To determine whether hepatitis B virus (HBV) transmission occurred among patients visiting a physician's office and to evaluate potential transmission mechanisms.

Serologic survey, retrospective cohort study, and observation of infection control practices.

Private medical office.

Those visiting the office between March 1 and December 26, 2001.

We identified 38 patients with acute HBV infection occurring between February 2000 and February 2002. The cohort study, limited to the 10 months before outbreak detection, included 91 patients with serologic test results and available charts representing 18 case-patients and 73 susceptible patients. Overall, 67 patients (74%) received at least one injection during the observation period. Case-patients received a median of 14 injections (range, 2-25) versus 2 injections (range, 0-17) for susceptible patients (P < .001). Acute infections occurred among 18 (27%) of 67 who received at least one injection versus none of 24 who received no injections (RR, 13.6; CI95, 2.4-undefined). Risk of infection increased 5.2-fold (CI95, 0.6-47.3) for those with 3 to 6 injections and 20.0-fold (CI95, 2.8-143.5) for those with more than 6 injections. Typically, injections consisted of doses of atropine, dexamethasone, vitamin B12, or a combination of these mixed in one syringe. HBV DNA genetic sequences of 24 patients with acute infection and 4 patients with chronic infection were identical in the 1,500-bp region examined. Medical staff were seronegative for HBV infection markers. The same surface was used for storing multidose vials, preparing injections, and dismantling used injection equipment.

Administration of unnecessary injections combined with failure to separate clean from contaminated areas and follow safe injection practices likely resulted in patient-to-patient HBV transmission in a private physician's office.