Current Concepts in Pediatric Septic Arthritis

Abstract

Septic arthritis continues to present challenges regarding the clinical diagnosis, workup, and definitive management. Urgent management is essential, so treating surgeons must efficiently work through differential diagnoses, identify concomitant infections, and do a timely irrigation and débridement. The incidence of methicillin-resistant Staphylococcus aureus is increasing, typically resulting in a more rapid progression of symptoms with more severe clinical presentation. The diagnostic utility of MRI has resulted in improved detection of concomitant septic arthritis and osteomyelitis, although MRI must not substantially delay definitive management. Early diagnosis followed by urgent irrigation and débridement and antibiotic therapy are essential for satisfactory long-term outcomes. Antibiotics should not be administered until blood cultures and arthrocentesis fluid are obtained, except in rare cases of a septic or toxic patient. Once cultures are obtained, empiric antibiotic therapy should commence and provide coverage for the most likely pathogens, given the patient's age. Laboratory markers, especially C-reactive protein, should be followed until normalization and correlate with resolution of clinical symptoms. Definitive antibiotic selection should be shared with a pediatric infectious disease specialist, who can help guide the duration of treatment.

Septic arthritis (SA) is a synovial joint infection that most commonly occurs in young children. Infection occurs from hematogenous spread, direct inoculation, or from nearby spread of osteomyelitis. The workup and diagnosis of SA can be challenging. It usually presents with acute onset in which the affected child is irritable, febrile, and will not bear weight on the involved extremity. However, it can also present with subtle signs and symptoms such as failure to feed in a neonate. The screening laboratory work may be normal early in the course of disease, and synovial fluid cultures may be culture negative even in the face of a severe infection. Treatment, including duration of antibiotics and surgical approach, is varied and with limited supporting evidence. Adding further complexity, SA is an orthopaedic urgency and must be worked up appropriately. Delay in diagnosis and treatment may result in irreversible joint damage.

Epidemiology

Incidence

The incidence of SA in children is 1 to 5 per 100,000 in developed countries.^1,2 SA is typically monoarticular, with hip and knee joints being the most commonly involved, but can occur in any joint.

Predispositions and Risk Factors

SA can occur in children several weeks after an upper respiratory infection such as strep throat after otitis media but most often occurs without an identifiable underlying cause. Nonetheless, several risk factors are described. In premature infants, a history of skin infections, bacteremia or candidemia, and invasive procedures such as umbilical catheterization or venous catheterization are reported as risk factors.³ In all age groups, immunodeficiency, sickle cell disease (associated with Salmonella but most common agent still Staphylococcus aureus), penetrating wounds (associated with anaerobic and pseudomonas), preceding trauma, and previous varicella and respiratory system infection (associated with Kingella) are risk factors for bone and joint infections.^4,5

Etiology

Several joints, including the hip, shoulder, elbow, and ankle, have an intraarticular metaphysis (not the knee). This is important, particularly in infants younger than 18 to 24 months because they have a transphyseal artery between the metaphysis and epiphysis, thereby allowing metaphyseal osteomyelitis to spread to the joint, leading to secondary SA. In addition, in these joints, a primary SA could also potentially spread to the metaphysis, leading to secondary osteomyelitis (Figure 1).

Figure 1: Figure demonstrating the hip joint of a neonate, with shared metaphyseal and epiphyseal blood supply and an intraarticular metaphysis, allowing metaphyseal osteomyelitis to spread into the epiphysis and into the hip joint causing secondary septic arthritis. By contrast, a child's knee is shown in which infection is constrained to the metaphysis because of extra-articular metaphysis of the knee and the development of the secondary ossification center with separate metaphyseal and epiphyseal blood supplies.

Once bacteria are in the joint, permanent articular damage can start within 8 hours, according to laboratory evidence.⁶ Bacteria and their cytotoxins activate an acute phase response, leading inflammatory cells to release proteolytic enzymes, ultimately resulting in chondral destruction.⁶ The buildup of leukocytes and pus increases the joint pressure and impairs blood supply, leading to further chondral damage.

Overall, S aureus is the primary causative organism. Sensitivities should be done to distinguish between methicillin-susceptible S aureus (MSSA) and methicillin-resistant S aureus (MRSA). MRSA has been increasingly reported and has been implicated in 30% to 40% of bone and joint infections.⁷

However, causative pathogens can vary depending on the child's age, comorbid diseases or immunodeficiency, socioeconomic factors, and vaccination status^2,8 (Table 1). S. aureus is the most common organism in infants younger than 2 months, often presenting in the neonatal intensive care unit setting, in patients with multiple lines. Streptococcus agalactiae (Group B Strep) and Gram-negative organisms such as Enterobacteriaceae are other potential etiologies, typically presenting in infants who return to the hospital after an uneventful birth, yet acquired these pathogens during labor and the perinatal period.^2,9 In children aged 2 months to 5 years, S aureus, Kingella kingae, and Streptococcus pneumoniae are the predominant causes. Although Haemophilus influenzae type B was the most common cause historically, immunization has fortunately all but eradicated it as an etiology of SA. K kingae is emerging as a common cause of bone and joint infections in children, with a recent study suggesting that Kingella is actually the most common agent in children younger than 5 years old.¹⁰ In adolescents, Neisseria Gonorrhoeae must be considered.⁹

Table 1 - Most Common Pathogens by Age in Septic Arthritis

Age	Most Common Pathogens
<3 mo	Staphylococcus aureus
	Escherichia coli and other Gram-negative bacteria
	Group B Streptococcus
	Candida albicans
	Neisseria gonorrhoeae (newborns)
3 mo to 5 yr	S aureus
	Kingella kingae
	Group A Streptococcus
	Streptococcus pneumoniae  (unimmunized children)
	Haemophilus influenzae type b  (unimmunized children)
>5 yr	S aureus
	Group A Streptococcus
	N gonorrhoeae (in sexually active  adolescents)

Refs: ^7,9,10.

History

Children with SA often appear ill, and symptoms can progress rapidly over hours. The classic presentation includes joint pain from inflammation that stretches the joint capsule, with limited range of motion of the affected joint, fever, and malaise. In the lower extremity, limping and refusal to bear weight are common. Joint effusion, pseudoparalysis, erythema, and warmth may be seen in SA of the knee, ankle, elbow, and smaller joints. SA of the hip will present with limping or inability to bear weight and the involved leg held in flexion and external rotation without any obvious swelling or skin changes (Figure 2). Fever may or may not be present. If present, fever is often especially high in patients with MRSA infection, although fever might be milder or even absent in K kingae infection—with only 15% of patients presenting with a fever.^7,10 On the other hand, MRSA SA can lead to extensive soft-tissue destruction, increased length of hospital stay, and increased risk for complications.⁷

Figure 2: Clinical photograph demonstrating classic posturing in septic arthritis of the hip. The affected right limb rests in a position of slight flexion, abduction, and external rotation.

The presentation in neonates requires special consideration because often little to no signs and symptoms are present. Often irritability and malaise are common, but refusal to eat or vomiting may be the only sign of infection. In neonates and infants, hypothermia may be seen instead of hyperthermia. Unlike older children, in neonates, multiple joints are often involved.

Diagnostic Workup

Laboratory Tests: Adding Meaning to Clinical Findings

Order a laboratory panel on any patient with suspected SA, including complete blood count with differential, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and blood cultures.⁹

Laboratory data combined with clinical symptoms are an essential component of the proper diagnosis. Kocher et al¹¹ described evidence-based criteria for differentiating between SA and transient synovitis (TS) of the hip. The 4 criteria to assess include fever >101.3°F, refusal to bear weight, leukocyte count >12,000, and ESR >40. The diagnostic sensitivity for SA was 93% when 3 criteria were positive and 99% when all 4 criteria were positive.

Although some authors have suggested that the Kocher criteria also can be applied to joints other than the hip, a recent study suggests that if the identical criteria were applied to the knee, 52% of septic knee cases could be missed.^12,13 Even in the hip joint, however, further research regarding the Kocher criteria has not been universal, with one study noting only a 59% predicted probability of having SA in patients with all 4 criteria being positive.¹⁴ Additionally important, organisms such as Kingella with a less severe presentation can be missed by a strictly parameter-based approach to SA.

Caird et al reported that CRP greater than 2 mg/dL is an independent risk factor for SA. Her team subsequently modified Kocher criteria, by including CRP.¹⁵ They also reported that a fever >38.5°C was the best predictor of SA, followed by an elevated CRP level, an elevated ESR, refusal to bear weight, and an elevated serum white blood cell (WBC) count.¹⁵ In this study, if 5 criteria were positive, patients had a 98% chance of having SA, those meeting 4 criteria had a 93% chance, and those meeting 3 criteria had an 83% chance.¹⁵

Another prospective study concluded that the sensitivity of using elevated ESR and CRP to diagnose acute bone and joint infections was 98%.¹⁶ The authors reported that in a cohort of patients who culture positive infections, 94% have had an ESR greater than 20 mm/hr, and 95% had CRP level greater than 2 mg/dL. It is also important to consider the acute time frame in which CRP responds to inflammation. An elevated CRP can be detectable within 6 to 8 hours after the infection started, and doubling time is 8 hours. Therefore, if no increase was noted in CRP after 6 to 8 hours, the risk of SA is low.¹⁷

Routine blood cultures should always be obtained in cases of suspected SA, although their sensitivity is quite low. It is absolutely essential to obtain blood cultures before initiating antibiotic therapy. Remember that negative cultures do not rule out an infection. In confirmed cases of SA, blood cultures are positive only 20% to 40% of the time.⁹

Although the above components of a laboratory panel are the most commonly assessed, it is important to understand the entirety of laboratory findings. In addition, the complete blood count differential is important to assess because it can rule out leukemia. Approximately 80% of patients with leukemia have anemia, neutropenia, or thrombocytopenia.¹⁶ However, leukopenia rather than leukocytosis may be seen in neonates with SA.¹⁸

Procalcitonin is another marker sensitive for bone and joint infections.¹⁹ A recent study showed that the diagnostic specificity of serum procalcitonin level in SA are higher than those of serum CRP. Procalcitonin cutoff value is 0.5 ng/ml.¹⁹ However, these data are more academic than clinically applicable because this laboratory test is not readily available in many facilities.

In addition to their diagnostic utility initially, laboratory findings can help assess for satisfactory treatment, antibiotic response, and resolution. CRP levels may increase initially after surgical intervention but should decrease by approximately 50% 1 or 2 days after a satisfactory irrigation and débridement.¹⁷ ESR level normalizes 2 to 3 weeks after an inflammatory insult, whereas CRP should normalize within 3 to 8 days.¹⁷

Initial Imaging

Begin with routine radiographs of the affected joint. If you suspect hip SA, order AP and frog pelvis films. Films are often nondiagnostic; however, in SA of the hip joint, a unilateral increased joint space of the affected hip may be appreciated on films. Regardless, plain radiographs are essential to rule out bone changes that can be associated with other differential diagnoses, such as a Brodie abscess, fracture, osteomyelitis, or tumor.

Ultrasonography is a highly sensitive and specific test for detecting joint effusions and potential SA. In addition to the diagnosis of a joint effusion, ultrasonography can identify periosteal elevation, cortical erosion, or a subperiosteal fluid collection.¹² Ultrasonography is used mainly for the hip and can be helpful to differentiate between a TS versus a SA. In hip ultrasonography, an anterior synovial fluid collection greater than 5 mm, or a 2 mm difference when compared with the asymptomatic contralateral hip, favors a pathologic hip effusion (Figure 3). Synovial capsule thickening and hypoechoic effusion might also be visualized and favor the diagnosis of SA.¹²

Figure 3: Ultrasound image of a normal right hip and left hip demonstrating an effusion consistent with septic arthritis

Once a diagnosis is established—or in cases with continued diagnostic ambiguity—you should consider the potential benefits of an MRI study, carefully weighing the diagnostic utility against the expected timeframe for the study and any delays in prompt treatment.

SA frequently does not occur in isolation. In infants, concomitant hip SA and osteomyelitis may be seen in up to 60% of cases and concomitant shoulder SA and osteomyelitis may be seen in up to 80% of cases.^20,21 Because of the potential for associated bone infections with SA, a preoperative MRI can add value and prevent second trips to the operating room (OR) by identifying concomitant infection (Figure 4, A and B, 4C). This has been demonstrated in the literature—a pretreatment MRI decreases the risk of return trips to the OR by identifying concomitant pathology that can be addressed at the time of the index procedure.^20,22 MRI is also superior to sonography for detecting marrow changes, subperiosteal abscess, and osteomyelitis. Nonetheless, the individual factors, acuity of illness, time from initial presentation, and time to study should be considered for each patient. Lengthy delays to definitive management in confirmed cases of SA should not be accepted because of the potential of chondral destruction.

Figure 4: Sagittal (A), axial (B), and coronal MRI images demonstrating septic arthritis of the knee with concomitant osteomyelitis of the tibia and subperiosteal abscess.

Microbiology Testing

Synovial fluid analysis by aspiration is essential when SA is considered. Synovial fluid should be sent for cell count, gram stain, culture, and antibiotic sensitivity. For fluid culture and Gram stain, a sterile container should be used. For WBC count and differential, a plain or lavender top tube should be used.²³ Synovial fluid with a cloudy or purulent synovial fluid appearance, WBC count >50,000 cells/mL with more than 90% polymorphonuclear neutrophils, and a glucose level 50 mg/dL less than serum glucose levels, all support the diagnosis of infection.¹⁶

Despite appropriate cultures, a notable proportion of cases (21% to 55%) remain culture negative.¹⁶ To detect the causes of culture negative SA, more sensitive diagnostic tests along with atypical organisms should be considered. For the purpose of detection of Kingella infection, synovial fluid should be cultured in aerobic blood culture bottles in addition to agar plates. Nucleic acid amplification methods (eg, conventional and real-time polymerase chain reaction [PCR]) increase the detection of bacteria not isolated by culture.²⁴ PCR can also provide results within hours.²⁵Kingella is the one of the common etiologic agents in young patients and is difficult to culture. In culture negative cases, Kingella is the most frequently identified bacteria via PCR.^19,26 PCR can also provide diagnostic clarity for the early differentiation between MSSA and MRSA, and in patients treated with antibiotics before culture.^25,26

Differential Diagnosis

Several diseases may have similar clinical manifestations to SA and should be considered especially in cases where the infection progresses insidiously or the infectious agent cannot isolated (Table 2).

Table 2 - Differential Diagnosis of Septic Arthritis

Differential Diagnosis
Transient synovitis
Juvenile idiopathic arthritis
Lyme disease
Hemophilia
Kawasaki disease
Osteomyelitis
Postviral reactive arthritis
Poststreptococcal reactive arthritis
Legg-Calve-Perthes disease
Slipped capital femoral epiphysis
Developmental hip dysplasia
Sickle cell anemia
Acute leukemia
Ewing sarcoma

Refs: ^9,11,27.

TS is one of the common differential diagnoses. TS is defined as a culture negative inflammation of joint synovium, most commonly affecting the hip joint, typically a few weeks after an upper respiratory system infection. A mild fever may accompany joint symptoms. The average age of diagnosis is 6 years; however, TS is very rare in children 3 years of age and younger.¹⁷ TS most commonly presents as acute, nonspecific hip pain or a subtle limp, although children can also refuse to bear weight. However, patients with TS have a less severe clinical presentation than those with SA. Laboratory markers may be slightly elevated, but ESR usually less than 20 mm/hr and CRP is usually less than 2 mg/dL, although a slight leukocytosis may be noted. Ultrasonography may still note an effusion. TS is usually self-limiting disease, which usually resolves in 48 to 72 hours without antibiotic therapy.^11,18

Lyme disease must also be considered, especially in the Northeast, Pacific Northwest, and upper Midwest United States. Over time, the geographic distribution of cases has been increasing.²⁷ As a part of your history, it is important to also ask families about recent travel or camping/hiking exposures. Classically, patients will report a history of a tick bite and resultant “bullseye” rash, although some patients may not recall either a bite or a rash.²⁷ Lyme disease is caused by Borrelia burgdorferi and transmitted by Ixodes tick. It often presents as monoarticular arthritis of the knee. Patients are less toxic in appearance compared with pyogenic SA. Lyme arthritis (LA) can begin 4 days to 2 years after the skin lesion and can be seen as persistent attacks of joint swelling and pain in one or multiple joints and the surrounding bursa or tendons.²⁸ Synovial fluid WBC count is usually less than SA and is typically 10,000 to 25,000 cells/mm³; however, synovial fluid findings may overlap between LA and SA, with one study reporting no notable difference synovial fluid WBC, absolute neutrophil count, and percent neutrophils in patients with LA compared with those with SA.²⁹

A 2018 study compared the reported laboratory values seen in pediatric patients presenting with TS, SA, or LA of the hip.²⁷ The authors noted a 95% confidence interval for ESR values of 21 to 33 mm/hr in those diagnosed with TS, 37 to 46 mm/hr for LA, and 44 to 64 mm/hr for SA. In conclusion, an ESR >40 mm/hr, CRP greater than 4.0 mg/L, patient-reported history of fever, and age younger than 2 years favor SA.^27,30 In clinically suspected cases, serum samples should be tested for antibodies to B burgdorferi by enzyme-linked immunosorbent assay.

Timing of Workup and Interventions

The release of proteolytic enzymes in the joint leads to articular damage within 8 hours of symptoms onset. Therefore, urgent irrigation and débridement coupled with the administration of empiric intravenous (IV) antibiotics is essential. If your history and clinical examination support the diagnosis of SA, immediately order plain films and the appropriate laboratory tests.

Involved hips with supporting clinical, laboratory, and/or radiographic findings should undergo a diagnostic ultrasonography, potentially coupled with an arthrocentesis based on the finding of an effusion.

In all cases where SA clinically suspected, in joints where an obvious effusion is noted, and in hips with ≥3/4 Kocher criteria and/or a CRP value >2 mg/dL, a joint aspirate should be obtained. If gross pus is seen in the aspirate or the aspirate analysis favors infection, patients should be prepared for urgent joint irrigation and débridement.

Antibiotics should be started immediately after blood cultures and joint aspirate are obtained.^2,9 Joint irrigation and débridement should proceed as soon as possible, once the diagnosis is confirmed. Early surgical intervention is particularly important in neonates and infants younger than 18 months of age with SA of the hip or shoulder joint because devastating sequalae are more commonly seen in this group of patients.^10,31

Finally, before surgery, consideration can be made for obtaining a preoperative emergent MRI. Sometimes, this can be accomplished under the same anesthesia care episode, going first to the scanner and then immediately to the OR. Concomitant OM in the setting of SA can require further surgery, if not appropriately recognized and treated before the index procedure.^25-27,32 One study showed that patients with a clinical suspicion of hip SA and ≥3/4 Kocher criteria positive have a 49% risk of concomitant SA and osteomyelitis.³⁰ Many authors recommend using MRI in patients with hip or shoulder SA.^25-27,32,33

A lack of consensus remains regarding which patients would benefit from a pretreatment MRI. Rosenfeld et al³⁴ developed a prediction criteria for the likelihood of adjacent infections in pediatric SA. They defined 5 risk factors—age ≥4, CRP >13.8 mg/L, duration of symptoms >3 days, platelet count <314,000 cells/mL, and absolute neutrophil count >8,600 cells/mL. The authors found that patients with ≥3 risk factors are at high risk for having associated adjacent infections and should undergo a preoperative MRI.³⁴ However, two additional studies were done to assess the clinical applicability of the Rosenfeld criteria. One study reported that the Rosenfeld algorithm showed reproducible predictive power, whereas the other found a lower sensitivity and specificity and higher false-positive rates.^35,36 Nonetheless, MRI should be considered before the OR, especially in patients with high inflammatory markers and severe clinical features, provided that obtaining the MRI will not result in a notable delay to surgical management (Figure 5).

Figure 5: Figure demonstrating the treatment approach decision-making algorithm for pediatric septic arthritis management. CBC = complete blood count, CRP = C-reactive protein, ESR = erythrocyte sedimentation rate, SA = septic arthritis, PCR = polymerase chain reaction, US = ultrasound, USG = ultrasonography, WBC = white blood cell.

Treatment

Surgical Intervention

Surgical intervention includes draining and irrigating the affected joint. Serial aspirations and lavage, arthrotomy, and arthroscopy have all been described as treatment modalities for patients with SA. Drainage clears the joint from bacteria, degradation materials, and enzymes, and decreases the intraarticular pressure, which can result in epiphyseal ischemia. An experimental study showed that toxins released by S aureus have a rapid and fatal action on chondrocytes, reducing chondrocyte viability with >45% chondrocyte death at 40 hours.³⁷

Classically, an open arthrotomy has been used to treat SA; however, arthroscopy has become more common, especially in the knee joint (Figure 6). Its limitations remain the patient age, the involved joint, and the treating surgeon's skill set. Arthroscopy is associated with shorter lengths of hospital stay, allows earlier range of motion, and may provide improved visualization of the joint space during treatment.³¹ One study found that 5 of 13 patients with knee SA treated with open surgery needed a secondary surgery, whereas 0 of 11 patients treated with arthroscopy required a second surgery.³⁸ In another recent study, patients who had hip SA treated with hip arthroscopy had a similar short-term complication rate and rate of return to the operation room comparable with the open arthrotomy cohort.³⁹

Figure 6: Illustration of hip anterior approach to irrigation and débridement of hip septic arthritis.

Although more historical in nature and not typically considered a definitive treatment, several studies nonetheless show that therapeutic serial arthrocentesis can be used successfully treat SA.^32,40 This requires a daily clinical examination and repeating of the arthrocentesis and lavage when necessary.⁴¹ In 2 studies of patients with hip SA, 5 of 43 and 4 of 28 patients ultimately needed an arthrotomy, but, the remainder of the patients were treated successfully by serial aspiration and lavage alone.^40,41 Journeau et al⁴¹ described several factors related to the need for arthrotomy, including symptoms lasting more than 5 days, CRP >10 mg/dL, >15,000 neutrophils/mm³, and ESR >50 mm/hr.

Antibiotic Management

After joint aspiration, empirical IV antibiotics may be started. S aureus is the most common cause of SA in pediatric patients >1 month of age. Empirical therapy should include therefore include S aureus coverage. Beta-lactams, such as first-generation cephalosporins (MSSA) and cloxacillin or other antistaphylococcal penicillins such as oxacillin, can be started as an empirical treatment.³¹ In regions where community-associated MRSA prevalence is 10% or higher and initial presentation, laboratory markers, and clinical symptoms severity are markedly elevated, anti-MRSA antibiotics should be included in empiric therapy. Clindamycin (most Gram-positive organisms and anaerobes) can be considered a suitable first choice in treatment, especially in settings with a high reported incidence of community-associated MRSA.⁴² In severe and invasive infections, vancomycin can also be considered a first choice in the patients with severe symptoms concerning for MRSA, with the guidance of a pediatric infectious disease specialist and after careful consideration of the potential difficulties of administration (eg, red man syndrome) and management of therapeutic levels.

Kingella should be considered in children younger than 5 years of age, especially in areas with high reported rates. Therefore, it is important to consider empiric antibiotics in this age group with coverage for both Kingella and Staph, including clindamycin and ceftriaxone (inpatient) and trimethoprim sulfamethoxazole (after discharge).

Neonates are at risk for Gram-positive organisms such as S aureus and group B Streptococci and Gram-negative organisms such as Escherichia coli. Penicillin, such as oxacillin (MSSA) or ampicillin-sulbactam (Gram-positive coverage), plus an aminoglycoside, such as gentamicin (Gram-negative coverage), may be used. Alternatively, vancomycin (MRSA and serious S aureus) with ceftriaxone (broad spectrum, Gram-negative coverage, lower efficacy against Gram-positive organisms) may also be used.

Children with SA should be hospitalized for initial empirical IV treatment for a minimum of 2 to 5 days while cultures and sensitivities are pending.³¹ Although final antibiotic decision-making should include in an infectious disease specialist, the general properties of these antibiotics should be considered (Table 3).

Table 3 - Empiric Antibiotic Recommendations for Pediatric Septic Arthritis

Age	Empirical IV Antibiotic
<1 mo	Ampicillin-sulbactam + gentamycin
1–3 mo	Vancomycin + ceftriaxone
Pediatric and  adolescent	Cefazolin or cefuroxime
	Clindamycin (high MRSA rate)
	Vancomycin (high MRSA rate, high  clindamycin resistance, severe  infection)

IV = intravenous, MRSA = methicillin-resistant Staphylococcus aureus

Refs: ^7,9,42.

Targeted Antibiotics

When bacteria are isolated, antibiotics should be adjusted according to culture and sensitivity or PCR results (Table 4).

Table 4 - Antibiotic Selection for Pathogens

Pathogen	Antibiotic Treatment
Staphylococcus  aureus	Penicillin, first generation  cephalosporins (MSSA)
	Clindamycin (MRSA sensitive)
	Vancomycin (MRSA sensitive)
	Daptomycin (vancomycin resistance)
	Linezolid (vancomycin resistance)
Streptococcus  pyogenes	Penicillin, ampicillin, or amoxicillin
Streptococcus  pneumoniae	Ampicillin and amoxicillin
	Second to third generation  cephalosporins
	Vancomycin and linezolid (high  resistance)
Haemophilus  influenzae type b	Amoxicillin-clavulanate
	Ampicillin-sulbactam
	Second generation cephalosporins
Kingella kingae	Penicillin, first generation cephalosporin
	Amoxicillin-clavulanate (beta- lactamase+)
	Ampicillin-sulbactam (beta-lactamase+)
Salmonella	Ceftriaxone or cefotaxime
Pseudomonas	Piperacillin-tozobactam
	Ciprofloxacin
Escherichia coli	Amoxicillin-clavulanate
Neisseria  gonorrhoeae	Ceftriaxone, cefotaxime, third generation  cephalosporine

MRSA = methicillin-resistant Staphylococcus aureus, MSSA = methicillin-susceptible S aureus

Refs: ^7,9,10,43.

Duration of Treatment

The length of total therapy, with IV and then potentially postoperative antibiotics, should be on average of 2 to 3 weeks for SA. The duration of treatment should be more prolonged in the case of MRSA or panton valentine leukocidin (PVL) producing MSSA infection, poor or slow response, and in newborns and young infants. Before discontinuation of treatment, most symptoms should be resolved and CRP should be normal (eg, <2 mg/dl).^2,31

In certain cases, oral antibiotics can be considered an alternative to prolonged IV therapy and may be associated with fewer complications.^9,43 Newborn and infants younger than 3 months may need a longer duration of IV treatment (10 to 14 days) and newborns should receive most of their antibiotic treatment via IV.⁴⁰ Overall, the duration of IV antibiotics should never be less than 2 to 5 days.^2,31,43 If children improve clinically, early transition to oral antibiotic treatment can be considered. The European Society for Pediatric Infectious Diseases has published bone and joint infection guidelines describing several factors of clinical improvement that can be used as criteria for switching oral antibiotics. These factors include an afebrile patient or decreased temperature for 24 to 48 hours, improvement of symptoms, decrease in CRP of about 30% to 50% from maximum value, no signs of complications such as bacterial spread to other areas (endocarditis, pneumonia, etc.) or DVT, and absence of virulent pathogens, especially MRSA or PVL+.^9,43

Complications

Complications most commonly arise from a delay in diagnosis. Delays in diagnosis have led to severe complications including longer hospital stays, more severe infections requiring multiple irrigation and debridements, devastating cartilage destruction resulting in end stage arthritis, osteonecrosis, and even leg length discrepancy from physeal destruction. These devastating complications are more likely to occur in cases of concomitant SA and osteomyelitis. A missed diagnosis in neonatal SA of the hip can be especially devastating, leading to osteonecrosis of the entire femoral head. In the acute phase of infection, a dislocation of the hip joint can also be seen among infants and may require immobilization with either a Pavlik harness or spica casting after drainage. Complications such as physeal arrest can develop slowly. Close follow-up of at least 1 to 2 years is recommended.

Summary

SA in children is rare but can have potentially devastating consequences including osteonecrosis, chondral destruction, and limb length discrepancies. The diagnosis is often most difficult in neonates because of lack of notable signs and symptoms. However, an urgent diagnosis acquired from a thorough history and examination along with careful review of laboratory findings, imaging, and joint aspiration. Appropriate treatment, including urgent irrigation and débridement and carefully checking antibiotic susceptibilities can lead to the best clinical result and long-term outcomes in children of all ages.