEARCH STUDIES
Research Studies
│
├── Observational (no intervention)
│ ├── Descriptive
│ │ ├── Case report
│ │ ├── Case series
│ │ └── Cross-sectional (prevalence study)
│ └── Analytical
│ ├── Case-control (retrospective; calculates OR)
│ └── Cohort (prospective; calculates RR, AR)
│
└── Experimental (intervention by researcher)
├── RCT (randomized; gold standard)
├── Non-randomized controlled trial
└── Community trial
17. CASE-CONTROL vs COHORT STUDY
| Feature | Case-Control | Cohort |
|---|---|---|
| Direction | Backward (effect → cause) | Forward (cause → effect) |
| Timing | Retrospective | Usually prospective |
| Start point | Disease present | No disease initially |
| Measure | Odds Ratio (OR) | RR, AR, Incidence |
| Bias | Recall bias | Loss to follow-up bias |
| For rare outcomes | Yes | Difficult (need huge n) |
| For rare exposures | Less ideal | Yes |
| Multiple exposures | Can study many | Less efficient |
| Multiple outcomes | Less efficient | Can study many |
| Cost/time | Low | High |
| Ethical issues | Less (no new exposure) | More (may expose to harm) |
18. META-ANALYSIS AND FOREST PLOT
Meta-analysis: Statistical technique that pools quantitative results from multiple studies to produce a single, more precise effect estimate.
Forest plot (exam essential):
- Each horizontal line = one study; length = confidence interval
- Box = point estimate (size proportional to study weight)
- Diamond at bottom = pooled effect estimate
- Vertical line at 1 (for ratios) or 0 (for differences) = "line of no effect"
- If CI of a study crosses the line of no effect → that study alone is non-significant
- If the diamond crosses the line of no effect → overall pooled result is non-significant
Heterogeneity:
- I² statistic: Measures % of variability due to true heterogeneity (not chance)
- I² < 25%: Low heterogeneity
- I² 25-75%: Moderate
- I² > 75%: High (pooling questionable)
- Cochran's Q test: p < 0.05 = significant heterogeneity
Statistical models:
- Fixed-effects model: All studies measure same true effect
- Random-effects model: True effect varies between studies (used when heterogeneity is present)
Publication bias: Assessed by funnel plot (asymmetric funnel = publication bias likely) and Egger's test.
19. NON-PARAMETRIC TESTS
When to use: When data is not normally distributed, ordinal data, or small samples.
| Parametric Test | Non-parametric Equivalent | Use |
|---|---|---|
| Independent t-test | Mann-Whitney U test | Compare 2 independent groups |
| Paired t-test | Wilcoxon signed-rank test | Compare 2 related groups |
| One-way ANOVA | Kruskal-Wallis test | Compare 3+ independent groups |
| Pearson correlation | Spearman's rank correlation | Correlation (non-normal data) |
Key features of non-parametric tests:
- Based on ranks, not raw values
- No assumption of normal distribution
- Less statistical power than parametric tests (if normality holds)
- More appropriate for small samples, skewed data, outliers
20. ANOVA (Analysis of Variance)
Use: Compare means of THREE or more groups simultaneously.
Why not multiple t-tests? Multiple comparisons inflate Type I error rate. ANOVA keeps overall α at 0.05.
Test statistic: F-ratio = Variance BETWEEN groups / Variance WITHIN groups
- Large F → Groups differ more than expected by chance
Types:
- One-way ANOVA: One independent variable (factor)
- Two-way ANOVA: Two independent variables + tests for interaction
- Repeated measures ANOVA: Same subjects measured multiple times
Post-hoc tests (done after significant ANOVA to find which groups differ):
- Tukey's HSD (most common)
- Bonferroni correction
- Scheffé test
Assumptions: Normality, homogeneity of variance (Levene's test), independence.
Non-parametric alternative: Kruskal-Wallis test.
21. EVIDENCE-BASED MEDICINE (EBM)
Definition: Conscientious, explicit, and judicious use of current best evidence in making decisions about individual patient care (Sackett, 1996).
Three pillars of EBM:
- Best available research evidence
- Clinical expertise of the clinician
- Patient values and preferences
Steps in EBM (5-step approach):
- Ask - Frame clinical question (using PICO)
- Acquire - Search literature (PubMed, Cochrane)
- Appraise - Critically evaluate evidence
- Apply - Implement in clinical practice
- Assess - Evaluate outcomes
22. LEVELS OF EVIDENCE AND GRADES OF RECOMMENDATION
Oxford Levels of Evidence (simplified):
| Level | Study Type |
|---|---|
| 1a | Systematic review of RCTs |
| 1b | Well-designed RCT |
| 2a | Systematic review of cohort studies |
| 2b | Individual cohort study |
| 3a | Systematic review of case-control studies |
| 3b | Individual case-control study |
| 4 | Case series |
| 5 | Expert opinion |
GRADE System Recommendations:
- Grade A: Consistent Level 1 evidence
- Grade B: Consistent Level 2-3 evidence or extrapolation from Level 1
- Grade C: Level 4 evidence or extrapolation from 2-3
- Grade D: Level 5 or troublingly inconsistent evidence
23. PRIMARY vs SECONDARY RESEARCH
| Feature | Primary Research | Secondary Research |
|---|---|---|
| Data | Original, newly collected | Already existing data |
| Examples | RCT, cohort study, case-control | Systematic review, meta-analysis, literature review |
| Time/Cost | High | Lower |
| Bias control | Researcher controls design | Limited by quality of original studies |
| Purpose | Generate new knowledge | Synthesize existing knowledge |
24. SAMPLING TECHNIQUES
| Type | Description | Use |
|---|---|---|
| Simple random | Every individual has equal chance (lottery) | Homogeneous population |
| Systematic random | Every nth person selected | Ordered lists |
| Stratified random | Divide into strata, then randomly sample from each | Heterogeneous population, ensure subgroup representation |
| Cluster | Groups (clusters) randomly selected; all in cluster included | Geographically dispersed populations |
| Multistage | Cluster sampling done in multiple stages | Large national surveys |
| Convenience | Whoever is available | Pilot studies (high bias) |
| Purposive | Researcher handpicks subjects | Qualitative research |
| Snowball | Existing subjects recruit next (chain referral) | Hard-to-reach populations |
25. RESEARCH ETHICS AND INFORMED CONSENT
Helsinki Declaration: Foundation of research ethics globally. Key principles:
- Participant welfare > scientific interest
- Informed consent mandatory
- Independent ethics committee review required
- Vulnerable populations require extra protection
- Protocol must be publicly registered
Components of valid informed consent:
- Disclosure (full information about study)
- Comprehension (patient understands)
- Voluntariness (no coercion)
- Competence (legal capacity)
- Decision (explicit agreement)
Ethical principles (Belmont Report):
- Respect for persons (autonomy)
- Beneficence (maximize benefits)
- Justice (fair distribution)
26. PUBLICATION ETHICS AND PLAGIARISM
Plagiarism: Presenting another's work/ideas as your own without proper attribution.
Types:
- Direct plagiarism (copy-paste)
- Self-plagiarism (republishing own prior work)
- Mosaic/patchwork plagiarism (mixing own and others' text)
- Inadvertent plagiarism
Prevention: Turnitin/iThenticate software; proper citation; paraphrasing with acknowledgment.
Predatory journals: Journals that charge publication fees without proper peer review. Beall's list identifies predatory publishers.
Authorship criteria (ICMJE): Must fulfill ALL of:
- Substantial contribution to conception/design or data
- Drafting or critically revising the manuscript
- Final approval of published version
- Accountability for all aspects of the work
Ghost authorship: Listed as author without contribution. Honorary/Gift authorship: Contributing without meeting criteria. Both are unethical.
QUICK REVISION TABLE
| Topic | Key Number/Formula |
|---|---|
| P-value threshold | p < 0.05 = significant |
| Type I error | α = 0.05 (false positive) |
| Type II error | β = 0.20 → Power = 80% |
| Sample size formula | N = Z²pq/d² |
| SD vs SE | SE = SD/√n |
| 95% CI | Mean ± 1.96 × SE |
| Normal distribution | 68-95-99.7 rule |
| OR formula | (ad)/(bc) in 2×2 table |
| Sensitivity | TP/(TP+FN) - SnNout |
| Specificity | TN/(TN+FP) - SpPin |
| AUC (perfect test) | 1.0 |
| AUC (useless test) | 0.5 |
| I² (low heterogeneity) | < 25% |
| Impact factor formula | Citations in year X / Articles published in (X-1 and X-2) |
---All 26 topics from your notes and the PDF are covered above. Here's a quick guide to how it's organized:
Must Do (Topics 1-7):
- P-value, Null Hypothesis, Type I/II errors, Odds Ratio/Risk Ratio, Sample Size, Randomization + Blinding, Sensitivity/Specificity
Very Likely (Topics 8-13):
- Impact Factor, SD vs SE, Confidence Intervals, Bias in trials, RCT structure, ROC curve
Recent (Topics 14-15):
- CONSORT guidelines + flow diagram, PICO framework
Extra/Additional (Topics 16-26):
- Study types, Case-control vs Cohort comparison table, Meta-analysis + Forest plot, Non-parametric tests, ANOVA, EBM, Levels of evidence, Primary vs Secondary research, Sampling techniques, Research ethics, Publication ethics
A few exam tips specific to DNB Anaesthesia:
- The 2x2 contingency table is the backbone - learn it once and you can derive OR, RR, sensitivity, specificity, PPV, NPV from it
- SnNout / SpPin mnemonics are frequently tested
- Forest plots - know how to read them (whether diamond crosses the line of no effect)
- For CONSORT, know the flow diagram stages by name
- I² > 75% = pooling data in meta-analysis is questionable
This is a shared conversation. Sign in to Orris to start your own chat.