1. Epidemiology
• Incidence: ~2× more frequent than acute haematogenous osteomyelitis in childhood
• Peak incidence: early first decade; bimodal distribution (infants & school-age children)
• Male predominance (approximately 2:1)
• Distribution: 90% lower extremity
◦ Hip: ~54% of all cases (Taiwanese cohort, n=186)
◦ Knee: second most common
◦ Shoulder, ankle, elbow: less frequent
• Overall incidence may be declining due to immunization programmes (H. influenzae Hib vaccine: 30% reduction in paediatric septic arthritis cases)
Age-specific Susceptibility
Neonates (<4 wk): Vessels from metaphysis penetrate into cartilaginous physis → simultaneous osteomyelitis + septic arthritis common; multifocal disease
Infants (<18 mo): Ogden-Lister blood supply to chondroepiphysis; 4 'at-risk' intracapsular metaphyses (proximal femur, proximal humerus, distal lateral tibia, proximal radius)
Older children: Classic haematogenous route; metaphysis outside joint; isolated septic arthritis more common
2. Microbiology
Causative Organisms by Age
Age Group | Primary Pathogens | Key Notes |
Neonates (<4 weeks) | Group B Streptococcus, S. aureus, Gram-negative rods (E. coli), N. gonorrhoeae | Multifocal disease common; often polyarticular |
Infants (4 wk – 24 mo) | Kingella kingae ★, S. aureus, Group A Streptococcus, H. influenzae (unimmunized) | K. kingae is #1 organism <24 mo in immunized cohorts |
Children (2 – 12 yr) | S. aureus ★★, Group A Streptococcus, S. pneumoniae | S. aureus in 60–90%; MRSA increasing |
Adolescents (>12 yr) | S. aureus ★★, N. gonorrhoeae (sexually active) | Gonococcal arthritis: polyarticular, tenosynovitis, rash |
Sickle Cell Disease | Salmonella spp., S. aureus | Salmonella: diaphysis; S. aureus: metaphysis |
Kingella kingae — High-Yield Points
• Fastidious gram-negative coccobacillus; oropharyngeal coloniser in young children
• Gains bloodstream access via upper respiratory infection or oral mucosal breach
• Transmitted child-to-child; outbreaks in day-care settings documented
• Predilection for heart and musculoskeletal system
• Culture pearls: inoculate directly into enriched blood culture media (BACTEC) → significantly improves recovery
• Resistance profile: often resistant to vancomycin and clindamycin; sensitive to β-lactams
• Diagnosis: real-time PCR assay most reliable (conventional culture frequently negative)
• Prognosis: typically responds well to β-lactams with few sequelae
MRSA — Clinical Significance
• MRSA isolated in ~30% of paediatric musculoskeletal infections (vs 0% two decades prior — UTSW data)
• Associated with higher rates of DVT, subperiosteal abscess, pyomyositis, ICU admission
• Antibiotic options: vancomycin (IV) or clindamycin (if susceptibility confirmed by D-zone test)
• Linezolid (oxazolidinone class): effective against MRSA/VRSA; limited paediatric data
• MSSA is now acquiring MRSA virulence factors → comparable disease severity being reported
3. Pathophysiology
Route of Infection
• Primary: haematogenous bacteraemia → synovial seeding (no basement membrane in synovium → bacteria freely enter joint)
• Secondary: contiguous spread from adjacent osteomyelitis (especially at intracapsular metaphyses)
• Direct inoculation: foot puncture wounds, joint injections (less common)
Joint Destruction Cascade
• Bacteraemia → synovial seeding → synovitis → fibrinous exudate → synovial necrosis
◦ Leukocytes, synovial cells, and chondrocytes release proteases, peptidases, collagenases
◦ Glycosaminoglycan (GAG) loss: first measurable change → begins within 8 hours
◦ Collagen destruction: follows GAG loss → visible cartilage damage
◦ Neovascularisation + persistent bacterial colonisation → chronic inflammatory infiltration
• Critical concept: once enzymes are released, cartilage destruction continues even after bacteria are eradicated
Acute Phase Response in Septic Arthritis
• Infection = continuous activation of the 'survival phase' of the acute phase response
• Results in: overwhelming SIRS, consumptive coagulopathy, thrombosis of epiphyseal vessels
• Thrombotic sequelae in the epiphysis: chondrolysis, joint resorption, physeal arrest
• Bacterial 'hijacking' of coagulation via coagulase (S. aureus) and streptokinase (S. pyogenes) amplifies tissue damage
Why the Hip is Most Vulnerable
• Femoral head blood supply enters at periphery of femoral neck → intracapsular pressure from purulent effusion → vascular occlusion → AVN
• Entire proximal femoral metaphysis is intracapsular → contiguous osteomyelitis communicates directly with joint
• Perlman et al.: signs of adjacent joint septic arthritis present in up to 40% of osteomyelitis cases
4. Clinical Presentation
History
• Pain is the most common symptom — young children may present as irritability, refusal to walk/use limb
• Fever >38°C: reported in only 36–74% (not reliably present)
• Preceding illness (URI, varicella): critical history — K. kingae, Group A Strep
• Antecedent trauma: present in 30–50% — can trigger osteomyelitis/septic arthritis or can confuse diagnosis
• Prior antibiotic use: blunts fever and lab values → maintain high index of suspicion
• Duration: symptoms of septic arthritis are progressive; traumatic symptoms typically improve
Physical Examination
• General: observe child unaware — posture, spontaneous movement, refusal to bear weight
• Hip position: flexion + abduction + external rotation (reduces intracapsular pressure)
• Pain with passive ROM: hallmark sign of septic arthritis
• Erythema and swelling: may appear 24–36 hours after onset
• Sympathetic effusion: suggests adjacent osteomyelitis (without joint irritability)
• Hip joint: inaccessible to direct assessment — Trendelenburg sign, log-roll test
• Axial joints (spine, SI): localise by percussion, compression, range of motion
Clinical Comparison: Septic Arthritis vs. Toxic Synovitis vs. JRA
Feature | Septic Arthritis | Toxic Synovitis | JRA |
Fever | >38°C (variable) | Low-grade or absent | Low-grade (systemic JRA: spiking) |
Pain severity | Severe, progressive | Moderate, fluctuating | Moderate; looks worse than it feels |
Weight bearing | Refused | Painful but may weight-bear | Usually maintained |
WBC (serum) | >12,000 (variable) | Normal | Variable |
ESR | Usually >40 mm/hr | Normal or mildly elevated | Often elevated |
CRP | Elevated (often >2 mg/dL) | Normal or mildly elevated | Elevated |
Synovial WBC | >50,000 (typical) | <15,000 | <100,000 (rarely >100,000) |
Joint appearance | Worse than function | Correlates with symptoms | Worse than function |
Resolution | Requires surgery ± ABx | Self-limiting (1–2 wk) | Chronic/relapsing |
5. Diagnostic Workup
Laboratory Tests
• Principle: only blood/tissue culture, antibody titer, or PCR confirm infection — all other tests measure the acute phase response
Test | Sensitivity | Key Points | Timing |
WBC count | 25–73% | Least sensitive; check manual differential to exclude leukaemia | Immediate |
ESR | 85–95% | Elevated in 85–95% SA; rises 24–48 hr after onset; peaks day 3–5; slow to normalise (2–4 wk) | Delayed (48–72 hr) |
CRP | 92–98% | Most useful: rises within 4–6 hr, t½ = 17 hr; normalises rapidly with effective Rx; fails to fall → treatment failure | Early |
Procalcitonin | Investigational | <0.3 ng/mL: low risk; >0.5 ng/mL: concern; rises earlier than CRP; more specific for bacterial vs. viral | Early |
IL-6 | Research only | First APR marker to rise; stored in MSK tissue; released immediately after injury; short half-life | Immediate |
Blood culture | 30–60% | Obtain before antibiotics; do NOT withhold antibiotics to obtain culture if child is unwell | Immediate |
Synovial aspirate culture | 30–80% | Inoculate enriched media (K. kingae); Gram stain positive in ~33% | Immediate |
Synovial Fluid Analysis — Critical Values
Parameter | Normal | Transient Synovitis | Septic Arthritis | JRA / Inflammatory |
Appearance | Clear, straw | Clear/xanthochromic | Turbid/purulent | Turbid/xanthochromic |
WBC count (/mL) | <200 | <15,000 | >50,000 (typical) | 15,000–100,000 |
PMN % | <25% | <50% | >75% ★ | Variable |
Glucose | = serum | = serum | <50% serum ★ | Decreased |
Culture | Negative | Negative | Positive 30–80% | Negative |
PMN >75% is highly suggestive of septic arthritis even if WBC is <50,000/mL | A WBC of ≤50,000/mL found in 55% of culture-proven cases (Fink & Nelson)
Imaging
Plain Radiographs
• Sensitivity 43–75%, Specificity 75–83%
• Soft tissue swelling visible within 3 days (earliest plain film finding)
• Joint space widening at hip: asymmetric vs. contralateral → suggestive of effusion (especially in neonates)
• Bone changes: not apparent for 7+ days → late sign
• Late untreated SA: joint space narrowing, bone destruction both sides of joint, osteonecrosis of femoral head
Ultrasound
• Detects effusion (sensitivity ~95% for effusion, but cannot distinguish SA from toxic synovitis)
• False-negative rate: 5% (especially in <24 hr symptoms or bilateral effusions — Gordon et al., n=132)
• Best use: guide hip aspiration in the ED/radiology department
• Cannot reliably differentiate pericapsular pyomyositis from septic arthritis
MRI — Imaging Modality of Choice for Complex Cases
• Sensitivity 88–100%, Specificity 75–100%, PPV 85%
• Detects: effusion, synovitis, cartilage damage, bone marrow oedema (↓T1, ↑T2), abscesses, AVN
• Gadolinium enhancement: differentiates osteomyelitis (irregular geographic enhancement) from bone infarct (thin linear rim) and neoplasm
• Yang et al.: SA more likely to show adjacent bone marrow signal abnormality + perisynovial soft tissue contrast enhancement (vs. toxic synovitis)
• Sensitivity of MRI vs. bone scintigraphy for S. aureus osteomyelitis: 98% vs. 53% (Texas Children's Hospital, n=199)
• MRI after surgery: not affected by recent surgical oedema; reliable for re-evaluation
• Disadvantage: cost, frequent need for sedation/GA in young children
Bone Scan (Tc-99m)
• Sensitivity 89–94%, Specificity 94%, Accuracy ~92%
• Best for: unknown/multiple sites of infection, fever of unknown source
• Limitation: cannot distinguish septic arthritis from osteomyelitis; 'equally increased uptake on both sides of joint' suggests SA
• Cold scan: paradoxically associated with more severe disease (100% PPV for OM); suggests vascular thrombosis
• Neonates: sensitivity only 30–86% — plain radiographs more helpful in this age group
6. Kocher Criteria — Septic Arthritis vs. Toxic Synovitis of the Hip
The Kocher algorithm is the most widely used clinical prediction tool for differentiating septic arthritis from transient synovitis of the hip.
Kocher Predictors (Original 4 Variables — 1999) |
• 1. Fever (history of fever) |
• 2. Non-weight bearing |
• 3. ESR ≥ 40 mm/hr |
• 4. WBC > 12,000/mL |
Predicted Probability of Septic Arthritis
Number of Predictors Present | Predicted Probability of Septic Arthritis |
0 | < 0.2% |
1 | 3.0% (original) → 9.5% (prospective validation) |
2 | 40.0% (original) → 35.0% (validation) |
3 | 93.1% (original) → 72.8% (validation) |
4 | 99.6% (original) → 93.0% (validation) |
Updated Kocher-Philadelphia Algorithm (5 Variables)
• CRP elevation added as 5th variable (Children's Hospital of Philadelphia)
• 5 factors present → 98% probability of septic arthritis
• 4 factors present → 93% probability of septic arthritis
Critical Limitation |
• Prospective validation at a different institution (Luhmann et al.): 4 predictors → only 59% probability (vs. 99.6% in original study) |
• The Kocher algorithm cannot replace clinical judgement — use as adjunct, not as sole decision-making tool |
• In the acutely ill child, err on the side of surgical exploration |
7. Differential Diagnosis
Condition | Differentiating Features | Urgency |
Transient Synovitis (Toxic Synovitis) | Self-limiting; recent URTI common; low-grade or no fever; WBC/CRP normal or mildly elevated; responds to NSAIDs | Outpatient |
Acute Haematogenous Osteomyelitis | Point tenderness over bone (metaphysis); joint irritability mild (sympathetic effusion); bone scan/MRI localises | Urgent |
Pericapsular Pyomyositis | Pericapsular muscle signal on MRI; effusion size smaller than SA; Vanderbilt: 2:1 pyomyositis:SA in 'rule out hip infection' | Urgent |
Juvenile Rheumatoid Arthritis (JRA) | Gradual onset; polyarticular; joint looks worse than it feels; remains ambulatory; ANA/RF may be positive | Elective |
Rheumatic Fever (ARF) | Jones criteria; exquisite pain out of proportion to effusion; migratory polyarthritis; recent strep infection | Medical |
Lyme Arthritis | Endemic area; Borrelia serology + synovial PCR; WBC 47,000–64,000/mL | Medical |
Psoas Abscess / Septic Sacroiliitis | Hip pain + groin/buttock pain; pain with SI compression; CT/MRI essential | Urgent |
Leukaemia | Metaphyseal bands on XR; anaemia, thrombocytopenia; atypical WBC on differential; bone pain often multifocal | Emergency |
PSRA (Post-streptococcal Reactive Arthritis) | Recent strep infection; does not fulfil Jones criteria; self-limiting; controversial role of prophylactic ABx | Outpatient |
8. Treatment
General Principles
• Septic arthritis = surgical + antibiotic combination — antibiotics alone are INSUFFICIENT
• Purulent joint fluid: proteolytic enzymes persist after bacterial clearance → joint must be decompressed
• Do NOT withhold antibiotics to obtain cultures in an acutely ill child with exuberant acute phase response
• Hip septic arthritis: treat as surgical emergency — risk of AVN with delay
Empiric Antibiotic Selection by Age
Age / Setting | Empiric Regimen | Rationale |
Neonates (<4 wk) | Ceftriaxone or Cefotaxime + Oxacillin | Cover: Group B Strep, S. aureus, gram-negatives, gonococci |
Infants <2 yr (unimmunized) | Cefuroxime or Ceftriaxone + Oxacillin | Cover: H. influenzae, K. kingae, Group A Strep, S. aureus |
Children 2–12 yr (immunized) | Oxacillin or Cefazolin | Cover: S. aureus, Streptococcus spp., S. pneumoniae |
Suspected MRSA | Vancomycin IV or Clindamycin (D-zone test negative) | Adjust when culture + sensitivity available |
Sickle Cell Disease | Cefotaxime/Ceftriaxone (Salmonella coverage) | Add antistaphylococcal agent if S. aureus not excluded |
Sexually active adolescent | Ceftriaxone (gonococcal coverage) | Consider additional MRSA coverage if no response |
Specific Antibiotic Dosing Reference
Antibiotic | IV Dose | Oral Equivalent | Notes |
Oxacillin | 150–200 mg/kg/d ÷ q6h | Dicloxacillin 100 mg/kg/d ÷ q6h | Preferred for MSSA; risk of skin sloughing if extravasation |
Cefazolin | 100–150 mg/kg/d ÷ q8h | Cephalexin 100–150 mg/kg/d ÷ q6h | Excellent MSSA coverage; more convenient dosing |
Vancomycin | 40–60 mg/kg/d ÷ q6h | Not available orally | Monitor troughs (10–15 μg/mL); infuse over ≥60 min; red man syndrome |
Clindamycin | 40 mg/kg/d ÷ q8h IV | 30–40 mg/kg/d ÷ q8h PO | Confirm MRSA susceptibility with D-zone test; excellent bone penetration |
Ceftriaxone | 50–100 mg/kg/d ÷ q12-24h | Not applicable | Gram-negative + gonococcal coverage; good for outpatient (once daily) |
Linezolid | 10 mg/kg q8h (<12 yr) | Available (same dose) | Reserve for VRSA/MRSA failures; monitor for thrombocytopenia |
IV-to-Oral Transition Criteria
Criteria for Safe Oral Step-Down (All must be met) |
• Afebrile (no fever for ≥24–48 hours) |
• Clinically improved: reduced tenderness, improved ROM, no limp, return of appetite |
• CRP declining (trending toward normal) — not necessarily normalised |
• Child can swallow and is expected to absorb oral medication |
• Compliant family/carer confirmed |
• Susceptible organism identified OR empiric oral agent achieves adequate serum bactericidal titre (≥1:8) |
Total Antibiotic Duration
• Septic arthritis: minimum 3 weeks total (IV + oral combined)
• Complicated cases (adjacent osteomyelitis, MRSA, abscess): 4–6 weeks
• ESR normalisation: NOT required before discontinuation (may take 4–8 wk — unnecessarily prolongs therapy)
• Use CRP as primary monitoring marker: short half-life (17 hr), responds within days of effective treatment
Monitoring During Treatment
• Weekly labs: CBC (neutropenia risk), CRP, ESR, AST, ALT, creatinine
• CRP serial monitoring: should decline daily with effective therapy; any plateau/rise → re-evaluate
• Vancomycin: monitor trough levels every 3–4 days; also monitor BUN/creatinine
• Clindamycin: monitor for C. difficile colitis, hepatotoxicity
• If no clinical improvement within 48 hours → consider abscess, inadequate debridement, resistant organism, alternative diagnosis
9. Surgical Treatment
Indications for Surgery
Surgical Indications |
• Synovial fluid WBC > 50,000/mL with clinical picture of septic arthritis → irrigate and debride |
• Septic arthritis of the hip → always surgical (emergency) |
• Failure to improve clinically within 48 hours of appropriate antibiotics |
• Concomitant subperiosteal/bone abscess requiring drainage |
• Gross purulence or positive Gram stain on aspiration |
• Neonate with septic arthritis (high risk of AVN, physeal damage) |
Drainage Options by Joint
Joint | Preferred Approach | Alternative | Notes |
Hip | Open arthrotomy (anterior approach) ★ | Arthroscopic washout | TRUE EMERGENCY; posterior approach risks posterior capsular instability |
Knee | Arthroscopic washout ★ or open arthrotomy | Serial needle aspiration (controversial) | Arthroscopic allows thorough debridement + lavage |
Shoulder | Open or arthroscopic washout | Serial aspiration (limited data) | Ensure decompression of subscapularis recess |
Ankle | Open or arthroscopic | Serial aspiration | Beware concurrent calcaneal/distal tibial OM |
Elbow | Open washout | Arthroscopic (technical challenge) | Avoid posterior capsular injury |
SI Joint / Pelvis | CT-guided percutaneous drainage or open | Image-guided aspiration | Often associated with psoas abscess |
Anterior Hip Drainage — Surgical Technique Key Points
• Patient position: supine with bump under affected hemipelvis (30–45° elevation)
• Prep entire leg and buttocks into field — allows full intraoperative hip motion
• Incision: transverse, ~1 fingerbreadth below ASIS (cosmetically superior to posterior)
• Interval: between sartorius (femoral n.) and tensor fascia lata (superior gluteal n.)
• Anterior capsulotomy: T- or H-shaped incision — avoid posterior capsule
• Thorough debridement: remove all purulent material + fibrinous debris
• Irrigation: copious normal saline lavage
• Closure: leave closed suction drain within joint, exiting percutaneously lateral to incision
• Arthrogram at closure to document capsular integrity
Arthroscopic vs. Open Drainage — Evidence Summary
• Experimental staphylococcal SA model (rabbits): arthrotomy + lavage → no collagen degradation; aspiration alone → collagen destruction persisted
• Knee: arthroscopic washout generally preferred — allows thorough visualisation and debridement of all compartments
• No high-level RCT data specifically comparing arthroscopic vs. open for paediatric septic arthritis
• 'The stakes are high, risking lifelong arthritis' — err on the side of surgical joint debridement (Lovell & Winter)
10. Complications & Sequelae
Short-term Complications | Long-term Sequelae |
SIRS / Sepsis | Avascular necrosis (AVN) of femoral head |
DVT (especially MRSA) | Physeal arrest → leg length discrepancy or angular deformity |
Pyomyositis (adjacent) | Joint stiffness / reduced ROM |
Adjacent osteomyelitis | Secondary osteoarthritis |
Pathological dislocation (hip) | Coxa magna / coxa breva |
Pulmonary embolism (rare) | Hip dysplasia (if treated in infancy) |
Neonatal Septic Arthritis — Special Considerations
• Highest risk of permanent sequelae due to vulnerability of cartilaginous epiphysis
• Ogden-Lister blood supply: metaphyseal vessels cross physis → simultaneous epiphyseal + joint infection
• Late presentations: hip dysplasia, subluxation, proximal femoral growth disturbance (as shown in Fig. 12-6, Lovell & Winter)
• Early treatment (within 24–48 hr): better functional outcomes; delay → high AVN rate
• Follow-up radiographs essential at 2 and 4+ years post-treatment (acetabular index, femoral head coverage)
11. Follow-up Protocol
Timepoint | Clinical Assessment | Laboratory | Imaging |
48 hr post-surgery | Fever resolution, ROM, wound check | CRP (should be declining) | None unless clinical concern |
Day 5–7 | Oral step-down decision | CRP, CBC, ESR | None routine |
Week 2 | Outpatient review | CRP, ESR, LFTs, renal function | Plain XR if no prior bony change |
Week 3–4 | Antibiotic discontinuation decision | CRP normalised? ESR trending | XR (hip: femoral head status) |
3 months | ROM, gait analysis | ESR (may still be elevated) | XR (growth, early AVN signs) |
6 months | Growth, limb length | None unless symptomatic | XR pelvis (AP) |
1–2 years | Long-term sequelae screen | None unless symptomatic | XR (AVN, physeal arrest, dysplasia) |
12. High-Yield Summary — Board & Fellowship Pearls
Must-Know Points |
• Cartilage destruction begins within 8 hours — time to surgery is critical |
• S. aureus remains #1 pathogen at all ages; MRSA increasing |
• K. kingae: #1 pathogen <24 months in immunized children; needs enriched blood culture media; PCR most reliable |
• Kocher criteria: 4 predictors → ~93–99% probability of SA (validate prospectively: 93% at BCH, 59% at external sites) |
• Hip septic arthritis = surgical emergency; anterior approach preferred over posterior |
• Synovial WBC >50,000/mL → strong indicator of SA, but 55% of proven SA have WBC ≤50,000/mL |
• PMN >75% in synovial fluid is highly suggestive even when total WBC is lower |
• CRP: most useful monitoring marker (t½ = 17 hr; rises within 4–6 hr; rapidly normalises with successful Rx) |
• ESR continues to rise for 3–5 days after initiating successful therapy — do NOT use for early response monitoring |
• Culture-negative SA: treat empirically as staphylococcal disease — outcomes equivalent to culture-positive |
• IV-to-oral step-down: safe when afebrile + clinically improved + CRP declining; ESR normalisation NOT required |
• MRSA: vancomycin or clindamycin (confirm susceptibility); MSSA now acquiring MRSA virulence factors |
• Neonatal SA: highest risk of AVN, physeal arrest, hip dysplasia — long-term follow-up mandatory |
• Adjacent osteomyelitis: present in up to 40% of SA cases (Perlman) — always examine neighbouring joints |
• Enzymes released into joint continue to destroy cartilage even after bacteria are eradicated |
References
1. Flynn JM, Weinstein SL, Crawford H (eds). Lovell and Winter's Pediatric Orthopaedics, 8th ed. Philadelphia: LWW/Wolters Kluwer, 2020. Chapter 12: Musculoskeletal Infections.
2. Tachdjian MO, Herring JA (eds). Tachdjian's Pediatric Orthopaedics, 5th ed. Philadelphia: Elsevier Saunders, 2014.
3. Kocher MS, Zurakowski D, Kasser JR. Differentiating between septic arthritis and transient synovitis of the hip in children. J Bone Joint Surg Am. 1999;81(12):1662–1670.
4. Kocher MS, Mandiga R, Zurakowski D, et al. Validation of a clinical prediction rule for the differentiation between septic arthritis and transient synovitis of the hip in children. J Bone Joint Surg Am. 2004;86(8):1629–1635.
5. Luhmann SJ, Jones A, Schootman M, et al. Differentiation between septic arthritis and transient synovitis of the hip in children with clinical prediction algorithms. J Bone Joint Surg Am. 2004;86(5):956–962.
6. Peltola H, Pääkkönen M, Kallio P, et al. Short- versus long-term antimicrobial treatment for acute hematogenous osteomyelitis of childhood. Pediatr Infect Dis J. 2010;29(12):1123–1128.
7. Yagupsky P, Dagan R. Kingella kingae: an emerging cause of invasive infections in young children. Clin Infect Dis. 1997;24(5):860–866.
8. Mignemi ME, Menge TJ, Cole HA, et al. Epidemiology, diagnosis, and treatment of pericapsular pyomyositis of the hip in children. J Pediatr Orthop. 2014;34(3):316–325.
9. Martínez-Aguilar G, Hammerman WA, Mason EO Jr, Kaplan SL. Clindamycin treatment of invasive infections caused by community-acquired, methicillin-resistant and methicillin-susceptible Staphylococcus aureus in children. Pediatr Infect Dis J. 2003;22(7):593–598.
10. Fink CW, Nelson JD. Septic arthritis and osteomyelitis in children. Clin Rheum Dis. 1986;12(2):423–435.