Bacterial Infections of the Bones and Joints

Osteomyelitis

Osteomyelitis occurring in the first 2 months of life is uncommon. During the worldwide pandemic of staphylococcal disease from the early 1950s to the early 1960s, pediatric centers in Europe, Australia, and North America reported the infrequent occurrence of neonatal osteomyelitis, accounting for only one or two admissions per year at each institution. An overall incidence of neonatal osteomyelitis has been reported as 1 in 5000 to 15,000 live births. With the introduction of invasive neonatal supportive care and the increased use of diagnostic and therapeutic procedures, there was concern that osteomyelitis and septic arthritis secondary to bacteremia might occur more frequently in the newborn. Yet subsequent experience in Europe, Canada, and the United States (J.D. Nelson, personal communication, 1987) during the decade 1970 to 1979 indicated little or no change in the incidence of this condition. Even in intensive care nurseries, despite an increasing problem with fungal ( Candida ) osteoarthritis, the overall rate of occurrence of nosocomial bone and joint infections remained low at equal to or less than 2.6 per 1000 admissions. Infections associated with invasive procedures, such as placement of intravascular catheters, may not appear (or be recognized) until days or weeks after the perinatal period, however. Although the incidence has not changed, causative organisms have become increasingly resistant to antibiotics, as exemplified by the increased incidence of Staphylococcus aureus infections resistant to oxacillin (methicillin-resistant S. aureus [MRSA]).

Little had been published on the relative incidence of neonatal osteomyelitis during the 1980s and 1990s. An ongoing review of nursery infections at a Kaiser Permanente hospital in southern California revealed only 3 cases of osteomyelitis among 67,000 consecutive live births from 1963 to 1993, and none occurred in the final years (A. Miller, personal communication, 1993). A similar survey performed at two pediatric referral centers in Texas showed no significant variation in the number of annual admissions for this condition from 1964 to 1986 (J.D. Nelson, personal communication, 1987). One to 3 cases of bone or joint infection per 1000 admissions remains a reported incidence, almost identical to that noted in years past.

In a review of more than 300 cases of neonatal osteomyelitis, male infants predominated over female infants (1.6:1). Premature infants acquire osteomyelitis with relatively greater frequency than term infants. In a series of osteomyelitis, 17 of 30 proven cases were in premature infants, 4 occurred in term infants receiving intensive care, and S. aureus was responsible for 23 of the proven cases of osteomyelitis (methicillin-sensitive strains in 16 cases and MRSA in 7 cases). Escherichia coli and group B streptococci (GBS) caused 3 and 2 cases, respectively. Risk factors for osteomyelitis and septic arthritis in premature infants have been mostly iatrogenic, including use of intravenous or intraarterial catheters, ventilatory support, and bacteremia with nosocomial pathogens.

Risk factors for osteomyelitis have been reviewed in detail in other publications. Osteomyelitis and septic arthritis today are largely associated with nosocomial bacteria acquired with systemic infection during neonatal intensive care. Narang and colleagues noted an incidence of 1 per 1500 in cases of bone and joint infections, with a mean gestational age of 34 weeks (range, 27-40) and mean birth weight of 2269 (range, 990-4750), emphasizing the high risk among premature infants. In addition to sepsis, both septic arthritis and osteomyelitis are risks for infection of the adjacent site, either the synovium or the bone. Box 8-1 reviews the known risk factors presented in most series. The neonate, and particularly the premature infant, is at much higher risk of multifocal osteomyelitis. For instance, Howan-Giles and Uren found that two peaks of multifocal osteomyelitis occurred with the first peak of less than 6 weeks, accounting for 38% of the cases of multifocal osteomyelitis. The association of sepsis and/or urinary tract infections or urinary tract anomalies has frequently been observed. The presence of certain focal or predisposing conditions has long been recognized, such as cephalohematomas and/or infection after fetal monitors with direct osteomyelitis of the skull. In addition, the associations of hip septic arthritis associated with breech delivery, and of brachial palsy with septic arthritis of the hip and shoulder, respectively, have been recognized for the complication of joint infection. Cieslak and Rajnik calculated a high incidence of septic arthritis or pelvic osteomyelitis among breech infants with an incidence of approximately 1 per 8000 or a relative risk of 4.1 compared with nonbreech infants. Abnormal uterine posture of the upper limb or associated radial nerve palsy has frequently been associated with septic arthritis.