Severity Adjusted Risk of Long-Term Adverse Sequelae Among Children with Osteomyelitis

    loading  Checking for direct PDF access through Ovid

Abstract

Background:

The purpose of this investigation is to evaluate the risk for long-term, adverse outcomes among children with osteomyelitis.

Methods:

Children with osteomyelitis were prospectively enrolled from 2012-2014. Care was accomplished by a multidisciplinary team according to an institutional algorithm. Data was collected to define the severity of illness during the initial hospitalization and assess short, intermediate and long-term outcomes. Clinical examination, radiographic assessment, and functional outcome survey administration were performed at a minimum of 2 year follow-up. A comparison cohort analysis was performed according to initial severity of illness score (SIS) of mild (0-2), moderate (3-6), and severe (7-10).

Results:

Of 195 children enrolled, 139 (71.3%) returned for follow-up at an average of 2.4 years (range 2.0 to 5.0 years). Children with severe illness were less likely to have normal radiographs (Severe – 4.0%; Moderate – 38.2%; Mild – 53.2%, p<0.0001), and more likely to have osteonecrosis, chondrolysis, or deformity (Severe – 32.0%; Moderate – 5.9%; Mild – 1.3%, p<0.0001). Functional outcome measures did not significantly differ between severity categories.

Results:

By regression analysis SIS, plus age less than 3 years and MRSA predicted severe sequelae with an area under the curve of 0.8617 and an increasing odds ratio of 1.34 per point of increase in severity score.

Conclusion:

Long term severe adverse outcomes among children with osteomyelitis occurred in 11 of 139 (7.9%) children and were predicted by initial severity of illness. Other risks that diminished the likelihood of complete resolution or increased the risk of severe sequelae included MRSA etiology and young age. The majority of children with osteomyelitis do not require long term follow-up beyond the initial treatment period.

Level of Evidence:

Level II, Prognostic, Prospective Cohort Comparison

Related Topics

    loading  Loading Related Articles