Aim. – To analyze the factors associated with the time to initiating tuberculosis contact investigations in the Somme department, France.
Methods. – All reported tuberculosis cases and all their contacts screened between 2007 and 2011 were retrospectively included. Univariate
and multivariate analyses were conducted to determine the factors associated with a ‘‘system delay’’  1 month and a ‘‘contact delay’’  0 days.
Results. – The mean time between the mandatory notification of a case of tuberculosis and the date set for the contact’s screening (system delay)
was 35.3 days and the average time between that date and when the contact was actually screened (contact delay) was 12.5 days. In multivariate
analysis, a smear-positive sputum sample (OR: 3.68; 95% CI: 1.63–8.30) and a diagnosis at the university hospital (OR: 2.61; 95% CI: 1.14–5.96)
were significantly associated with a system delay  1 month. A smear-positive sputum sample (OR: 1.35; 95% CI: 1.08–1.69), male gender (OR:
1.21; 95%CI: 1.01–1.49), being born in a foreign country (OR: 1.31; 95%CI: 1.02–1.69), being a family member (OR: 1.37; 95%CI: 1.05–1.77), or
being another type of close contact of the case (OR: 2.47; 95% CI: 1.81–3.36) were significantly associated with a contact delay  0 days.
Conclusion. – System and contact delays were longer than recommended, and the factors associated with the lengthening of these delays need
to be taken into account.

Tuberculosis; TB screening; Contact investigation; Delay; France