// Evidence Grader
Claim grades, A through F
Every clinical claim attached to a biomarker in this atlas gets a transparent grade. Audit what's guideline-established, what's preliminary, and what's hype.
A
Guideline-established
B
Supported, context-specific
C
Promising, preliminary
D
Plausible, unproven
E
Popular, weak support
F
Misleading or false
A
Restrictive transfusion (Hb <7) is non-inferior to liberal in stable adults
TRICC, TRISS, FOCUS, multiple meta-analyses.
Hemoglobin
F
Hemoglobin alone is sufficient to diagnose iron deficiency
Iron deficiency precedes anemia by months. Use ferritin and TSAT.
Hemoglobin
D
Borderline-high Hb in a smoker should be 'watched' rather than worked up
Reasonable up to a point, but persistent Hct >49% (men) or >48% (women) warrants EPO and JAK2.
Hemoglobin
A
MCV >100 mandates B12 measurement
B12, folate (and MMA if borderline) are first-line.
Mean Corpuscular Volume
B
Mentzer index (MCV/RBC) <13 strongly suggests thalassemia trait over IDA
Useful screening rule with reasonable PPV in appropriate populations.
Mean Corpuscular Volume
B
High RDW with low MCV favors iron deficiency over thalassemia trait
RDW is normal in most thalassemia trait carriers.
Red Cell Distribution Width
B
RDW predicts mortality across many disease states
Reproducible association; mechanism is non-specific (likely inflammation/oxidative stress).
Red Cell Distribution Width
A
Platelet count <50 contraindicates most invasive procedures without correction
Specialty-specific thresholds vary; widely accepted.
Platelet Count
C
PLT/lymphocyte ratio is a useful general inflammation marker
Reproducible association in oncology and cardiology cohorts; not yet actionable at the bedside.
Platelet Count
A
ANC drives infection risk during chemotherapy
Foundational, guideline-supported.
White Blood Cell Count
B
Neutrophil-to-lymphocyte ratio predicts outcomes in many diseases
Consistent association; specificity is low.
White Blood Cell Count
A
Ferritin <30 µg/L is essentially diagnostic of absolute iron deficiency
Multiple guidelines (BSG, AGA, WHO).
Ferritin
B
Iron supplementation improves restless legs syndrome when ferritin <75
RLS guidelines support IV iron when ferritin <75 and TSAT <20%.
Ferritin
C
Targeting ferritin to 'optimal' 70–100 in non-anemic patients improves general fatigue
Plausible, several small RCTs positive in women; not universally replicated.
Ferritin
B
Routine ferritin screening detects pre-clinical hemochromatosis
Combined with TSAT and HFE genotype; cost-effective in high-prevalence populations.
Ferritin
A
ADA fasting glucose thresholds are appropriate for diabetes diagnosis
Evidence basis is microvascular complication risk; thresholds are pragmatic.
Fasting glucose
D
Fasting glucose 90–99 with normal A1c is harmless
Rising fasting glucose in this range correlates with insulin resistance and future diabetes risk.
Fasting glucose
A
HbA1c ≥6.5% diagnoses diabetes when assay is standardized
ADA, IDF, WHO.
Glycated hemoglobin
A
Tight glycemic control (<7%) reduces microvascular complications in type 2
UKPDS.
Glycated hemoglobin
D
Aggressive A1c targets (<6.5%) reduce macrovascular risk
ACCORD found increased mortality with aggressive glycemic targets.
Glycated hemoglobin
B
Fasting insulin detects insulin resistance years before fasting glucose rises
Reproducible in cohorts; assay variability limits clinical use.
Fasting insulin
B
HOMA-IR is a clinically useful insulin-resistance metric
Validated against clamp; population-specific cutoffs.
Fasting insulin
B
Lower upper-normal ALT thresholds (<30M/<19F) catch more clinically relevant disease
Prati 2002 and AASLD support.
Alanine aminotransferase
A
Statin-induced ALT elevations <3× ULN do not require statin discontinuation
FDA and major lipid guidelines.
Alanine aminotransferase
A
Lowering LDL-C reduces ASCVD events approximately linearly to <40 mg/dL
CTT meta-analyses, FOURIER, ODYSSEY.
LDL cholesterol
A
Statins are net-beneficial in primary prevention at 10-y risk ≥7.5%
USPSTF, ACC/AHA.
LDL cholesterol
E
LDL-C alone is sufficient — apoB adds nothing clinically
Discordance between LDL-C and apoB is common in insulin resistance and predicts residual risk.
LDL cholesterol
F
Raising HDL-C with niacin reduces ASCVD events
AIM-HIGH and HPS2-THRIVE were neutral or harmful.
HDL cholesterol
B
ApoB outperforms LDL-C as a CV risk predictor
Multiple cohorts including Sniderman meta-analyses.
Apolipoprotein B
A
Lp(a) is a causal, independent ASCVD risk factor
Mendelian randomization, multiple cohorts.
Lipoprotein(a)
E
Niacin or aspirin lower Lp(a)-mediated CV risk
No outcome trial supports this.
Lipoprotein(a)
C
Pelacarsen (antisense oligonucleotide) reduces Lp(a) ~80% — trial pending for events
HORIZON trial ongoing.
Lipoprotein(a)
B
hs-CRP refines CV risk stratification (JUPITER population)
Useful especially in intermediate-risk patients.
C-reactive protein
B
CRP-guided antibiotic stewardship reduces unnecessary prescribing in primary care
Cochrane review supports modest reduction.
C-reactive protein
A
TSH is the best initial test for thyroid dysfunction
All major endocrine guidelines.
Thyroid-stimulating hormone
E
Subclinical hypothyroidism (TSH 4.5–10) should always be treated
TRUST and other RCTs found no symptomatic benefit in mild cases.
Thyroid-stimulating hormone
B
Vitamin D supplementation prevents fractures and falls in older adults
Modest effect, especially with calcium.
25-OH vitamin D
C
Vitamin D supplementation reduces COVID-19 severity
Inconsistent RCT data.
25-OH vitamin D
E
Vitamin D prevents type 2 diabetes
VITAL and D2d trials largely negative.
25-OH vitamin D
B
PSA screening reduces prostate cancer-specific mortality.
ERSPC: NNS ~570 to prevent one death over 16 y; absolute benefit modest, harms (overdiagnosis) substantial.
Prostate-specific antigen
B
PSA density (PSA / prostate volume) outperforms total PSA for biopsy decisions.
Useful >0.15 ng/mL/cc threshold.
Prostate-specific antigen