- Email: [email protected]
Optimizing arthroscopic knots using braided or monofilament suture
Optimizing Arthroscopic Knots Using Braided or Monofilament Suture Todd D. Loutzenheiser, B.S., M.E., Douglas T. Harryman II, M.D., Dean W. Ziegler, M.D., and ShingWai Yung, M.D.
Summary: Surgeons need to know how the material properties of a suture affect the security of a surgical knot. The purpose of this study was to compare the security of some clinically important arthroscopic knots when tied using a braided multifilament suture and to draw comparisons with results of similar knots tied with monofilament suture. Permanent braided polyester suture was used to test 10 knot configurations. Eight of the knots included (1) two types of initial cinching knots followed by (2) one of four combinations of half-hitches. We also tested the taut-line hitch locked with half-hitches and the original Revo knot. Each knot was subjected to cyclic loading followed by an ultimate load to failure. Clinical failure was defined as the maximum force that resulted in 3 mm of loop displacement. Force versus displacement data were obtained, and the maximal loop holding capacities were compared statistically. The Duncan loop with switched-post half-hitches and the Revo knot (Linvatec, Largo, FL) showed the highest knot-holding capacities (mean, 87N and 92N, respectively) when compared with all other configurations (P ⬍ .0001) for braided suture. A similar knot-holding capacity was described for monofilament suture using the Duncan loop locked with switched-post, reverseddirection half-hitches (mean, 81 N). All knots without post switching slipped completely at significantly lower loads than knots with post switching (monofilament, P ⬍ .001; braided, P ⬍ .0001). When compared with results of knots tied with monofilament suture, the braided switched-post configurations had smaller cyclic displacements (braided, 0.7 mm; monofilament, 1.7 mm). Although the Revo knot showed good strength for braided suture, it was significantly weaker than other configurations when tied with monofilament suture. Therefore, it is important to test the knot strength for a given suture material before applying it clinically. Key Words: Arthroscopic knots—Braided—Monofilament—Cyclic and failure loads.
D
isruption of a suture repair may result from failure of tissue, suture material, or the knot.1 In a previous study, we compared the strength characteristics of eight arthroscopic knot configurations using absorbable polydioxanone monofilament suture.2 AlFrom the Medical College of Vermont, Burlington, Vermont (T.D.L.); the Department of Orthopaedics, The University of Washington, Seattle, Washington (D.T.H.); Blount Orthopaedic Group, Ltd., Milwaukee, Wisconsin (D.W.Z.); and the Department of Orthopaedic Surgery, Singapore General Hospital, Singapore (S.-W.Y). Address correspondence and reprint requests to Douglas T. Harryman III, M.D., Department of Orthopaedics Box356500, The University of Washington, Seattle, WA 98195, U.S.A. r 1998 by the Arthroscopy Association of North America 0749-8063/98/1401-1677$3.00/0
though some of the knots tested appeared similar, we found that significant differences existed simply because of minor variations in the complexity of the initial cinching knot or the pattern of locking halfhitches. Because only one type of suture material was used, it was not necessary to consider differences in holding capacity based on suture material properties such as tensile strength, structural shape, ductility, pliability, coefficient of friction, and others. Properties such as these affect the ultimate strength of a knot and the efficacy of various knot configurations to resist slippage under load.3-5 Arthroscopic surgeons select an absorbable monofilament suture and others a braided material for
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 14, No 1 (January-February), 1998: pp 57–65
57
58
T. D. LOUTZENHEISER ET AL.
specific reasons. Absorbable monofilament is often selected for arthroscopic procedures because it can be advanced through arthroscopic suturing devices, slides easily for tying slip knots and it maintains reasonable strength while resorption and healing occurs. Under loaded conditions, however, absorbable monofilament can elongate up to 30% of its length and slips more easily than braided suture.6 Braided suture is often preferred to avoid knot slippage, suture stretch, or breakage, which would otherwise risk loss of tissue apposition. A braided polyester suture is also chosen because it remains in situ, is able to resist a load application during all phases of healing, and because it has been effective in many open procedures7 (e.g., Bankart and rotator cuff repair). The purpose of this study was to determine whether (1) the principles and methods of tying monofilament suture documented in our previous study apply to braided suture, (2) current clinical methods of tying braided arthroscopic knots provide secure fixation, (3) the holding capacities of braided knots tied with a knot pusher are comparable to those tied by hand, and (4) the strongest monofilament and braided suture knots have comparable holding capacities for the same size suture material. METHODS Suture and Knot Configurations All knots in this study were tied using No. 1 braided suture (Ethibond; Ethicon, Somerville, NJ). This suture was selected because braided polyester is a standard suture used for both open and arthroscopic repairs, and its size is comparable to the previously tested No. 1 polydioxanone, (PDS II; Ethicon), which would allow comparison of holding capacities under cyclic and clinical failure displacement loads to the authors previous study. The current study was completed in 1994 following monofilament suture testing using nearly the same methods and examiners as the present study. One investigator hand-tied a total of 10 different knot configurations, 8 of which were identical to those tied in our previous study. In the previous study, the knots consisted of 2 different initial cinching knots, (Duncan loop [DL], overhand throw [OT], Fig 1A), followed by 4 different configurations of locking half-hitches (Fig 1B, 1C). These combinations resulted in a total of 8 configurations (Table 1). In this study, 2 additional configurations were tested. The first consisted of the taut-line hitch (TLH) initial cinching knot locked with three nonidentical switched-post halfhitches (TLH⫽S//xS//xS Fig 1A, 1C, Table 1). The
Tera & Aberg coding system was extended to include the TLH initial cinching knot (TLH, Table 1).4 The second new knot, the original Revo configuration (Linvatec, Largo, FL), consisted of two pairs of nonidentical half-hitches connected via a switchedpost (Fig 1D, Table 1). The Revo knot is represented by the standard coding system (SxS//SxS - Table 1). After running statistical analyses on the hand-tied data, four configurations with the highest holding capacities were selected for the multiple tyer experiment. High strengths for the DL, TLH, and OT, all locked with switched-posts and reversed loops as well as the original Revo knot qualified each configuration for the multiple tyer experiment. Tying and Testing Procedure The tying and testing techniques used in this study were identical to those described in our previous study.2 The apparatus, calibrations, tying methods, strand tensioning, knot preloading, load applications, and quantification of loop displacement for cyclic, clinical failure, and ultimate failure were duplicated for the braided and additional monofilament knots in this study. Force versus displacement data were recorded for 10 knots of each configuration, and compared statistically using Dunnett’s t-test to adjust for multiple comparisons. Values of P ⬍ .05 were considered statistically significant. Clinical failure at 3 mm of displacement was defined as the minimum force required to distract the loop to a displacement of 3 mm. The ultimate force was defined as the minimum force required to cause total failure either by complete slippage or material breakage. Each of the four throws in the REVO knot were tied and tightened separately (SxS//SxS, Table 1, Fig 1D) = FIGURE 1. (A) The slip knot configurations shown above were used initially to remove slack from the suture loop. The OT configuration designates the simple OT, the TLH configuration designates the intermediate TLH, and the DL configuration designates the complex DL configuration. (B) The half-hitch configurations shown above represent two combinations that do not incorporate post switching. The labels ‘‘under’’ and ‘‘over’’ indicate the direction for the initial pass around the post when making each half-hitch loop. Note that the direction of the half-hitch loops are reversed for the SxSxS configuration (see Table 1 for an explanation of the Tera and Aberg coding system). (C) The half-hitch configurations shown above represent two combinations that incorporate post switching and reversal of the loop direction. Note that the S//xS//xS configuration combines post switching and loop direction reversal (see Table 1 for an explanation of Tera and Aberg coding). (D) The Original Revo knot configuration shown above represents a combination with reversed direction half-hitches with a switched-post after the second throw. SxS//SxS (see Table 1 for an explanation of Tera and Aberg coding).
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
B
59
60
T. D. LOUTZENHEISER ET AL. TABLE 1. Knot Configurations: Tera and Aberg Notation13
Symbol OT DL TLH S ⫽ x // //x
Description Overhand-throw sliding knot Duncan loop sliding knot Tautline Hitch Single sliding throw (half-hitch) knots OT, DL, TLH slip knots ⫹3 half-hitches Identical throw, loop direction same Non-identical throw, loop reversed Alternating post between throw Alternating & reversing posts
Knot Configuration OT⫽S⫽S⫽S DL⫽S⫽S⫽S TLH⫽S⫽S⫽S* OT⫽SxSxS DL⫽SxSxS OT⫽S//S//S DL⫽S//S//S OT⫽S//xS//xS DL⫽S//xS//xS TLH⫽S//xS//xS SxS//SxS (Revo)
Description Identical half-hitches, loop in same direction around a single post. Non-Identical half-hitches, loop in opposite direction around same post. Identical half-hitches, post is alternated after each throw. Non-identical (reversed-throw), half-hitches looped around alternating posts. Revo Knot: two pairs of nonidentical half-hitches with one alternated post.
NOTE. ‘‘//x’’ symbol is equivalent to the ‘‘#’’ symbol used by other authors.1 *Not tested.
using the same technique for placing the locking half-hitches of the other configurations. The TLH, Table 1, Fig 1A) was used as the initial cinching knot with locking half-hitches in the TLH⫽S//xS//xS configuration. For the multiple tyer experiment, three practiced coauthor examiners, tied a total of 10 knots for each of the four configurations. The knots were tied with a single-hole knot pusher through a 7-mm cannula. The best knots for monofilament and braided suture were compared for each group using ANOVA and an F-test. RESULTS Hand Tied Knots Cyclic Loading: All configurations with nonswitched-post half-hitches applied after the OT cinching knot (OT⫽S⫽S⫽S, OT⫽SxSxS) slipped completely during the first few repetitions of the cyclic load test (Table 2, Fig 2). Knots with the more complex DL initial cinching knot and non-switched-post halfhitches (DL⫽S⫽S⫽S, DL⫽SxSxS) had displacements ⱕ 1.1 mm (Table 2 , Fig 2). All switched-post configurations (S//S or S//xS), distracted less than 1 mm during cyclic loading resulting in no significant differences between the cyclic displacements for these knots (P ⬎ .2). Load to Clinical Failure (3 mm of Displacement): There were no significant differences in strength between configurations with reversed half-hitches (designated by ‘x’) or without reversed direction (designated by ‘⫽’) half-hitches (e.g., OT⫽S⫽S⫽S v OT⫽SxSxS, DL⫽S⫽S⫽S v DL⫽SxSxS, OT⫽S// S//S v OT⫽S//xS//xS, DL⫽S//S//S v DL⫽S//xS//xS; Table 3, P ⬎ .2). Configurations without switched-posts, (i.e., no ‘//’
designation; DL⫽S⫽S⫽S, DL⫽SxSxS, OT⫽S⫽S⫽S, OT⫽SxSxS), and the knots with the less complex OT initial cinching knots were significantly weaker than those with the DL (Table 3 and Fig 3, P ⬍ .0001). Similarly, the switched-post configurations with the OT, (OT⫽S//S//S, OT⫽S//xS//xS), were significantly weaker than all other configurations with switchedposts (Table 3, Fig 3, P ⬍ .0001). All configurations containing at least one switchedpost were significantly stronger than configurations without a switched-post (Table 3, Fig 3). There were no significant differences in holding capacities beTABLE 2. Cyclic Loading Data for Braided Suture Loop Displacement (mm) Knot Configurations
15-N Load (mean ⫾ SD)
30-N Load (mean ⫾ SD)
⌬30-15 N (mean ⫾ SD)
OT⫽S⫽S⫽S OT⫽SxSxS OT⫽S//S//S OT⫽S//xS//xS OT⫽S//xS//xS-P DL⫽S⫽S⫽S DL⫽SxSxS DL⫽S//S//S DL⫽S//xS//xS DL⫽S//xS//xS-P TLH⫽S//xS//xS TLH⫽S//xS//xS-P SxS//SxS SxS//SxS-P
* * 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.2 ⫾ 0.2 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.2 ⫾ 0.1 0.2 ⫾ 0.1 0.2 ⫾ 0.1
* * 0.6 ⫾ 0.2 0.7 ⫾ 0.2 0.5 ⫾ 0.1 0.8 ⫾ 0.4 1.1 ⫾ 0.7* 0.6 ⫾ 0.2 0.5 ⫾ 0.2 0.6 ⫾ 0.1 0.7 ⫾ 0.2 0.6 ⫾ 0.1 0.8 ⫾ 0.2 0.7 ⫾ 0.1
* * 0.5 ⫾ 0.2 0.6 ⫾ 0.2 0.3 ⫾ 0.1 0.7 ⫾ 0.4 1.0 ⫾ 0.7* 0.5 ⫾ 0.2 0.4 ⫾ 0.2 0.4 ⫾ 0.1 0.6 ⫾ 0.2 0.4 ⫾ 0.1 0.6 ⫾ 0.2 0.5 ⫾ 0.1
NOTE. n ⫽ 10 knots per configuration. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. Standard deviations approaching or equivalent to the mean reflect loop displacements recorded at low cyclic loads and the limit of accurate measurement rounded to the nearest tenth. *Denotes complete slippage or outliers causing a wide standard deviation.
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
FIGURE 2. Displacement of braided knots under cyclic loads of 15 and 30 N. There were no significant differences among the knots that displaced less than 0.8 mm.
tween all configurations with switched-post halfhitches (DL⫽S//S//S, DL⫽S//xS//xS, TLH⫽S//xS// xS, SxS//SxS; P ⬎ .2). Ultimate Failure Load: All knots ultimately failed after a displacement of greater than 3 mm. Of all the knots tested, only the Revo knot failed by consistent breakage (average displacement of 4 ⫾ 0.6 mm, SxS//SxS, Table 3). All other configurations demonstrated a high percentage of failure via complete slippage. At ultimate failure, the Revo knot (SxS//SxS) had
61
FIGURE 3. Clinical failure is defined as the loop-holding capacity (maximum load in Newtons), sustained by the braided sutured loop at 3 mm of displacement. All knots with OT⫽S⫽S⫽S and OT⫽SxSxS configurations slipped under cyclic loading.
the highest holding capacity and was statistically different than all other configurations (P ⬍ .0001; Table 3). The configuration with the OT and switchedpost half-hitches was significantly weaker than all other configurations with switched-posts (Table 3, P ⬍ .0001). Pusher Tied Knots Multiple Examiner: The OT configuration (OT⫽S//xS//xS), tied by the most experienced investigator (examiner B), was weaker than all other knots tied by examiners A or C for this specific configuration
TABLE 3. Displacement and Ultimate Failure Data of Braided Knots (Single Examiner)
Knot Configurations OT⫽S⫽S⫽S OT⫽SxSxS OT⫽S//S//S OT⫽S//xS//xS OT⫽S//xS//xS-P DL⫽S⫽S⫽S DL⫽SxSxS DL⫽S//S//S DL⫽S//xS//xS DL⫽S//xS//xS-P TLH⫽S//xS//xS TLH⫽S//xS//xS-P SxS//SxS SxS//SxS-P
Failure Force (N) at 3 mm of displacement (mean ⫾ SD)
Ultimate Failure Force (N) (mean ⫾ SD)
Ultimate Failure Displacement (mm) (mean ⫾ SD)
Failure by Slippage (%)
15.0 ⫾ 0 15.0 ⫾ 0 61.8 ⫾ 4.6 59.0 ⫾ 6.1 65.6 ⫾ 5.0 45.5 ⫾ 8.0 45.2 ⫾ 3.7 92.0 ⫾ 12.7 87.1 ⫾ 12.3 89.0 ⫾ 7.1 88.4 ⫾ 9.8 88.9 ⫾ 10.1 91.8 ⫾ 10.3 85.0 ⫾ 8.6
15.0 ⫾ 0* 15.0 ⫾ 0* 61.7 ⫾ 4.6 59.0 ⫾ 6.1 65.6 ⫾ 5.0 46.5 ⫾ 8.6 46.0 ⫾ 4.0 92.4 ⫾ 12.9 90.9 ⫾ 17.8 96.4 ⫾ 14 90.2 ⫾ 10.3 97.6 ⫾ 19.2 114.1 ⫾ 7.1 100.0 ⫾ 12.0
† † † † † † † † 3.5 ⫾ 1.3 5.1 ⫾ 1.4 3.4 ⫾ 0.5 3.9 ⫾ 1.0 4.0 ⫾ 0.6 4.2 ⫾ 1.2
100 100 100 100 100 100 100 90 80 70 70 40 0 0
NOTE. N ⫽ Newton: 1 lb is equivalent to 4.43 N. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. *Ultimate failure occurred by slippage during cyclical loading for all knots of these configurations. †Complete slippage. n ⫽ 10 knots per configuration.
62
T. D. LOUTZENHEISER ET AL.
(Table 4, P ⬍ .0001). There were no other significant differences in holding capacities between examiners using other configurations (DL⫽S//xS//xS, TLH⫽S// xS//xS, SxS//SxS, Table 4, Fig 4). Hand-tied Versus Pusher-tied Configurations: There were no significant differences in holding capacities between configurations tied by hand and those tied with a knot pusher (Table 3, P ⬎ .2). When comparing the hand-tied with the pusher-tied knots, all configurations showed similar slip failure percentages. Comparative Results of the Best Monofilament and Braided Knots Comparison of Cyclic Loading and Clinical Failure: We collated the mean maximal holding capacity results for monofilament and braided suture at 3 mm of displacement among four knot configurations (Table 5). The values in this table represent the mean holding capacity from a total of 10 hand-tied knots by the most experienced investigator. We did not consider the results of pusher-tied knots because (1) the strand tension for each throw could not be consistently controlled and (2) the results of pusher-tied knots varied significantly among investigators. To complete this table, additional data were acquired for monofilament suture using the TLH (TLH⫽S//xS//xS) and the Revo (SxS//SxS) knots that were not reported previously. The average mean displacement on cyclic loading among the best knots for monofilament PDS II suture was 1.68 ⫾ 0.6 mm (range, 0.08 to 2.8 mm). On comparing clinical failure results for monofilament suture, the DL configuration (DL⫽S//xS//xS), held the greatest average load, (81 N at ⱕ 3 mm of displace-
ment), but statistically, there were no differences among all monofilament knot configurations. Next, we compared the hand-tied results for the identical knot configurations using braided suture. The average mean displacement on cyclic loading among the best knots for braided Ethibond suture was 0.73 mm (range, 0.02 to 1.1 mm). The Revo knot had the highest strength at clinical failure (mean, 92 N) for braided suture; however, there were no significant differences in holding capacities between it and two other knot configurations with switched-post halfhitches (DL⫽S//xS//xS, TLH⫽S//xS//xS, SxS//SxS, Fig 5). The braided OT configuration (OT⫽S//xS//xS), was significantly weaker than all other braided knot configurations under cyclic and clinical failure loads (Table 5, P ⬍ .0001). DISCUSSION In a previous study, we investigated the characteristics of various knot configurations using a monofilament absorbable polydioxanone suture (No. 1 PDS II) under a variety of loading conditions. We found that polydioxanone suture exhibited significant loop elongation under the application of a low magnitude cyclic load. This unwanted lengthening could potentially place a ligamentous or tendon repair at risk for failure during rehabilitation exercises. Therefore, a repair secured with polydioxanone may fail clinically by internal cinching or slippage of the knot or suture-loop stretch elongation under load. To reduce the possibility of a failed repair by loss of suture fixation, surgeons have recommended using a
TABLE 4. Inclusive Failure Data for Pusher-Tied Braided Knots Among Three Examiners
Knot Configuration OT⫽S//xS//xS-P
DL⫽S//xS//xS-P
TLH⫽S//xS//xS-P
SxS//SxS-P
Knot Examiner
Failure at 3 mm Force (N) mean ⫾ SD
Ultimate Failure Force (N) mean ⫾ SD
Ultimate Failure Displacement (mm) ⫾ SD
Failure by Slippage (%)
A B C A B C A B C A B C
82.9 ⫾ 17.6 65.6 ⫾ 4.0 76.7 ⫾ 14.2 99.6 ⫾ 21.9 89.0 ⫾ 7.1 92.5 ⫾ 12.2 98.4 ⫾ 11.7 88.8 ⫾ 10.2 94.4 ⫾ 7.5 97.0 ⫾ 6.7 85.0 ⫾ 8.5 79.9 ⫾ 8.2
82.9 ⫾ 17.6 65.6 ⫾ 4.1 80.0 ⫾ 19.8 105.0 ⫾ 24.4 96.4 ⫾ 14 97.6 ⫾ 10.7 109.7 ⫾ 8.4 97.6 ⫾ 19.2 102.3 ⫾ 14.0 98.6 ⫾ 9.7 100.0 ⫾ 12.0 94.5 ⫾ 16.7
2.1 ⫾ 0.5 * * 3.3 ⫾ 1.3 5.1 ⫾ 1.5 3.5 ⫾ 0.4 3.4 ⫾ 0.9 3.9 ⫾ 0.7 3.7 ⫾ 0.4 3.1 ⫾ 0.5 4.0 ⫾ 1.1 5.6 ⫾ 0.7
80 100 100 70 70 70 60 40 60 40 0 50
NOTE. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. *Complete slippage. n ⫽ 10 knots per examiner per configuration.
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
63
FIGURE 4. The loads recorded at 3 mm of displacement for the pusher-tied configurations.
permanent, nonabsorbable suture material for repairs subject to loading. We used our previous experimental methods to test the holding capacities of knots tied with braided polyester suture to see whether the principles of securing monofilament suture applied to braided suture material. We chose Ethibond, a coated multifilament, nonabsorbable, braided polyester suture. Braiding suture strands increases pliability, ductility, and the overall strength of the final suture construct over similarly sized monofilament suture. Braiding roughens the surface thereby adding to knot security; however, this effect is partly offset by surface coating that reduces friction between strands. The coating allows knot throws to slide down easier but lowers the coefficient of friction, thus increasing knot slippage.8 It is essential for surgeons who wish to perform arthroscopic repairs to tie secure knots to achieve results comparable with open repairs. Previously, we found that when using monofilament suture, loop TABLE 5. Comparative Results of Monofilament and Braided Suture for Hand-Tied Knot Configurations That Held the Greatest Loads and Exhibited the Least Suture Loop Displacement Clinical Failure Force (N) at ⱕ3 mm of Displacement (mean ⫾ SD) Knot Configuration
Monofilament
Braided
OT⫽S//xS//xS DL⫽S//xS//xS TLH⫽S//xS//xS SxS//SxS
71.7 ⫾ 9.0 80.9 ⫾ 10.8 74.2 ⫾ 9.6 61.0 ⫾ 8.9†*
59.0 ⫾ 6.1†* 87.1 ⫾ 12.3 88.4 ⫾ 9.8 91.8 ⫾ 10.3
†Weaker than other knot configurations for the same suture type, P ⬍ .0001 (except for monofilament SxS//SxS v OT⫽S//xS//xS). *Weaker than corresponding knot configuration for alternate suture type, P ⫽ .0002.
FIGURE 5. The above configurations represent the knots with the best strength when tied with braided suture. In the prior study using monofilament suture, the DL⫽S//xS//xS knot configuration showed the best strength (see Table 1 for an explanation of Tera and Aberg coding).
displacement was significantly minimized by tying a complex cinching slip knot secured with nonidentical (reversed direction) half-hitches on alternate strands.2 However, the results of the present study indicate that, for braided suture, reversing the direction of halfhitches does not enhance knot security. When immediate loading of a repair for rehabilitation is desirable, small displacements at the repair site may become critically important. Therefore, the surgeon must select a suture that minimizes material stretch, knot cinching, and slippage. We compared the displacements measured for braided suture loops in this study with those previously reported for similarly sized PDS II monofilament suture. A remarkable difference was found on comparison of cyclic load displacements for half-hitch locking of the DL (Fig 6, P ⫽ .0003). The average displacement among the best knots for braided Ethibond suture (0.73 mm) was less than one half that for monofilament PDS II suture (1.68 mm)! We attribute this difference to knot cinching and slippage and not to stretch elongation of monofilament suture because the cyclic loads were of low magnitude (3 to 6 lb of pull).
64
T. D. LOUTZENHEISER ET AL.
FIGURE 6. This graph provides an example of the relative difference in mean suture loop displacement between monofilament and braided suture for DLs locked with half-hitch combinations and cyclically loaded (alternating 15 and 30 N), under identical test conditions.
We also compared the results of clinical failure for monofilament and braided hand-tied knots. When using PDS II, the DL with switched-post, reversed direction, half-hitch configuration had the highest holding capacity at 3 mm of displacement (mean, 81 N). This configuration, as opposed to others, showed consistent strength characteristics when applied to monofilament and braided suture. In contrast, the Revo knot was the strongest configuration (mean, 92 N) with braided suture but was significantly weaker when tied with PDS II (mean, 61 N, see Table 5 for P values). A knot that fails by shearing of suture material can be just as detrimental as slippage. Investigators have shown that the tensile strength of a suture is much greater than its shear strength.1 Material failure occurs at a point at which the shear stress is the highest. Thus, a knot configuration prone to shear suture material may fail at a lower load than a configuration with minimal shear points. For this reason, we cannot conclude from our data that the Revo knot, which failed by breakage instead of slippage, is an improvement over the load values for ultimate failure recorded for the best TLH and DL configurations locked with half-hitches. However, it does help to explain why certain knots have similar ultimate strength values and yet a different failure mechanism. On review of results for tying monofilament and braided suture, the data presented here and by others confirm that switching the post after each successive half-hitch is the most critical factor affecting the holding capacity of a knot.5,9 In our studies, the switched-post configurations were significantly stron-
ger than non-post-switched configurations regardless of suture material. Since the original Revo knot achieved excellent knot security with only a single switched-post, we conclude that having at least one switched-post is essential for knot security. Adding additional throws with switched-posts, therefore, would only further augment knot security.2-5,9,10 Configurations with the OT were typically weaker than all other knots tied with monofilament and braided suture. When using polydioxanone, the strength of the knots that used the OT were comparable to those with the DL as long as reversed direction half-hitch throws were used to lock the knot. On the other hand, OT configurations showed the most interexaminer variability in both the braided and PDS suture material when tied with a single-hole pusher. Lastly, when working with braided or monofilament suture, the initial cinching knot significantly affected knot strength. We do not recommend using the OT cinching knot. Under our test conditions, the loop-holding capacity of hand tied knots was superior to identical knots tied using a pusher. Similarly, a major portion of the pusher tied knots slipped when tested to ultimate failure, a finding that never occurred for identical hand tied knots. However, these comparisons should be considered in the context of our inability to quantitate and control the magnitude and direction of the force applied for knots tied using a pusher. This fact may partially account for the significant variability between the overhand knots tied by each examiner. Because arthroscopic surgeons must tie their knots with a pusher, the best way to optimize knot security is by careful attention to the technique and sequence of knot throws. Tying knots arthroscopically can be technically challenging and requires considerable practice in a nonsurgical simulated enviornment. Arthroscopic knot tying depends on a multitude of factors such as the material properties, tensile strength, coating of the suture material, the tension applied to make a tight knot, the knot configuration, the tying instruments, and most importantly, the surgeons skill. In this study, we found that large differences in knot security relate to knot configuration yet a specific knot that works well with monofilament suture may not perform well with a braided suture and vice versa. A previous study analyzed five different slip knots tied with five different suture materials.11 None of the knots were strong in all five types of suture. The investigators were unable to find a universally strong knot configuration. Therefore, it is important to compare the
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT strengths of an arthroscopic knot configuration among various types of sutures. A good universal arthroscopic knot would be strong when tied with any suture material. Given this criterion and the available data, the DL locked with half-hitches stands as the best approximation of a universal knot. However, because it is a sliding knot that demands dragging the post strand through tissue, bone, or an anchor eyelet to cinch the loop, it cannot be applied to all clinical repairs. Dragging a suture to cinch a slip knot may also abrade or cut suture fibers or the tissue being repaired. Therefore, we recommend that arthroscopists learn to apply the DL and the Revo knot configurations and to test all other knots before using them clinically. Recognizing the factors that affect knot strength is the first step in developing reliable tying methods. Learning how to apply principles of knot tying appropriately and practicing knot tying technique will maximize the quality of an arthroscopic repair. The following list summarizes our findings to date: 1. Braided suture shows much smaller displacements than monofilament suture when the suture loop is subjected to low level repetitive cyclic loads. 2. When using braided suture, it is not necessary to use reverse-direction half-hitches to lock a cinching knot. 3. The strength of hand-tied and pusher-tied knots was no different for braided suture but differences have been shown for monofilament suture. 4. At least one switched-post is necessary to avoid slippage for all knot configurations regardless of suture material. Adding additional switched-posts augments knot security. 5. The OT cinching slip-knot configuration is not recommended for either monofilament or braided suture. 6. When using monofilament or braided suture, the DL slip knot, locked with switched-post reversed-
65
direction half-hitches will provide reliable holding capacities. (The DL requires that the suture can slide through tissue, bone, or an anchor without sticking or causing damage.) 7. The Revo knot holds well when using braided suture and avoids damage to the suture or tissue that may result from using slip knots that drag the strand through tissue, bone, or anchor eyelet. 8. It is important to tie and test new knot configurations or a knot’s holding capacity with a new suture material before using them clinically.
REFERENCES 1. Dinsmore RC. Understanding surgical knot security: A proposal to standardize the literature. J Am Coll Surg 1995;180:689699. 2. Loutzenheiser TD, Harryman DT II, Yung SW, France MP, Sidles JA. Optimizing arthroscopic knots. Arthroscopy 1995;11: 199-206. 3. Trimbos JB. Security of various knots commonly used in surgical practice. Obstet Gynecol 1984;64:274-280. 4. Tera H, Aberg C. Tensile strengths of twelve types of knot employed in surgery, using different suture materials. Acta Chir Scand 1976;142:1-7. 5. Trimbos JB, Van Rijssel EJC, Klopper PJ. Performance of sliding knots in monofilament and multifilament suture material. Obstet Gynecol 1986;68:425-430. 6. Ray JA, Doddi N, Regula D, Williams JA, Melveger A. Polydioxanone (PDS), a novel monofilament synthetic absorbable suture. Surg Gynecol Obstet 1981;153:497-503. 7. Matsen III FA, Lippitt SB, Sidles JA, Harryman II DT. Practical evaluation of management of the shoulder. Philadelphia: WB Saunders, 1994:19-109. 8. Trimbos JB, Niggebrugge A, Trimbos R, Van R-EJ. Knotting abilities of a new absorbable monofilament suture: poliglecaprone 25 (Monocryl). Eur J Surg 1995;161:319-322. 9. Brouwers JE, Oosting H, de Haas D, Klopper PJ. Dynamic loading of surgical knots. Gynecol Obstet 1991;173:443-447. 10. Herrmann JB. Tensile strength and knot security of surgical suture materials. Am Surgeon 1971;37:209-217. 11. Shimi SM, Lirici M, Vander V-G, Cuschieri A. Comparative study of the holding strength of slipknots using absorbable and non absorbable ligature materials. Surg Endosc 1994;8):12851291.
Summary: Surgeons need to know how the material properties of a suture affect the security of a surgical knot. The purpose of this study was to compare the security of some clinically important arthroscopic knots when tied using a braided multifilament suture and to draw comparisons with results of similar knots tied with monofilament suture. Permanent braided polyester suture was used to test 10 knot configurations. Eight of the knots included (1) two types of initial cinching knots followed by (2) one of four combinations of half-hitches. We also tested the taut-line hitch locked with half-hitches and the original Revo knot. Each knot was subjected to cyclic loading followed by an ultimate load to failure. Clinical failure was defined as the maximum force that resulted in 3 mm of loop displacement. Force versus displacement data were obtained, and the maximal loop holding capacities were compared statistically. The Duncan loop with switched-post half-hitches and the Revo knot (Linvatec, Largo, FL) showed the highest knot-holding capacities (mean, 87N and 92N, respectively) when compared with all other configurations (P ⬍ .0001) for braided suture. A similar knot-holding capacity was described for monofilament suture using the Duncan loop locked with switched-post, reverseddirection half-hitches (mean, 81 N). All knots without post switching slipped completely at significantly lower loads than knots with post switching (monofilament, P ⬍ .001; braided, P ⬍ .0001). When compared with results of knots tied with monofilament suture, the braided switched-post configurations had smaller cyclic displacements (braided, 0.7 mm; monofilament, 1.7 mm). Although the Revo knot showed good strength for braided suture, it was significantly weaker than other configurations when tied with monofilament suture. Therefore, it is important to test the knot strength for a given suture material before applying it clinically. Key Words: Arthroscopic knots—Braided—Monofilament—Cyclic and failure loads.
D
isruption of a suture repair may result from failure of tissue, suture material, or the knot.1 In a previous study, we compared the strength characteristics of eight arthroscopic knot configurations using absorbable polydioxanone monofilament suture.2 AlFrom the Medical College of Vermont, Burlington, Vermont (T.D.L.); the Department of Orthopaedics, The University of Washington, Seattle, Washington (D.T.H.); Blount Orthopaedic Group, Ltd., Milwaukee, Wisconsin (D.W.Z.); and the Department of Orthopaedic Surgery, Singapore General Hospital, Singapore (S.-W.Y). Address correspondence and reprint requests to Douglas T. Harryman III, M.D., Department of Orthopaedics Box356500, The University of Washington, Seattle, WA 98195, U.S.A. r 1998 by the Arthroscopy Association of North America 0749-8063/98/1401-1677$3.00/0
though some of the knots tested appeared similar, we found that significant differences existed simply because of minor variations in the complexity of the initial cinching knot or the pattern of locking halfhitches. Because only one type of suture material was used, it was not necessary to consider differences in holding capacity based on suture material properties such as tensile strength, structural shape, ductility, pliability, coefficient of friction, and others. Properties such as these affect the ultimate strength of a knot and the efficacy of various knot configurations to resist slippage under load.3-5 Arthroscopic surgeons select an absorbable monofilament suture and others a braided material for
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 14, No 1 (January-February), 1998: pp 57–65
57
58
T. D. LOUTZENHEISER ET AL.
specific reasons. Absorbable monofilament is often selected for arthroscopic procedures because it can be advanced through arthroscopic suturing devices, slides easily for tying slip knots and it maintains reasonable strength while resorption and healing occurs. Under loaded conditions, however, absorbable monofilament can elongate up to 30% of its length and slips more easily than braided suture.6 Braided suture is often preferred to avoid knot slippage, suture stretch, or breakage, which would otherwise risk loss of tissue apposition. A braided polyester suture is also chosen because it remains in situ, is able to resist a load application during all phases of healing, and because it has been effective in many open procedures7 (e.g., Bankart and rotator cuff repair). The purpose of this study was to determine whether (1) the principles and methods of tying monofilament suture documented in our previous study apply to braided suture, (2) current clinical methods of tying braided arthroscopic knots provide secure fixation, (3) the holding capacities of braided knots tied with a knot pusher are comparable to those tied by hand, and (4) the strongest monofilament and braided suture knots have comparable holding capacities for the same size suture material. METHODS Suture and Knot Configurations All knots in this study were tied using No. 1 braided suture (Ethibond; Ethicon, Somerville, NJ). This suture was selected because braided polyester is a standard suture used for both open and arthroscopic repairs, and its size is comparable to the previously tested No. 1 polydioxanone, (PDS II; Ethicon), which would allow comparison of holding capacities under cyclic and clinical failure displacement loads to the authors previous study. The current study was completed in 1994 following monofilament suture testing using nearly the same methods and examiners as the present study. One investigator hand-tied a total of 10 different knot configurations, 8 of which were identical to those tied in our previous study. In the previous study, the knots consisted of 2 different initial cinching knots, (Duncan loop [DL], overhand throw [OT], Fig 1A), followed by 4 different configurations of locking half-hitches (Fig 1B, 1C). These combinations resulted in a total of 8 configurations (Table 1). In this study, 2 additional configurations were tested. The first consisted of the taut-line hitch (TLH) initial cinching knot locked with three nonidentical switched-post halfhitches (TLH⫽S//xS//xS Fig 1A, 1C, Table 1). The
Tera & Aberg coding system was extended to include the TLH initial cinching knot (TLH, Table 1).4 The second new knot, the original Revo configuration (Linvatec, Largo, FL), consisted of two pairs of nonidentical half-hitches connected via a switchedpost (Fig 1D, Table 1). The Revo knot is represented by the standard coding system (SxS//SxS - Table 1). After running statistical analyses on the hand-tied data, four configurations with the highest holding capacities were selected for the multiple tyer experiment. High strengths for the DL, TLH, and OT, all locked with switched-posts and reversed loops as well as the original Revo knot qualified each configuration for the multiple tyer experiment. Tying and Testing Procedure The tying and testing techniques used in this study were identical to those described in our previous study.2 The apparatus, calibrations, tying methods, strand tensioning, knot preloading, load applications, and quantification of loop displacement for cyclic, clinical failure, and ultimate failure were duplicated for the braided and additional monofilament knots in this study. Force versus displacement data were recorded for 10 knots of each configuration, and compared statistically using Dunnett’s t-test to adjust for multiple comparisons. Values of P ⬍ .05 were considered statistically significant. Clinical failure at 3 mm of displacement was defined as the minimum force required to distract the loop to a displacement of 3 mm. The ultimate force was defined as the minimum force required to cause total failure either by complete slippage or material breakage. Each of the four throws in the REVO knot were tied and tightened separately (SxS//SxS, Table 1, Fig 1D) = FIGURE 1. (A) The slip knot configurations shown above were used initially to remove slack from the suture loop. The OT configuration designates the simple OT, the TLH configuration designates the intermediate TLH, and the DL configuration designates the complex DL configuration. (B) The half-hitch configurations shown above represent two combinations that do not incorporate post switching. The labels ‘‘under’’ and ‘‘over’’ indicate the direction for the initial pass around the post when making each half-hitch loop. Note that the direction of the half-hitch loops are reversed for the SxSxS configuration (see Table 1 for an explanation of the Tera and Aberg coding system). (C) The half-hitch configurations shown above represent two combinations that incorporate post switching and reversal of the loop direction. Note that the S//xS//xS configuration combines post switching and loop direction reversal (see Table 1 for an explanation of Tera and Aberg coding). (D) The Original Revo knot configuration shown above represents a combination with reversed direction half-hitches with a switched-post after the second throw. SxS//SxS (see Table 1 for an explanation of Tera and Aberg coding).
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
B
59
60
T. D. LOUTZENHEISER ET AL. TABLE 1. Knot Configurations: Tera and Aberg Notation13
Symbol OT DL TLH S ⫽ x // //x
Description Overhand-throw sliding knot Duncan loop sliding knot Tautline Hitch Single sliding throw (half-hitch) knots OT, DL, TLH slip knots ⫹3 half-hitches Identical throw, loop direction same Non-identical throw, loop reversed Alternating post between throw Alternating & reversing posts
Knot Configuration OT⫽S⫽S⫽S DL⫽S⫽S⫽S TLH⫽S⫽S⫽S* OT⫽SxSxS DL⫽SxSxS OT⫽S//S//S DL⫽S//S//S OT⫽S//xS//xS DL⫽S//xS//xS TLH⫽S//xS//xS SxS//SxS (Revo)
Description Identical half-hitches, loop in same direction around a single post. Non-Identical half-hitches, loop in opposite direction around same post. Identical half-hitches, post is alternated after each throw. Non-identical (reversed-throw), half-hitches looped around alternating posts. Revo Knot: two pairs of nonidentical half-hitches with one alternated post.
NOTE. ‘‘//x’’ symbol is equivalent to the ‘‘#’’ symbol used by other authors.1 *Not tested.
using the same technique for placing the locking half-hitches of the other configurations. The TLH, Table 1, Fig 1A) was used as the initial cinching knot with locking half-hitches in the TLH⫽S//xS//xS configuration. For the multiple tyer experiment, three practiced coauthor examiners, tied a total of 10 knots for each of the four configurations. The knots were tied with a single-hole knot pusher through a 7-mm cannula. The best knots for monofilament and braided suture were compared for each group using ANOVA and an F-test. RESULTS Hand Tied Knots Cyclic Loading: All configurations with nonswitched-post half-hitches applied after the OT cinching knot (OT⫽S⫽S⫽S, OT⫽SxSxS) slipped completely during the first few repetitions of the cyclic load test (Table 2, Fig 2). Knots with the more complex DL initial cinching knot and non-switched-post halfhitches (DL⫽S⫽S⫽S, DL⫽SxSxS) had displacements ⱕ 1.1 mm (Table 2 , Fig 2). All switched-post configurations (S//S or S//xS), distracted less than 1 mm during cyclic loading resulting in no significant differences between the cyclic displacements for these knots (P ⬎ .2). Load to Clinical Failure (3 mm of Displacement): There were no significant differences in strength between configurations with reversed half-hitches (designated by ‘x’) or without reversed direction (designated by ‘⫽’) half-hitches (e.g., OT⫽S⫽S⫽S v OT⫽SxSxS, DL⫽S⫽S⫽S v DL⫽SxSxS, OT⫽S// S//S v OT⫽S//xS//xS, DL⫽S//S//S v DL⫽S//xS//xS; Table 3, P ⬎ .2). Configurations without switched-posts, (i.e., no ‘//’
designation; DL⫽S⫽S⫽S, DL⫽SxSxS, OT⫽S⫽S⫽S, OT⫽SxSxS), and the knots with the less complex OT initial cinching knots were significantly weaker than those with the DL (Table 3 and Fig 3, P ⬍ .0001). Similarly, the switched-post configurations with the OT, (OT⫽S//S//S, OT⫽S//xS//xS), were significantly weaker than all other configurations with switchedposts (Table 3, Fig 3, P ⬍ .0001). All configurations containing at least one switchedpost were significantly stronger than configurations without a switched-post (Table 3, Fig 3). There were no significant differences in holding capacities beTABLE 2. Cyclic Loading Data for Braided Suture Loop Displacement (mm) Knot Configurations
15-N Load (mean ⫾ SD)
30-N Load (mean ⫾ SD)
⌬30-15 N (mean ⫾ SD)
OT⫽S⫽S⫽S OT⫽SxSxS OT⫽S//S//S OT⫽S//xS//xS OT⫽S//xS//xS-P DL⫽S⫽S⫽S DL⫽SxSxS DL⫽S//S//S DL⫽S//xS//xS DL⫽S//xS//xS-P TLH⫽S//xS//xS TLH⫽S//xS//xS-P SxS//SxS SxS//SxS-P
* * 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.2 ⫾ 0.2 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.1 ⫾ 0.1 0.2 ⫾ 0.1 0.2 ⫾ 0.1 0.2 ⫾ 0.1
* * 0.6 ⫾ 0.2 0.7 ⫾ 0.2 0.5 ⫾ 0.1 0.8 ⫾ 0.4 1.1 ⫾ 0.7* 0.6 ⫾ 0.2 0.5 ⫾ 0.2 0.6 ⫾ 0.1 0.7 ⫾ 0.2 0.6 ⫾ 0.1 0.8 ⫾ 0.2 0.7 ⫾ 0.1
* * 0.5 ⫾ 0.2 0.6 ⫾ 0.2 0.3 ⫾ 0.1 0.7 ⫾ 0.4 1.0 ⫾ 0.7* 0.5 ⫾ 0.2 0.4 ⫾ 0.2 0.4 ⫾ 0.1 0.6 ⫾ 0.2 0.4 ⫾ 0.1 0.6 ⫾ 0.2 0.5 ⫾ 0.1
NOTE. n ⫽ 10 knots per configuration. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. Standard deviations approaching or equivalent to the mean reflect loop displacements recorded at low cyclic loads and the limit of accurate measurement rounded to the nearest tenth. *Denotes complete slippage or outliers causing a wide standard deviation.
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
FIGURE 2. Displacement of braided knots under cyclic loads of 15 and 30 N. There were no significant differences among the knots that displaced less than 0.8 mm.
tween all configurations with switched-post halfhitches (DL⫽S//S//S, DL⫽S//xS//xS, TLH⫽S//xS// xS, SxS//SxS; P ⬎ .2). Ultimate Failure Load: All knots ultimately failed after a displacement of greater than 3 mm. Of all the knots tested, only the Revo knot failed by consistent breakage (average displacement of 4 ⫾ 0.6 mm, SxS//SxS, Table 3). All other configurations demonstrated a high percentage of failure via complete slippage. At ultimate failure, the Revo knot (SxS//SxS) had
61
FIGURE 3. Clinical failure is defined as the loop-holding capacity (maximum load in Newtons), sustained by the braided sutured loop at 3 mm of displacement. All knots with OT⫽S⫽S⫽S and OT⫽SxSxS configurations slipped under cyclic loading.
the highest holding capacity and was statistically different than all other configurations (P ⬍ .0001; Table 3). The configuration with the OT and switchedpost half-hitches was significantly weaker than all other configurations with switched-posts (Table 3, P ⬍ .0001). Pusher Tied Knots Multiple Examiner: The OT configuration (OT⫽S//xS//xS), tied by the most experienced investigator (examiner B), was weaker than all other knots tied by examiners A or C for this specific configuration
TABLE 3. Displacement and Ultimate Failure Data of Braided Knots (Single Examiner)
Knot Configurations OT⫽S⫽S⫽S OT⫽SxSxS OT⫽S//S//S OT⫽S//xS//xS OT⫽S//xS//xS-P DL⫽S⫽S⫽S DL⫽SxSxS DL⫽S//S//S DL⫽S//xS//xS DL⫽S//xS//xS-P TLH⫽S//xS//xS TLH⫽S//xS//xS-P SxS//SxS SxS//SxS-P
Failure Force (N) at 3 mm of displacement (mean ⫾ SD)
Ultimate Failure Force (N) (mean ⫾ SD)
Ultimate Failure Displacement (mm) (mean ⫾ SD)
Failure by Slippage (%)
15.0 ⫾ 0 15.0 ⫾ 0 61.8 ⫾ 4.6 59.0 ⫾ 6.1 65.6 ⫾ 5.0 45.5 ⫾ 8.0 45.2 ⫾ 3.7 92.0 ⫾ 12.7 87.1 ⫾ 12.3 89.0 ⫾ 7.1 88.4 ⫾ 9.8 88.9 ⫾ 10.1 91.8 ⫾ 10.3 85.0 ⫾ 8.6
15.0 ⫾ 0* 15.0 ⫾ 0* 61.7 ⫾ 4.6 59.0 ⫾ 6.1 65.6 ⫾ 5.0 46.5 ⫾ 8.6 46.0 ⫾ 4.0 92.4 ⫾ 12.9 90.9 ⫾ 17.8 96.4 ⫾ 14 90.2 ⫾ 10.3 97.6 ⫾ 19.2 114.1 ⫾ 7.1 100.0 ⫾ 12.0
† † † † † † † † 3.5 ⫾ 1.3 5.1 ⫾ 1.4 3.4 ⫾ 0.5 3.9 ⫾ 1.0 4.0 ⫾ 0.6 4.2 ⫾ 1.2
100 100 100 100 100 100 100 90 80 70 70 40 0 0
NOTE. N ⫽ Newton: 1 lb is equivalent to 4.43 N. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. *Ultimate failure occurred by slippage during cyclical loading for all knots of these configurations. †Complete slippage. n ⫽ 10 knots per configuration.
62
T. D. LOUTZENHEISER ET AL.
(Table 4, P ⬍ .0001). There were no other significant differences in holding capacities between examiners using other configurations (DL⫽S//xS//xS, TLH⫽S// xS//xS, SxS//SxS, Table 4, Fig 4). Hand-tied Versus Pusher-tied Configurations: There were no significant differences in holding capacities between configurations tied by hand and those tied with a knot pusher (Table 3, P ⬎ .2). When comparing the hand-tied with the pusher-tied knots, all configurations showed similar slip failure percentages. Comparative Results of the Best Monofilament and Braided Knots Comparison of Cyclic Loading and Clinical Failure: We collated the mean maximal holding capacity results for monofilament and braided suture at 3 mm of displacement among four knot configurations (Table 5). The values in this table represent the mean holding capacity from a total of 10 hand-tied knots by the most experienced investigator. We did not consider the results of pusher-tied knots because (1) the strand tension for each throw could not be consistently controlled and (2) the results of pusher-tied knots varied significantly among investigators. To complete this table, additional data were acquired for monofilament suture using the TLH (TLH⫽S//xS//xS) and the Revo (SxS//SxS) knots that were not reported previously. The average mean displacement on cyclic loading among the best knots for monofilament PDS II suture was 1.68 ⫾ 0.6 mm (range, 0.08 to 2.8 mm). On comparing clinical failure results for monofilament suture, the DL configuration (DL⫽S//xS//xS), held the greatest average load, (81 N at ⱕ 3 mm of displace-
ment), but statistically, there were no differences among all monofilament knot configurations. Next, we compared the hand-tied results for the identical knot configurations using braided suture. The average mean displacement on cyclic loading among the best knots for braided Ethibond suture was 0.73 mm (range, 0.02 to 1.1 mm). The Revo knot had the highest strength at clinical failure (mean, 92 N) for braided suture; however, there were no significant differences in holding capacities between it and two other knot configurations with switched-post halfhitches (DL⫽S//xS//xS, TLH⫽S//xS//xS, SxS//SxS, Fig 5). The braided OT configuration (OT⫽S//xS//xS), was significantly weaker than all other braided knot configurations under cyclic and clinical failure loads (Table 5, P ⬍ .0001). DISCUSSION In a previous study, we investigated the characteristics of various knot configurations using a monofilament absorbable polydioxanone suture (No. 1 PDS II) under a variety of loading conditions. We found that polydioxanone suture exhibited significant loop elongation under the application of a low magnitude cyclic load. This unwanted lengthening could potentially place a ligamentous or tendon repair at risk for failure during rehabilitation exercises. Therefore, a repair secured with polydioxanone may fail clinically by internal cinching or slippage of the knot or suture-loop stretch elongation under load. To reduce the possibility of a failed repair by loss of suture fixation, surgeons have recommended using a
TABLE 4. Inclusive Failure Data for Pusher-Tied Braided Knots Among Three Examiners
Knot Configuration OT⫽S//xS//xS-P
DL⫽S//xS//xS-P
TLH⫽S//xS//xS-P
SxS//SxS-P
Knot Examiner
Failure at 3 mm Force (N) mean ⫾ SD
Ultimate Failure Force (N) mean ⫾ SD
Ultimate Failure Displacement (mm) ⫾ SD
Failure by Slippage (%)
A B C A B C A B C A B C
82.9 ⫾ 17.6 65.6 ⫾ 4.0 76.7 ⫾ 14.2 99.6 ⫾ 21.9 89.0 ⫾ 7.1 92.5 ⫾ 12.2 98.4 ⫾ 11.7 88.8 ⫾ 10.2 94.4 ⫾ 7.5 97.0 ⫾ 6.7 85.0 ⫾ 8.5 79.9 ⫾ 8.2
82.9 ⫾ 17.6 65.6 ⫾ 4.1 80.0 ⫾ 19.8 105.0 ⫾ 24.4 96.4 ⫾ 14 97.6 ⫾ 10.7 109.7 ⫾ 8.4 97.6 ⫾ 19.2 102.3 ⫾ 14.0 98.6 ⫾ 9.7 100.0 ⫾ 12.0 94.5 ⫾ 16.7
2.1 ⫾ 0.5 * * 3.3 ⫾ 1.3 5.1 ⫾ 1.5 3.5 ⫾ 0.4 3.4 ⫾ 0.9 3.9 ⫾ 0.7 3.7 ⫾ 0.4 3.1 ⫾ 0.5 4.0 ⫾ 1.1 5.6 ⫾ 0.7
80 100 100 70 70 70 60 40 60 40 0 50
NOTE. Knot configuration codes with a ⫺P at the end indicate knots tied with a pusher. *Complete slippage. n ⫽ 10 knots per examiner per configuration.
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT
63
FIGURE 4. The loads recorded at 3 mm of displacement for the pusher-tied configurations.
permanent, nonabsorbable suture material for repairs subject to loading. We used our previous experimental methods to test the holding capacities of knots tied with braided polyester suture to see whether the principles of securing monofilament suture applied to braided suture material. We chose Ethibond, a coated multifilament, nonabsorbable, braided polyester suture. Braiding suture strands increases pliability, ductility, and the overall strength of the final suture construct over similarly sized monofilament suture. Braiding roughens the surface thereby adding to knot security; however, this effect is partly offset by surface coating that reduces friction between strands. The coating allows knot throws to slide down easier but lowers the coefficient of friction, thus increasing knot slippage.8 It is essential for surgeons who wish to perform arthroscopic repairs to tie secure knots to achieve results comparable with open repairs. Previously, we found that when using monofilament suture, loop TABLE 5. Comparative Results of Monofilament and Braided Suture for Hand-Tied Knot Configurations That Held the Greatest Loads and Exhibited the Least Suture Loop Displacement Clinical Failure Force (N) at ⱕ3 mm of Displacement (mean ⫾ SD) Knot Configuration
Monofilament
Braided
OT⫽S//xS//xS DL⫽S//xS//xS TLH⫽S//xS//xS SxS//SxS
71.7 ⫾ 9.0 80.9 ⫾ 10.8 74.2 ⫾ 9.6 61.0 ⫾ 8.9†*
59.0 ⫾ 6.1†* 87.1 ⫾ 12.3 88.4 ⫾ 9.8 91.8 ⫾ 10.3
†Weaker than other knot configurations for the same suture type, P ⬍ .0001 (except for monofilament SxS//SxS v OT⫽S//xS//xS). *Weaker than corresponding knot configuration for alternate suture type, P ⫽ .0002.
FIGURE 5. The above configurations represent the knots with the best strength when tied with braided suture. In the prior study using monofilament suture, the DL⫽S//xS//xS knot configuration showed the best strength (see Table 1 for an explanation of Tera and Aberg coding).
displacement was significantly minimized by tying a complex cinching slip knot secured with nonidentical (reversed direction) half-hitches on alternate strands.2 However, the results of the present study indicate that, for braided suture, reversing the direction of halfhitches does not enhance knot security. When immediate loading of a repair for rehabilitation is desirable, small displacements at the repair site may become critically important. Therefore, the surgeon must select a suture that minimizes material stretch, knot cinching, and slippage. We compared the displacements measured for braided suture loops in this study with those previously reported for similarly sized PDS II monofilament suture. A remarkable difference was found on comparison of cyclic load displacements for half-hitch locking of the DL (Fig 6, P ⫽ .0003). The average displacement among the best knots for braided Ethibond suture (0.73 mm) was less than one half that for monofilament PDS II suture (1.68 mm)! We attribute this difference to knot cinching and slippage and not to stretch elongation of monofilament suture because the cyclic loads were of low magnitude (3 to 6 lb of pull).
64
T. D. LOUTZENHEISER ET AL.
FIGURE 6. This graph provides an example of the relative difference in mean suture loop displacement between monofilament and braided suture for DLs locked with half-hitch combinations and cyclically loaded (alternating 15 and 30 N), under identical test conditions.
We also compared the results of clinical failure for monofilament and braided hand-tied knots. When using PDS II, the DL with switched-post, reversed direction, half-hitch configuration had the highest holding capacity at 3 mm of displacement (mean, 81 N). This configuration, as opposed to others, showed consistent strength characteristics when applied to monofilament and braided suture. In contrast, the Revo knot was the strongest configuration (mean, 92 N) with braided suture but was significantly weaker when tied with PDS II (mean, 61 N, see Table 5 for P values). A knot that fails by shearing of suture material can be just as detrimental as slippage. Investigators have shown that the tensile strength of a suture is much greater than its shear strength.1 Material failure occurs at a point at which the shear stress is the highest. Thus, a knot configuration prone to shear suture material may fail at a lower load than a configuration with minimal shear points. For this reason, we cannot conclude from our data that the Revo knot, which failed by breakage instead of slippage, is an improvement over the load values for ultimate failure recorded for the best TLH and DL configurations locked with half-hitches. However, it does help to explain why certain knots have similar ultimate strength values and yet a different failure mechanism. On review of results for tying monofilament and braided suture, the data presented here and by others confirm that switching the post after each successive half-hitch is the most critical factor affecting the holding capacity of a knot.5,9 In our studies, the switched-post configurations were significantly stron-
ger than non-post-switched configurations regardless of suture material. Since the original Revo knot achieved excellent knot security with only a single switched-post, we conclude that having at least one switched-post is essential for knot security. Adding additional throws with switched-posts, therefore, would only further augment knot security.2-5,9,10 Configurations with the OT were typically weaker than all other knots tied with monofilament and braided suture. When using polydioxanone, the strength of the knots that used the OT were comparable to those with the DL as long as reversed direction half-hitch throws were used to lock the knot. On the other hand, OT configurations showed the most interexaminer variability in both the braided and PDS suture material when tied with a single-hole pusher. Lastly, when working with braided or monofilament suture, the initial cinching knot significantly affected knot strength. We do not recommend using the OT cinching knot. Under our test conditions, the loop-holding capacity of hand tied knots was superior to identical knots tied using a pusher. Similarly, a major portion of the pusher tied knots slipped when tested to ultimate failure, a finding that never occurred for identical hand tied knots. However, these comparisons should be considered in the context of our inability to quantitate and control the magnitude and direction of the force applied for knots tied using a pusher. This fact may partially account for the significant variability between the overhand knots tied by each examiner. Because arthroscopic surgeons must tie their knots with a pusher, the best way to optimize knot security is by careful attention to the technique and sequence of knot throws. Tying knots arthroscopically can be technically challenging and requires considerable practice in a nonsurgical simulated enviornment. Arthroscopic knot tying depends on a multitude of factors such as the material properties, tensile strength, coating of the suture material, the tension applied to make a tight knot, the knot configuration, the tying instruments, and most importantly, the surgeons skill. In this study, we found that large differences in knot security relate to knot configuration yet a specific knot that works well with monofilament suture may not perform well with a braided suture and vice versa. A previous study analyzed five different slip knots tied with five different suture materials.11 None of the knots were strong in all five types of suture. The investigators were unable to find a universally strong knot configuration. Therefore, it is important to compare the
ARTHROSCOPIC KNOTS: BRAIDED OR MONOFILAMENT strengths of an arthroscopic knot configuration among various types of sutures. A good universal arthroscopic knot would be strong when tied with any suture material. Given this criterion and the available data, the DL locked with half-hitches stands as the best approximation of a universal knot. However, because it is a sliding knot that demands dragging the post strand through tissue, bone, or an anchor eyelet to cinch the loop, it cannot be applied to all clinical repairs. Dragging a suture to cinch a slip knot may also abrade or cut suture fibers or the tissue being repaired. Therefore, we recommend that arthroscopists learn to apply the DL and the Revo knot configurations and to test all other knots before using them clinically. Recognizing the factors that affect knot strength is the first step in developing reliable tying methods. Learning how to apply principles of knot tying appropriately and practicing knot tying technique will maximize the quality of an arthroscopic repair. The following list summarizes our findings to date: 1. Braided suture shows much smaller displacements than monofilament suture when the suture loop is subjected to low level repetitive cyclic loads. 2. When using braided suture, it is not necessary to use reverse-direction half-hitches to lock a cinching knot. 3. The strength of hand-tied and pusher-tied knots was no different for braided suture but differences have been shown for monofilament suture. 4. At least one switched-post is necessary to avoid slippage for all knot configurations regardless of suture material. Adding additional switched-posts augments knot security. 5. The OT cinching slip-knot configuration is not recommended for either monofilament or braided suture. 6. When using monofilament or braided suture, the DL slip knot, locked with switched-post reversed-
65
direction half-hitches will provide reliable holding capacities. (The DL requires that the suture can slide through tissue, bone, or an anchor without sticking or causing damage.) 7. The Revo knot holds well when using braided suture and avoids damage to the suture or tissue that may result from using slip knots that drag the strand through tissue, bone, or anchor eyelet. 8. It is important to tie and test new knot configurations or a knot’s holding capacity with a new suture material before using them clinically.
REFERENCES 1. Dinsmore RC. Understanding surgical knot security: A proposal to standardize the literature. J Am Coll Surg 1995;180:689699. 2. Loutzenheiser TD, Harryman DT II, Yung SW, France MP, Sidles JA. Optimizing arthroscopic knots. Arthroscopy 1995;11: 199-206. 3. Trimbos JB. Security of various knots commonly used in surgical practice. Obstet Gynecol 1984;64:274-280. 4. Tera H, Aberg C. Tensile strengths of twelve types of knot employed in surgery, using different suture materials. Acta Chir Scand 1976;142:1-7. 5. Trimbos JB, Van Rijssel EJC, Klopper PJ. Performance of sliding knots in monofilament and multifilament suture material. Obstet Gynecol 1986;68:425-430. 6. Ray JA, Doddi N, Regula D, Williams JA, Melveger A. Polydioxanone (PDS), a novel monofilament synthetic absorbable suture. Surg Gynecol Obstet 1981;153:497-503. 7. Matsen III FA, Lippitt SB, Sidles JA, Harryman II DT. Practical evaluation of management of the shoulder. Philadelphia: WB Saunders, 1994:19-109. 8. Trimbos JB, Niggebrugge A, Trimbos R, Van R-EJ. Knotting abilities of a new absorbable monofilament suture: poliglecaprone 25 (Monocryl). Eur J Surg 1995;161:319-322. 9. Brouwers JE, Oosting H, de Haas D, Klopper PJ. Dynamic loading of surgical knots. Gynecol Obstet 1991;173:443-447. 10. Herrmann JB. Tensile strength and knot security of surgical suture materials. Am Surgeon 1971;37:209-217. 11. Shimi SM, Lirici M, Vander V-G, Cuschieri A. Comparative study of the holding strength of slipknots using absorbable and non absorbable ligature materials. Surg Endosc 1994;8):12851291.