The limping child: an algorithm to outrule musculoskeletal sepsis

Abstract

Background: The acutely limping child presents a significant diagnostic challenge.

Aim: The purpose of this study was to create a clinically useful algorithm to allow exclusion of 'musculoskeletal sepsis' as a differential diagnosis in the child presenting with limp.

Methods: Data were collected on all 286 limping children admitted to our centre over a 3-year-period. Using logistic regression analysis, the predictive model was constructed, to exclude infection.

Results: Duration of symptoms, constitutional symptoms, temperature, white cell count and ESR were significantly different in children with musculoskeletal infection (P < 0.05). Multivariate analysis demonstrated that when all three variables of duration of symptoms >1, <5 days; temperature >37.0 degrees C; and ESR >35 mm/h were present, the predicted probability of infection was 0.66, falling to 0.01 when none were present.

Conclusion: This multivariate model enables us to rule out musculoskeletal infection with 99% certainty in limping children with none of these three presenting variables.