Abstract
OBJECTIVE
To determine whether axial twisting within an ending loop negatively impacts maximum load to failure and failure mode of suture knots.
SAMPLES
525 knots (15 samples each of 7 different suture types/sizes tested in 5 knot-twist configurations each).
PROCEDURES
Each suture type (polydioxanone [PDO], Monoderm [polyglecaprone 25], and Nylon) and size (1, 0, 2-0, 3-0) were used to create a starting square knot, and each of the following ending square knot configurations: 0 twists, 1 twist, 4 twists, and 10 twists. Each suture was tested for failure using a universal testing machine (Instron, Instron Corp) with a 100 kg load cell at a speed of 100 mm/min. Each suture and knot was evaluated for a mode of failure using gross evaluation of the knots and video footage recorded during testing. Maximum load at failure (P-value set at .005) and failure mode (p-value set at 0.003) were recorded for each group.
RESULTS
Maximum load at failure was decreased in knots tied within ending loops containing more twists for some types and sizes of the suture. With 4 twists, 0-PDO, 1 PDO, and 2-0 Nylon was more likely to fail at the knot than knots with 0 twists. All sutures containing 10 twists, except 3-0 Monoderm, were more likely to fail at the knot than knots with 0 twists.
CLINICAL RELEVANCE
The number of twists within the ending loop may not increase the risk of failure at the knot; however, it can decrease the maximum load to failure at a knot, particularly as the suture size increases.
Monofilament absorbable and non-absorbable sutures are used to close various types of tissues inside and outside the body. There are various benefits of monofilament sutures over braided sutures, including less tissue drag, lower risk of infection, and lower knot tie-down resistance.1,2 However, they have relatively higher bending stiffness and memory, leading to difficulty in forming a stable knot.2–4
Normal passage of suture and movement of a surgeon’s wrist while suturing can lead to axial twisting longitudinally along the suture strand.5 Depending on the number of twists that occur, this can result in decreased tensile strength of the suture.5,6 A prior study determined that a 4- to 5-cm strand of 6-0 polypropylene suture can be axially twisted up to 4 times without losing tensile strength;5 however, when twisted 10 times or greater, as shown in a separate study using gauge 1 polydioxanone, polypropylene, or nylon suture, significant decreases in tensile strength can occur.6 Based on these findings, it is possible that twisting within an ending loop when tying off a continuous pattern could also affect the maximum load at failure and failure mode of various suture types and sizes.
Numerous studies have been performed to determine the knot security of different suture types and sizes under various conditions and knot configurations.7–9 These studies have shown that in various-sized sutures and knot configurations, maximum load at failure and knot security depend on suture material, tying technique, number of throws, and the location of the knot (starting knot vs ending knot), but was not dependent on size or the type of knot tied (surgeon’s vs square).7–9 To the best of the author’s knowledge, no study has been performed to test the security of the monofilament suture ending square knot where the ending loop has been twisted. The objective of this study is to determine the maximum load at failure and failure mode of ending square knots created from commonly used monofilament suture types and sizes when the ending loop has been twisted on itself and to compare these results with starting square knots and ending square knots without twists. We hypothesized that the maximum load at failure for knots with twists would be less than for those with no twists, but that failure mode would not differ.
Materials and Methods
Three suture types were provided by TruStitch Surgical Specialties Corporation and were chosen for this study because they are commonly used to close the body wall (2-0, 0, 1 polydioxanone [PDO]), the subcutaneous and dermal tissues (3-0 PDO, 2-0 and 3-0 (poliglecaprone 25 [Monoderm]), and the skin (2-0 and 3-0 Nylon) in small animal patients. For each suture type and size, 15 knots were created for each of 5 different knot variations: (1) starting square knots, (2) ending square knots with 0 twists within the ending loop, (3) ending square knots with 1 twist, (4) ending square knots with 4 twists, and (5) ending square knots with 10 twists (Figure 1).
Illustration of how sutures were tied around PVC pipe: (A) Ending loop with no twists before being tied. (B). Ending loop with single twist with arrow indicating how suture is manually twisted. (C) Ending loop with 4 twists with arrow indicating how suture is manually twisted. (D) The final product for each knot conformation.
Citation: American Journal of Veterinary Research 84, 7; 10.2460/ajvr.23.01.0003
All knots were created by a single individual (MAW) using a 3.175-cm-diameter, 30-cm-long PVC pipe with each end embedded into the center of a block of floral craft foam. The suture pattern was started by tying a 4-throw starting square knot, which was left on the apparatus. The suture was then wrapped around the testing apparatus twice, and a 5-throw ending square knot was performed by tying the free suture end to the second loop of the suture, as described by Upchurch et al.10 When twists were applied to the ending square knot, after the suture was wrapped around the apparatus, the loop was grasped by hand and tied (zero twists), or manually twisted 1 time, 4 times, or 10 times. When tying the knot, the ending loop was firmly grasped with the surgeon’s non-dominant hand (a finger secured within the loop) to maintain the twists while the needle drivers were wrapped once with the free end of the suture and then the knot was tied in a standard fashion. Each knot was performed using instrument ties without the use of a tension gauge. The suture was removed from the apparatus and the suture connecting the 2 knots was cut and tags were measured and cut to 3 mm, as previously recommended.11 Each of the ending knots, as well as the first 15 starting knots tied for each suture, were saved for testing.
Knot testing was performed using a universal testing machine (Instron, Instron Corp) with a 100 kg load cell. The universal testing machine was connected to a computer equipped with standard software (Bluehill Software Suite version 2; Illinois Tool Works, Inc) that recorded all data as each test was performed.12 The machine contained a single motor unit that supplied vertical displacement in a direction away from the stationary unit, as well as 2 horizontal, metal rods, measuring 4 cm long and 1 cm in diameter, with 1 attached to both the motor and to the stationary unit. The beams started at 3 cm apart and the suture samples were looped around each rod (Figure 2; Supplementary Video S1). The motor pulled the horizontal rods in opposite directions at a rate of 100 mm/min until failure of the sample occurred, as previously described.10 The maximum load at failure, defined as the force required for the suture to fail, was recorded and the mode of failure was recorded for each suture size and knot type. A video camera (Sony FDR-AX33, Sony) was used to record each test and to allow evaluation of each knot before testing.
The universal testing machine (Instron, Instron Corp) with the suture loop wrapped around the mobile beam (top) and the stationary beam (bottom) just before distraction.
Citation: American Journal of Veterinary Research 84, 7; 10.2460/ajvr.23.01.0003
After testing, all sutures were evaluated using video footage and gross evaluation of the suture and knot to determine where the suture had failed. Suture failure was classified into 4 categories: (1) unraveling of the knot; (2) failure/breakage of the knot; (3) knot intact, suture breakage occurring 1 mm from knot; and (4) knot intact, suture breakage occurring > 2-mm from the knot.
Statistical analysis
All statistical analysis was performed using standard software.13 Maximum load at failure was recorded for each suture and knot type. To test for normality, a Kolmogorov-Smirnov test was performed. None of the variables were normally distributed. A Kruskal-Wallis test was performed to analyze variables (significance set at P < .05). For significant variables, Dunn post hoc analysis was performed. Bonferroni correction was performed, and the significance value was set to P < .005. To assess the failure mode for each different suture and knot type, a chi-squared test was performed. For significant variables, a post hoc analysis was conducted by assessing adjusted residuals. Significance was set at P < .003 with Bonferroni correction. Sutures were assessed first within their distinct suture type and size, then later compared with other sutures of the same size and knot type within the dataset.
Results
Overall, there is no clear trend between any suture type, size, or number of twists and maximum load. Sutures tied with PDO tended to have a lower maximum load at failure when tied with more twists; however, this was not a consistent finding. Monoderm and Nylon had no clear trends in maximum load at failure based on the number of twists present. The average maximum load at failure (N) for each suture size and knot configuration is listed (Table 1). Within each data set, if no significant difference was noted in the maximum load at failure, the results were not discussed.
Evaluation of average maximum load at failure (N) and standard deviation for each suture size and knot configuration.
| Suture | Twists | Average maximum load at failure (N) | Standard deviation (±N) |
|---|---|---|---|
| 1 PDO | 0 | 118 | 13.9 |
| 1 | 123 | 7.8 | |
| 4 | 116 | 4.5 | |
| 10 | 113 | 19.9 | |
| 0 PDO | 0 | 93 | 3.9 |
| 1 | 78 | 24.5 | |
| 4 | 87 | 5.4 | |
| 10 | 87 | 5.8 | |
| 2-0 PDO | 0 | 71 | 7.3 |
| 1 | 70 | 4.5 | |
| 4 | 67 | 9.5 | |
| 10 | 63 | 3.9 | |
| 3-0 PDO | 0 | 51 | 4.4 |
| 1 | 49 | 4.3 | |
| 4 | 43 | 1.9 | |
| 10 | 45 | 2.9 | |
| 2-0 Monoderm | 0 | 90 | 9.9 |
| 1 | 93 | 8.1 | |
| 4 | 87 | 7.1 | |
| 10 | 84 | 6.6 | |
| 3-0 Monoderm | 0 | 61 | 4.5 |
| 1 | 60 | 4.1 | |
| 4 | 52 | 6.0 | |
| 10 | 57 | 5.1 | |
| 2-0 Nylon | 0 | 58 | 4.1 |
| 1 | 57 | 7.2 | |
| 4 | 58 | 2.6 | |
| 10 | 60 | 2.3 | |
| 3-0 Nylon | 0 | 40 | 5.2 |
| 1 | 36 | 3.4 | |
| 4 | 40 | 2.2 | |
| 10 | 38 | 3.5 |
PDO = polydioxanone.
Within the PDO groups (Figure 3), 3-0 PDO had a significantly lower maximum load at failure in the 4-twist group compared with the starting group (P < .001), the 0-twist group (P < .001), and the 1-twist group (P < .001). The maximum load at failure for the 10-twist group was significantly lower when compared with the starting group (P = .003) and the 0-twist group (P = .001. For 2-0 PDO, the maximum load at failure was significantly lower in the 10-twist group when compared with the starting (P < .001), the 0-twist (P = .002), and 1-twist (P = .003) groups. With 0 PDO, the maximum load at failure was significantly lower in the 1-twist (P < .001), 4-twist (P = .004), and 10-twist (P = .001) groups compared with the 0-twist group. Within 1 PDO, the maximum load at failure was significantly lower in the 4-twist group compared with the 1-twist group (P = .005). There were no other significant differences in maximum load at failure between knot types.
Comparison of maximum load at failure of all PDO sizes for each of the 5-knot variations: starting square knot, ending square knot with 0 twists within the ending loop, ending square knot with 1 twist within the ending loop, ending square knot with 4 twists within the ending loop, or ending square knot with 10 twists within the ending loop. Different superscripts represent significantly different maximum loads. The whiskers for each box plot depict the interquartile range for that group. The “x” within each box is the mean for that subgroup and the line through each box is the median value (second quartile), while the outer lines of the box indicate the first and third quartiles. The whiskers of the box plot indicate the minimum and maximum values within each group. PDO = polydioxanone.
Citation: American Journal of Veterinary Research 84, 7; 10.2460/ajvr.23.01.0003
For 3-0 Monoderm (Figure 4), the maximum load at failure was significantly lower in the 4-twist group, when compared with the starting (P = .004), 0-twist (P < .001), and 1-twist groups (P < .001). For 2-0 Monoderm, the maximum load at failure was not significantly different between any group.
Comparisons of maximum load at failure of all monoderm (A) and nylon (B) suture sizes for each of 5 different knot variations described in Figure 3. See Figure 3 for the key.
Citation: American Journal of Veterinary Research 84, 7; 10.2460/ajvr.23.01.0003
In the 3-0 Nylon group (Figure 4), maximum load at failure was significantly lower in the 0-twist (P < .001), 1-twist (P < .001), 4-twist (P = .001), and the 10-twist (P < .001) groups, when compared with starting group. For 2-0 Nylon, maximum load at failure was significantly lower in the starting group when compared with the 0-twist (P < .001), 1-twist (P < .001), 4-twists (P < .001), or 10-twists (P < .001).
When all suture types and sizes were compared with each other, there was no significant difference in maximum load at failure (P = .411). However, when all 3-0 sutures were compared, there was a significant difference in maximum load at failure between the number of twists (P < .001). The maximum load at failure was lower for 4 twists and 10 twists when compared with the starting group (P < .01). When all 2-0 sutures were compared, no significant difference existed in maximum load at failure.
Regardless of suture type or size, failure of the starting knot always occurred by breakage 1 mm from the knot. Knots with 0 twists, regardless of type or size, failed most commonly by breakage > 2 mm from the knot, except for 2-0 Nylon, which failed more commonly by breakage at the knot (8/15, 53%). Knots with 1 twist, regardless of size or type, were more likely to fail by breakage > 2 mm from the knot. Knots tied with 4 or 10 twists did not have a consistent failure pattern. For knots with 4 twists, 0 PDO (10/15, 67%), 1 PDO (12/15, 80%), and 2-0 Nylon (9/15, 60%) were more likely to fail by breakage at the knot (Table 2). The remaining knots with 4 twists were more likely to fail > 2 mm from the knot. Knots with 10 twists were more likely to fail by breakage at the knot, except 3-0 Monoderm, which more commonly failed by breakage > 2 mm from the knot (10/15, 67%). All failure modes and percentages are depicted in Table 2.
Evaluation of failure mode within each group.
| Failure mode x/15 (%) | ||||
|---|---|---|---|---|
| Suture | Twists | Knot | < 1 mm from the knot | > 2 mm from the knot |
| 1 PDO | 0 | 5/15 (33) | 10/15 (67) | |
| 1 | 4/15 (27) | 2/15 (13) | 9/15 (60) | |
| 4 | 12/15 (80) | 3/15 (20) | ||
| 10 | 12/15 (80) | 3/15 (20) | ||
| 0 PDO | 0 | 1/15 (7) | 14/15 (93) | |
| 1 | 3/15 (20) | 12/15 (80) | ||
| 4 | 10/15 (67) | 5/15 (33) | ||
| 10 | 15/15 (100) | |||
| 2-0 PDO | 0 | 5/15 (33) | 3/15 (20) | 7/15 (47) |
| 1 | 3/15 (20) | 12/15 (80) | ||
| 4 | 3/15 (20) | 8/15 (53) | 4/15 (27) | |
| 10 | 8/15 (53) | 7/15 (47) | ||
| 3-0 PDO | 0 | 4/15 (27) | 11/15 (73) | |
| 1 | 5/15 (33) | 10/15 (67) | ||
| 4 | 2/15 (13) | 13/15 (87) | ||
| 10 | 8/15 (53) | 7/15 (47) | ||
| 2-0 Monoderm | 0 | 2/15 (13) | 13/15 (87) | |
| 1 | 4/15 (27) | 1/15 (7) | 10/15 (67) | |
| 4 | 1/15 (7) | 14/15 (93) | ||
| 10 | 6/15 (40) | 9/15 (60) | ||
| 3-0 Monoderm | 0 | 1/15 (7) | 14/15 (93) | |
| 1 | 4/15 (27) | 11/15 (73) | ||
| 4 | 5/15 (33) | 10/15 (67) | ||
| 10 | 6/15 (40) | 9/15 (60) | ||
| 2-0 Nylon | 0 | 8/15 (53) | 7/15 (47) | |
| 1 | 6/15 (40) | 9/15 (60) | ||
| 4 | 9/15 (60) | 6/15 (40) | ||
| 10 | 9/15 (60) | 6/15 (40) | ||
| 3-0 Nylon | 0 | 6/15 (40) | 3/15 (20) | 6/15 (40) |
| 1 | 4/15 (27) | 11/15 (73) | ||
| 4 | 4/15 (27) | 3/15 (20) | 8/15 (53) | |
| 10 | 8/15 (53) | 7/15 (47) | ||
No listed number indicates that no failure occurred by that type.
PDO = polydioxanone.
Discussion
The purpose of this study was to determine whether axial twisting within the ending loop of a simulated continuous pattern would decrease the maximum load at failure. There are no clear trends within the dataset; however, it suggests that maximum load at failure is decreased when more twists are added to a loop when compared with knots without twists, particularly in larger sutures. The first part of our hypothesis was that the load at failure would be less for sutures with more twists. Based on our data, we partially accept this hypothesis. The second part of our hypothesis was that mode of failure would not be different between knot configurations and we reject this hypothesis. The primary mode of failure of starting knots or knots with 0 or 1 twist within the knot was breakage of the suture away from the knot, regardless of size. However, the primary mode of failure occurred at the knot more commonly for larger suture sizes containing 4 or 10 twists within the ending loop.
Previous literature reports that axial twisting of the suture during normal wrist motion of the surgeon can lead to the weakening of the suture strand.5 Twisting up to 4 times was not thought to weaken a suture strand; however, twisting up to 10 times was shown to decrease the tensile strength of the suture.5,6,11 By the natural motion of the surgeon’s wrist while suturing, the suture can start to form curls or can twist on itself, and these may result in the twisting of the loop of the ending square knot of a continuous suture strand. In some cases, the suture loop is not unwound, and the twisted suture is tied into the ending knot. We suspect that the decrease in maximum load at failure of the suture knots containing more twists could be due to the axial twisting of the ending loop, leading to compromise of the suture stand, similar to that seen in prior studies.5,6
Research devoted to understanding various knot configurations is lacking. A study comparing square and surgeon knots with sliding knots showed that the former had superior knot security.14 When performing the current study, it was thought that because of the twists within the ending loop, the ending square knot would not be as secure as a standard starting knot or a knot without any twists in the ending loop due to the inability of the knot to lay down on itself appropriately. As it may be possible that the twisting within the loop decreases the contact between the suture strands and friction within the knot itself. Our data showed that increased twisting of the ending loop may lead to a higher likelihood of unraveling the knot or failure at the knot during testing, particularly for larger suture sizes. The data in this report found that only 0 PDO, 1 PDO, and 2-0 Monoderm with 10-twists were more likely to fail at the knot, while the remaining suture and knot types were more likely to have a comparable maximum load at failure and breakage of the suture strand. The sutures were more likely to fail at the knot when more twists are applied and had a larger gauge and because monofilament, synthetic suture tends to have more memory than multifilament sutures, this may lead to decreased ability to tighten down the knot securely,15,16 leading to an increased likelihood of failure at the knot. Thus, when twists are added, this may further decrease knot security due inability of tightening the knot appropriately. Suture with a smaller bore, although twisted, still has relatively strong knot security, as it is easier to tighten the knot appropriately. Because Nylon is a non-absorbable monofilament, it will undergo very little to no plastic deformation before failure.17 This likely contributed to the decreased failure at the knot as the suture will not undergo deformation within the knot, as may happen with monofilament absorbable sutures.
A major limitation of this study was difficulty in mimicking the axial twisting of the suture, as would occur during surgery. Therefore, although our data may suggest that an increased number of twists within the ending loop of a larger suture will decrease maximum load at failure and increase the likelihood of failure at the knot, it is assumed that a suture that has been axially twisted is weaker than suture that has not been twisted. Because the current study was simply focused on axial twisting within an ending loop, we cannot comment on the effect of axial twisting on the strength of the suture. Therefore, ending loops that are unraveled may still be weaker than those without any twisting. An additional limitation may be the removal of the suture and knot from the PVC pipe before testing, as the movements necessary to remove the suture from the pipe may make the knot more prone to slippage.18 Cyclic testing was not considered in this study as single distraction tests are more commonly performed and the selection of testing protocol for this study was based on prior studies.5,6,10,16 However, cyclic loading may be a more accurate way of testing sutures and should be considered for future studies. Lastly, a tension gauge was not used when tying sutures, so sutures were not tied with the exact same tension applied to more closely reflect what occurs in a clinical setting.
In conclusion, when finishing a continuous pattern, the number of twists within the ending loop may not increase the risk of failure at the knot; however, it can decrease the maximum load to failure at a knot, particularly as the suture size increases. Therefore, care should be taken not to twist the suture when handling.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
All sutures were donated by TruStitch Surgical Specialties Corporation. This company did not have any involvement in the study design, data analysis, or writing of this manuscript.
The authors declare that there were no conflicts of interest.
The authors would like to thank Anne Lovett for the illustrations.
References
- 1.↑
Abellán D, Nart J, Pascual A, Cohen R, Sanz-Moliner J. Physical and mechanical evaluation of five suture materials on three knot configurations: an in vitro study. Polymers. 2016;8(4):147. doi:10.3390/polym8040147
- 2.↑
Kim JC, Lee YK, Lim BS, Rhee SH, Yang HC. Comparison of tensile and knot security properties of surgical sutures. J Mater Sci Mater Med. 2007;18(12):2363–2369. doi:10.1007/s10856-007-3114-6
- 3.
Gallup DG, Nolan TE, Smith RP. Primary mass closure of midline incisions with a continuous polyglyconate monofilament absorbable suture. Obstet Gynecol. 1990;76(5):872–875. doi:10.1097/00006250-199011000-00031
- 4.↑
Tomihata K, Suzuki M, Tomita N. Handling characteristics of poly(L-lactide-co-epsilon-caprolactone) monofilament suture. Biomed Mater Eng. 2005;15(5):381–391.
- 5.↑
Dobrin PB. Surgical manipulation and the tensile strength of polypropylene sutures. Arch Surg. 1989;124(6):665. doi:10.1001/archsurg.1989.01410060027005
- 6.↑
Hennessey DB, Carey E, Simms CK, Hanly A, Winter DC. Torsion of monofilament and polyfilament sutures under tension decreases suture strength and increases risk of suture fracture. J Mech Behav Biomed Mater. 2012;12:168–173. doi:10.1016/j.jmbbm.2012.02.001
- 7.↑
Silver E, Wu R, Grady J, Song L. Knot security- how is it affected by suture technique, material, size, and number of throws? J Oral Maxillofac Surg. 2016;74(7):1304–1312. doi:10.1016/j.joms.2016.02.004
- 8.
Marturello DM, McFadden MS, Bennett RA, Ragetly GR, Horn G. Knot security and tensile strength of suture materials: knot security and tensile strength of suture materials. Vet Surg. 2014;43(1):73–79. doi:10.1111/j.1532-950X.2013.12076.x
- 9.↑
Muffly TM, Boyce J, Kieweg SL, Bonham AJ. Tensile strength of a surgeon’s or a square knot. J Surg Educ. 2010;67(4):222–226. doi:10.1016/j.jsurg.2010.06.007
- 10.↑
Upchurch DA, Wang Y, Chen S, Roccabianca S, Roush JK. Assessment of time to completion, number of errors, and knot-holding capacity of square knots and Aberdeen knots tied by veterinary students and student perceptions of knot security and knot-tying difficulty. J Am Vet Med Assoc. 2020;256(2):230–238. doi:10.2460/javma.256.2.230
- 11.↑
Thacker JG, Rodeheaver G, Moore JW, et al. Mechanical performance of surgical sutures. Am J Surg. 1975;130(3):374–380. doi:10.1016/0002-9610(75)90408-0
- 14.↑
Schubert DC, Unger JB, Mukherjee D, Perrone JF. Mechanical performance of knots using braided and monofilament absorbable sutures. Am J Obstet Gynecol. 2002;187(6):1438–1442. doi:10.1067/mob.2002.129156
- 16.↑
Gillen AM, Munsterman AS, Hanson RR. In vitro evaluation of the size, knot holding capacity, and knot security of the forwarder knot compared to square and surgeon’s knots using large gauge suture: in vitro evaluation of forwarder knot using large gauge suture. Vet Surg. 2016;45(8):1034–1040. doi:10.1111/vsu.12556
- 17.↑
McFadden MS. Suture materials and suture selection for use in exotic pet surgical procedures. J Exot Pet Med. 2011;20(3):173–181. doi:10.1053/j.jepm.2011.04.003
- 18.↑
Schaaf O, Glyde M, Day RE. In vitro comparison of secure Aberdeen and square knots with plasma- and fat-coated polydioxanone: comparison of secure Aberdeen and square knots. Vet Surg. 2010;39(5):553–560. doi:10.1111/j.1532-950X.2009.00640.x
