Secondary Logo

Journal Logo

Regional Anesthesia and Acute Pain Medicine: Research Report

Comparison of Changes in Tensile Strength in Three Different Flexible Epidural Catheters Under Various Conditions

Gonzalez Fiol, Antonio MD*; Horvath, Robert PhD, MBA; Schoenberg, Catherine BSN, CCRC*; Ahmed, Nubyra PhD; Dhar, Sunil Kumar PhD; Le, Vanny MD*

Author Information
doi: 10.1213/ANE.0000000000001305

Epidural catheter fracture or breakage is an infrequent,1 or at least, underreported event. The fragments of epidural catheters are considered to be inert,2 but in rare instances, they have resulted in nerve injuries.2,3 Therefore, the insight we acquire from the mechanical characteristics of epidural catheters, such as their tensile strength, is important to aid in the safe removal of entrapped catheters. Studies on tensile strength have been used to determine external factors that can affect epidural catheters.1,4,5 Some of these factors include the duration of the catheter within the epidural space4 and the use of hemostats for grasping the catheter.5 Factors that decrease the tensile strength of a catheter pose an increased risk of breakage upon catheter removal.4

Understanding some of the mechanical properties of epidural catheters, such as how the environment (i.e., temperature) could aid physicians when making decisions if faced with a trapped or difficult-to-remove catheter. Some authors have recommended changing the anatomic position of the patient, waiting 30 to 60 minutes, injecting normal saline (NS) through the catheter during removal, and withdrawing the inner metal coil from an amputated catheter.1,6 Studies on tensile strength have not been used to confirm the mechanical benefits, if any, from following these last 2 recommendations.

To measure the strength of the 3 types of reinforced catheters used at our institution, we elected to use the tensile strength test to evaluate several current recommendations for the removal of entrapped catheters in an ideal simulated environment. In the present study, we evaluated some of the current recommendations (e.g., injecting saline through the catheter and wire removal) and also whether a temperature of 37°C representing core body temperature would affect tensile strength. Our primary outcomes were comparing the tensile strength of the individual brands with their control after replicating 2 clinical recommendations and exposure to simulated core body temperature. The secondary outcomes were any differences in tensile strength among the 3 tested brands. We hypothesized that the current recommendations would result in an increase in tensile strength, thereby theoretically improving our ability to safely remove an entrapped catheter.


The tensile strength of 10 catheters from 3 brands of flexible epidural catheters (polyurethane, FlexTip Plus®, open tip catheter, uniport [Arrow International, Reading, PA]; Pebax nylon, Duraflex®, closed tip, multiport [Smiths Medical, St. Paul, MN]; polyamide nylon, Perifix® FX Springwound, closed tip, multiport [B-Braun Medical, Melsungen GA, Germany]) was recorded as the control group. These were then compared with the 3 experimental groups: catheters after wire removal, at 37°C (T37), and after injection of NS.

Tensile strength was measured by an independent laboratory in the Department of Materials Science and Engineering at Rutgers University. The reinforced epidural catheters were affixed to opposing, specially designed tensile test fixtures and then installed in an Applied Test System tensile test apparatus (Fig. 1). The fixtures were designed making use of a curved (rubber-coated) attachment point, thus minimizing any induced flaws associated with gripper type attachments. The tensile load was applied at a rate of 200 mm/min, consistent with guidelines from American Society for Testing and Materials (ASTM D638 “Tensile Testing of Plastics”).7 This procedure was then repeated using 10 epidural catheters from each brand in the experimental groups.

Figure 1.
Figure 1.:
The especially designed tensile test fixtures and a catheter installed in the Applied Test System tensile test apparatus.

In the experimental group of catheters after wire removal, the inner coil was exposed by trimming the end of the catheter by using a razor tool. The wire was then stretched and drawn carefully out from the catheter to avoid causing any damage to the inner surface of the catheter. Tensile strength was measured according to the protocol described previously.

To achieve a 37°C environment, a residential heater was modified to operate under the precision control of a digital temperature controller and readout. A thermocouple fine wire (type K), which has a similar mass as an epidural catheter, was used as the temperature sensor input into the temperature controller. This probe was highly indicative of the actual temperature of the catheters during testing without the need to physically contact the catheter and affect tensile test data. The test area was enclosed on 3 sides using Plexiglas to minimize temperature fluctuations. The epidural catheters were installed in the test fixture, brought up to 37 ± 1°C, and allowed to stabilize for 1 minute at that temperature before initiating the tensile testing. Throughout the testing, the catheter temperatures were monitored and controlled to be between 37 ± 1°C.

To measure the tensile strength of the last experimental group of NS, epidural catheters were saturated with NS at room temperature.

The saturation process consisted of the injection of NS through the catheter and a minimal soak time of 1 hour for all catheters. We added the soaking time to allow for capillary filling of saline into the catheters and to ensure that all catheters were completely saturated and tested under similar conditions.

Statistical Methods and Analysis

A power analysis was performed using the exact F test of a 1-way analysis of variance (ANOVA) procedure with fixed mean values and pooled sample variance from our pilot study (5 catheters from each brand). The pilot study was conducted comparing the tensile strength from 3 different brands with and without the inner coil (6 treatments). Estimates of group means were obtained from the pilot data, and it is the difference between these means that we sought to distinguish statistically. The results from the pilot study revealed the mean tensile strength (in kg) for the 3 brands with and without the wire as: 2.06, 1.96, 2.39, 2.316, 2.78, 3.01, with a pooled SD of 0.521; therefore, groups of 10 catheters per brand and experimental groups would be necessary to achieve a 99% power at a significance level of 0.05.

Our data were tested for normality and homogeneity of variance using the Shapiro-Wilk test and the Levene, Brown, Forsythe, and Bartlett tests, respectively (Supplemental Digital Content, Because the equality of variance was not met for the experimental T37 ANOVA model, a weighted ANOVA was performed for this group. Comparisons were made by 1-way ANOVA, using a 3 (number of brands) × 2 (control group versus experimental group) factor ANOVA, where each group had n = 10. This analysis was repeated for each group (catheters with wire versus after wire removal). Because the F test showed that there were significant differences between the groups, a post hoc Tukey test was performed to detect whether there were differences between the different brand catheters and the external variables tested (temperature, injection of NS, and presence or absence of inner metal coil). Statistical analysis was performed using SAS system 9.4 for Windows (SAS Institute Inc., Cary, NC). P values are expressed using the Tukey corrected results, and P < 0.05 was considered statistically significant.


We measured the tensile strength of 120 catheters from 3 different brands and materials (Table 1). The order in tensile strength for the control group expressed as the mean in kilograms (kg) was Arrow (2.85) > Smith (2.33) > B-Braun (2.17). When comparing within the control groups, statistical difference was noted between Arrow and B-Braun (P < 0.0001) and between Arrow and Smith (P = 0.0005). No statistical difference was noted between Smith and B-Braun (P = 0.39).

Table 1.
Table 1.:
Summary of Tensile Strength Results (n = 10 Catheters in Each Group)

When comparing the Arrow, B-Braun, and Smith control catheters with their counterpart after wire removal, no statistical significance was noted (P = 0.86, 1.0, and 1.0, respectively). The Arrow catheters after wire removal (3.02 kg) showed superior tensile strength when compared with B-Braun (2.17 kg, P < 0.0001) and Smith (2.33 kg, P = 0.0005) control groups and after their wire removal (2.18 kg and 2.35 kg; P < 0.0001 and P = 0.0008, respectively). The Arrow control group showed superior tensile strength when compared with the other 2 brands after wire removal (B-Braun, P < 0.0009, and Smith, P = 0.026; Table 2).

Table 2.
Table 2.:
Comparison Between Control Groups and Catheters After Wire Removal

Measuring the tensile strength at 37 ± 1°C, a decrease in tensile strength was noted only when comparing the B-Braun control group with its experimental T37 group (1.53 kg, P < 0.0001). Statistical difference also was noted between the Arrow control group, Arrow T37, Smith control group, and Smith T37 over the B-Braun T37 (P < 0.0001 for all comparisons in question; Table 3).

Table 3.
Table 3.:
Comparison Between Control Groups and Catheters at 37 ± 1°C
Table 4.
Table 4.:
Comparison Between Control Groups and Catheters After Normal Saline (NS) Injection

The NS study group showed a decrease in tensile strength for the catheters from B-Braun NS (1.58 kg) and Arrow NS (2.33 kg) when compared with their control groups (P = 0.0001 and P = 0.0010, respectively). The B-Braun NS group showed a statistically significant lower tensile strength when compared with the Arrow and Smith control groups (P < 0.0001 and P < 0.0001, respectively). The latter 2 brands NS groups (2.33 and 2.61 kg, respectively) showed superior tensile strength when compared with the B-Braun NS group (P < 0.0001 and P < 0.0001, respectively). In addition, the Smith NS group showed a statistically significant superiority over the B-Braun control group (P = 0.0060; Table 4).


Epidural catheter breakage or fragmentation is a rare but potentially serious complication of neuraxial anesthesia, despite the general notion that epidural catheters are inert.2,6 There is some evidence of morbidity related to retained catheters, mainly from case reports, suggesting that epidural catheters could potentially result in radicular pain2 or complicate spinal stenosis.3 For this reason, we think that it is important to understand how different epidurals behave under certain conditions to optimize the chances of removing catheters safely and intact.

Despite the tensile strength differences noted between the brands, all tested catheters exceeded the force required for epidural catheter removal (0.04–1.17 kg).1,4 The lowest tensile strength reported in our control group was 2.17 kg, which is almost twice the maximal force needed to remove a catheter. The Arrow brand was the only open-ended uniport catheter tested and the strongest of the tested catheters. It is possible that the manufacturing process for open-ended catheters confers some advantage over the multiorifice closed-tip catheters.

In some instances, the epidural catheter may fracture leaving the catheter and the wire within it exposed at the level of the skin. In such a scenario, the catheter should be removed, surgically if necessary. This type of fractured catheter would form an iatrogenic fistula between the epidural space and the skin, potentially exposing the patient to infections.2 Asai et al.1 described a similar case in which they removed the wire and then the catheter. We questioned whether removing the wire added any benefit in terms of tensile strength. In our study, it did not. Hence, there would be no advantage with this practice or technique. In a clinical scenario, the removal of the wire from a stretched catheter could further compromise its integrity. The only potential benefit that we see from removing the wire from the catheter is for avoidance of the theoretical risk of thermal injury or dislodgement6 secondary to the ferromagnetic component if magnetic resonance imaging is considered.

Clinically, the distal end of the catheter (average of 3–5 cm into the epidural space) would be almost exclusively exposed to the patient’s core temperature; hence, there is a possibility that this portion of the catheter might reach temperatures close to core (37°C). For this reason, we decided to evaluate the effect of temperature over the tensile strength. A decrease in tensile strength was noted only for B-Braun catheters when tested at 37 ± 1°C and compared with their control group. Our results are in accordance with the expected behavior of polymers at elevated temperatures.8 To the best of our understanding, only one previous study was designed to evaluate the tensile strength of catheters after exposure to 37°C.9 Their results did not reach statistical significance, perhaps because their study was underpowered. However, if the temperatures at the distal end of catheters reach temperatures close to core, their tensile strength could be compromised, at least for the B-Braun catheters.

Some authors have described the use of NS through the catheter to help remove an entrapped catheter, proposing that the injection of saline may increase the turgor of an entrapped epidural catheter, making its removal easier.6,10 Our study shows that there is no benefit, at least in terms of catheter strength, to injecting saline when faced with an entrapped catheter. Clinically speaking, we cannot oppose its use because there could be other benefits, such as lubrication or expansion of the space where the catheter is trapped, which may indeed facilitate its removal. Another potential benefit may be related to temperature. If the distal end of a catheter does come close to core temperature, the use of saline (at room temperature or colder) might bring the temperature closer to room temperature and improve the strength of the catheter.

The clinical significance of our results should be interpreted with caution. All tensile strength measurements in our study were performed using new catheters and under ideal conditions. However, clinicians should acknowledge that the integrity and tensile strength of catheters in the clinical setting may have been compromised because the mechanical properties of a catheter could be affected during placement by the Tuohy needle1 or by bony structures (i.e., osteophytes). This type of damage might result in a decrease in tensile strength to as low as 0.5 to 0.1 kg.4,5 Moreover, Kim et al.4 showed that epidural catheter strength decreases with use (new—2.04–2.55 kg to 1.73–1.94 kg—used). Contrary to our results, Asai et al.1 reported that Arrow catheters have a tensile strength lower than the maximal force needed to remove a catheter (982 ± 28 g). This marked difference might be attributed to the fact that they fixed their catheters with forceps. This methodology decreases the tensile strength of catheters.5,11

There are a few limitations to this study. There are differences in the composition of the brands studied, which make generalization and comparisons between brands difficult. In terms of design, there was a wide range of variability in the displacement between samples within each of the groups. These variations could be explained by slight differences in the starting length of the epidural catheter secondary to minor slack at the attachment to the test fixtures. Some of this variability is expected when testing polymers because of their ability to stretch. Another limitation is that we measured tensile strength by applying a tensile load at 200 mm/min, which might not clinically resemble the load exerted at the time of removal of an entrapped catheter. As mentioned previously, all our catheters were tested under ideal circumstances, that is, we were using new or undamaged catheters. In addition, a power analysis was only performed in our pilot study comparing the control groups with their respective after wire removal group. It is possible that more catheters would be needed to detect statistical differences within brands in the other 2 experimental groups.

In conclusion, the current recommendation of injecting saline through an entrapped catheter might further compromise the tensile strength of 2 of the tested brands (Arrow and B-Braun). The B-Braun catheters showed a decrease in tensile strength when tested at 37 ± 1°C. The Smith catheter showed no significant changes when tested under the 2 experimental conditions described previously. There seems to be no benefit, at least in terms of tensile strength, in removing the wire or inner coil from any of the tested epidural catheters. Further research is necessary to explore recommendations that may result in an increase in tensile strength and, hence, improve the safe removal of entrapped catheters.


Name: Antonio Gonzalez Fiol, MD.

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

Attestation: Antonio Gonzalez Fiol has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Name: Robert Horvath, PhD, MBA.

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

Attestation: Robert Horvath has reviewed the analysis of the data and approved the final manuscript.

Name: Catherine Schoenberg, BSN, CCRC.

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

Attestation: Catherine Schoenberg has reviewed and approved the final manuscript.

Name: Nubyra Ahmed, PhD.

Contribution: This author helped with statistical analysis.

Attestation: Nubyra Ahmed has reviewed the analysis of the data and approved the final manuscript.

Name: Sunil Kumar Dhar, PhD.

Contribution: This author helped with sample size and power analysis, arrive at the most appropriate statistical model, and write the manuscript.

Attestation: Sunil Kumar Dhar has reviewed the analysis of the data and approved the final manuscript.

Name: Vanny Le, MD.

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

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

This manuscript was handled by: Terese T. Horlocker, MD.


1. Asai T, Yamamoto K, Hirose T, Taguchi H, Shingu KBreakage of epidural catheters: a comparison of an arrow reinforced catheter and other nonreinforced catheters.Anesth Analg2001922468
2. Blanchard N, Clabeau JJ, Ossart M, Dekens J, Legars D, Tchaoussoff JRadicular pain due to a retained fragment of epidural catheter.Anesthesiology19978715679
3. Staats PS, Stinson MS, Lee RRLumbar stenosis complicating retained epidural catheter tip.Anesthesiology19958311158
4. Kim P, Meyer U, Schüpfer G, Rukwied R, Konrad C, Gerber HTensile strength decreases and perfusion pressure of 3-holed polyamide epidural catheters increases in long-term epidural infusion.Reg Anesth Pain Med2011361515
5. Nishio I, Sekiguchi M, Aoyama Y, Asano S, Ono ADecreased tensile strength of an epidural catheter during its removal by grasping with a hemostat.Anesth Analg2001932102
6. Mitra R, Fleischmann KManagement of the sheared epidural catheter: is surgical extraction really necessary?J Clin Anesth2007193104
7. ASTM Standards D638-14 A: Standard Test Method for Tensile Properties of Plastics2014West Conshohocken, PAASTM International
8. Vettegren VI, Kulik VB, Bronnikov SVTemperature dependence of the tensile strength of polymers and metals at elevated temperatures.Tech Phys Lett20053196972
9. Tsui BC, Finucane BTensile strength of 19- and 20-gauge arrow epidural catheters.Anesth Analg20039715246
10. Gadalla FRemoval of a tenacious epidural catheter.Anesth Analg19927510712
11. Hobaika ABBreakage of epidural catheters: etiology, prevention, and management.Rev Bras Anestesiol20085822733

Supplemental Digital Content

Copyright © 2016 International Anesthesia Research Society