Diagnostic delay in pediatric solid tumors: a population based study on determinants and impact on outcomes.

Background: Despite socioeconomic and clinical progress, pediatric tumors continue to present in advanced stage, and may be due to delays in diagnosis. This study aimed to identify factors associated with diagnostic delay (time between symptom onset and diagnosis) in pediatric tumors in a population-based study, and to assess the impact of delay on subsequent outcome.

Procedure: Natural logarithm of delay was retrospectively described for 390 newly diagnosed tumors reported to the Singapore Childhood Cancer Registry from 1997 to 2007. Delay was correlated with socio-demographic, disease and healthcare-system factors using multivariate linear regression, and with event-free survival (EFS) using Cox regression analysis.

Results: Total median delay was 5.3 weeks (range 0.1-283.1). Shorter delay was independently associated with younger patient age (P = 0.006), abdominal and pelvic sites (P < 0.001 and P = 0.036, respectively), incidental diagnoses by healthcare staff (P = 0.002), and when pediatric emergency units were the first contacted healthcare facilities and the first to raise suspicion of malignancy (P = 0.034, and P = 0.018, respectively). These factors explained only a small percentage of variance in delay times (21%). Delay was not associated with EFS and disease stage, with 24% of tumors presenting in stage 4.

Conclusions: Diagnostic delay was independently associated with age and site of presentation, and points of first symptom detection, first healthcare contact, and first suspicion of malignancy. The broad range of clinical variables analyzed could only account in a small way for differences in delay times observed. While overall delay times were short, they did not influence disease stage at presentation and eventual outcome.