The incidence of device-related infections depends directly on the definition employed. The lack of precision is also compounded by the latency between the initiation and manifestation of the infection. It is not rare for there to be some erythema at the incision site during the first week of healing, and it is not clear that this represents infection. Less frequent, but still common, there can be a small, superficial stitch abscess, which will respond to local measures. When the diagnosis of device system infection is made, it should be made on the basis of pocket cellulitis, erosion, abscess, persistent bacteremia, or endocarditis with or without vegetation on the lead.