Presented in part at the ACVS Veterinary Symposium, San Antonio TX, October 2013.
Original Article: Research
Knot security and tensile strength of suture materials
Article first published online: 19 NOV 2013
© Copyright 2013 by The American College of Veterinary Surgeons
Volume 43, Issue 1, pages 73–79, January 2014
How to Cite
Marturello, D. M., McFadden, M. S., Bennett, R. A., Ragetly, G. R. and Horn, G. (2014), Knot security and tensile strength of suture materials. Veterinary Surgery, 43: 73–79. doi: 10.1111/j.1532-950X.2013.12076.x
- Issue published online: 2 JAN 2014
- Article first published online: 19 NOV 2013
- Manuscript Accepted: 8 OCT 2013
- Manuscript Received: 17 MAY 2013
To evaluate knot security and tensile failure load of suture tied in simple interrupted, beginning continuous, and ending continuous patterns for 11 suture materials commonly used in small animal surgery.
For each of 11 suture material types, and 5 knot sizes (2, 3, 4, 5, and 6 throws) 2 surgeons each tied 6 knots (n = 12 for each knot size in 11 suture materials). Three types of patterns were evaluated: a simple interrupted square knot, a square knot beginning a simple continuous pattern, and the knot ending a simple continuous pattern. All knots were incubated in healthy canine donor plasma at 40°C for a minimum of 24 hours. Sutures were evaluated for knot security (knots untied, suture failed by breaking, suture slipped from the clamps, or suture untied before testing) and maximum load carried before knot slippage or knot failure (termed tensile failure load).
Significant differences were found in knot security and tensile failure load among suture types. There was no significant difference between the simple interrupted knots and the knots at the beginning of a simple continuous pattern; however, both were significantly less likely to fail than the knots tied at the end of a simple continuous pattern. The number of throws per knot had a significant effect for knot security and tensile failure load. Surgeon experience had a significant effect on failure mode and tensile failure load.
Suture type, number of throws per knot (knot size), suture pattern, and surgeon experience play an important role in knot security and should be considered when performing surgery.
All suture materials are evaluated on basic characteristics such as handling properties, degree of inflammatory reaction induced in tissues, cost, initial tensile strength, temporal loss of tensile strength, and knot security. In most surgical procedures, tying knots with suture material is an essential component of maintaining tissue apposition. Security of a knot is crucial to holding tissues together until they have healed, and a secure knot is defined as one that does not untie or slip open before the suture line breaks.
Tensile breaking strength is commonly defined as the force a suture can withstand before breaking. The type of suture and the type of knot that are used affect knot security. Monofilament sutures are less pliable and more susceptible to damage from crushing or nicking, and are also smoother, which makes a knot more likely to slip. Therefore, more throws are typically required when using a monofilament suture. Braided sutures have greater strength and pliability making a knot less likely to slip, meaning fewer throws could result in a secure knot. Square knots are more secure than granny or half hitch knots which are more prone to slipping.
For different suture materials, the type of knot requires varying numbers of throws to create a secure knot. Since that report, different suture materials are commonly used in surgical practice and have not been evaluated mechanically to assess differences in knot security. The earlier study did not evaluate tensile failure load or the effect of surgeon experience on knot security.
Our purpose was to determine the minimum number of throws necessary to create a secure knot with 11 suture material types and 3 suture patterns independent of the experience of the person tying the knot. A secure knot was defined as a knot that failed by breaking rather than untying. We hypothesized that the security of the knot would increase as the number of throws increased but that 2 square knots would be sufficient to cause the suture to break instead of failing because of the knot untying. Our 2nd hypothesis was that knots tied with multifilament sutures would require fewer throws than monofilament sutures to form a secure knot. Lastly, we hypothesized that there would not be a difference between a 4th year veterinary student and an ACVS Diplomate with regard to knot security or tensile load to failure.
MATERIALS AND METHODS
Suture materials tested were 3–0 polyglactin 910 (Vicryl™; Ethicon, Johnson and Johnson, Somerville, NJ), 3–0 lactomer (Polysorb™, Covidien, Mansfield, MA), 3–0 polydioxanone (PDS II™, Ethicon), 3–0 polyglyconate (Maxon™, Covidien), 3–0 polyglecaprone 25 (Monocryl™, Ethicon), 3–0 glycomer (Biosyn™, Covidien), 3–0 polyglytone 6211 (Caprosyn™, Covidien); 3–0 monofilament nylon (Ethilon™, Ethicon; Monosof™, Covidien) and 3–0 polypropylene (Prolene™, Ethicon; Surgipro™ Covidien).
Square knots (2 throws) were created by tying suture around a 55 mm diameter aluminum cylinder (1A,B). Subsequent throws were then tied in the same manner to create knot with 3, 4, 5, and 6 throws. All knots were made using instrument ties by 2 authors (D.M.M., M.S.M.) and non-sterile latex gloves were worn whenever handling suture materials. The suture ends (tails) were cut to 3 mm length (measured with a ruler before cutting). For each surgeon (n = 2), 6 samples of each suture type (n = 11), and each knot size (n = 5) were created.
For each suture type and each surgeon 3 different square knots were evaluated: (1) a simple interrupted knot, (2) a knot at the beginning of a simple continuous pattern with 1 tail cut to a minimum of 10 mm (1C), and (3) the knot at the end of a simple continuous pattern using the free loop to tie the knots (1D). Suture loops and knots were then placed in a 7 mL glass vial (BD Vacutainer® red top blood collection tubes, BD Vacutainer, Franklin Lakes, NJ) that were filled with healthy canine donor plasma and incubated at 40°C for a minimum of 24 hours.
A tabletop materials testing machine (Instron Mini 44 Tensile Tester, Instron, Norwood, MA) with a 500 N load cell, was used for mechanical testing of all samples. Just before testing, each suture loop was removed from the Vacutainer® and cut directly opposite the knot creating 2 strands of equal length, which were inserted into the Instron Mini 44's pneumatic grips that clamped the suture ends with air pressure at 45 psi. This pressure was chosen to minimize the risk of sutures slipping from the grips as well as breaking in the grips because of excessive pressure (based on a pilot study with these suture materials). The grips were set to a distance such that the suture material was taut, with minimal force (<0.1 N) applied before testing began. The screw driven load frame was operated in position control with a continuous distraction rate of 20 mm/min applied throughout the test. Displacement and force data were continuously recorded at a sampling rate of 1 Hz using data collection software (LabView, National Instruments, Austin, TX).
All samples were tested in random order (by block randomization) by 1 author (D.M.M.). Each sample was evaluated for knot security (knot untied before testing, knot untied while under load, suture slipping from Instron clamps, or suture failure by breaking under load). Tensile load at which failure occurred was recorded. A unifying tensile failure load was defined in this study as the maximum load the knot carried before slippage or the maximum load the knot carried before breakage. If the knot untied before testing or slipped in the clamps, that sample was not included in the calculations of mean tensile failure load.
Statistical analysis was performed using software (SPSS Statistics for Windows, Version 19.0. Armonk, NY). A multivariate general linear model was used to evaluate the effect of the independent variables (suture type, pattern type, number of throws, and surgeon) and their interactions on the dependent variables failure mode (breaking vs. slipping) and tensile failure load in kg and Newtons (N). Post hoc analyses were performed using Bonferroni tests. P < .05 was considered significant. A 2nd analysis was used to evaluate the effect of the independent variables for each suture type. Using the same software, an ANOVA was used to evaluate the effect of the independent variables (knot type, number of throws and surgeon) for each suture type. Post hoc analyses were performed using a Tukey HSD to evaluate the effect of the number of throws for a simple interrupted knot for each suture type.
All independent variables (suture type, number of throws (knot size), pattern type, and surgeon) and their interactions had a significant effect on failure mode and tensile failure load (N) with P < .001. Ending continuous knots failed and untied before testing more often than the other 2 patterns (P < .001).
Tensile Failure Load
Suture material type affected mean tensile failure load (Table 1). Braided lactomer (Polysorb™) had the greatest force to breaking (19.6 N), whereas polyglytone 6211 (Caprosyn™) had the lowest (7.92 N). There was no statistically significant difference among the forces required to break polypropylene (Surgipro™) and the 2 nylon sutures (Ethilon™ and Monosof™; Table 1).
|Suture Type||Mean Load to Failure (N)||SD|
The number of throws used to construct the knot also had a significant effect (P < .001) on tensile failure load. For all suture types tested (Table 2, the highest tensile failure load was found with 5 throws (16.60 N), which was significantly different from all but 6 throws. For all suture types tested, the lowest tensile failure load was found with 2 throws, which is significantly different from all other sizes.
|Number of Throws||Mean tensile Failure Load (N)||SD|
Suture pattern used and surgeon experience had a significant effect on tensile failure load. Knots ending in a simple continuous pattern had a lower tensile failure load than simple interrupted knots (P < .001) and knots beginning a simple continuous pattern (P < .001). The Diplomate (M.S.M.) had a mean force to failure of 14.72 N whereas the student (D.M.M.) had a mean force to failure of 11.60 N (Table 3).
|Surgeon||Mean ± SD Load to Failure||Failure Mode|
|Mean Number of Knots That Untied||Mean Number of Knots That Broke|
|D.M.M.||11.60 ± 7.44||380||558|
|M.S.M.||14.72 ± 7.79||246||693|
Suture material type significantly affected knot security (Table 4). Polyglytone 6211 was statistically different from all other sutures excluding glycomer (Biosyn™) and it also untied before testing more often than any other suture (3). Polyglyconate (Maxon™) and polyglecaprone 25 (Monocryl™) had the greatest percentage of knots (74% each) that broke rather than untied (Table 4). There was no significant difference in knot security when comparing polydioxanone, polyglecaprone 25, polyglactin 910, Prolene™, Ethilon™, Surgipro™, or lactomer (Polysorb™; 2).
|Suture Type||Failure Mode|
|Knot Untied||Suture Failed (Secure Knot)||Suture Slipped From Clamp||Suture Untied Before Testing|
Among suture materials, the number of throws (knot size) significantly affected knot security. For polyglactin 910, Ethilon™, polyglytone 6211, Surgipro™, and Monosof™, 3, 4, 5, and 6 throws were all significantly more secure compared with 2 throws (P < .01). For polydioxanone, Prolene™, polyglecaprone 25, lactomer, and glycomer 4, 5, and 6 throws were significantly more secure than those tied with 2 and 3 throws (P < .01). For all suture materials tested, 2 throws always untied before breaking and thus never created a secure knot (Table 5).
|Suture||2 Throws||3 Throws||4 Throws||5 Throws|
|Polydioxanone||P < .000||P = .001||4 ◊ 5 = 0.073||5 ◊ 6 = 0.398|
|Polyglecaprone 25||P < .000||P < .000||4 ◊ 5 = 0.377||5 ◊ 6 = 1|
|4 ◊ 6 = 0.316|
|Polyglactin 910||P < .000||P = 1||P = 1||P = 1|
|Polypropylene (Prolene™)||P < .000||3 ◊ 4 = 0.295||P = 1||P = 1|
|3 ◊ 5 = 0.176|
|3 ≠ 6 = 0.002|
|Nylon (Ethilon™)||P < .000||P = 1||P = 1||P = 1|
|Polyglytone 6211||P < .000||P = 1||P = 1||P = 1|
|Glycomer||P < .000||P = .059||P = 1||P = 1|
|Polyglyconate||P < .000||P = 1||4 ◊ 5 = 0.115||P = .05|
|4 ◊ 6 = 1|
|Lactomer||P < .000||3 ◊ 4 = 0.508||4 ◊ 5 = 0.543||P = 1|
|3 ≠ 5 = 0.016||4 ◊ 6 = 1|
|3 ◊ 6 = 0.584|
|Polypropylene (Surgipro™)||P < .000||P = 1||P = 1|
|Nylon (Monosof™)||P < .000||P = 1||4 ◊ 5 = 0.373||P = 1|
|3 ◊ 4 = 0.068|
|3 ◊ 5 = 1||4 ◊ 6 = 0.098|
|3 ◊ 6 = 1|
Suture pattern type significantly affected knot security with knots ending a simple continuous pattern being significantly less secure than knots with a simple interrupted pattern (P < .001) and knots beginning a simple continuous pattern (P < .001). For all suture types and knot sizes tested, simple interrupted knots and beginning simple continuous knots untied a mean of 141 times and 161 times, respectively, whereas ending simple continuous knots untied a mean of 324 times (Table 6). There was no statistical difference between the simple interrupted pattern and the beginning of the simple continuous pattern (P = 1).
|Pattern Type||Mean ± SD Tensile Failure Load (N)||Failure Mode|
|Mean Number of Knots That Untied||Mean Number of Knots That Broke|
|Simple interrupted||14.67 ± 7.33||141||491|
|Beginning simple continuous||14.30 ± 7.18)||161||468|
|Ending simple continuous||10.48 ± 8.09)||324||292|
Surgeon experience was also statistically significant (P < .001) with regard to knot security (Table 3). The percentage of secure knots for M.S.M. was 73% compared with 59% for D.M.M.
Knot security is influenced by a variety of factors including the type of suture, the number of throws used in knot construction, and the type of pattern used. We found significant differences in knot security between suture material types, supporting our hypothesis. We also found that for 5 of 11 sutures, 3 throws created a knot that was as secure as a knot with 6 throws. Thus, for these 5 suture materials (polyglactin 910, Ethilon™, polyglytone 6211, Surgipro™, and Monosof™), the hypothesis that a minimum of 4 throws were needed to create a secure knot was not supported. We found that knots either failed (untied) or the suture broke before the knot untied. When knot failure or suture breakage occurred, it did so immediately adjacent to the knot itself in 74% of cases. The knot has previously been reported to be the weakest part of the suture loop because of the shear stresses that exist at the point between the loop and the 1st throw of the knot.[4, 7] Factors found to influence knot security were suture type, number of throws used to create the knot, pattern type, and surgeon experience.
An important surgical principle is to leave the minimum amount of foreign material in tissue while still maintaining the integrity of the suture and the security of the knot. Suture materials contribute to local inflammation and bacterial colonization.[8-10] The most pronounced tissue reaction occurs at the knot component of the suture line because this is where the highest density of foreign material is located. Therefore, identifying the minimum number of throws required to produce a secure knot for a given suture material has important implications. It may be possible to tie a secure knot with only 3 throws using Polyglactin 910, Ethilon™, Polyglytone 6211, Surgipro™ or Monosof™ thus leaving less material in the tissue.
The type of suture used affected both knot security and tensile failure load. Suture selection depends on various factors. One of these is suture material composition because this affects handling characteristics, tensile strength and duration the suture will hold tissues in apposition. Non-absorbable suture undergoes little to no plastic deformation before breaking and retains most of its original tensile strength for several years (some with no appreciable loss at 5 years). The 2 polypropylene sutures and Ethilon™ had similar knot security (Table 4); all 3 sutures have the same indications in surgery: soft tissue approximation or ligation, cardiovascular surgery, ophthalmologic surgery, and neurologic surgery. These non-absorbable synthetic sutures had fairly poor knot security in our study, untying a mean of 40% of the time during testing. Although, the smooth surface of these sutures creates little tissue drag, the high yield point and increased memory creates poor knot security.
Absorbable sutures lose most of their tensile strength within 60 days, with most retaining 50% or less 4 weeks after implantation.[9, 12] The type of monomer used to create synthetic absorbable sutures affects the flexibility, strength, and rate at which it is absorbed. Polyglytone 6211 was different from all sutures except glycomer for failure mode. Polyglytone 6211 and glycomer untied before testing more often than any other suture material (2). One potential reason for this is that these sutures have a trimethylene carbonate component (polyglyconate also has this copolymer and was similar to glycomer). Glycomer is composed of glycolide and dioxanone with alternating segments of dioxanone and trimethylene carbonate. Polyglytone 6211 is composed of glycolide, lactide, trimethylene carbonate, and caprolactone. Glycolide and lactide contribute to the strength of the suture material, trimethylene carbonate contributes to flexibility, and dioxanone provides strength and flexibility. Once implanted, polyglytone 6211 loses 80% of its tensile strength in 10 days and glycomer loses 60% in 3 weeks.[8, 9]
Braided sutures have greater tensile strength than monofilament sutures, making it logical that they can withstand more force before breaking. We found a significant difference between lactomer and polyglactin 910 for load to failure. Lactomer failed at 19.6 N whereas polyglactin 910 failed at 14.14 N. This difference likely reflects difference in the properties of the materials used by each company. Lactomer is made from glycolide and lactide, and is coated with caprolactone/glycolide and calcium stearoyl lactylate. Polyglactin 910 is made from glycolide and lactide and coated with calcium stearate and a 2nd copolymer of lactide and glycolide. It is also plausible to consider that the different coatings may change the frictional interaction within the knot thus affecting the shear stresses that exist at the point between the loop and the 1st throw of the knot.
We tested sutures made by Ethicon and Covidien. Recently, other sutures made by different companies with identical chemical composition were compared with sutures made by Ethicon. In that study, 3 different types of suture were tested by comparing load and stiffness at different time points after incubation in bovine serum. Although chemical composition was identical, there were significant differences between the two products when comparing load and stiffness of the suture material. Similarly, we found differences between nylon suture made by 2 different companies. Whereas the tensile failure load was not different between Ethilon™ and Monosof™, Ethilon™ had more knots that were secure when compared with Monosof™ (Table 4). For polypropylene sutures, knot security was similar, but the tensile failure load for Prolene™ was higher. It is important to keep this in mind and recognize that sutures of the same material and size do not necessarily have the same properties.
We found that the number of throws affected knot security. For 5 suture types tested (polyglactin 910, Ethilon™, polyglytone 6211, Surgipro™, and Monosof™), 3 throws were as secure as 6 throws, though the tensile failure load was greater with 6 throws. It was reported by Schaaf et al in 2010 that the minimum number of throws for a secure knot using polydioxanone suture is 4 throws. This finding was confirmed by our study, and for each of the suture materials tested, there was no significant difference in knot security with more than 4 throws. Rosin reported that knots made with polydioxanone were secure with 5 throws at the beginning of a continuous pattern and 7 throws at the end. Unlike Rosin's study, our results showed no significant difference in knot security with more than 4 throws for any pattern type using polydioxanone. Rosin's study did not comment on the tensile failure load of the suture. We found that 4 throws of polydioxanone was as secure as 6 throws; however, knots with 6 throws had a higher tensile failure load than knots with 4 throws. Rosin's study also did not evaluate the effect of surgeon experience on knot security.
The type of pattern used affected knot security. There were no significant differences between the simple interrupted pattern and the beginning of the simple continuous pattern. These 2 knots are formed in the same manner, and the only difference exists in the extra length of the continuous end of the simple continuous pattern. There was, however, a significant difference in knot security for the ending of a simple continuous pattern. The ending simple continuous pattern knot would be expected to have different properties because it is formed using a loop (from the continuous portion) tied to the other end of the suture. The additional strand of suture from the loop arising and ending on different sides of the incision would make the knot more likely to untie because it is more bulky and cannot be pulled as tightly.
Surgeon experience and training was a significant factor affecting knot security. The percent of knots that were secure when tied by an experienced surgeon was 73%, compared with 59% for a 4th year veterinary student. Tensile failure load was also different between knots tied by an experienced surgeon and a student. A likely explanation is the tension applied when creating the knots. To determine if the tensile failure load was related to the force applied to the suture when tying the knots, a pull spring scale could be used as in Rosin and Robinson.
To our knowledge, this is the 1st study to look at a large range of suture materials and their knot security. For many of the suture types tested, this is the first study to report the effect of knot size on knot security, surgeon experience on knot security, and report on the tensile failure load at the knot. Factors in this study affecting the security of a knot were suture type, number of throws, pattern type, and surgeon experience. All of these factors are important considerations when performing surgery and selecting a suture material.
We used 3–0 suture. Recent reports testing knot security evaluated sutures of different sizes.[16, 17] Friction between throws of large gauge suture is higher than for small gauge. Muffly et al showed that for 0 suture, 5 throws was the minimum for a secure knot and adding additional throws did not change the tensile failure load. We found that knots with 5 and 6 throws had a higher tensile failure load, supporting the findings of Muffly et al. Tidwell et al determined that 4–0 sutures required fewer throws to form a secure knot than 2–0 suture. The number of knots that untied in their study also increased with a larger gauge suture.
Other factors evaluating suture strength and knot security not evaluated in our study include intracorporeal effects on suture strength and knot security like ischemia, inflammation, and overall tissue health. We used healthy donor serum which did not allow for evaluation of tissue condition. More studies are needed, as there is not currently a good model for testing in vivo effects on suture. We found significant differences between suture materials, knot size, suture pattern type, and surgeon experience for knot security and tensile failure load. A secure knot with 3 throws may be possible using polyglactin 910, Ethilon™, polyglytone 6211, Surgipro™, or Monosof™. Comparing these biomechanical results with in vivo testing would be an interesting follow up study.
The authors declare no financial or other conflicts of interest related to this report.
- 2Biomaterials, suturing, and hemostasis, in Welch Fossum T (ed): Small animal surgery (ed 3). St. Louis, MO, Mosby Inc., 2007, pp 57–78
- 3Suture material, tissue staples, ligation devices, and closure, in Tobias K (ed): Veterinary surgery: small animal. St Louis, MO, Elsevier, 2012, pp 187–200
- 5Tissue reaction and surgical knots: the effect of suture size, knot configuration, and knot volume. Obstet Gynecol 1989; 74:64–68, , , et al:
- 9Evaluation of the histologic reactions to commonly used suture materials in the skin and musculature of ball pythons (python regius). Am J Vet Res 2011; 72:1397–1406, , , et al:
- 11Covidien Suture™, Product Information. Covidien, Mansfield, MA
- 12Suture materials, tissue adhesives, staplers, and ligating clips, in Slatter D (ed): Textbook of small animal surgery (ed 2). Philadelphia, PA, Saunders, 1993, pp 204–212