Utility of Inflammatory Markers in Hospitalized Children With Skin Erythema.

BACKGROUND

There is overlap in the clinical presentations of superficial skin infections (eg, cellulitis) and deeper infections (eg, osteomyelitis). Inflammatory markers are frequently obtained in hospitalized patients with cellulitis. However, it is unknown whether inflammatory markers discriminate between superficial and deeper infections.

METHODS

We performed a retrospective review of children hospitalized with erythema on the skin and diagnoses of cellulitis, bursitis, myositis, osteomyelitis, and/or tenosynovitis. Patients were grouped into superficial infection (cellulitis and bursitis) and deeper infection (myositis, tenosynovitis, osteomyelitis). We compared C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) values between the two groups and constructed receiver operating characteristic curves to determine optimal cutoff values.

RESULTS

Among 168 children hospitalized with skin erythema, 122 patients had superficial and 46 patients had deeper infections. Mean CRP and ESR were significantly higher in patients with deeper infections (CRP: 8.3 [±9.6] mg/dL versus 2.8 [±3.5] mg/dL; P < .001; ESR: 48.3 [±28.6] mm/h versus 21.8 [±16.0] mm/h; P < .001). Neither WBC nor fever was significantly different between groups. CRP ≥2 mg/dL and ESR ≥30 mm/h were associated with deeper infections (CRP: odds ratio: 4.77; 95% confidence interval: 2.24-10.15; ESR: odds ratio: 7.93; 95% confidence interval: 3.46-17.67). When CRP and ESR were below these cutoffs, the negative predictive value was 89%.

CONCLUSIONS

Among patients presenting with skin erythema, CRP ≥2 mg/dL and ESR ≥30 mm/hr were significantly associated with deeper infection, whereas values below these cutoffs were reassuring against deeper infection. Future study of inflammatory markers in skin and soft tissue infections may help develop tailored testing strategies.