The Epinephrine Test Dose in Obstetrics

Despite an apparent decline in the frequency of fatal systemic toxic reactions to local anesthetics since 1980 [1], this complication represents the most serious anesthetic risk to the parturient during epidural anesthesia. Many safety steps were advocated after the concern over bupivacaine toxicity in 1979, including the epinephrine test dose introduced by Moore and Batra [2]. They found that epinephrine 15 [micro sign]g IV created a tachycardia sufficient to serve as a reliable indicator of unintentional vascular placement of an epidural needle or catheter. Guinard et al. [3] later refined this test by showing that a heart rate change of 20 bpm was a sensitive and specific threshold to indicate intravascular injection. Although this test is limited in the presence of beta-adrenergic blockade, advanced age, and general anesthesia, it has remained a useful tool for the surgical anesthesiologist. Attempts to transfer this tool to the obstetrical suite have been less successful. Pain from ongoing uterine contractions results in maternal heart rate variations during labor, which confuse interpretation of a heart rate change after the injection of an epinephrine test dose. In addition, the pregnant patient has a different sensitivity to catecholamines, and the threshold for a positive heart rate response to epinephrine differs in obstetric patients. When Leighton et al. [4] applied surgical criteria (an increase in heart rate of 25 bpm within 2 min of injection) to laboring women who received epinephrine 15 [micro sign]g IV, only 5 of 10 responded with such a tachycardia. Of the 10 study patients, 9 had a positive response when an alternative (retrospectively derived) definition of a positive response (an increase in heart rate of >10 bpm over the maximal heart rate observed during the 2 min before injection) was applied. Colonna-Romano et al. [5] later confirmed that the threshold of 10 bpm is appropriate in the laboring patient, but only if attention is paid to the rate of increase in the heart rate, which must be distinguished from the tachycardia associated with contractions [6]. Finally, in addition to concerns about reliability, safety is a concern in the pregnant patient. Animal data reveal that the alpha-adrenergic effects of epinephrine can cause a reduction in uterine blood flow [7]. Further, either exaggerated hypertensive or tachycardiac effects of IV epinephrine are feared in patients with hypertension, preeclampsia, or cocaine intoxication. The lack of specificity, sensitivity (depending on the definition of positive heart rate response used), and potential risks of IV epinephrine as a test for IV epidural catheter placement has led many to avoid this test in parturients. In this issue of Anesthesia & Analgesia, Colonna-Romano and Nagaraj [8] present further data regarding the use of an epinephrine test dose in laboring women. Two hundred nine unmedicated laboring women participated in a study of the efficacy of 45 mg of lidocaine containing 15 [micro sign]g of epinephrine as a test dose for epidural catheter placement. All women were in active labor and received appropriate fluids IV before epidural catheter placement. The patients were monitored with a pulse oximeter, and heart rate data were displayed on a strip chart recorder. A single orifice epidural catheter was placed using the loss of resistance to saline technique, and aspirated. If aspiration was positive for blood or cerebrospinal fluid, the catheter was replaced. If the aspiration was negative, the patient received the test dose during uterine diastole, provided that her heart rate was stable. A positive maternal heart rate response, which suggests an IV injection of epinephrine, was defined as a sudden increase in maternal heart rate of more than 10 bpm above the baseline within 1 min of the administration of the test dose. The test dose was invalidated and repeated if the heart rate response was equivocal or if a uterine contraction occurred. If a positive response was recorded and catheter aspiration remained negative, the patient received an injection of 100 mg of lidocaine via the catheter. If subjective symptoms were present (tinnitus or metallic taste), the catheter was presumed to be IV. At first glance, the results are reassuring. First, there were no complications among the patients, and fetal heart rate monitoring revealed no worrisome changes. Of the 198 patients in whom the epinephrine test dose was judged to be valid, there were 176 true negatives (defined as negative catheter aspiration and a negative test dose followed by effective epidural analgesia from 6-12 mL of bupivacaine), 0 false negatives, 14 true positives (defined as a positive epinephrine test dose followed by either positive catheter aspiration or a positive response to 100 mg of lidocaine), and 8 false positives (defined as a positive epinephrine test dose following by a negative 100 mg of lidocaine test). The authors determined the epinephrine test dose in this population to be 100% sensitive and 96% specific, with a negative predictive value of 100% and a positive predictive value of 63%. These results support the safety of relying on the epinephrine test dose in obstetrics. That is, using the author's criteria, the test identified all IV catheters. However, to apply these results to clinical practice, the criteria used in this study must be carefully analyzed. The qualifiers used in this study must be clearly understood and underscored to prevent misinterpretation of the results. Although 209 patients participated in the study, the results of only 198 patients are presented. Why are the results from 5% of the patients not used in the final analysis? These patients are accounted for in the article, but their presence must be recognized. Five patients never received the epinephrine test dose because of safety concerns raised by the anesthesiologist (preeclampsia, maternal heart disease, use of propranolol, nonreassuring fetal heart rate pattern, and maternal tachycardia). Technical problems precluded the initial use of the epinephrine test dose in three patients (unavailability of the test dose, difficulties with reliable pulse oximeter monitoring, and frequent uterine contractions). In three additional patients, the epinephrine test dose was injected but was invalidated (and not repeated) because of frequent uterine contractions. Therefore, in clinical practice, 5% of patients may not be eligible for this test of IV epidural catheter placement. Thirteen patients required two epinephrine test doses: in 3, the first injection resulted in an equivocal result; in the other 10, a uterine contraction occurred immediately after the injection. Among these patients is one who did not have a tachycardic response with the initial test dose injection, had a uterine contraction (which invalidated the test dose), and then had a second test dose injected, which resulted in tachycardia (and catheter aspiration was positive for blood). If the strict criteria of this study (invalidating the results of the test dose if a uterine contraction immediately follows injection) had not been applied to the care of this patient, this would have been a false-negative result. Certainly, appropriate subsequent care with incremental injections of a local anesthetic would have probably eventually identified the misplacement of this catheter, but inappropriate care (such as bolus injection of anesthetic in preparation for cesarean delivery) could have been disastrous. There is also the issue that eight patients (4.5%) would have had their catheter removed unnecessarily because of a false-positive test. Although this is admittedly less a risk than a false-negative test, it leads to another way of viewing the data. Of the 209 patients, 32 did not benefit directly from the initial epinephrine test dose (8 with safety/technical concerns, 13 invalidated by a contraction, 3 with equivocal responses, and 8 false positives). Thus, it seems that only 85% of the subjects experienced this test as a simple and useful technique (although it is 100% sensitive). The results of this observational study rely heavily on unique monitoring practices. The authors had previously determined that the tachycardic response to IV epinephrine occurs within 1 min and has "an acceleratory phase of more than 1.2 bpm" [6]. Is the occasional obstetric anesthesiologist aware of this unique feature, and is he or she able to apply these criteria to a typical laboring patient? How many laboring patients have any formal monitor of heart rate applied during the administration of epidural anesthesia? Although the authors state that they cannot estimate how important evaluating a hard copy of maternal heart rate changes was to their success rate with the epinephrine test dose, it is worth underscoring this feature of the diagnostic criteria used to evaluate the efficacy of the test dose in this study. The authors also point out that there is no practical "gold standard" for identifying IV catheters. It would be unreasonable even to suggest that patients travel to the radiology suite for injection of contrast material! Thus, the test used in this study to confirm a positive epinephrine test dose (injection of 100 mg of lidocaine and observing the patient for tinnitus or metallic taste [9]) is subject to observer bias. The clinicians participating in this study were, by definition, aware of the epinephrine test dose results, and they administered the lidocaine test dose based on these results. A desire to prove the reliability of the epinephrine test dose could have easily led to the clinician suggesting the symptoms to the patient. This source of bias could have exaggerated the number of true positives. However, from a safety standpoint, this is much less of a concern than not recognizing a catheter that is truly IV. As mentioned before, this would lead to unnecessary removal of a properly placed epidural catheter. Two other outcomes from this study are worth noting. The most important is the lack of reliability of catheter aspiration for either blood or cerebrospinal fluid. When catheter aspiration is positive, it provides useful information; when catheter aspiration is negative, it means nothing. In addition, part of any evaluation of proper epidural catheter location is an assessment of possible intrathecal placement. Both catheter aspiration of cerebrospinal fluid and lower extremity strength (ability to maintain a straight-leg raise for 5 s at 4 min after the test dose injection was defined as a negative test result) were used in this study. Only one patient had an intrathecal catheter (determined from aspiration of cerebrospinal fluid). These data add to our criteria for assessing possible intrathecal catheter placement. This study contributes to our understanding of the proper use of the epinephrine test dose in laboring patients. It confirms the stance of many that there are limitations to this test for IV catheter placement in obstetric patients. This study supports the conclusion that the proper and safe use of the epinephrine test dose in obstetrics requires several modifications: 1) exclusion of patients at risk of complications from intravascular epinephrine injection, 2) lowering the threshold of suspicion to a 10-bpm increase (rather than the surgical level of 20 bpm), 3) intensive monitoring systems able to detect rapid changes in maternal heart rate, 4) careful timing to avoid uterine contractions, and 5) most importantly, considerable clinical judgment about the reliability of the test result. This leads to the obvious question: is there something better to use in this patient population? There might be. Injecting 1-2 mL of air through the catheter and listening for evidence of IV air with a Doppler placed over the maternal precordium has been suggested [10]. The obvious benefit of this test is the ability to use it in virtually all patients (although the theoretical concern of shunting of the air across an undiagnosed patent foramen ovale has been raised, we suggest that few are so careful with the administration of IV fluids to avoid inadvertent injection of a comparable amount of IV air in many patients). The drawback of the test is the need to have a suitably placed Doppler, an educated ear, and the absence of other IV air sources (such as the IV line). We suggest that these issues are no more complicated than the limitations of the epinephrine test dose in an obstetric patient. The authors of the current study have previously studied use of 100 mg of lidocaine as a test for IV catheter placement [9]. Why not only use this test? Truly, what is needed is a comparison of the reliability of the available tests. Given the absence of a practical gold standard, the design of such a study is challenging. What we can say is that we still do not have a single reliable, perfect test for detecting IV epidural catheters in laboring women. Like all clinical tests, substantial judgment must accompany the application and interpretation of the result of an epinephrine injection.