Secondary Logo

Journal Logo

The Learning Curve Associated with the Epidural Technique Using the Episure™ AutoDetect™ Versus Conventional Glass Syringe: An Open-Label, Randomized, Controlled, Crossover Trial of Experienced Anesthesiologists in Obstetric Patients

Carabuena, Jean M. MD; Mitani, Aya M. MPH; Liu, Xiaoxia MS; Kodali, Bhavani S. MD; Tsen, Lawrence C. MD

doi: 10.1213/ANE.0b013e31826c7cad
Obstetric Anesthesiology: Research Reports
Free
SDC

BACKGROUND: The Episure™ AutoDetect™ (spring-loaded) syringe has been observed to successfully identify the epidural space in 2 pilot studies. In this study we evaluated the impact of the spring-loaded syringe on the establishment of successful epidural labor analgesia (primary outcome), elapsed time for catheter placement, and learning curve (cumulative summary analysis, i.e., Cusum) of experienced anesthesiologists.

METHODS: Fourteen attending and fellow anesthesiologists were randomized to perform 50 consecutive epidural technique attempts using a spring-loaded or conventional glass syringe. Ten participants completed an additional 50 attempts with the alternate syringe in a crossover design.

RESULTS: A total of 1200 epidural placement attempts were performed. Use of the spring-loaded syringe was associated with a nonsignificant difference of estimated success rate in obtaining analgesia success (absolute difference of 1.0% 95% confidence interval, CI: −8.9% to 10.8%), shorter elapsed mean time to epidural catheter placement (ratio of 0.92 95% CI, 0.89–0.96); P = 0.003) and similar Cusum curves when compared with a conventional glass syringe. Analgesia success was more common with attending versus fellow anesthesiologists (absolute difference of 34.6% 95% CI, 14.9% to 54.3%; P < 0.001), and when the initial preferred technique was loss-of-resistance to continuous saline versus intermittent air (absolute difference of 33.8% 95% CI, 12.6% to 55.0%; P < 0.001). Shorter elapsed mean times were also observed in the group exposed to the spring-loaded syringe first (ratio of 0.65 95% CI, 0.62–0.67; P = 0.02).

CONCLUSIONS: When used by experienced obstetric anesthesiologists, the spring-loaded syringe was associated with a similar overall rate for establishing successful epidural labor analgesia, a shorter elapsed time to epidural catheter insertion, particularly when the anesthesiologist was randomized to use the novel syringe first, and a similar Cusum curve when compared with a conventional glass syringe. Attending versus fellow anesthesiologists and an initial technique preference for loss-of-resistance to continuous saline were associated with greater analgesia success with the novel syringe.

Published ahead of print December 7, 2012 Supplemental Digital Content is available in the text.

From the Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA.

Accepted for publication July 12, 2012.

Published ahead of print December 7, 2012

Funding: Departmental funds.

Conflicts of Interest: See Disclosures at the end of the article.

Reprints will not be available from the authors.

Address correspondence to Lawrence C. Tsen, MD, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis Street, Boston, MA 02115. Address e-mail to ltsen@zeus.bwh.harvard.edu.

The obstetric ward is a primary environment where anesthesiologists learn and refine the epidural analgesia/anesthesia technique.1 The intensity and progression of obstetric pain, coupled with the potential for patient movement and the consequences of technique failure, make rapid and successful epidural catheter placement a valid concern, even among experienced anesthesiologists.

The Episure™ AutoDetect™ Syringe is a spring-loaded “loss-of-resistance” syringe that allows for 2-handed advancement of the epidural needle and visual confirmation that the epidural space has been identified.2 In an observational pilot study, the use of the device was reported to confer increased success in epidural space identification, primarily with resident anesthesiologists.3 However, how quickly consistent, successful results (e.g., the “learning curve”) can be obtained among anesthesia care providers using the spring-loaded syringe, particularly among those who are already proficient with another technique, has not been evaluated.

Few studies have examined the learning curves associated with manual anesthesia techniques4–8; even fewer have specifically assessed the adoption of a novel epidural technique by experienced anesthesiologists.9 The purpose of this study was to evaluate the use of the Episure™ AutoDetect™ Syringe by experienced anesthesiologists, with no prior experience with the device, in laboring obstetric patients in a randomized, controlled, crossover trial. We hypothesized that the spring-loaded syringe, in comparison with a conventional glass syringe, would result in a similar overall success rate for providing satisfactory epidural labor analgesia, but with a reduction in elapsed time from epidural needle insertion to removal after catheter placement. Moreover, we hypothesized that a cumulative summary (Cusum) curve indicating an 80% rate of analgesia success would be similar between the 2 syringes.

Back to Top | Article Outline

METHODS

After IRB/human research committee approval, written informed consent was obtained from attending and fellow obstetric anesthesiologists. All anesthesiologists completed a brief questionnaire about their epidural technique experience and preferences. Anesthesiologists who had performed a minimum of 200 epidural catheter placements in obstetric patients were invited to participate in the study. Change in epidural technique was defined as when the anesthesiologist’s standard technique preference was not a loss-of-resistance approach to continuous pressure with saline. The anesthesiologists were randomized by a computer-generated random numbers table to perform 50 consecutive epidural technique attempts with either the 5-mL Episure™ AutoDetect™ syringe (Indigo-Orb, Irvine, CA) or a 5-mL Luer-Slip conventional glass syringe (Spectra Medical, Wilmington MA) using a loss-of-resistance approach to continuous pressure with 3 mL sterile saline. Anesthesiologists randomized to the spring-loaded syringe had 1 training session to examine the syringe and view a training video produced by the manufacturer. No practice attempts with the spring-loaded syringe on patients or with inert materials were allowed; the first 50 consecutive uses of the syringe were included in the study.

The epidural space was identified with the assigned syringe in laboring parturients in the sitting position using a 17-gauge Tuohy–Weiss needle with a midline approach. After placement of a 19-gauge epidural catheter with a single, open-end hole (Arrow FlexTip Plus®, Arrow International, Reading, PA) 5 cm into the epidural space, 20 mL of bupivacaine 1.25 mg/mL with fentanyl 2μg/mL was administered over 5 minutes through the catheter in 3-mL divided doses.

An epidural attempt was defined as the insertion and removal of the epidural needle from the selected skin insertion site, regardless of whether the insertion of the epidural catheter was successful. Each attempt was therefore further classified as to whether the catheter was successfully inserted (catheter success) and produced labor analgesia (analgesia success). Analgesia success was defined as analgesia that resulted in a verbal analog pain score (VAPS) <3 (0 = no pain and 10 = worst pain imaginable) within 30 minutes of the conclusion of epidural catheter dosing. All epidural attempts occurred sequentially, with each attempt timed by a study observer using a stopwatch laser tuned for 0.01 second accuracy (220 Sport Timer, Sportline, Yonkers, NY). The “elapsed time” for each epidural attempt started when the syringe contacted the hub of the epidural needle and stopped when the tip of the epidural needle was withdrawn from the skin.

The study observer recorded patient demographic characteristics, number of epidural attempts, catheter success, elapsed time, VAPS, and any related complications (i.e., intrathecal or intravascular needle or catheter, or need for catheter replacement). The anesthesiologist performing the epidural attempt graded the clarity of the loss-of-resistance (0 = unable, 1 = equivocal, 2 = clear) and ease of epidural catheter threading (0 = unable, 1 = needle adjusted to thread, 2 = easy). After each spring-loaded syringe use, the anesthesiologist performing the epidural attempt recorded ease of syringe use (0 = difficult, 1 = equivocal, 2 = easy), the degree of satisfaction if the syringe were to be the only one available (0 = no, 1 = equivocal, 2 = yes), and whether the syringe would be a useful learning tool (0 = no, 1 = equivocal, 2 = yes).

A subset of attending and fellow anesthesiologists, determined by those who successfully completed the first 50 consecutive attempts within 7 months, gave written informed consent to complete an additional 50 epidural attempts with the alternative syringe (i.e., crossover design). The same variables were identified and compared; in addition, the effect of the exposure order to the spring-loaded syringe versus conventional syringe was evaluated. The sequence of syringe exposure was either AC (spring-loaded syringe then conventional syringe) or CA (conventional syringe then spring-loaded syringe).

The primary outcome of the study was the ability to achieve analgesia success. The number of epidural attempts (n = 50) needed to achieve analgesia success was determined by assuming a 20% failure rate for lumbar epidural catheter placement at the interspace first chosen by attending anesthesiologists, as observed by de Oliveira Filho.6 On the basis of the report, we determined that a sample size of 28 procedures would be sufficient to detect an effect size of 0.8, using a 2-sided z test of a binomial proportion with a 0.1 significance level and 90% power (SAS 9.2, Cary, NC; Proc Power). The number of epidural attempts was increased to evaluate Cusum curves associated with a novel device. Schuepfer et al., in a group of staff and resident anesthesiologists attempting a novel (caudal) technique for the first time, calculated Cusum curves using a least-square fit model and 95% confidence intervals using a bootstrapping technique to mimic a large statistical population.9 They determined that the average number of attempts to reach a mean success rate of 80% was 32 procedures.9 Grau et al., in a group of residents performing their first epidural techniques with ultrasound guidance, noted a 90% success rate after 20 attempts, but maintenance of this success rate after 45 procedures.7 We consequently selected a sample size of 50 attempts to encompass these aforementioned procedure numbers.

Data from all epidural attempts, including those associated with the subset of participants who completed an additional 50 epidural attempts, were combined for most analyses as the use of the spring-loaded syringe was novel to all participants. However, a separate analysis was performed to determine whether the sequence of syringe exposure (i.e., CA or AC sequence) was of importance.

Analgesia success, recorded as success or failure, was analyzed using generalized estimating equation (GEE) models with the logit link function to account for the binomial distribution of the response variable and for the within-subject correlation (PROC GENMOD, SAS 9.2, SAS Institute, Cary, NC). Univariate analyses were conducted to estimate the effects of syringe type (spring-loaded versus conventional), sequence of syringe exposure (AC versus CA), experience/training (attending versus fellow), and preferred technique (continuous saline versus intermittent air) on analgesia success, adjusted for number of attempts made. The fully adjusted model included all the aforementioned variables. The elapsed time to placement was natural log-transformed to account for the log-normal distribution of time to event with skewness of 3.25 and Kolmogorov–Smirnov goodness of test statistic of 0.20 (P < 0.01; see the Appendix). The new response variable, log of elapsed time to placement, is distributed with mean of 3.79 and SD of 0.53.

The difference in elapsed log time between the 2 syringes was analyzed by generating a linear mixed-effects model with a random subject effect (PROC MIXED). After comparing the Akaike Information Criterion (AIC) among AR,1 exponential, Toeplitz, and unstructured correlation structures, the compound symmetry structure was chosen because of the smallest AIC. Univariate analyses were conducted for each independent variable adjusted only for the number of epidural attempts. In the full model, estimates of the elapsed time for epidural catheter placement were adjusted for syringe type, sequence of syringe exposure, preferred technique, provider (attending or fellow), analgesia success, and number of epidural attempts.

The predicted means of analgesia success and their 95% confidence intervals were estimated using the LSMEANS statement from univariate and full-model analyses and converted into probabilities using the inverse-logit. The estimated means and 95% confidence intervals of elapsed time were calculated using generalized pivotal methods10 (see the Appendix) under univariate and full-model analyses, respectively. The differences in predicted means of elapsed time between 2 groups were presented as a ratio of estimated elapsed mean time. The 95% confidence intervals of such ratios were also computed11 (see the Appendix).

The ability to achieve loss-of-resistance and the ease of threading the epidural catheter were analyzed separately. Cochran–Armitage Trend Test (PROC FREQ) was used to provide trend analysis between the 2 syringe types. When the anesthesiologist answered “not applicable (NA),” because of the inability to locate the epidural space, the ease of threading catheter was removed from the analysis. A P value of <0.05 was considered statistically significant.

Learning curves were constructed using the cumulative summation (Cusum) technique, which is a method of sequential analyses that represents successive epidural attempt failures or successes as positive or negative increments to a cumulative score.12 The acceptable (p0) and unacceptable (p1) failure rates for each epidural attempt in producing analgesia success were set at 20% and 40%, respectively, with α and β errors of 0.16,12,13; these parameters were established through epidural attempts by attending anesthesiologists.6 These values represented the 20% failure rate identified for the sample size analysis to allow the calculation and plotting of acceptable (h1) and unacceptable (h0) boundary limits and the interval s, by which the Cusum value decreases by s for success and increases by 1-s for failure.12,13 Acceptable performance is denoted by a Cusum line, which is horizontal or down-sloping.12 Therefore, a Cusum line that crosses the unacceptable (h0; i.e., upper) line indicates performance that is worse than the 20% failure rate. By contrast, a Cusum value that crosses the acceptable (h1; i.e., lower) line indicates a performance that is better than the 20% failure rate.

Back to Top | Article Outline

RESULTS

The study was conducted between January 2007 and November 2008. Fourteen (8 attending and 6 fellow) obstetric anesthesiologists confirmed that they had performed a minimum of 200 obstetric epidural techniques before enrollment, and all rated themselves as “very comfortable” (on a scale of not comfortable, comfortable, very comfortable) performing the epidural technique. Anesthesiologist rank (attending or fellow), baseline technique preferences (intermittent versus continuous pressure, air versus saline loss-of-resistance), sequence of syringe exposure, and participation in the crossover portion were recorded (Table 1). Participation in the study involved an initial change in “preferred technique” to a continuous saline loss-of-resistance technique in 7 of the 14 (50%) anesthesiologists, and 6 of the 10 (60%) anesthesiologists who participated in the crossover portion of the study.

Table 1

Table 1

A total of 1200 epidural placement attempts occurred in 1018 women. The mean (SD) height and weight for the entire patient population was 163.6 (6.9) cm, 80.3 (14.2) kg, respectively, and did not differ between groups (P = 0.93 and P = 0.47, respectively); however, the mean age of parturients who underwent epidural attempts with the spring-loaded syringe was older than that of those who underwent conventional syringe attempts (31.5 ± 5.4 years vs 30.9 ± 5.6 years, respectively; P = 0.04). Eight attending and 6 fellow obstetric anesthesiologists completed 700 epidural attempts (350 with the spring-loaded syringe, 350 with the conventional syringe) in the initial study. A subgroup of 7 attending and 3 fellow obstetric anesthesiologists completed an additional 500 attempts with the syringe to which they were not initially randomized (250 with the spring-loaded syringe, 250 with the conventional syringe) as participants in the crossover portion of the study. Of the 1200 epidural placement attempts, 87.8% obtained analgesia success at the initial needle insertion site (Fig. 1). Ultimately, all women received an epidural catheter, with 99% obtaining analgesia success. The initial and 30-minute mean (SD) VAPS scores were 8.4 (1.6) and 1.1 (0.5), respectively, for the spring-loaded syringe, and 8.4 (1.5) and 1.2 (1.6), respectively, for the conventional syringes; a 30-minute VAPS score <3 was not obtained in 11 women (5 and 6 women with the Episure™ AutoDetect™ and conventional syringes, respectively; Fig. 1). The estimated adjusted analgesia success did not differ by syringe type or order of syringe exposure; however, the adjusted analgesia success was higher for attending versus fellow anesthesiologists and in anesthesiologists with an initial preference for a loss-of-resistance to continuous saline technique (Table 2).

Table 2

Table 2

Figure 1

Figure 1

Regardless of syringe type, all epidural analgesia attempts, including those performed in the crossover group, had more rapid elapsed times from the first to the last attempt (P = 0.002; Fig. 2). The spring-loaded syringe had more rapid times than did the conventional syringe for epidural attempts (P = 0.003), including when the exposure sequence involved the spring-loaded syringe first (P = 0.02; Table 3). These findings remained significant when adjusted for anesthesiologist rank, preferred technique, and analgesia success in a multivariate model (Table 3). Because the correlation of repeated measurements in response variables was relatively weak when fitting the mixed model for elapsed time, a regular log-linear model was adopted so that the method by Tian and Wu14 could be applied to calculate the confidence intervals (Table 3).

Table 3

Table 3

Figure 2

Figure 2

Cusum curves indicated that the unacceptable failure rate (upper boundary limit; h1) was crossed by 2 individuals, one with the spring-loaded syringe only, and another with both syringes (Figs. 3 and 4); these crossings occurred on attempt 31 with the spring-loaded syringe, and attempts 16 and 25 with the conventional syringe. Both of these individuals were attending anesthesiologists whose preferred technique was an intermittent loss-of-resistance to air.

Figure 3

Figure 3

Figure 4

Figure 4

There were no significant differences between the syringes in the epidural attempts that did not result in analgesia success. There were 15 cases of equivocal loss-of-resistance in which the catheter threaded easily, 7 and 8 with the spring-coiled and conventional syringes, respectively. Unintentional dural punctures occurred with 3 and 2 attempts with the spring-loaded syringe and conventional syringes, respectively; in 4 of these cases, an epidural catheter was placed intrathecally and successfully used to provide labor analgesia. The unintentional dural punctures were made by 2 attending anesthesiologists (3 and 1 dural punctures, respectively) and 1 fellow anesthesiologist (1 dural puncture); these anesthesiologists were not the individuals who crossed the unacceptable (upper) failure line on the Cusum curves.

In the majority of epidural analgesia attempts, there was no difficulty in achieving loss-of-resistance or obtaining catheter success (Table 4); however, the spring-loaded syringe was less successful in achieving a clear loss-of-resistance (P = 0.015). Users of the spring-loaded syringe rated the syringe as “easy,” “equivocal,” or “difficult” to use in 82.7%, 17.2%, and 0.2% of epidural attempts, respectively. However, the majority (55.5%) of spring-loaded syringe users indicated that they would be dissatisfied if it were the only available syringe, and 55.7% were equivocal as to its value as a learning tool.

Table 4

Table 4

Back to Top | Article Outline

Crossover Study Analysis

Cusum curves of the individuals who participated in the crossover study, regardless of syringe type, identified 2 individuals who crossed the unacceptable boundary limit (h1; Fig. 5). All remaining crossover study participants were able to remain within the study boundary limits, with the majority performing better (crossing below) than the acceptable boundary limit (h0).

Figure 5

Figure 5

Back to Top | Article Outline

DISCUSSION

The principal findings of this study with experienced obstetric anesthesiologists are that the spring-loaded syringe is associated with a similar incidence of successful epidural labor analgesia, more rapid elapsed time from needle insertion to catheter placement, and similar Cusum curves when compared with a conventional glass syringe. Analgesia success was greater for attending versus fellow rank, and an initial preferred loss-of-resistance technique for continuous saline versus intermittent air.

The introduction of a new technique is commonly associated with a learning period characterized by increased failures or complications and a longer duration of time necessary to perform the technique.15 This effect was partially observed with the spring-loaded syringe in our study, with a lower incidence of “clear” loss-of-resistance during needle insertion being observed when compared with a conventional syringe; however, the analgesia success and incidence of dural punctures were not different. Although our results contrasted with an observational pilot study with the same syringes by Habib et al.,3 which indicated greater analgesia success with fewer inadvertent dural punctures with a spring-loaded syringe, these differences may be attributable to the predominant (90%) resident population being evaluated in their study. This may serve to confirm the clinical impression that experienced practitioners have high rates of analgesia success and low rates of inadvertent dural punctures despite alterations in equipment.

Cusum curves indicated a rapid learning curve and continued overall improvement in achieving analgesia success with subsequent placements for the majority of participants with both syringes. Although the improvement was more rapidly and consistently observed with the conventional glass syringe, which may indicate the comfort practitioners have with their traditional equipment used in routine practice, adoption of the novel syringe was not associated with an extended Cusum curve. However, 2 attending anesthesiologists, both of whom had previously used an intermittent loss-of-resistance to air technique, had unacceptable analgesia success rates (i.e., crossed the unacceptable boundary limit (h1)) with the conventional syringe; one of these individuals also displayed a similar outcome with the spring-loaded syringe. The failures are consistent with the report by Kestin,13 in which a change in loss-of-resistance technique from air to saline was directly attributed to being responsible for decreasing success in performing the epidural technique. Failures may also highlight the traditional role of the attending anesthesiologist as an observer with fewer opportunities for hands-on experience; this merits further study. Finally, had less rigorous definitions for an epidural attempt and analgesia success been adopted,5,9 it is possible that these failures may not have been observed.

The elapsed time for an epidural attempt was reduced with the spring-loaded syringe. This finding is similar to that of Habib et al.,3 who noted that the median elapsed time with the spring-loaded syringe versus a conventional glass syringe was 20 (range, 11 to 28) and 40 (range, 25 to 58) seconds, respectively (P < 0.001). In comparison, the mean elapsed times reported in our study were longer; this difference likely reflects the different end time in the 2 studies (identification of the epidural space versus identification, catheter placement and needle withdrawal) as well as how the results were analyzed (median versus mean). Greater experience level (attending versus fellow) and exposure to the syringes over the course of the study resulted in a reduction in elapsed time with both syringes.

The reduction in elapsed times for an epidural attempt with the spring-loaded syringe is of interest. A loss of haptic (sense of touch) sensation is traditionally associated with significantly slower performance times, for example, with robotic versus manual surgical procedures.16,17 However, it is possible that epidural space identification may be similar to certain procedures in which vision dominates over haptic stimuli.18 Habib et al.3 speculated that the spring-loaded syringe resulted in less variation in the pressure applied to the syringe, resulting in less subjectivity in the loss-of-resistance; this would be consistent with the faster performance observed with robotic-assisted surgery, versus conventional endoscopy surgery, due to the automation of certain technical elements.16 Riley and Carvalho2 also observed that the spring-loaded syringe presented the advantage of having both hands able to “advance and steady” the epidural needle. Both of these reasons may be relevant.

There are several limitations to our study. First, despite randomization, it is possible that faster or more technically proficient participants were randomized to 1 group; indeed, more rapid elapsed times with both syringes were observed in the AC sequence group. However, regardless of sequence of syringe exposure, all participants in the crossover study experienced more rapid elapsed times with the spring-loaded syringe in comparison with their epidural attempts performed with the conventional syringe. Second, the elapsed time and analgesic success for each epidural attempt may have been manipulated either by pulling the epidural needle from the skin to reduce elapsed time or by performing repeated passes within a selected insertion site to improve analgesia success. Although these competing actions may have affected our outcomes, the decision to abandon or persist with an epidural attempt reflects common clinical practice. Third, our study was intentionally limited to experienced obstetric anesthesiologists at a busy obstetric institution; thus, our results may not be applicable to all providers or practices. Finally, whether the reduction in elapsed time (<5 seconds) with the spring-loaded syringe versus conventional syringe is a meaningful clinical difference is debatable, particularly given the total time from patient positioning to analgesia success.

We conclude that with experienced obstetric anesthesiologists, the spring-loaded syringe is associated with a similar success rate for establishing successful epidural labor analgesia, a more rapid elapsed time to epidural catheter insertion, and a similar Cusum curve when compared with a conventional glass syringe. Analgesia success with the new syringe was greater for attending than for fellow anesthesiologists and an initial technique preference for loss-of-resistance to continuous saline versus intermittent air.

Back to Top | Article Outline

APPENDIX: COMPUTATION OF THE ESTIMATED MEAN OF ELAPSED TIME AND RATIO OF THE TWO MEANS

Our response variable, elapsed time to placement (X) is log-normally distributed (see Figs. 6 and 7: Web supplement), such that log(X) ~ N(μ, σ2), where log() is the natural logarithm function. Then the mean value of X depends on both μ and σ:

Figure 6

Figure 6

Figure 7

Figure 7

For example, in Table 3, the unadjusted means (standard deviations) of log-transformed elapsed time to placement for the spring-loaded group and conventional group were 3.751 (0.5318) and 3.833 (0.5318), respectively. Substituting these numbers in formula,1 we obtain the means of elapsed time to placement (in seconds): 49.05 and 53.22, respectively. The ratio of these means is 0.92.

Next, we describe the algorithm to obtain the confidence interval for mean of elapsed time and confidence interval for the ratio of 2 means.

Because the correlation of repeated measurements in response variable is relatively weak when fitting the mixed model for elapsed time, we have adopted a regular log-linear model for elapsed time and applied the method by Tian and Wu (2007)18 to calculate the generalized confidence intervals for the elapsed time of epidural attempts. Using Monte Carlo simulation, we obtained 1000 generalized pivotal quantities for the mean response of elapsed time at specific covariate values. The corresponding 95% confidence interval was then obtained by using the 5th and 95th percentiles of 1000 generalized pivotal quantities as the lower and upper bounds of the confidence interval. Ideally, we would have liked to apply a method similar to that of Tian and Wu (2007)18 that could calculate the confidence interval of log normally distributed response variable but also accommodate the mixed effects. However, such a method is not currently available.

To obtain the 95% confidence limits for this ratio, we used the algorithm developed by Ledolter, Dexter, and Epstein.10 For example, to obtain the ratio of the means of elapsed times to placement for spring-loaded group X1 and for conventional group X2, we used the following steps:

For i = 1 to 2 (1 represents spring-loaded group and 2 represents conventional group)

For j = 1 to 1000

Generate

and

.

where

and si are the sample means and sample standard deviations of the log-transformed elapsed times to placement for spring-loaded group or for conventional group.

Next j

Next i

Then, the 5th and 95th percentiles of

for j = 1 to 1000 are the lower and upper bounds of generalized confidence interval for the ratio of the means.

Back to Top | Article Outline

DISCLOSURES

Name: Jean M. Carabuena, MD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Jean M. Carabuena has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: This author consulted for Indigo-Orb, Inc., in 2006 and received unexercised stock options at that time, and their value is unknown.

Name: Aya M. Mitani, MPH.

Contribution: This author helped design the study, analyze the data, and write the manuscript.

Attestation: Aya M. Mitani has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: This author has no conflict of interest to declare.

Name: Xiaoxia Liu, MS.

Contribution: This author helped analyze the data and write the manuscript.

Attestation: Xiaoxia Liu has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: This author has no conflict of interest to declare.

Name: Bhavani S. Kodali, MD.

Contribution: This author helped design the study and conduct the study.

Attestation: Bhavani S. Kodali has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: This author has no conflict of interest to declare.

Name: Lawrence C. Tsen, MD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Lawrence C. Tsen has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Conflicts of Interest: This author consulted for Indigo-Orb, Inc., in 2006 and received unexercised stock options at that time, and their value is unknown.

This manuscript was handled by: Cynthia A. Wong, MD.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The authors would like to thank the attending anesthesiologist members of the Brigham and Women’s Obstetric Anesthesia Study Group who participated in this study (Eric Cappiello, Miriam Harnett, David Hepner, McCallum Hoyt, and Khadija Khan), the fellows, and our obstetric nursing and physician colleagues.

Back to Top | Article Outline

REFERENCES

1. Kopacz DJ, Neal JM. Regional anesthesia and pain medicine: residency training–the year 2000. Reg Anesth Pain Med. 2002;27:9–14
2. Riley ET, Carvalho B. The Episure syringe: a novel loss of resistance syringe for locating the epidural space. Anesth Analg. 2007;105:1164–6
3. Habib AS, George RB, Allen TK, Olufolabi AJ. A pilot study to compare the Episure Autodetect syringe with the glass syringe for identification of the epidural space in parturients. Anesth Analg. 2008;106:541–3
4. Kopacz DJ, Neal JM, Pollock JE. The regional anesthesia “learning curve.” What is the minimum number of epidural and spinal blocks to reach consistency? Reg Anesth. 1996;21:182–90
5. Konrad C, Schupfer G, Wietlisbach M, Gerber H. Learning manual skills in anesthesiology: is there a recommended number of cases for anesthetic procedures? Anesth Analg. 1998;86:635–9
6. de Oliveira Filho GR. The construction of learning curves for basic skills in anesthetic procedures: an application for the cumulative sum method. Anesth Analg. 2002;95:411–6
7. Grau T, Bartusseck E, Conradi R, Martin E, Motsch J. Ultrasound imaging improves learning curves in obstetric epidural anesthesia: a preliminary study. Can J Anaesth. 2003;50:1047–50
8. Sites BD, Gallagher JD, Cravero J, Lundberg J, Blike G. The learning curve associated with a simulated ultrasound-guided interventional task by inexperienced anesthesia residents. Reg Anesth Pain Med. 2004;29:544–8
9. Schuepfer G, Konrad C, Schmeck J, Poortmans G, Staffelbach B, Johr M. Generating a learning curve for pediatric caudal epidural blocks: an empirical evaluation of technical skills in novice and experienced anesthetists. Reg Anesth Pain Med. 2000;25:385–8
10. Ledolter J, Dexter F, Epstein RH. Analysis of variance of communication latencies in anesthesia: comparing means of multiple log-normal distributions. Anesth Analg. 2011;113:888–96
11. Krishnamoorthy K, Mathew T. Inferences on the means of lognormal distributions using generalized p-values and generalized confidence intervals. J Stat Plan Infer. 2003;115:103–21
12. Bolsin S, Colson M. The use of the Cusum technique in the assessment of trainee competence in new procedures. Int J Qual Health Care. 2000;12:433–8
13. Kestin IG. A statistical approach to measuring the competence of anaesthetic trainees at practical procedures. Br J Anaesth. 1995;75:805–9
14. Tian L, Wu J. Inferences on the mean response in a log-regression model: the generalized variable approach. Stat Med. 2007;26:5180–8
15. Bull C, Yates R, Sarkar D, Deanfield J, de Leval M. Scientific, ethical, and logistical considerations in introducing a new operation: a retrospective cohort study from paediatric cardiac surgery. BMJ. 2000;320:1168–73
16. Maniar HS, Council ML, Prasad SM, Prasad SM, Chu C, Damiano RJ Jr. Comparison of skill training with robotic systems and traditional endoscopy: implications on training and adoption. J Surg Res. 2005;125:23–9
17. GarciaRuiz A, Gagner M, Miller JH, Steiner CP, Hahn JF. Manual vs robotically assisted laparoscopic surgery in the performance of basic manipulation and suturing tasks. Arch Surg. 1998;133:957–61
18. Hecht D, Reiner M. Sensory dominance in combinations of audio, visual and haptic stimuli. Exp Brain Res. 2009;193:307–14
© 2013 International Anesthesia Research Society