Methicillin-resistant Staphylococcus aureus (MRSA) continues to be a major source of morbidity in pediatric patients. Ju et al. developed an evidence-based clinical prediction algorithm to differentiate MRSA osteomyelitis from methicillin-sensitive Staphylococcus aureus (MSSA) osteomyelitis5. In a retrospective review involving 129 children with culture-proven Staphylococcus aureus osteomyelitis, the authors found that a temperature of >38°C, a hematocrit of <34%, a white blood-cell count of >12,000, and C-reactive protein of >13 mg/L were independent multivariate predictors of MRSA osteomyelitis. The predicted probability of MRSA osteomyelitis was 92% when all four predictors were present, 45% when three were present, 10% when two were present, 1% when one was present, and 0% when none were present. These clinical criteria can help to guide patient management and provide criteria for antibiotic management.
Belthur et al. reviewed pathologic fractures in children with acute infection6. Seventeen patients who had a pathologic fracture in the setting of Staphylococcus aureus osteomyelitis were compared with a control group of forty-nine patients without fracture. The patients in the fracture group had a longer hospital stay (mean, 25.6 versus 12.5 days), longer durations of parenteral antibiotics (mean, 7.1 versus 5.4 days) and oral antibiotics (mean, 42.1 versus 11.5 days), a higher rate of multiple surgical procedures (mean, 65.0% versus 31.0%), and more surgical and medical complications (mean, 5.5 versus 1.3). In addition, magnetic resonance imaging (MRI) studies at the time of admission demonstrated a greater prevalence and size of subperiosteal abscesses as well as a sharp zone of abnormally diminished enhancement of the bone marrow. Patients with fractures were commonly found to have the USA300 pulsotype.
The findings of these studies suggest that patients with more severe disease presentation should be managed more conservatively in terms of activity and weight-bearing status until resolution of the infection as demonstrated by clinical, laboratory, and radiographic findings.
Developmental Dysplasia of the Hip
The overall effect of swaddling has been controversial for centuries, and its negative effect on the development of the hip is of great concern. Wang et al.13 conducted a prospective study in which 112 neonatal rats were divided into a control group and three experimental groups that were swaddled with use of surgical tape in a manner simulating the human practice for the first five days of life (early swaddling), the second five days (late swaddling), and the first ten days (prolonged swaddling). Rats in the prolonged-swaddling group had the highest prevalence of hip dysplasia (thirty-six of forty-four), followed by the early-swaddling group (twenty-one of forty-four). Most of the dysplastic hips in the prolonged-swaddling group were dislocated, whereas subluxation dominated in the late-swaddling group.
That study demonstrated that straight-leg swaddling increased the prevalence of developmental dysplasia of the hip in an animal model, especially if the swaddling was early or prolonged. The severity of hip impairment varied, with both early and prolonged swaddling leading to more dislocations than subluxations. The clinical relevance of the paper by Wang et al. is related to the possible harmful effects of traditional infant swaddling with the lower extremities extended and adducted, as is currently recommended to new mothers by their pediatricians.
The Pavlik harness is more likely to stabilize a hip with developmental dysplasia when treatment is started before the age of six months, so most clinicians are reluctant to try the Pavlik harness after that age. Also, the harness is usually abandoned if treatment is unsuccessful after four weeks because of concerns about causing greater problems if the harness is used on a dislocated hip that stays dislocated. van de Sande and Melisie14, in a study of thirty-one infants with a mean age of eight months (range, five months to thirteen months) at the time of diagnosis, reported that twenty (65%) of the thirty-one hips were successfully reduced with the Pavlik harness. An average of seven weeks (range, three to eleven weeks) were needed to obtain the reduction. Additional treatment with an abduction brace was then used for three to six months. The rate of successful reduction was higher for hips with less severe dislocations (81% for Tönnis type-2 hips, compared with 25% Tönnis type-3 and 4 hips). Five patients (15.0%) demonstrated radiographic signs of osteonecrosis. The authors concluded that prolonged use of the Pavlik harness for the treatment of late-diagnosed hip dislocation (developmental dysplasia of the hip) is acceptably safe and potentially successful for the infants in this older age group. From a practical standpoint, it is often very hard to manage a child who is more than six months old with a Pavlik harness because parental compliance with the use of the harness is often problematic.
There are growing concerns about radiation exposure in children. In spite of major advances in the field of imaging, radiographs remain the primary tool for evaluating hip dysplasia in children older than four months and are commonly used for treatment decisions. Upasani et al.15 assessed the reliability and reproducibility of the radiographic measurements that are routinely used to assess dysplasia in young children. Fifty radiographs from twenty-one children with developmental dysplasia of the hip were reviewed. The authors concluded that it is difficult to reliably measure three-dimensional pelvic shape on a frontal radiograph, especially when important pelvic landmarks have yet to ossify. This article outlines the limitations of using radiographic measurements as the sole method for the evaluation of developmental dysplasia of the hip in infancy and childhood, yet this method remains the best method currently available.
Accurate reduction of the hip is important for children with dislocated hips. Gould et al.16 reported on the use of MRI to confirm reduction in patients with developmental dysplasia of the hip. They reviewed thirty-four consecutive MRI studies that had been performed without sedation after spica cast placement in twenty-four patients with developmental hip dysplasia. Ninety-seven percent of the studies were diagnostic. T2-weighted fast spin echo sequences had the best overall scores and were performed in less than three minutes. T1 and fat-suppressed T2-weighted fast spin echo sequences did not score as well and also required less than three minutes. Sequences such as single-shot fast spin echo and three-dimensional gradient recovery sequences scored poorly. The authors concluded that MRI is a useful tool for evaluating the hip without radiation exposure and without sedation in infants and toddlers following the reduction of a subluxed or dislocated hip. Orthopaedic surgeons can request these MRI sequences for accurate assessment of concentric reduction with a potential study time of fifteen minutes. The stationary position of the child immobilized in a hip spica cast obviates the need for sedation.
In spite of good neonatal screening programs and a global push toward early diagnosis of developmental dysplasia of the hip in infants, children with developmental dysplasia of the hip and dislocated hips that need surgical intervention are still seen. Holman et al.17 reviewed 148 patients (179 dislocated hips) who were managed with open reduction for the treatment of developmental dysplasia of the hip over a forty-year period. Of these, fifty-three patients (sixty-six dislocated hips) were available for follow-up. Almost half of the children needed a second surgical procedure. Redislocation and osteonecrosis of the femoral head predicted poor outcomes. The authors also noted that the contralateral hip may develop dysplasia even though it appeared normal initially. This report outlines the unpredictable success rate following surgical procedures related to the delayed treatment of hip dysplasia.
Legg-Calvé-Perthes Disease
Larson et al.18 sought to prospectively document pain and function in a cohort of adults who had previously been managed nonoperatively for Legg-Calvé-Perthes disease. Patients were enrolled between 1984 and 1991 as part of a multicenter prospective trial and were managed with hip range-of-motion exercises or bracing. Fifty-six patients (fifty-eight hips) were examined at a mean of 20.4 years (range, 16.3 to 24.5 years). The authors noted that pain, arthritis, and ongoing hip dysfunction are common in patients with Legg-Calvé-Perthes disease who are managed nonoperatively. Hips that were rated as Stulberg type III or IV more frequently had poor or fair outcomes according to the Iowa Hip Score and the Nonarthritic Hip Score (61% and 72% for type-III hips and 77% and 60% for type-IV hips). Patients with a Herring lateral pillar type of B, B/C, or C frequently had pain and radiographic evidence of osteoarthritis. Clinical signs of femoroacetabular impingement were associated with pain and with lower functional scores. Their findings are in contrast to those reported by Boyer et al.19 and may reflect contemporary concepts, including increasing obesity in our culture, as well as a more aggressive method of questioning patients and providing “hip scores” in the “hip impingement” era.
Recent advances in surgical techniques have permitted surgical dislocation of the hip, allowing for correction of femoroacetabular impingement even in patients with Legg-Calvé-Perthes disease. Shore et al.20 conducted a study to characterize the location and number of lateral epiphyseal arteries supplying the femoral head in children with healed Legg-Calvé-Perthes disease. High-resolution contrast-enhanced MRI scans were made for nineteen children (twenty-two hips) with a diagnosis of Legg-Calvé-Perthes disease and a matched control group of seventeen children (twenty hips) with developmental hip dysplasia. The authors noted that the lateral epiphyseal arteries reliably inserted on the posterior-superior aspect of the femoral neck from a superior-anterior to a superior-posterior position in both groups. The authors concluded that reperfusion of the medial femoral circumflex artery does occur in patients with Legg-Calvé-Perthes disease; however, the overall number of vessels is decreased compared with patients with developmental hip dysplasia.
Slipped Capital Femoral Epiphysis
The most important objective of clinical classifications of slipped capital femoral epiphysis is to identify hips that are associated with a high risk for osteonecrosis; however, traditional closed surgery for the treatment of slipped capital femoral epiphysis has made confirmation of physeal stability difficult.
Ziebarth et al. retrospectively reviewed eighty-two patients with slipped capital femoral epiphysis who were managed with open surgery between 1996 and 200921. They categorized the clinical stability of all hips on the basis of the onset of symptoms (acute, acute-on-chronic, chronic) and the Loder system (stable, unstable). They also categorized the intraoperative stability as intact or disrupted. They then determined the sensitivity and specificity of two classification systems to determine intraoperative stability. Complete physeal disruption at the time of open surgery was seen in twenty-eight (34%) of the eighty-two hips. With classification as acute, acute-on-chronic, and chronic, the sensitivity for disrupted physes was 82% and the specificity was 44%. According to the classification system of Loder et al., the values were 39% and 76%, respectively. The authors concluded that current clinical classification systems are limited in terms of accurately diagnosing physeal stability in patients with slipped capital femoral epiphysis.
Further advances were made in the area of pediatric spine surgery, with continued multicenter and basic science and clinical collaboration with an emphasis on patient safety.
Sponseller et al. performed a retrospective review of 1912 patients with adolescent idiopathic scoliosis to determine the associations of patient age, sex, primary curve magnitude, and Scoliosis Research Society (SRS)-22 outcome scores with the Lenke curve type26. The authors found that Lenke type-1 curves were the most common and that type-4 curves were the least common. The Lenke curve type also varied by sex (with males having more major thoracic than major thoracolumbar/lumbar curves), age (with type-5 curves occurring in the oldest patients), and primary curve magnitude (with type-4 curves having the largest mean magnitude and type-1 and 5 curves having the smallest). Patients with type-4 curves had lower preoperative SRS-22 scores for self-image than patients with type-1 curves.
Roye et al. compared the ScoliScore (Axial Biotech, Salt Lake City, Utah), a saliva-based genetic test that can potentially predict adolescent idiopathic scoliosis curve progression, with traditional clinical estimates of curve progression (the Risser sign and Cobb angle) in a study of ninety-one adolescent patients with idiopathic scoliosis and curves of between 10° to 25°27. When the risk of curve progression was assessed with use of the ScoliScore system, 36% of the patients were classified as low risk, 55% were classified as intermediate risk, and 9% were classified as high risk. In contrast, when the risk of curve progression was based on clinical estimates, 2% of the patients were classified as low risk, 51% were classified as intermediate risk, and 47% were classified as high risk. These findings indicated that the ScoliScore predicted sixteen times more low-risk and five times fewer high-risk patients than the clinical estimates. In addition, only the Cobb angle positively correlated with the ScoliScore.
Diab et al. examined the outcomes associated with the use of wound drains after spinal fusion in patients with adolescent idiopathic scoliosis28. In this multicenter, retrospective analysis, 324 patients who had been managed with drains and 176 patients who had been managed without drains were evaluated for two years postoperatively. Complication rates did not differ between the two cohorts of patients. Of note, patients who had been managed with a drain were more likely to receive a postoperative transfusion (43.0%) than those who had not (22.0%).
Phillips et al. performed a retrospective chart review of 165 surgical procedures in twenty-eight patients with early-onset scoliosis with the diagnoses of congenital scoliosis, syndromic/chromosomal abnormalities, cerebral palsy, and spinal muscular atrophy29. The authors found an 84% complication rate and an 18% mortality rate, largely as a result of pulmonary failure. This alarmingly high mortality rate suggests that additional study is necessary to decrease the complication rate associated with surgical intervention in this population.