Most medical headlines report Relative Risk Reduction (RRR), which magnifies small effects when baseline disease incidence is low. To evaluate any clinical trial or environmental cohort accurately, always convert relative percentages into Absolute Risk Reduction (ARR = Control Event Rate minus Experimental Event Rate) and Number Needed to Treat (NNT = 1 divided by ARR) over a defined follow-up horizon.
Comparing media-reported relative risk figures against true per-1,000 absolute differences and methodological certainty across major medical specialties.
| Clinical Domain & Intervention | Headline Relative Metric | Baseline vs. Treated Event Rate | True Absolute Difference (per 1,000) & NNT | Methodological Verdict |
|---|---|---|---|---|
| Primary Prevention Statin Low-to-moderate CV risk (5-yr RCT meta-analysis) |
RRR -25% to -30% Major vascular events |
Control: 4.0% Treated: 2.9% |
ARR 1.1% (11 per 1,000) NNT = 91 over 5 years |
HIGH CERTAINTY Hard endpoint benefit; absolute gain depends heavily on 10-yr baseline ASCVD risk. |
| GLP-1 Agonist (Semaglutide) SELECT Trial (Overweight + established CVD, 39.8 mo) |
HR 0.80 (RRR -20%) 3-point MACE |
Control: 8.0% Treated: 6.5% |
ARR 1.5% (15 per 1,000) NNT = 67 over 3.3 years |
HIGH CERTAINTY Proven secondary CV reduction; 16.6% vs 8.2% discontinued due to GI adverse events. |
| Metastatic Solid Tumor Approval Typical accelerated approval on PFS surrogate |
HR 0.68 (PFS gain) Median +2.4 months |
Median PFS: 4.8 vs 7.2 mo Overall Survival: NS |
OS ARR: 0 per 1,000 Grade 3/4 Toxicity: +220 / 1,000 |
SURROGATE ONLY Radiographic delay in tumor growth of 2.4 months without survival or QoL extension. |
| Urban Road Traffic Noise Exposure >53 dB Lden (WHO Europe & MedRead MS-946) |
RR 1.08 (+8% per 10 dB) Ischemic heart disease |
Baseline IHD: 10.0 / 1,000 At 63 dB Lden: 10.8 / 1,000 |
ARI +0.8 per 1,000 / +10 dB High population burden |
MODERATE · COHORT Small individual relative ratio translates to millions of DALYs across dense peri-urban corridors. |
Primary sources: NEJM 2023 (SELECT Trial) · The Lancet 2012 (CTT Statin Meta-Analysis) · JAMA Internal Medicine 2015 (Oncology Surrogates) · WHO Environmental Noise Guidelines 2018.
Without the baseline event rate in the untreated control group, a Relative Risk Reduction (RRR) is mathematically incomplete.
Concealed randomization balances both known and unknown baseline confounders. When an adequately powered multi-center RCT demonstrates a statistically and clinically meaningful reduction in all-cause mortality, organ failure, or validated patient quality of life under Intention-to-Treat (ITT) analysis, it warrants practice-changing clinical confidence.
Audit checklist: 10-Minute Study Appraisal Guide · Cardiometabolic RCT Ledger
Many FDA and EMA accelerated approvals rely on laboratory or radiographic surrogates: shrinking a tumor on CT scan (Progression-Free Survival), lowering HbA1c by 0.5%, or clearing amyloid plaques. Across oncology and cardiology history, over half of surrogate improvements fail to extend Overall Survival (OS) or improve patient symptoms due to off-target toxicity.
Audit checklist: Oncology OS vs. PFS Decoder · Surrogate vs. Hard Endpoints
For exposures that cannot ethically be randomized, such as urban road traffic noise (dB Lden), air particulate matter (PM2.5), or occupational toxins, large prospective cohorts provide vital dose-response curves. However, single-food or supplement observational studies with Hazard Ratios between 0.80 and 1.30 are frequently driven by socioeconomic and healthy-user confounding.
Audit checklist: WHO Environmental Noise Monograph · Confounding Bias Table
An open-access PDF indexed in Google Scholar or deposited in PubMed Central (PMC) via author manuscript upload is not automatically vetted by MEDLINE's Literature Selection Technical Review Committee. Watch for turnaround times under 14 days from submission to acceptance, missing ClinicalTrials.gov (NCT) registration, and post-hoc subgroup p-values.
Audit checklist: Open-Access Peer-Review Verification Guide · 2019-2022 Archival Index
med-read.org published open-access clinical and environmental monographs cited by the World Health Organization EMRO (who.int), Heriot-Watt University (hw.ac.uk), and SCIRP. Rebuilt as a non-commercial appraisal mentor and permanent citation archive.
A step-by-step clinical checklist for dissecting randomized trials, observational cohorts, hazard ratios, confidence intervals, and p-hacking red flags.
Why indexing in PubMed Central (PMC) differs from MEDLINE curation, how to verify Crossref DOIs and DOAJ standing, and how to spot paper-mill fabrications.
Dedicated acoustic epidemiology monograph honoring the historical Heriot-Watt University and SCIRP citations on peri-urban noise exposure (dB Lden / Lnight) and cardiovascular risk.
Side-by-side comparison of surrogate biomarkers (PFS, ORR, LDL-C, HbA1c) versus hard patient-centered outcomes (Overall Survival, MACE, Grade 3/4 toxicity).
Candid, plain-language answers grounded in CONSORT 2025, GRADE, WHO guidelines, and per-1,000 outcome arithmetic.
Relative risk compares the ratio of probabilities between two groups (for example, cutting risk from 2% to 1% is a 50% relative reduction), whereas absolute risk difference subtracts one rate from the other (2% minus 1% equals a 1% absolute reduction, or 10 fewer events per 1,000 people). Only absolute risk tells you your actual likelihood of benefiting.
Source: NCBI Bookshelf (Harding Risk Literacy) · Study Appraisal Guide
Between 2019 and 2022, MedRead Publishing Group published open-access articles on environmental noise exposure, pediatric nutrition integration, and polypharmacy that were cited by the World Health Organization (WHO EMRO), Heriot-Watt University, SCIRP, and Auctores. Maintaining permanent semantic concordance anchors prevents broken academic citations and updates those findings with 2024-2026 evidence.
Source: 2019-2022 Citation Concordance · Environmental Noise Monograph
No. A p-value below 0.05 only indicates that the observed difference would occur less than 5% of the time if the null hypothesis were true and there were zero bias. In large trials, tiny clinically irrelevant changes can achieve p < 0.05, while in small or p-hacked studies, selective reporting produces false-positive p-values.
Source: American Statistical Association Statement on P-Values (2016)