Osteoarticular infections can be limb or life-threatening. Timely diagnosis and appropriate treatment are paramount to minimizing complications and optimizing outcomes. The current review focuses on recent developments in the cause, epidemiology, diagnosis, and treatment of pediatric septic arthritis and osteomyelitis.
In many cases infection may coexist in more than one anatomic location and more than one tissue type . Illness severity has been linked to the primary infection location, and presence of infection in adjacent tissues. Other factors, including patient age and comorbidities, and the species and strain of the responsible microbe, also affect illness severity and response to treatment.
In developed countries, recent reports of osteomyelitis rates are 2 to 13 per 100 000 children [1,2]. Higher rates have been reported in developing countries [3,4▪]. Osteomyelitis is approximately twice as common as septic arthritis. There is a near equal distribution amongst male and female children. The mean age at diagnosis is 6.6 years [4▪].
Some studies have reported a decreasing rate of musculoskeletal infections over time, whereas others have reported a significant increase [4▪]. A recent study reported a 2.8-fold increase in osteomyelitis over the past 20 years, whereas septic arthritis rates have remained constant over this period . MRI is being used more frequently than in the past , and it may be that rates of osteomyelitis may not have changed considerably, but detection methods are better than in the past.
The pathogen most commonly responsible for septic arthritis and osteomyelitis depends on the child's age, comorbidities, socioeconomic, and immune and vaccination status. There are also significant geographic variations in prevalence of different bacterial species and resistances.
Staphylococcus aureus is typically described as the most common cause of musculoskeletal infections in children . However, gram-negative bacteria may account for 60% of musculoskeletal infections in children under the age of 4, and, in a recent Swiss study, Kingella kingae caused 82% of all osteoarticular infections in children under 4 .
Hemophilus influenza B (HiB) was previously a common cause of osteoarticular infection in young children. In the 1990s widespread vaccination programs were initiated, and HiB infections dropped significantly. A retrospective Canadian study found that 5% of osteomyelitis and 41% of septic arthritis was caused by HiB prior to the vaccination program, whereas no cases were reported in the following years . HiB should still be considered in the differential in unvaccinated children and in developing countries.
Kingella kingae is emerging as a common cause of osteoarticular infections. In the Swiss study in which Kingella kingae was isolated in 82% of osteoarticular infections in children under 4, the authors reported a more benign presentation and course, with only 15% febrile during admission, and 39% having a normal C-reactive protein (CRP) . A recent retrospective review found that age of diagnosis in children with Kingella kingae was significantly younger than in children with Staphylococcus aureus septic arthritis, and that the clinical course was more benign, but that otherwise all presenting features and clinical markers were equivalent . A predictive algorithm has been produced for Kingella, and this includes temperature below 38°C, CRP below 55 mg/l, white blood cells (WBC) below 14 000/mm, and bands below 150/mm [10▪]. Kingella kingae is notoriously difficult to culture. In one series, all cases of Kingella kingae osteoarticular infection were gram stain and culture negative, but subsequently identified by real-time PCR assay specific to the Kingella kingae RTX toxin .
Methicillin-resistant Staphylococcus aureus (MRSA) has been increasingly reported in recent years. There is significant geographic variation in MRSA prevalence, with no MRSA found in osteomyelitis in Saudi Arabia  or osteoarticular infections in Finland , whereas in the US MRSA has been implicated in 30–40% of osteoarticular infections .
Panton-Valentine leukocidin (PVL) is a necrotizing toxin that is secreted by some forms of methicillin-susceptible Staphylococcus aureus (MSSA) and MRSA. It is associated with more invasive osteoarticular disease including higher rates of septic shock, longer hospital stays, and need for prolonged antibiotic and more surgical interventions . It is more commonly associated with MRSA than MSSA, and is thought to be responsible for the increased virulence seen in many cases of MRSA [15▪].
Altered immune status can make children susceptible to different pathogens. In a recent study comparing osteoarticular infections in HIV-positive with HIV-negative children [16▪], Streptococcus pneumoniae was the causative organism in 67% of HIV-positive children, whereas it caused only 10% of infections in HIV-negative children. The authors recommended that broader empiric antibiotic coverage be used before culture results are available in children known to be HIV-positive, or in areas with a high prevalence of HIV.
Children with septic arthritis or osteomyelitis typically present with localized pain and swelling, limited range of motion, and difficulty or refusal to weight bear or use an extremity (pseudo paralysis). In a recent systematic review including 12 000 patients with acute and sub-acute osteomyelitis [4▪], 81% presented with pain, 70% showed localized signs and symptoms, 62% presented with fever, 50% had reduced range of motion, and 50% had reduced weight-bearing.
There is no single test that can confirm or rule out septic arthritis or osteomyelitis. A combination of careful history, physical exam, imaging, laboratory tests, and aspiration/biopsy is typically required to make a definitive diagnosis.
Radiographs should be the first imaging study performed, although they are frequently negative in septic arthritis and early osteomyelitis. Ultrasound is useful for assessing effusion in suspected septic arthritis. If increased fluid is identified, arthrocentesis can be performed under image guidance. Ultrasound of potential septic joints in the emergency room has been shown to have high sensitivity and specificity [17,18]. Ultrasound is less helpful in the diagnosis of osteomyelitis, although abscesses and periosteal abnormalities may be visualized. Bone scan can be useful when potential infection is poorly localized or multifocal disease is suspected. However, false-negatives have been reported in up to 50% of cases .
Early MRI is helpful in the diagnostic process to better classify the location and extent of disease and to plan surgical interventions . A recent study has shown that community-acquired MRSA (CA-MRSA) frequently affects the nonossified cartilage of bones, and that on standard MRI sequences this involvement is frequently missed. However, gadolinium enhancement enables identification of these involved areas, and is recommended in suspected cases [20▪]. Another recent study has reported that MRI can be used to help differentiate between Kingella kingae osteoarticular infections and those caused by other pathogens . In this study, epiphyseal cartilage abscesses were found only in the Kingella kingae infection group. Soft tissue and bone reaction were significantly less in infections caused by Kingella. PET/computed tomography (CT) has been described as superior to MRI in monitoring response to treatment for osteomyelitis, as it is better at distinguishing between ongoing infection and reparative activity [22▪]. However, radiation exposure and access to PET/CT may limit its practical use.
LABORATORY STUDIES: BLOOD
Blood tests are used to assess for potential osteoarticular infections. Peripheral blood is sent for cell count, erythrocyte sedimentation rate (ESR), CRP, gram stain, aerobic and anaerobic culture. In children with osteomyelitis, 36% of patients had elevated WBC count, 91% had elevated ESR, and 81% had elevated CRP. In patients with both osteomyelitis and septic arthritis, 100% had elevated ESR and CRP [4▪]. In a combined series of children with osteomyelitis and/or septic arthritis, ESR was above 20 mm/h in 94% and CRP was above 20 mg/l in 95% of patients. ESR and CRP normalized in 24 and 10 days, respectively. Optimal sensitivity was found when combining ESR and CRP (98%). Due to more rapid normalization of CRP, this marker provides a better option for monitoring response to treatment .
LABORATORY STUDIES: JOINT FLUID OR BONE SPECIMEN
Joint aspirate or bone specimen should ideally be obtained prior to starting antibiotics. For joints and bones that are easily accessible, aspiration/arthrocentesis may be done under conscious sedation in the emergency room. Joints such as the ankle and hip that are difficult to aspirate, and areas of osteomyelitis that are difficult to localize by clinical examination alone, may need to be obtained by interventional radiology. Alternatively, an orthopedic surgeon may obtain specimens intraoperatively by open or image guided techniques.
ANALYSIS OF SPECIMENS
Recent studies have reported that a causative pathogen is not identified (negative cultures) in 33%  to 55%  of osteoarticular infections. When unusual pathogens are suspected the physician must communicate these concerns with the laboratory, as some pathogens are difficult to culture and may require specific media, growth conditions or prolonged culture time.
Kingella kingae is known to culture poorly on solid media. Placing joint aspirate in a blood culture tube (purple top) has been shown to increase positive culture yields . Use of an enriched liquid medium instead of plating on solid media may also increase yields. Other studies have suggested that patients suspected of having Kingella kingae should undergo nasopharyngeal culture, as patients with Kingella musculoskeletal infection typically have nasopharyngeal colonization, and throat culture returns positive cultures six times more than joint aspirate [26▪]. Alternatively, Kingella can be identified by PCR [27▪], although this technique is not yet available at many facilities.
PCR can provide results within 3 h [28▪]. In addition to identifying difficult-to-culture pathogens, it may also be helpful in early differentiation between MSSA and MRSA, thus ensuring appropriate early antibiotic treatment, minimizing prolonged and expensive broad-spectrum therapy, and potentially decreasing the iatrogenic contribution to development of resistant bacteria.
DECREASING BACTERIAL LOAD
Mechanical removal of bacteria from a septic joint is typically performed by open or arthroscopic means. A recent small series reported a trend towards improved outcomes and significantly shorter hospitalization times in children who underwent arthroscopic debridement for septic hips, compared with open debridement . Unfortunately, arthroscopy is an advanced skill, and may not be available to all patients on an emergent basis. Needle aspiration and irrigation under general anesthesia for the hip has been described , although a proportion of patients treated in this manner have been shown to eventually need formal irrigation and debridement. Another recent study has reported an even less invasive strategy, in which pediatric septic arthritis of the hip was treated with a single aspiration and appropriate antibiotics . Good clinical results were reported, with formal irrigation and debridement performed only when there was sub-optimal clinical and laboratory response to antibiotic treatment. In this series invasive surgery was avoided in 81% of patients. Despite these reports, definitive treatment with needle aspiration and irrigation has not been widely adopted.
For patients with acute and sub-acute osteomyelitis, irrigation and debridement are not typically required unless there is insufficient clinical response to antibiotic therapy. Chronic osteomyelitis typically does require irrigation and debridement. Clinical response (temperature, pain, use of the extremity) and CRP levels guide the need for ongoing treatment, and need for further surgical interventions.
Patients with osteoarticular infections are typically admitted to hospital and started on empiric antibiotics as soon as a joint or bone specimen has been obtained and sent for analysis. Empiric therapy is selected to cover the most likely pathogens, which are determined primarily by local prevalence of infectious agents and resistance levels, the age of the child, and early lab results such as gram stain if available. Selecting a cephalosporin (e.g. Ancef) that covers MSSA and Kingella is typical. A study out of Finland showed success with clindamycin as initial empiric therapy . However, in many locations there is MSSA resistance to clindamycin, and/or high prevalence of Kingella, which is also typically resistant, and in these scenarios clindamycin would not be the optimal empiric therapy. The decision to cover empirically for MRSA (e.g. vancomycin) is controversial due to concerns over developing resistances, costs and potential complications with therapy. If vancomycin is selected, a cephalosporin or penicillin to which Kingella is sensitive should be added (Kingella is typically resistant to vancomycin). Coverage for gram-negatives must be added for neonates and young children.
LENGTH AND ROUTE OF TREATMENT
The length and route of treatment depend on the location of the infection, as well as the clinical and laboratory response to treatment.
Although osteomyelitis has traditionally been treated with 4–6 weeks of antibiotics, studies supporting shorter treatment times are emerging. A prospective study of septic arthritis and osteomyelitis in children found that 59% of patients could be converted to oral antibiotics by 3 days, and 86% by 5 days . Patients who were converted to oral antibiotics by 5 days received an additional 3 weeks of oral therapy. The mean in-hospital treatment time was 5 days. There were no complications at 1-year follow-up. Recently a randomized controlled trial (RCT) provided 2–4 days of i.v. antibiotics followed by 20 versus 30 days of oral therapy for acute osteomyelitis . No differences were found amongst treatment groups, and the authors recommend that 20 days of oral antibiotics is sufficient provided there is clinical improvement and CRP is normalizing.
A retrospective study of conversion times from i.v. to oral antibiotics for treatment of septic arthritis showed no difference in outcomes when patients were treated with 7.5 versus 18.5 days of i.v. antibiotics with a total course of 4 weeks of antibiotics . A recent RCT treated patients with 2–4 days of i.v. empiric therapy followed by conversion to oral therapy, with a mean total antibiotic treatment duration of 10 versus 30 days . No difference in outcome was identified between groups. This antibiotic regimen was reported as successful, with most patients undergoing reduction of bacterial load within the joint by aspiration only. Only 12% required surgical irrigation/debridement of the joint, based on clinical and laboratory response to treatment.
Recently a randomized controlled trial [36▪] of children with septic arthritis compared treatment with i.v. dexamethasone, standard antibiotic therapy, and surgical debridement with standard antibiotic therapy and surgical debridement alone. Dexamethasone was associated with less fever and local inflammatory signs, lower acute-phase reactants, and shorter treatment time with i.v. therapy, although total antibiotic treatment time did not differ. No long-term differences in outcomes were identified.
A recent study out of Thailand [37▪] found complications in 29% following pediatric osteoarticular infections. Complications included avascular necrosis, limb-length discrepancy, and pathologic fractures. The authors identified symptoms more than 1 week at presentation, neonatal age at presentation, infection of the hip joint, MRSA infection, and more than 3 days’ delay to appropriate antibiotics as predictors of complications.
Although MSSA has been found to have a similar course to other causative pathogens , a number of complications have been reported following septic arthritis and osteomyelitis, particularly related to CA-MRSA . In a series of 27 patients with CA-MRSA osteoarticular infections, 12 required ICU admission, four had acute multisystem failure, seven developed deep vein thrombosis (DVT) and septic pulmonary emboli, and all required at least one surgical intervention. Prompt recognition and treatment of MRSA osteoarticular infections are of particular importance. Despite appropriate intervention, potential for significant morbidity still exists.
In a series of children with osteomyelitis, DVT was identified in 10%, with all cases occurring in patients with MSSA or MRSA . Another retrospective review of DVT complications associated with osteoarticular infections has identified that this complication occurs most frequently in boys, more frequently occurs with Staphylococcus aureus, and is often associated with pulmonary complications .
Fractures following osteomyelitis have been reported more frequently in patients with MRSA [42▪]. Patients who went on to fracture were found to have more sub-periosteal abscess and larger area of infection at the time of initial MRI compared with patients who did not fracture. The USA300–0114 pulsotype was associated with an increased risk of fracture.
The cause and epidemiology of septic arthritis and osteomyelitis have changed over recent years and likely will continue to evolve with changes in immunization practices and bacterial resistance patterns. More widespread use of MRI and implementation of bacterial RNA/DNA detection by PCR have improved our diagnostic ability. Recent prospective and randomized trials have shown that shorter antibiotic regimens can successfully treat uncomplicated pediatric musculoskeletal infections, and that addition of steroids has improved short-term outcomes for patients with septic arthritis. Future research could focus on further prospective assessments of shorter antibiotic treatment regimens, and diagnostic protocols, such as PCR, that may allow earlier determination of causative organisms and the appropriate antibiotic therapy.
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
- ▪ of special interest
- ▪▪ of outstanding interest
Additional references related to this topic can also be found in the Current World Literature section in this issue (p. 151).
1. Gafur OA, Copley LA, Hollmig ST, et al. The impact of the current epidemiology of pediatric musculoskeletal infection on evaluation and treatment guidelines. J Pediatr Orthop 2008; 28:777–785.
2. Riise OR, Kirkhus E, Handeland KS, et al. Childhood osteomyelitis-incidence and differentiation from other acute onset musculoskeletal features in a population-based study. BMC Pediatr 2008; 8:45.
3. Rossaak M, Pitto RP. Osteomyelitis in Polynesian children. Int Orthop 2005; 29:55–58.
4▪. Dartnell J, Ramachandran M, Katchburian M. Haematogenous acute and subacute paediatric osteomyelitis: a systematic review of the literature. J Bone Joint Surg Br 2012; 94:584–595.
This review of pediatric osteomyelitis includes studies involving a total of 12 000 patients.
5. Copley LA. Pediatric musculoskeletal infection: trends and antibiotic recommendations. J Am Acad Orthop Surg 2009; 17:618–626.
6. Chen WL, Chang WN, Chen YS, et al. Acute community-acquired osteoarticular infections in children: high incidence of concomitant bone and joint involvement. J Microbiol Immunol Infect 2010; 43:332–338.
7. Ceroni D, Cherkaoui A, Ferey S, et al. Kingella kingae osteoarticular infections in young children: clinical features and contribution of a new specific real-time PCR assay to the diagnosis. J Pediatr Orthop 2010; 30:301–304.
8. Howard AW, Viskontas D, Sabbagh C. Reduction in osteomyelitis and septic arthritis related to Haemophilus influenzae
type B vaccination. J Pediatr Orthop 1999; 19:705–709.
9. Basmaci R, Lorrot M, Bidet P, et al. Comparison of clinical and biologic features of Kingella kingae and Staphylococcus aureus arthritis at initial evaluation. Pediatr Infect Dis J 2011; 30:902–904.
10▪. Ceroni D, Cherkaoui A, Combescure C, et al. Differentiating osteoarticular infections caused by Kingella kingae from those due to typical pathogens in young children. Pediatr Infect Dis J 2011; 30:906–909.
This study presents a prediction model for differentiation of Kingella kingae infection from osteoarticular infections caused by other pathogens.
11. Al Zamil F, Al Saadi M, Bokhary NA. The clinical profile of childhood osteomyelitis: a Saudi experience. J Pediatr Infect Dis 2008; 3:235–240.
12. Peltola H, Paakkonen M, Kallio P, Kallio MJ. OM-SA Study GroupClindamycin vs. first-generation cephalosporins for acute osteoarticular infections of childhood: a prospective quasi-randomized controlled trial. Clin Microbiol Infect 2012; 18:582–589.
13. Arnold SR, Elias D, Buckingham SC, et al. Changing patterns of acute hematogenous osteomyelitis and septic arthritis: emergence of community-associated methicillin-resistant Staphylococcus aureus
. J Pediatr Orthop 2006; 26:703–708.
14. Dohin B, Gillet Y, Kohler R, et al. Pediatric bone and joint infections caused by Panton-Valentine leukocidin-positive Staphylococcus aureus
. Pediatr Infect Dis J 2007; 26:1042–1048.
15▪. Otto M. MRSA virulence and spread. Cell Microbiol 2012; 14:1513–1521.
This is a comprehensive review of MRSA.
16▪. Robertson AJ, Firth GB, Truda C, et al. Epidemiology of acute osteoarticular sepsis in a setting with a high prevalence of pediatric HIV infection. J Pediatr Orthop 2012; 32:215–219.
This study highlights the importance of unusual pathogens in children with altered immunity, specifically Streptococcus pneumoniae, in HIV-positive children.
17. Vieira RL, Levy JA. Bedside ultrasonography to identify hip effusions in pediatric patients. Ann Emerg Med 2010; 55:284–289.
18. Tsung JW, Blaivas M. Emergency department diagnosis of pediatric hip effusion and guided arthrocentesis using point-of-care ultrasound. J Emerg Med 2008; 35:393–399.
19. Browne LP, Mason EO, Kaplan SL, et al. Optimal imaging strategy for community-acquired Staphylococcus aureus
musculoskeletal infections in children. Pediatr Radiol 2008; 38:841–847.
20▪. Browne LP, Guillerman RP, Orth RC, et al. Community-acquired staphylococcal musculoskeletal infection in infants and young children: necessity of contrast-enhanced MRI for the diagnosis of growth cartilage involvement. Am J Roentgenol 2012; 198:194–199.
This study highlights the importance of gadolinium for optimal MRI of Staphylococcus aureus osteoarticular infections.
21. Kanavaki A, Ceroni D, Tchernin D, et al. Can early MRI distinguish between Kingella kingae and Gram-positive cocci in osteoarticular infections in young children? Pediatr Radiol 2012; 42:57–62.
22▪. Warmann SW, Dittmann H, Seitz G, et al. Follow-up of acute osteomyelitis in children: the possible role of PET/CT in selected cases. J Pediatr Surg 2011; 46:1550–1556.
This study introduces the utility of PET/CT in differentiating between ongoing infection and signs of healing.
23. Paakkonen M, Kallio MJ, Kallio PE, Peltola H. Sensitivity of erythrocyte sedimentation rate and C-reactive protein in childhood bone and joint infections. Clin Orthop Relat Res 2010; 468:861–866.
24. Chometon S, Benito Y, Chaker M, et al. Specific real-time polymerase chain reaction places Kingella kingae as the most common cause of osteoarticular infections in young children. Pediatr Infect Dis J 2007; 26:377–381.
25. Weiss S, Geiss H, Kommerell M, et al. Improving the diagnosis of septic arthritis by use of a pediatric blood culture system [in German]. Orthopade 2006; 35:456458–462.
26▪. Basmaci R, Ilharreborde B, Bidet P, et al. Isolation of Kingella kingae in the oropharynx during K. kingae arthritis in children. Clin Microbiol Infect 2012; 18:E134–E136.
This study presents the use of oropharyngeal cultures in patients with suspected Kingella kingae infection.
27▪. Lehours P, Freydiere AM, Richer O, et al. The rtxA toxin gene of Kingella kingae: a pertinent target for molecular diagnosis of osteoarticular infections. J Clin Microbiol 2011; 49:1245–1250.
PCR detection of the Kingella kingae toxin is discussed.
28▪. Carpenter CR, Schuur JD, Everett WW, Pines JM. Evidence-based diagnostics: adult septic arthritis. Acad Emerg Med 2011; 18:781–796.
PCR for detection of osteoarticular infections is discussed.
29. El-Sayed AM. Treatment of early septic arthritis of the hip in children: comparison of results of open arthrotomy versus arthroscopic drainage. J Child Orthop 2008; 2:229–237.
30. Journeau P, Wein F, Popkov D, et al. Hip septic arthritis in children: assessment of treatment using needle aspiration/irrigation. Orthop Traumatol Surg Res 2011; 97:308–313.
31. Paakkonen M, Kallio MJ, Peltola H, Kallio PE. Pediatric septic hip with or without arthrotomy: retrospective analysis of 62 consecutive nonneonatal culture-positive cases. J Pediatr Orthop B 2010; 19:264–269.
32. Jagodzinski NA, Kanwar R, Graham K, Bache CE. Prospective evaluation of a shortened regimen of treatment for acute osteomyelitis and septic arthritis in children. J Pediatr Orthop 2009; 29:518–525.
33. Peltola H, Paakkonen M, Kallio P, Kallio MJ. Osteomyelitis-Septic Arthritis Study GroupShort- versus long-term antimicrobial treatment for acute hematogenous osteomyelitis of childhood: prospective, randomized trial on 131 culture-positive cases. Pediatr Infect Dis J 2010; 29:1123–1128.
34. Ballock RT, Newton PO, Evans SJ, et al. A comparison of early versus late conversion from intravenous to oral therapy in the treatment of septic arthritis. J Pediatr Orthop 2009; 29:636–642.
35. Peltola H, Paakkonen M, Kallio P, Kallio MJ. Osteomyelitis-Septic Arthritis (OM-SA) Study GroupProspective, randomized trial of 10 days versus 30 days of antimicrobial treatment, including a short-term course of parenteral therapy, for childhood septic arthritis. Clin Infect Dis 2009; 48:1201–1210.
36▪. Harel L, Prais D, Bar-On E, et al. Dexamethasone therapy for septic arthritis in children: results of a randomized double-blind placebo-controlled study. J Pediatr Orthop 2011; 31:211–215.
This RCT discusses the use of dexamethasone in treatment of septic arthritis. Improved outcomes are reported.
37▪. Sukswai P, Kovitvanitcha D, Thumkunanon V, et al. Acute hematogenous osteomyelitis and septic arthritis in children: clinical characteristics and outcomes study. J Med Assoc Thai 2011; 94 (Suppl 3):S209–S216.
Short and long-term complications following septic joints are presented.
38. Paakkonen M, Kallio PE, Kallio MJ, Peltola H. Management of osteoarticular infections caused by Staphylococcus aureus
is similar to that of other etiologies: analysis of 199 staphylococcal bone and joint infections. Pediatr Infect Dis J 2012; 31:436–438.
39. Vander Have KL, Karmazyn B, Verma M, et al. Community-associated methicillin-resistant Staphylococcus aureus
in acute musculoskeletal infection in children: a game changer. J Pediatr Orthop 2009; 29:927–931.
40. Bouchoucha S, Benghachame F, Trifa M, et al. Deep venous thrombosis associated with acute hematogenous osteomyelitis in children. Orthop Traumatol Surg Res 2010; 96:890–893.
41. Mantadakis E, Plessa E, Vouloumanou EK, et al. Deep venous thrombosis in children with musculoskeletal infections: the clinical evidence. Int J Infect Dis 2012; 16:e236–e243.
42▪. Belthur MV, Birchansky SB, Verdugo AA, et al. Pathologic fractures in children with acute Staphylococcus aureus
osteomyelitis. J Bone Joint Surg Am 2012; 94:34–42.