Pediatric Septic Arthritis Surgery 

Septic arthritis in infancy and childhood is a true clinical emergency (see image below). Delays in the diagnosis and treatment of septic arthritis can result in disastrous complications, including complete destruction of the articular cartilage and the underlying epiphysis, loss of the adjacent growth plate, and dislocation of the joint. With prompt treatment, the condition may be cured and sequelae avoided. The clinician must be alert to a diagnosis that can be subtle, particularly in the young child or infant. Pain with passive motion is the most consistent finding in septic arthritis, and the diagnosis must be considered in any joint with this presentation.

This article focuses on the most common presentations of septic arthritis as it occurs in infancy and childhood. Important variations from these common presentations are seen in septic arthritis in the neonate and with gonococcal arthritis. Discussions of these 2 entities appear in the sections below, as appropriate.

Prior to the advent of antibiotics, the consequences of septic arthritis were usually severe and often fatal. Surgical drainage was the only treatment available. With the development of effective antibiotics, it became possible to cure septic arthritis, but, except in a very few select situations, surgical drainage must still be performed. The most recent change in the prevention and treatment of septic arthritis has occurred with the institution of routine immunizations against Haemophilus influenzae type b.

From the 1970s until the development of the vaccine, H influenzae type b was responsible for the majority of joint infections in children aged 1-4 years. Since the institution of this vaccination, joint infections with H influenzae have virtually disappeared in the United States, and Staphylococcus aureus has become the most common joint pathogen in all age groups.

The chief concern with septic arthritis in childhood and infancy is the potential for severe complications. The condition must be expeditiously diagnosed and appropriate treatment begun without delay. The diagnosis must be considered whenever painful joint dysfunction is present.

Frequency

Septic arthritis is relatively common in infancy and childhood, to the extent that all primary care physicians caring for children can expect to encounter this disease. The exact incidence is difficult to estimate because all series are retrospective and lack denominators. Septic arthritis is known to be about twice as common as osteomyelitis in childhood, and the incidence of joint infection in all age groups peaks in the early years of the first decade of life.

In all age groups, the most common infecting organism for septic arthritis is Staphylococcus aureus. Other common pathogens include Streptococcus species, Pseudomonas aeruginosa, pneumococci, Neisseria meningitidis (with or without an associated meningitis), Escherichia coli, Klebsiella species, and Enterobacter species. Newborns can acquire Neisseria gonorrhoeae from an infected birth canal. Gonococcal arthritis is more common in sexually active teenagers, and it may be seen in younger children in association with sexual abuse.^[1]

A neonate aged 5 weeks or younger is susceptible to infection as a result of a wide range of organisms that are unlikely pathogens in children with more developed immune systems. S aureus is still the most common pathogen in this age group; group B streptococcus is the next most common pathogen. Gram-negative organisms may be seen in as many as 15% of joint infections affecting neonates in a neonatal intensive care setting.^[2]Candida albicans may also be present in these patients, as well as in patients who have received prolonged antibiotic therapy.

In the past, H influenzae was the dominant pathogen causing septic arthritis in children younger than 3 years of age. A vaccine against this organism was introduced in 1985, and more effective conjugated vaccines against H influenzae are now extensively used in the United States. As a consequence, H influenzae has nearly disappeared as a pathogen of osteoarticular infections in young children.^[3]

Kingella kingae is a fastidious aerobic gram-negative microorganism that was first described in 1960. As the incidence of H influenzae has decreased, the incidence of K kingae as a pathogen of osteoarticular infection in children younger than 3 years of age has dramatically increased. Because it is fastidious, this relatively new organism may be difficult to grow in culture. It is now common enough that its presence must be suspected and investigated in children in this age group.^{[4, 5]}

The incidence of community-associated methicillin-resistant S aureus (MRSA) is increasing in North America. In regions where this trend has been observed, consideration should be given to the inclusion of coverage for MRSA in the empirical coverage for patients with septic arthritis pending culture results.

In approximately one third of all cases of septic arthritis, identification of the responsible organism may be impossible.

Although it is uncommon, a penetrating wound may result in septic arthritis. It also may develop by contiguous spread from an adjacent cellulitis. Most commonly, however, the pathogenesis of septic arthritis is hematogenous. Transient episodic bacteremia is common, even in healthy individuals. It may occur with something as simple as toothbrushing. In almost all cases, the body's defense mechanisms rapidly eliminate the threat; occasionally, bone or joint infection may result.

Hematogenous septic arthritis may develop directly through the synovial blood vessels. Another common route is from an adjacent hematogenous metaphyseal osteomyelitis. Just under the metaphyseal side of the growth plate in a growing child, vascular loops nourish the bone that forms in association with enchondral ossification. Blood flow in these loops is thought to be relatively slow, and this region is somewhat poorly defended by the reticuloendothelial system. These loops form the site of origin for hematogenous osteomyelitis in infancy and childhood.

An abscess forms in the metaphysis in association with localized bone destruction. In the infant (from birth to 18 months), the infection may spread into the epiphysis through blood vessels that cross the cartilaginous physis. From the epiphysis, the infection may then break directly into the adjacent joint, resulting in septic arthritis. This mechanism of spread directly into the epiphysis is unique to infancy.

The blood vessels that cross the physis disappear by age 18 months, and the cartilaginous growth plate becomes a barrier to the spread of infection. In the older child, hematogenous osteomyelitis spreads within the metaphysis until it breaks through the metaphyseal cortex. For most of the long bones, the bone infection breaks either into the subperiosteal space or through the periosteum into the adjacent soft tissues. Joint infection does not result, because the joint capsule is firmly anchored to the epiphysis. In the proximal femur, the proximal humerus, the distal lateral tibia, the distal fibula, and the proximal radius, however, the joint capsules attach to the metaphyses; therefore, in these locations, hematogenous osteomyelitis may decompress directly into the hip, shoulder, ankle, and elbow joints, respectively.

Whatever the mechanism, once bacteria enter the joint, the space effectively becomes a closed abscess. A variety of enzymes capable of degrading articular cartilage are released by leukocytes and by certain bacteria, such as S aureus and some gram-negative organisms. Significant articular damage can occur in as little as 8 hours. Increased pressure within the joint can interrupt the blood supply to the epiphysis, causing bone destruction and loss of the adjacent growth plate. If the infection remains untreated, the ligaments of the capsule will be destroyed and the joint may dislocate. For example, the extreme result in a septic hip would be complete destruction of the femoral head, dislocation of the proximal femur from the acetabulum, and loss of 30% of all future growth of the affected femur.

Neonates are special in that their immune systems still are immature. They are susceptible to a wide range of organisms that are unlikely pathogens in an older individual, and they are less capable of mounting an inflammatory response to infection. Premature infants in the neonatal intensive care unit are particularly at risk. They often are debilitated with other illnesses, and they typically present with multiple ports for bacterial entry. Bone and joint infections in these patients commonly involve multiple sites.

Neisseria gonorrhoeae may cause a distinct syndrome of a disseminated gonococcal infection, with skin lesions, tenosynovitis, and polyarthralgias, rather than frank arthritis. Less commonly, it may cause a suppurative arthritis resembling septic arthritis caused by other bacteria isolated to 1 or 2 joints.

A developing septic arthritis is generally accompanied by the onset of fever, malaise, and prominent localizing signs at the affected joint. In the distal extremities, swelling, erythema, and tenderness are prominent; the findings may be less evident with deeper joints, such as the hip. The most consistent sign is pain with passive motion. The patient will generally hold the joint in the position that maximizes intracapsular volume. For the hip, these positions are flexion, abduction, and external rotation, as seen in the image below. With a septic knee, the joint is most comfortable when moderately flexed. It is not unusual for a young child or infant to appear completely comfortable, so long as the affected joint remains immobile in the position of comfort. Any attempt by the examiner to passively move the joint, however, dramatically reveals the pathology.^[6]

Refusal of the child to move the affected joint is called pseudoparalysis. This sign is often mistaken for a neurologic problem. An isolated true paralysis, however, is far less common than a septic arthritis; when it does occur, it is rarely associated with pain with passive motion. The inability of a child to bear weight on a lower extremity or to spontaneously move any joint must be considered a sign of septic arthritis until proven otherwise.

The differential diagnosis of septic arthritis includes noninfectious inflammatory arthritides, including juvenile rheumatoid arthritis (JRA) and acute rheumatic fever. JRA usually presents a more gradual onset, as well as symptoms and signs that are less dramatic. Acute rheumatic fever is often associated with migratory polyarthralgias.^[7]Legg-Calvé-Perthes disease may present with pain in the hip with weight bearing, but rarely is the pain as disabling as the pain associated with a true infection.

Lyme disease commonly appears with arthritis, and the clinical presentation may be difficult to distinguish from septic arthritis. With Lyme arthritis, there is generally less pain with joint motion and, when lower extremity joints are affected, ambulation is usually possible; however, in regions where Lyme disease is endemic, this condition should be included in the differential diagnosis.^[8]

The clinical presentation of an abscess of the psoas muscle may be indistinguishable from that of septic arthritis of the hip.^[9] Atypical features, such as bladder irritability or a femoral nerve neurapraxia, may point toward this rare disorder, but in their absence joint aspiration should be performed. If the aspiration culture is negative, obtain a computed tomography (CT) scan, as this is generally diagnostic for a psoas abscess.

Transient synovitis (also called toxic synovitis) is the most common condition that mimics septic arthritis. Possibly of viral origin, it is a self-limited condition that most commonly involves the juvenile hip. While transient synovitis usually causes less pain with passive joint motion than septic arthritis, it may be clinically indistinguishable from septic arthritis. Patients with transient synovitis are less likely to have fever or elevated C-reactive protein levels, but differentiation from septic arthritis may still require examining the joint aspirate.

Neonates with septic arthritis typically present with a blunted immune response. They may have little or no fever. They may have minimal or no increase in their serum WBC counts. Joint involvement is less obvious in neonates than in other patients because neonates are less active to begin with. Septic arthritis is also more difficult to diagnose in neonates, but making an early diagnosis is more important in these patients. If 1 infected joint is diagnosed in a neonate, look for other sites of bone or joint infection. Consider obtaining a bone scan to assist in this search.

Immunocompromised patients with septic arthritis are also likely to present with a blunted immune response. Symptoms, physical signs, and laboratory parameters may appear deceptively benign. Such patients warrant an increased index of suspicion.

With gonococcal infections, a disseminated infection may, in addition to fever and chills, cause a constellation of clinical signs quite distinct from septic arthritis, including a variety of skin lesions, such as macules, pustules, and bullae; tenosynovitis, most commonly of the dorsum of the wrists and hands; and polyarthralgias. Less commonly, the presentation may be clinically indistinguishable from septic arthritis caused by other organisms. The knee is the most commonly affected joint, but any joint may be involved. If this diagnosis is suspected (for example, in a sexually active teenager), cultures and antigen examinations of all potential mucosal sites of infection should be combined with cultures of joint aspirate and blood. The physician should keep in mind that gonococcal arthritis may be seen in younger patients in association with sexual abuse.

An aggressive workup is indicated whenever signs and symptoms suggest septic arthritis. This evaluation must include blood cultures and joint aspiration. If the aspirate results are consistent with a septic joint, antibiotics must be begun on an empirical basis immediately after the cultures have been obtained. Unless the joint is easily accessible to repeated aspirations and the signs are minimal and improving with treatment, formal operative drainage is required. The relevant principles that apply to all surgical infections should be followed. A septic joint should be considered an abscess, and medical treatment alone is inappropriate.

The clinical presentations of other conditions can closely mimic that of septic arthritis. Occasionally, the clinical signs are prominent, but the clinician may fail to obtain pus by aspiration. Decision-making in these situations should be analogous to that used in the management of an acute abdomen. The risks of an unnecessary arthrotomy are minimal compared with the certainty of permanent joint damage that accompanies a neglected septic arthritis.