Diagnostic accuracy of an intrathecal test dose in epidural analgesia

The authors evaluated the accuracy of an intrathecal test dose (TD) to rule out unintentional subarachnoid injection in a prospective and double-blind study. Twenty-nine unmedicated patients scheduled to receive continuous spinal analgesia for their surgical procedures entered the study. After placement of non-invasive monitoring (ECG, BP and SpO2), an intrathecal catheter was placed in the lumbar area. In the supine position 15 patients in the study group received lidocaine 45 mg + epinephrine 15 mg and 14 patients in the control group received normal saline 3 ml intrathecally. After the injection, a blinded observer assessed the presence of the following signs at one minute intervals for five minutes: sensory loss by pin-prick (P) over the lumbar and sacral dermatomes, ability to raise legs (R), and subjective feeling of warmth (W) and heaviness (H) over the lower limbs. For each sign the sensitivity (SN), specificity (SP), and negative (-) and positive (+) predictive value (PV) were calculated. At four minutes SN reached 100% (CI 78-100%) for R and H, while only R obtained a SP of 93% (CI 66-100%). Using R as detector of intrathecal injection the -PV was 100% and the +PV 25% (assuming a 2% prevalence of unintentional spinal). We conclude that, with our test dose; (a) four minutes are needed to recognize signs of intrathecal injection; (b) leg-raising is a reliable sign; and (c) inquiring about other signs may only decrease the diagnostic accuracy of this test.