Evaluating the child with acute hip pain ("irritable hip") in a Lyme endemic region.

OBJECTIVES

To estimate the prevalence of Lyme infection among children presenting with acute, nontraumatic hip pain in a Lyme endemic region and to investigate predictors of Lyme disease among children with suspected transient synovitis.

STUDY

DESIGN

Retrospective cross-sectional study of children with unilateral hip pain who were brought to an academic pediatric emergency department. Cases were identified by specific discharge diagnoses or radiologic imaging. Lyme infection was determined by serologic criteria, and a minimum prevalence was estimated for the entire study population; maximum estimate was determined for those who had Lyme testing. Multivariate regression was used to identify discriminating clinical findings for Lyme disease among those with nonseptic arthritis.

RESULTS

Three hundred eighty-five children with a median age of 5.4 years were studied; 15% of children had fever ≥38.0°C and 40% had pain for less than 24 hours at evaluation. Lyme infection was identified in 5.2% (95% CI 3.2%-7.9%). A maximum estimate of Lyme disease was calculated to be 8.0% (95% CI 4.9%-12.0%). Regression analysis did not identify any practical clinical predictors of Lyme infection.

CONCLUSIONS

Lyme infection occurred in approximately 5% of children with acute, nontraumatic hip pain who were evaluated in a pediatric emergency department in a Lyme endemic region. Based on this estimate, we do not recommend routine Lyme testing when transient synovitis is suspected; however, Lyme testing should be considered in children having laboratory studies obtained for alternative diagnoses such as septic/pyogenic arthritis and for those with an atypical clinical course for transient synovitis.