Multiport Epidural Catheters: Does the Air Test Work?

Three hundred women in labor requesting epidural analgesia gave oral consent and participated in this study, which was approved by the Thomas Jefferson University Institutional Review Board. Patients sat while we used the air loss-of-resistance technique to identify the epidural space at the L2–L3 or L3–L4 interspace using an 18-gauge Hustead needle. The study was discontinued if cerebrospinal fluid flowed or could be aspirated through the epidural needle. We threaded a blunt-tip, three-orifice, 20-gauge, polyamide epidural catheter (B. Braun Medical) 5 cm into the epidural space and taped the catheter to the patient’s back while she was sitting. We attempted to aspirate blood or cerebrospinal fluid from each catheter. If no fluid could be aspirated, we injected 3 ml lidocaine, 2%, or 3 ml bupivacaine, 0.25%, waited 3 min, and questioned the patient about changes in lower extremity sensation and motor strength. If we aspirated cerebrospinal fluid or obtained signs of spinal anesthesia with local anesthetic injection, we documented intrathecal placement, performed no further testing, and managed the catheter as clinically appropriate.

We performed an air test on all nonintrathecal catheters using a previously described technique. 2Briefly, we placed the Doppler probe of a Hewlett Packard (Palo Alto, CA) HP8040A fetal heart rate monitor over the lower maternal sternum (i.e. , over the right ventricle) and positioned the probe so clear maternal heart sounds could be heard. We injected 1 ml air through the epidural catheter and listened for maternal heart sound changes for 15 s. We considered the air-test result to be positive if a loud swishing sound, lasting at least 5 s, was heard in addition to or in place of the normal “lub-dub” heart sounds. Because the increased circulating blood volume during a uterine contraction can change maternal heart sounds, we repeated the air test if the patient’s uterus contracted during the observation period. If no blood was aspirated and no heart tone changes were heard, we injected 10 ml bupivacaine, 0.125%, with 50 μg fentanyl. We recorded whether a sensory analgesic band was present 20 min after epidural injection. If a sensory band was present, we continued epidural analgesia by continuously infusing 8–12 ml/h bupivacaine, 0.05%, with 1.45 μg/ml fentanyl and 1:700,000 epinephrine.

Epidural catheters through which blood initially was aspirated were air-tested. If the first air test was negative, the catheter was tested a second time, with the air injected as fast as possible. Then, at the discretion of the attending anesthesiologist, the catheter was either replaced immediately or withdrawn 2 cm and air-tested again. If blood was aspirated or the air test produced heart tone changes after catheter repositioning, the catheter was replaced. Otherwise, we injected the catheter with bupivacaine–fentanyl, as described previously.

If the air test was positive but blood could not be aspirated from the catheter, we injected 5 ml lidocaine, 2%, and questioned the patient about perioral numbness, tinnitus, and diplopia. If the patient reported symptoms of local anesthetic intravenous injection, we replaced the catheter. Otherwise, we injected the catheter with bupivacaine–fentanyl and used the catheter to provide analgesia if an appropriate sensory band developed.

If no positive results are obtained in a case series of 300 patients, one can state with 95% confidence that the true incidence of the phenomenon is less than 1%. 3Therefore, we enrolled 300 patients.

Testing results are summarized in table 1.

We found that neither the air test nor blood aspiration detected all cases of intravascularly placed multiorifice epidural catheters. We previously reported that the air test detected all intravascularly placed single-orifice catheters in a 313-patient case series, suggesting that the air test is more sensitive for detection of single- rather than multiorifice catheters. 2However, a study showing a difference between 0 and 1% false-negative rates would necessitate approximately 770 patients per group (assuming α= 0.05 and 1−β= 0.8). 4 

Blood was aspirated more frequently from multiple- than from single-orifice epidural catheters in three large studies 5–7and with equal frequency in two studies. 8,9There may be a higher incidence of vascular cannulation with a multiport design. Alternatively, aspiration may detect a higher fraction of intravenous multiport catheters. Our aspiration results are similar to those of Norris et al. , 10who reported that aspiration was 98% sensitive in detecting clinically evident intravascular multiorifice epidural catheters. However, Beck et al.  11found that aspiration detected only 54% of the intravascular catheters.

One cannot assume that an aspiration-negative, multiorifice catheter lies completely outside the vascular system. Different holes of a multiorifice catheter can be in different body compartments. Beck et al.  11obtained postoperative catheter epidurograms in 113 consecutive nonpregnant patients who underwent placement of multiorifice catheters. Fourteen catheters were located in two body compartments: 13 (12%) epidurovascular and 1 (1%) epidurointrathecal. The one epidurointrathecal and 3 of 13 epidurovascular catheters provided adequate anesthesia for surgery. Blood was never aspirated, no epidural anesthesia developed, and no symptoms of local anesthetic toxicity were elicited after administration of 112 mg bupivacaine and 100 mg etidocaine in 1 of 13 epidurovascular catheters. In the current study, nine catheters showed no signs of epidural, intrathecal, or intravascular placement. The results of Beck et al.  11suggest that some of these catheters may have been at least partially intravascular rather than in an innocuous but ineffective location.

Patients with unrecognized epidurovascular or epidurointrathecal catheters are at risk if a large local anesthetic dose is rapidly injected through a catheter previously used only for low-speed injection. Ward et al.  12reported a fatality during such conditions. In obstetric anesthesia, patients are at risk if a multiport catheter previously used only for continuous infusion labor analgesia is rapidly injected to provide anesthesia for an urgent cesarean delivery. The difficulty of testing multiport catheters underscores the need for careful local anesthetic dose fractionation, even if cesarean delivery is urgently needed.

Differential fluid flow through multiport orifices depends on catheter diameter and injection pressure. At low injection pressures, fluids preferentially exit the most proximal catheter orifice. Above a threshold pressure, flow is similar through all holes. This threshold pressure increases with increasing catheter diameter. Power and Thorburn 13reported a threshold local anesthetic injection pressure of 133 kilopascals (kPa) for 16-gauge and 93.1 kPa for 18-gauge catheters from the same manufacturer. Because injection pressure increases with injection speed, a test dose is more likely to exit via  all three ports the more rapidly it is injected.

Air may be an inappropriate multiport epidural catheter test dose. The minimum injection speed for flow through all three holes during air injection is fast (2–4 ml/s). 14Much slower, and clinically much more practical, injection speeds permit fluid to flow through all three catheter holes (0.006–0.1 ml/s). Even if fluid is the appropriate state of matter, however, the correct fluid test dose must be determined. The effectiveness of epinephrine, local anesthetic, and other fluid test doses has not been adequately tested with multiorifice catheters.

In conclusion, we found that aspiration and the air test detected 91 and 82%, respectively, of clinically evident, intravascularly placed, multiorifice epidural catheters. No false-negative air-test results were obtained in a previous study 14with single-orifice catheters. Air may be less reliable in multiport than in single-orifice catheters because gases, to a greater extent than liquids, preferentially exit the most proximal catheter hole. Our results underscore the importance of always fractionating the local anesthetic dose when injecting through multiorifice epidural catheters.