C-Reactive Protein (CRP) Testing: When to Order and How to Interpret
Order CRP testing in metabolically stable patients for cardiovascular risk stratification, or in acute settings to help differentiate bacterial infection from viral/non-infectious causes when combined with clinical assessment—but never use it as a standalone diagnostic tool.
When to Order CRP
Cardiovascular Risk Assessment
For cardiovascular risk stratification in metabolically stable patients, obtain two CRP measurements (fasting or nonfasting) optimally 2 weeks apart and average the results 1. This approach reduces within-individual variability and provides a more reliable baseline.
Critical caveat: If CRP is ≥10 mg/L, repeat the test and examine the patient for sources of infection or inflammation, as this level suggests acute inflammation rather than chronic cardiovascular risk 1.
Suspected Bacterial Infection
Outpatient Pneumonia
In adults with acute cough and suspected pneumonia, measure CRP when fever ≥38°C, dyspnea, pleural pain, tachypnea, or new chest examination findings are present 2. CRP strengthens both diagnosis and exclusion of pneumonia when combined with clinical features.
ICU/Sepsis Setting
In critically ill patients with new fever but no clear focus of infection, measure either CRP or procalcitonin to help rule out bacterial infection 3. However, do not routinely use biomarkers in established sepsis or septic shock 3—the evidence for benefit is uncertain and cost considerations are significant.
Pediatric Fever
In febrile children 1-36 months with temperature ≥39°C and no clinically detectable source, CRP can help identify serious bacterial infection, though it should never delay antibiotic treatment when clinically indicated 4.
How to Interpret Results
Cardiovascular Risk Stratification
Use these three risk categories 1, 5:
- Low risk: <1.0 mg/L
- Average risk: 1.0-3.0 mg/L
- High risk: >3.0 mg/L
The high-risk category carries approximately 2-fold increased relative risk compared to low-risk. These tertiles are based on population distributions from 15 populations involving 40,000 individuals 5.
Important limitations: Other inflammatory conditions produce mildly-to-moderately elevated CRP including inflammatory bowel disease, rheumatoid arthritis, and chronic alcoholism 1.
Infection Assessment
Outpatient Pneumonia
- CRP ≥30 mg/L plus suggestive symptoms/signs increases likelihood of pneumonia 2
- CRP <10 mg/L or 10-50 mg/L without dyspnea and daily fever makes pneumonia less likely 2
- CRP rises 12-24 hours after inflammatory insult, peaking at 48 hours 3
Critical Care Setting
- Typical cutoff: 10 mg/L for distinguishing infection from non-infectious inflammation 3
- Normal levels (<5 mg/L) in healthy individuals 3
- CRP has moderate diagnostic accuracy for sepsis (area under curve 0.73) with sensitivity 80% and specificity 61% 3
- Procalcitonin shows slightly better diagnostic accuracy (area under curve 0.85) with higher specificity (77%) than CRP 3
Pediatric Infections
- CRP <5 mg/dL effectively rules out serious bacterial infection (likelihood ratio 0.087, post-test probability 1.9%) 4
- CRP ≥7 mg/dL maximizes sensitivity (79%) and specificity (91%) for serious bacterial infection 4
- In neonatal sepsis, two CRP measurements <10 mg/L taken 24 hours apart are useful for excluding sepsis 6
Critical Pitfalls to Avoid
Never use CRP alone for diagnosis—it lacks sufficient sensitivity and specificity as a standalone test 7, 8. Always integrate with clinical assessment.
Do not exclude values >10 mg/L reflexively—while traditionally considered indicative of acute infection, chronic conditions (obesity, smoking, heritable factors) can produce sustained elevations 9. Consider the clinical context and potentially screen for infection symptoms or check liver enzymes.
Timing matters—CRP rises 12-24 hours after inflammatory insult and peaks at 48 hours 3. Early measurement may miss evolving infection.
Do not delay antibiotics waiting for CRP results in clinically suspected serious infections like meningitis, neonatal sepsis, or septic shock 7.
Serial measurements are more valuable than single values for monitoring treatment response 10, 11. CRP that fails to decrease or continues rising after 48 hours of appropriate antibiotics suggests treatment failure 6.
Reporting Standards
Express CRP results only as mg/L to one decimal point 1. This standardization improves consistency across laboratories and clinical interpretation.