Centor Score

The Centor Score (also called Centor Criteria) is a clinical decision rule developed in 1981 by Dr. Robert Centor to estimate the probability of Group A Streptococcal (GAS) pharyngitis (strep throat) and guide testing and antibiotic treatment decisions. It uses four clinical criteria, each worth 1 point: (1) Tonsillar exudates (white or yellow patches on tonsils), (2) Tender anterior cervical lymphadenopathy (painful swollen lymph nodes in front of neck), (3) Fever >38°C (100.4°F) by history or measurement, and (4) Absence of cough. Total score ranges 0-4 points. Probability of strep throat by score: Score 0: 1-2.5% likelihood (strep extremely unlikely). Score 1: 5-10% likelihood (strep unlikely). Score 2: 11-17% likelihood (strep possible, testing recommended). Score 3: 28-35% likelihood (strep likely, testing recommended). Score 4: 51-53% likelihood (strep very likely, consider empiric treatment). The score helps avoid unnecessary antibiotic use in viral pharyngitis (70-85% of sore throats) while identifying patients needing further testing or treatment. Modified Centor Score (McIsaac Score) adds age adjustment: Age 3-14 years: add 1 point (higher strep prevalence). Age 15-44 years: 0 points. Age ≥45 years: subtract 1 point (lower strep prevalence). This extends score range to -1 to +5. Clinical validation shows the Centor Score has 75-80% sensitivity and 55-60% specificity for GAS pharyngitis. While not perfect, it effectively stratifies risk and reduces unnecessary rapid strep tests or throat cultures by 30-40% in low-risk patients while ensuring appropriate testing in moderate-to-high-risk patients. The score does not replace clinical judgment—patients with severe symptoms, immunocompromise, or known strep exposure may warrant testing regardless of score.

Tonsillar exudates (1 point if present): Use tongue depressor and light source (penlight or otoscope light) to visualize posterior pharynx. Look for discrete white, yellow, or gray patches on tonsillar surface or tonsillar pillars. True exudates are patchy and adherent (not easily wiped away), distinguishing them from post-nasal drip or food debris. Strep exudates are typically creamy-white to yellow and may cover entire tonsillar surface in severe cases. Do not give point for: Generalized pharyngeal erythema alone, clear mucus, or debris that easily wipes away. Note: Viral pharyngitis (especially EBV infectious mononucleosis and adenovirus) can also cause exudates, so presence doesn't confirm strep but increases suspicion. Exudates present in 50-70% of strep cases but also 20-30% of viral pharyngitis. Tender anterior cervical lymphadenopathy (1 point if present): Palpate anterior cervical chain (along anterior border of sternocleidomastoid muscle) from angle of jaw to clavicle bilaterally. Positive finding requires enlarged (>1 cm) AND tender lymph nodes. Technique: Use fingertips with gentle circular pressure, ask patient if tender. Strep typically causes anterior nodes; posterior cervical or generalized lymphadenopathy suggests viral infection (EBV, CMV). Do not give point for: Non-tender nodes (may be chronic), posterior cervical nodes only, or nodes <1 cm. Tender anterior cervical adenopathy present in 50-80% of strep cases but only 10-20% of viral pharyngitis. Fever >38°C/100.4°F (1 point if present): Use documented temperature measurement or reliable fever history within past 24 hours. Oral, temporal, or tympanic measurements acceptable. Give point if: Current measured temperature ≥38.0°C OR patient reports documented fever ≥38.0°C at home within 24 hours. Do not give point for: Subjective fever without measurement, mild temperature elevation 37.5-37.9°C, or fever >48 hours ago. Fever >38°C present in 40-60% of strep cases versus 20-30% of viral pharyngitis. Note that antipyretic use (acetaminophen, ibuprofen) may mask fever—if patient took antipyretics within 4 hours, use pre-medication temperature or history. Absence of cough (1 point if absent): Ask patient directly: "Do you have a cough?" Give point if patient denies cough. Do not give point if: Any cough present (productive, non-productive, intermittent). Minor throat-clearing doesn't count as cough. Rationale: Cough strongly suggests viral upper respiratory infection. Cough present in 70-80% of viral pharyngitis but only 15-25% of strep pharyngitis. Absence of cough is one of strongest predictors, increasing strep likelihood 2-fold. Examination context: Complete exam in well-lit room with tongue depressor, thermometer, and hand sanitizer. Document all four criteria explicitly. Typical strep presentation: abrupt onset severe sore throat, odynophagia (painful swallowing), fever, tonsillar exudates, anterior cervical adenopathy, absence of URI symptoms (rhinorrhea, cough, conjunctivitis). Typical viral presentation: gradual onset, associated cough, rhinorrhea, hoarseness, conjunctivitis, absence of exudates.

Score 0-1 (low risk, 1-10% strep probability): Recommendation: No testing, no antibiotics. Treat symptomatically with analgesics (acetaminophen, ibuprofen), throat lozenges, warm salt water gargles. Provide reassurance—sore throat likely viral, will resolve in 3-7 days. Rationale: Testing yields more false positives than true positives at this prevalence (positive predictive value only 10-15% even with positive test). Empiric antibiotics cause more harm (side effects, cost, resistance) than benefit. Follow-up: Return if symptoms worsen or persist >7 days, or if new symptoms develop (difficulty breathing, drooling, severe pain). Score 2-3 (moderate risk, 11-35% strep probability): Recommendation: Perform rapid antigen detection test (RADT) or throat culture. If RADT positive: Treat with antibiotics. If RADT negative in adults: No further testing, treat symptomatically (adult throat culture not recommended by IDSA due to low GAS complication rate). If RADT negative in children: Send backup throat culture (GAS can cause rheumatic fever in children, so higher sensitivity needed). Testing specifics: Rapid strep test (RADT): Results in 5-10 minutes, sensitivity 85-95%, specificity 95-98%. Sample posterior pharynx and both tonsillar areas with swab. Throat culture: Gold standard, sensitivity 95-98%, results in 24-48 hours. Indicated if RADT negative in children or if high clinical suspicion despite negative RADT. Score 4 (high risk, 51-53% strep probability) or Modified Centor ≥4: Two acceptable strategies: Option 1 (preferred in most settings): Perform RADT, treat if positive. Still recommended to confirm before antibiotics as 47-49% will not have strep. Option 2 (empiric treatment): Start antibiotics without testing in high-resource settings where follow-up assured, patient preference for immediate treatment, or limited access to testing. Antibiotic regimens (treat for 10 days): First-line: Penicillin V 500 mg PO twice daily (or 250 mg four times daily) OR Amoxicillin 500 mg twice daily (or 875 mg once daily). Benefits: narrow spectrum, cheap ($4-10), effective against GAS. Penicillin allergy: Cephalexin 500 mg twice daily (safe if non-severe penicillin allergy), Azithromycin 500 mg day 1, then 250 mg days 2-5 (use if severe penicillin allergy), or Clindamycin 300 mg three times daily. Avoid: Fluoroquinolones (resistance, unnecessary broad spectrum), amoxicillin-clavulanate (broader than needed), shorter courses than 10 days (higher relapse rates). Benefits of treating confirmed strep: Reduces symptom duration by 1-2 days, prevents suppurative complications (peritonsillar abscess, cervical lymphadenitis, rare in antibiotic era), prevents acute rheumatic fever (primary reason to treat—reduces risk from 3% to <0.3%), prevents post-streptococcal glomerulonephritis (questionable benefit), reduces transmission (patient non-contagious after 24 hours of antibiotics), allows return to school/work sooner. Harms of unnecessary antibiotics: Adverse effects (diarrhea 5-10%, rash 1-5%, C. difficile <1%, anaphylaxis <0.1%), antibiotic resistance, cost, medicalization of self-limited illness. The Centor Score balances these benefits and harms by targeting treatment to those most likely to benefit.

Clinical limitations: Moderate predictive accuracy: Even at highest score (4 points), only 51-53% have strep—nearly half of patients scored as high-risk have viral pharyngitis. Conversely, 1-2.5% with score 0 still have strep. This necessitates testing even at high scores rather than empiric treatment in most scenarios. The score provides probability, not diagnosis. Overlap with other conditions: Infectious mononucleosis (EBV) frequently mimics high Centor scores—presents with fever, exudates, adenopathy, absence of cough, scoring 3-4 points. However, EBV typically has posterior cervical adenopathy (not anterior), generalized lymphadenopathy, splenomegaly, and more severe fatigue. Consider monospot test if clinical picture suggests EBV, especially in adolescents/young adults. Treating EBV with amoxicillin causes morbilliform rash in 80-90% of cases. Other mimics: Fusobacterium necrophorum pharyngitis (Lemierre's syndrome precursor—similar presentation, needs different antibiotics), acute HIV infection, diphtheria (rare in vaccinated populations). Age-related performance: Original Centor Score developed in adults (ages 15-50). Performance differs in children: Children <3 years: GAS pharyngitis rare (<5% of cases), presentation atypical (may have nasal symptoms, cough), rheumatic fever risk very low. Score not validated in this age group—don't use Centor. Children 3-14 years: Higher GAS prevalence (20-37% of pharyngitis), higher rheumatic fever risk if untreated. Modified Centor (McIsaac) adds 1 point for this age, improving performance. Testing threshold lower—test and treat more liberally. Adults >45 years: Lower GAS prevalence (5-10%), complications rare. Modified Centor subtracts 1 point, reducing unnecessary testing. When NOT to use Centor Score: Severe or complicated pharyngitis: Peritonsillar abscess (unilateral tonsillar bulge, uvula deviation, trismus—needs drainage and broad-spectrum antibiotics), retropharyngeal abscess (muffled voice, drooling, neck stiffness—needs CT and ENT referral), epiglottitis (drooling, tripod positioning, respiratory distress—medical emergency), Lemierre's syndrome (internal jugular vein thrombosis—needs IV antibiotics). Immunocompromised patients: HIV, chemotherapy, immunosuppressive therapy—atypical presentations, broader differential (fungal pharyngitis, CMV, HSV), lower threshold for testing and treatment. Asymptomatic pharyngeal GAS carriage: 5-15% of population (20-30% of children) carry GAS without infection. Positive test in asymptomatic person or person with viral URI symptoms doesn't require treatment—treat clinical pharyngitis, not colonization. Recent strep exposure/outbreak: Known household contact with strep, or community outbreak setting may warrant lower threshold for testing/treatment regardless of score. Patient requesting antibiotics: Score doesn't replace shared decision-making. Discuss score results, strep probability, antibiotic benefits/risks. Some patients strongly prefer immediate antibiotics despite low score; others prefer avoiding antibiotics even with positive test. International applicability: Score developed in US. GAS prevalence, rheumatic fever risk, and antibiotic resistance patterns vary internationally, affecting score performance and treatment thresholds.

Modified Centor Score (McIsaac Score, 1998): Adds age modification to original Centor: Age 3-14 (+1 point), Age 15-44 (0 points), Age ≥45 (-1 point). Score range -1 to +5. Advantages: Better performance across age spectrum, validated in both children and adults. Thresholds: Score ≤0: 1-2% strep probability, no testing. Score 1: 5-10%, no testing. Score 2-3: 11-35%, perform RADT. Score ≥4: 50-60%, RADT or empiric treatment. Studies show Modified Centor slightly better than original Centor (AUC 0.69 vs. 0.66) but difference modest. Widely adopted in primary care guidelines. FeverPAIN Score (UK, 2013): Five criteria: Fever in last 24h (1 point), Purulence/exudate (1 point), Attend rapidly <3 days (1 point), severely Inflamed tonsils (1 point), No cough/coryza (1 point). Score 0-1: 13-18% strep, no antibiotics. Score 2-3: 34-40% strep, delayed prescribing (wait 3-5 days, use if not improving). Score 4-5: 62-65% strep, immediate antibiotics. Comparison to Centor: Similar overall accuracy (AUC 0.69-0.71), but FeverPAIN includes symptom acuity and tonsillar inflammation severity (not just exudate presence). Some studies show FeverPAIN identifies high-risk patients better (62-65% vs 51-53% at highest score). However, "severely inflamed tonsils" more subjective than Centor criteria. Clinical prediction rules vs. biomarkers: C-reactive protein (CRP): CRP >50 mg/L suggests bacterial infection. Sensitivity 60-70%, specificity 70-75%. Adding CRP to Centor slightly improves accuracy but requires point-of-care testing (not universally available). Cost-effectiveness debated—CRP test costs $10-20 versus RADT $8-12. White blood cell count: WBC >12,000 suggests bacterial infection but poor specificity (many viral infections cause leukocytosis). Not recommended for routine pharyngitis evaluation. RADT alone vs. clinical scoring: Universal RADT (testing all pharyngitis regardless of clinical features) achieves highest sensitivity but substantially increases cost and antibiotic use. Centor-based selective testing reduces tests by 30-40% with minimal impact on patient outcomes. Cost-effectiveness models favor Centor-guided testing in adults, modified Centor in children. "Treat all" vs. "test and treat" strategies: Empiric antibiotics for all pharyngitis (historical approach pre-1980s) results in 70-85% unnecessary antibiotic use (most pharyngitis is viral). Causes significant antibiotic-related adverse events, drives resistance, increases costs. Modern "test and treat" approach (Centor/RADT-guided) reduces antibiotic use by 60-70% while maintaining low rates of complications. Comparative cost-effectiveness: From payer perspective in US, strategies ranked by cost-effectiveness: (1) Modified Centor-guided RADT (most cost-effective—$45-80 per strep case diagnosed), (2) Original Centor-guided RADT ($50-90 per case), (3) FeverPAIN-guided delayed prescribing ($55-95 per case), (4) RADT for all comers ($80-120 per case), (5) Empiric antibiotics for all pharyngitis ($150-250 per case, least cost-effective). Conclusion: Centor/Modified Centor combined with RADT represents optimal balance of accuracy, cost-effectiveness, antibiotic stewardship, and patient outcomes.

Suppurative (bacterial spread) complications (early, within 1-2 weeks): Peritonsillar abscess (quinsy): Most common complication, occurs in 1-2% of untreated strep cases (0.1% if treated). Collection of pus between tonsillar capsule and pharyngeal muscles. Presents with worsening unilateral throat pain, trismus (difficulty opening mouth), muffled "hot potato" voice, uvular deviation away from abscess. Treatment: needle aspiration or incision and drainage plus antibiotics (amoxicillin-clavulanate or clindamycin). Retropharyngeal abscess: Rare complication (0.1% untreated, <0.01% treated), more common in children <5 years. Infection in potential space behind pharynx. Presents with severe dysphagia, drooling, neck stiffness, respiratory distress. Requires CT diagnosis, ENT consultation, often surgical drainage plus IV antibiotics. Life-threatening if airway obstruction or mediastinitis develops. Cervical lymphadenitis: Bacterial infection of cervical lymph nodes, occurs in 1-2% untreated cases. Presents with progressively enlarging, tender neck mass, fever, may develop fluctuance requiring drainage. Otitis media and sinusitis: Secondary spread to middle ear or sinuses, seen in 5-10% of untreated pediatric cases. Antibiotic effect on suppurative complications: Treatment reduces risk by 50-75%, from 2-3% to 0.5-1%. Benefit modest because overall risk is low and many develop despite treatment. Non-suppurative (immune-mediated) complications (delayed, 2-3 weeks post-infection): Acute rheumatic fever (ARF): Most serious complication and primary reason to treat strep pharyngitis. Immune response to GAS antigens cross-reacts with cardiac, joint, CNS, and skin tissues. Occurs in 0.3-3% of untreated GAS pharyngitis (risk higher in developing countries, certain populations). Presents 2-4 weeks after pharyngitis with Jones Criteria—major: carditis (pancarditis with permanent valve damage, especially mitral regurgitation), polyarthritis (migratory, affecting large joints), Sydenham chorea (involuntary movements), erythema marginatum (circular rash), subcutaneous nodules. Minor: fever, arthralgia, elevated ESR/CRP, prolonged PR interval on ECG. Diagnosis requires 2 major or 1 major + 2 minor criteria plus evidence of recent GAS infection. Antibiotic effect: Treatment within 9 days of symptom onset reduces ARF risk by 90-95%, from 3% to <0.3%. Critical window: Treatment effective even if started up to 9 days after symptom onset—allows time for test results without sacrificing ARF prevention. This is why rapid symptom relief is not the primary treatment goal; preventing ARF is. ARF now rare in developed countries (<1 case per 100,000 children) due to widespread antibiotic treatment, but remains significant in developing countries (20-50 per 100,000). Post-streptococcal glomerulonephritis (PSGN): Immune-mediated kidney inflammation, occurs 1-3 weeks after GAS pharyngitis in 0.5-1% of cases. Presents with hematuria (tea-colored urine), edema, hypertension, elevated creatinine. Most cases resolve spontaneously over 1-2 months, but 1-5% progress to chronic kidney disease. Antibiotic effect: Minimal to none—antibiotics do not prevent PSGN. Reason unknown, possibly because antibiotic kills bacteria but doesn't prevent immune response to bacterial antigens already present. PANDAS (Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcus): Controversial entity—sudden onset OCD and/or tics following GAS infection in children. Causality debated, optimal treatment unclear. Some experts recommend treating GAS to potentially prevent/improve symptoms, others question association. Estimated incidence if real: 1 in 200-1000 GAS infections. Overall treatment benefit: Antibiotics provide greatest benefit for preventing rheumatic fever (number needed to treat = 50-100 to prevent 1 ARF case), modest benefit for suppurative complications (NNT = 100-200), symptom relief by 1-2 days (NNT = 4-8 for 50% faster resolution), and reduced transmission (non-contagious after 24 hours of antibiotics versus 2-3 weeks untreated).