Bone and Joint Infections

INTRODUCTION The European Society for Paediatric Infectious Diseases (ESPID) Bone and Joint Infection Guidelines (ESPID Guidelines) are intended for use by health providers who take care of children with bone and joint infection (BJI). Although BJI can include a diverse range of presentations, these guidelines will focus on “acute, hematogenous BJI in children,” with an emphasis on bacterial infections. ESPID Guidelines are consensus-based practice recommendations developed in a systematic manner that aim to be clear, valid and reliable, and presented with clinical applicability. Because evidence from large randomized controlled trials is rare or lacking, practice statements and recommendations provided here frequently reflect our expert consensus process based on best current practice. Although these guidelines include evidence-based and opinion-based recommendations for the diagnosis and management of children with BJI, these guidelines may not provide the best clinical solution and are not intended to serve as a substitute for the clinical judgment of physicians in individual cases or to establish a protocol valid for all children with these infections. Consequently, they do not represent the “only” appropriate approach for children with this kind of infection. We kindly refer to the full version available online (Supplemental Digital Content, https://links.lww.com/INF/C729) for more information on sources used, literature search strategies, guideline development methodology and the ESPID Review Team. The authors of these ESPID Guidelines have made considerable efforts to ensure that the information upon which they are based is accurate and up-to-date. Users of these guidelines are strongly recommended to confirm that the information contained within them, especially drug doses, is correct by way of independent sources. ESPID and the authors of these guidelines accept no responsibility for any inaccuracies, information perceived as misleading or the outcome of any treatment regimen detailed in the guidelines. 2. SUMMARY OF BJI RECOMMENDATIONS There is a paucity of clinical trial or prospective cohort study data to inform the diagnosis and management of BJI in children. Most data are derived from retrospective, observational studies of variable quality. Therefore, ESPID decided to apply a simple grading of the practice statements in this guideline (see notes below). BJI more frequently affects children younger than 5 years of age, and the infection more often involves joints and bones of the lower extremities (IIA). Staphylococcus aureus is the most prevalent microorganism involved in BJI in children at all ages. In addition, Kingella kingae is a common causative pathogen in children <5 years old in some regions (IIA). C-reactive protein (CRP) and erythrocyte sedimentation rate for the diagnosis of BJI have a high sensitivity, which is slightly increased by combining the 2 tests, whereas the specificity is low (IIB). Ultrasound (US) has a high sensitivity for the diagnosis of septic arthritis (SA), whereas magnetic resonance imaging (MRI) is the most reliable imaging study for the diagnosis of BJI overall (IIA). The isolation of a microorganism from the bone, joint or blood with a clinical or radiologic syndrome compatible with BJI is the gold standard for diagnosis in children (IIA). Empirical antibiotic therapy should be started as soon as possible after collecting appropriate samples for microbiologic analysis upon suspecting BJI in children (IIA). Empirical therapy should include an antibiotic with appropriate coverage against methicillin-sensitive S. aureus (MSSA) and against methicillin-resistant S. aureus (MRSA) in geographical areas with more than 10%–15% prevalence of this bacterium (IIA). Empirical therapy in young children needs to include appropriate coverage for K. kingae in relevant areas (IIA). First-generation cephalosporins, anti-staphylococcal penicillins (ASPs) and clindamycin are the antibiotics most studied in BJI in children (IIA). If MRSA infection is suspected and the patient is not critically ill, empirical therapy should include clindamycin if the rate of clindamycin-resistant S. aureus is less than 10%–15%. A glycopeptide or other appropriate antibiotic for MRSA, such as linezolid, should be included if local clindamycin-resistant MRSA rates are high (IIIB). SA in children should be treated with joint drainage by arthrocentesis, arthrotomy or arthroscopy, depending on the preference and experience of the treating clinicians and surgeons. Arthrocentesis may be appropriate as the only invasive procedure in most uncomplicated cases of SA in children (IIB). Short intravenous (IV) therapy followed by oral therapy is appropriate in the majority of children with uncomplicated BJI based on absence of complications and favorable outcome (IA). Follow-up oral antibiotic therapy should be guided by the antibiotic susceptibilities of the bacteria if isolated; if susceptible, the antibiotics of choice are first-generation cephalosporins and clindamycin (IIA). The minimum total duration of antibiotic therapy should be 2–3 weeks for SA and 3–4 weeks for osteomyelitis (OM) (IA). Complicated or high-risk BJI such as those produced by Salmonella, MRSA or Panton–Valentine leukocidin (PVL)-positive strains, developing in young infants, or with slow clinical improvement, may need to receive longer duration of both IV and oral therapy (IIB). Risk factors associated with sequelae include young infants and newborns, infections caused by MRSA or PVL-positive strains, longer duration of symptoms before initiation of therapy and hip involvement. Thus, children with BJI who have any of these risk factors should be followed more closely and for a longer time to rule out or treat sequelae (IIB). A multidisciplinary team should follow children with BJI until osteoarticular function is restored and sequelae are resolved. If bone growth is the only concern, an orthopedic specialist will suffice. Infants with BJI in hip or with any physis involvement should be followed for extended periods of time (IIB). Notes – Quality of evidence I = Good evidence: Randomized placebo controlled trials; other studies appropriately randomized; good meta-analysis and systematic reviews of randomized controlled trials. II = Moderate evidence: Well designed but not randomized studies, cohort and case control studies. III = Poor evidence: Expert opinion, case series. – Strength of recommendation—team consensus based on calculation of votes for A, B or C by the team members: A = strong recommendation; B = moderate recommendation and C = weak recommendation. 3. EPIDEMIOLOGY Musculoskeletal infections involve bones, muscles and joints and are a significant cause of morbidity, and mortality in certain circumstances or settings, in children worldwide.1,2 Acute hematogenous BJI in children may clinically manifest as OM, SA, both combined (OM-SA) or pyomyositis. Pediatric spondylodiscitis is uncommon and accounts for 1%–2% of all children with OM. Pyomyositis may complicate BJI and can also be a primary infection without the coexistence of BJI. Acute OM is an inflammatory process in the bone with bone destruction usually resulting from bacterial infection.3 In high-income settings, the time from onset of symptoms to presentation for medical care is usually <5 days, and rarely more than a week.4,5 Half of the children with acute hematogenous OM are under 5 years of age.1 SA is an acute infection of the joint that occurs most commonly in young children, mainly monoarticular.4,6 (See Section 5 “Clinical Features.”) Spondylodiscitis is characterized by infection involving the intervertebral disc and adjacent vertebrae. Early in the disease, differentiation between discitis and vertebral OM may be difficult. The pathogens implicated in discitis are similar to those in other BJI.3 It occurs mainly in children <5 years of age.2,7 Vertebral OM is more common in older children and usually involves the anterior vertebral body.7 In these instances, infectious agents such as Mycobacterium tuberculosis and Salmonella should be considered. Pyomyositis is frequently seen with pelvic involvement and may be related to MRSA or PVL production.8–11 3.1. European Guidelines Europe is a group of countries with great differences in population, culture, wealth and health services. All variations of disease are impacted by differing epidemiology of pathogens and bacterial resistance, differences in presentation of reported cohorts between regions, medical approaches of infectious diseases, possibilities of medical care, etc. To deal with variations in resource availability, this document aims to provide choices of diagnostic tools and options for treatment. Perhaps, see Table 3 in the full, online version (Supplemental Digital Content, https://links.lww.com/INF/C729) for BJI incidence in several European countries (between 1.4 and 22 per 100,000 people). Differences in incidence may also be related to dissimilar capacity to reach etiologic diagnoses and surveillance methods. 3.2. Predispositions/Risk Factors Most BJI do not have a predisposed condition and occur in primarily healthy children. In specific situations, the following associations have been described. Upper respiratory infection (K. kingae)12,13 Preceding trauma,14 although some recent papers question this since trauma is very common in children15 Wounds,3 erosions and varicella infection (group A Streptococcus)3 Sickle cell disease (Salmonella spp.)3,16 Immunodeficiency—for example, chronic granulomatous disease (Serratia and Aspergillus)17 Penetrating wounds—for example, through the sole of a shoe or sandal (anaerobes and Pseudomonas)2 Living conditions, occupation—for example, animal handling and laboratory work in cases of infection caused by Brucella and Coxiella spp.18,19 Contact with pulmonary tuberculosis or living in endemic areas (tuberculosis BJI) Newborns: prematurity, skin infections, bacteremia or candidemia and previous central venous catheter20,21 4. ETIOLOGY AND PATHOGENESIS Most BJI in children are of a hematogenous origin, and it is the focus of these guidelines. Much less frequently than in adults, BJI in children can be secondary to an adjacent infection, prosthetic material or traumatism. For practical reasons, “acute” and “subacute” are usually considered those BJI with a history of <2 weeks and 2 weeks to 3 months, respectively. 4.1. Causative Agents and Bacterial Resistance The prevalence of different pathogens encountered in various European countries is the main factor influencing the antibiotic regimen in BJI (Table 14, Supplemental Digital Content, https://links.lww.com/INF/C729). Some important points are a higher incidence of community-acquired MRSA (CA-MRSA) in some countries such as Romania or Greece, or important differences in K. kingae incidence within some countries (ie, very low in Scandinavia and quite high in Spain, or A recent European study of invasive S. aureus disease has a prevalence of of Table the most common pathogens by in acute Most by in Acute OM and SA are most commonly caused by S. followed by K. kingae or group A depending on and other risk or geographical In some studies, K. kingae is the the most common after S. aureus in children <5 years of has been involved less frequently in these infections are Salmonella, B and are important pathogens in In certain a variable but considerable of cases are caused by The of BJI is of or and of or Table 2 a of the most and symptoms of children with of BJI by and There is considerable in the symptoms of OM, SA and OM frequently has a more SA more frequently with and range of in such as or and of the may also be very to symptoms are as or to use range of Acute or onset of SA and OM is in of In and young infants only symptoms A systematic literature of studies of OM and symptoms range of In children with BJI, the infection can any bone, or Most commonly the bones and joints of the lower are (Table OM is seen in of infants and young in OM to be more and are more common in it involves muscles the hip can of BJI in Table for a of recommendations for the diagnosis of for In case of suspected BJI, the following are blood and erythrocyte sedimentation this evidence of the clinical of can be very both for and the bone is especially important the may be is not considered with other with appropriate should be before of blood for and bone in recent years has in the of K. kingae as of the most common of BJI in children <5 years of in regions or In recent and have also the of bacteria not by may be very important use of antibiotics diagnostic to after antibiotic or for a pathogen in which diagnostic kingae is mainly the specific may the sensitivity of to etiologic diagnosis is recommended S. aureus is common that an empirical treatment usually especially for children years of Although most cases of BJI can be treated with empirical it is important to establish a microbiologic diagnosis to therapy and to rule out of the has a aim in SA (see the need for a bone for a suspected uncomplicated OM is more this procedure not to the outcome of these Table for a of microbiologic approach to BJI. imaging is considered an important in all for of if disease not and to rule out other Table for a of diagnostic Acute at imaging of or after onset of frequently seen can be with which usually for at 2–3 weeks especially intervertebral of the vertebral Vertebral of a vertebral anterior bone may be in suspected spondylodiscitis and vertebral or pelvic OM. is the most imaging for OM, it can within of disease it of the bone and the of or associated or venous and can the orthopedic to the most appropriate for diagnostic may not be in certain other clinical and diagnostic tools are strongly of the It may be in clinical conditions, are the or a is may be as Although not it may be if is Thus, in a recent of children with acute OM a Spondylodiscitis and vertebral may be a if these infections are suspected for bone and involvement and to rule out and sensitivity and especially for the hip and may be as need of or in young children and is a with some and certain of is not it is less with in and children to high It may be in is not for guided such as or and may not need of the time or is most for SA it has a high sensitivity for the diagnosis of joint although with a lower It should be in all suspected SA by may be for OM, mainly in the diagnosis of and and it may provide for diagnostic or may provide of a high in the Bone or Bone is to involvement and to document the of OM local symptoms are It has a high sensitivity but less and both are lower in It may also in and with pathogens may the sensitivity of bone the is In some bone is and more than involves a significant of range to see also Section in Supplemental Digital Content, and the Society specificity may with or individual cases may be with an Table Supplemental Digital Content, for a of imaging studies in BJI in children. infections and may have similar clinical to BJI should be especially the infection not appropriately and no infectious is of infection, disease or are the most common or important that may BJI. Table Supplemental Digital Content, for the most common diagnosis of BJI. Table 5 for a of recommendations for the management of for of or and BJI (See for The treatment in most cases of OM, SA and can be from the regimen in Early diagnosis and treatment are to factors in the management approach are prevalence of and of the management drainage of of for microbiologic studies and initiation of antibiotic The choice of therapy is based on the most causative pathogens to patient age, disease, and other clinical and prevalence of Most children are at the of the infection as IV therapy is may be especially important in regions with a high rate of MRSA or PVL-positive S. clinical and in high-risk such as infants and There is no evidence that BJI can be treated with oral therapy the of the disease, although children with infections without risk factors for a outcome may have a favorable outcome on with the current this approach by some IV antibiotics are for specific is the of a central for antibiotic treatment at IV therapy may be associated with complications oral therapy not to be with a higher risk of treatment with IV therapy in children with Empirical IV empirical therapy should include coverage of S. prevalence is 10%–15% or this pathogen should be included in the choice of are to the best empirical therapy (Supplemental Digital Content, https://links.lww.com/INF/C729) a of pathogens with geographical The of may also lower the to therapy or other Table in the full, online version (Supplemental Digital Content, https://links.lww.com/INF/C729) for empirical therapy in different European empirical therapy are as such as first-generation cephalosporins and or other are the of choice for good experience and is a especially in with high rate of may be an although no data are available and a higher reported rate of with against Kingella should be considered in children <5 years of age, especially in areas with high Table empirical therapy for BJI to Empirical by of MRSA or PVL-positive S. aureus can be if is a possible Although some authors in the case of have good experience with clindamycin in this and venous as as resistance, may be out before treating children with BJI with Some may treatment of BJI with clindamycin if MRSA is to may be combined with a to until bacterial sensitivity is It is important to PVL-positive S. aureus disease if infection to to empirical is or associated with a In case of infection or are a concern, is recommended by the Infectious Society of at high good data for in children in clinical outcome should be evidence of the of in BJI is and other antibiotic may be or especially if no or minimum to may be to all but with options may be or experience in Table the empirical therapy to rate or Empirical and of MRSA 3 of cases or a antibiotic such as or may be Although data are this is considered for in and in children and with PVL S. aureus in In case of PVL-positive infections, treatment with first-generation cephalosporins or clindamycin be in most situations, the clinicians do not have the PVL to the therapy of BJI. There are some and in studies the use of intravenous on S. aureus BJI infections, but is not evidence to therapy should be a microorganism has been and sensitivity Table most antibiotic therapy to specific bacterial and to Resistance In case of to the options are as and The best to the of Kingella infection are and may be to and may be for although recent studies have a in to the therapy following IV treatment has been as to IV therapy and may be associated with to IV Early oral has been if the is clinical is evidence and variable which may include the or for of with and in of from of such as or of such as Salmonella, MRSA or blood if In infections, the recommendation is to with an oral antibiotic similar to the in IV treatment. In high MRSA clindamycin in younger for clindamycin may be or In low MRSA is a good especially in children years may be an but evidence is and the is In infections, follow the recommendations in Table to are no good data for younger infants and need IV Most treat newborns, in and young infants, for example, with IV therapy and for a longer total duration is some experience in to after a minimum duration of IV therapy the of The of total IV should be on of 2–3 weeks for SA and 3–4 weeks for OM. Although the evidence is lower for weeks of total therapy a of IV may be appropriate for this In the following situations, longer therapy may be practice some may to or pathogens MRSA, and and young infants (ie, or infections of or or in children most symptoms should have and the should be <2 with disease, symptoms or need trial has that therapy for and with in large the acute of are is of Although some a placebo controlled to have a in children with SA, of is not recommended until prospective studies are may the diagnosis of in OM that to of with an OM can be with treatment of especially antibiotics are the of the onset of is usually not if is for in the case of and in some cases is on the and for In the the following is to antibiotic for example, of or and and and of the such as in to a growth may have good outcome with only or other suspected complications of MRSA or S. aureus may the need for OM or of prosthetic material in Joint drainage and is recommended after the diagnosis of SA is A in may be associated with and antibiotic therapy should be within of the onset of SA to a more favorable to some may be more important in and infants of with SA of the hip or drainage by arthrotomy has been but or arthroscopy, depending on the local may be in a of cases of these are invasive with Some orthopedic arthrotomy more can be studies, prospective and the retrospective, have that may be an appropriate approach for SA therapy in children, and hip are In some of SA such as those in the and and hip without risk are by arthrocentesis, with and if more than 2–3 have to be may be considered in some SA involving the hip or in young children longer duration of symptoms at presentation and with more pathogens or the rate of developing complications and sequelae may be Some studies have an between SA of the hip and higher development of some authors arthrotomy this joint is has been associated with of and may provide of the joint for after is no need for for control or upon risk of although some orthopedic especially after hip SA to a of the There is evidence to a in If considered to the of infection or in should be for as as is a very important in the management of BJI, and especially in SA and after Although to the involved should be is for the of complications such as on the and of the OM, some of may the development of a is considered in the management for control for the and longer (ie, in case of spondylodiscitis may be BJI management is often a multidisciplinary approach with and therapy should be on a with Follow-up and Early diagnosis and appropriate treatment are associated with outcome and of chronic and development of and sequelae are as chronic and chronic in the absence of an infectious (Table and by and with experience especially infants, and physis is recommended at 2 3 and after longer in children with involvement of the the and or if the physis is especially infants and younger children. is an important before from should clinical only or as The of MRSA or bacteria may be related with higher rate of complications although not all studies have S. aureus or may also be associated with higher in Some authors that MRSA may be related to PVL other PVL is more commonly in MRSA than in It is important to out for in S. aureus OM and especially In case of it is recommended to the best treatment options with a may be started and until the is resolved. For with antibiotics are for longer periods of although is no evidence of which be the most appropriate of therapy for this refer to the full, online version (Supplemental Digital Content, https://links.lww.com/INF/C729) of this guideline for the of pathogens in BJI with geographical prevalence (Table of antibiotic recommendations in BJI and in this guideline Review team information and