Background and Objectives Epinephrine-containing test doses for detection of intravascular injection during epidural anesthesia may be unreliable or hazardous in beta-blocked, elderly, or pregnant patients. Subtoxic injections of lidocaine have been used as an alternative marker of intravascular injection in unpremedicated patients. We studied two groups of premedicated patients and unpremedicated subjects to evaluate the reliability of the local anesthetics bupivacaine (B) and 2-chloroprocaine (2-CP) as test dose injections.

Methods Thirty ASA I and II subjects received blinded randomized injections of 2-CP, B, or normal saline via a peripheral vein.

Results In group I, 10 healthy unpremedicated volunteers universally recognized injection of 90 mg 2-CP or 25 mg B, but did not reliably detect the injection of 60 mg 2-CP. In group II, 20 patients premedicated with 1 μg/kg fentanyl and 30 μg/kg midazolam could not reliably detect similar injections. Sixteen responded to the injection of 90 mg 2-CP, while 13 detected the 25 mg B test dose. A blinded observer rated only 12 of the subjective reports as conclusive of detection of intravascular injection after each drug. There were no false-positive reports in any group.

Conclusions While 90 mg 2-CP or 25 mg B may be reliable alternatives to epinephrine test doses in unpremedicated subjects, they are unreliable indicators of intravascular injection in the premedicated patient.