- Email: [email protected]
Accuracy of Visual Estimates of Partial Flexor Tendon Lacerations
SCIENTIFIC ARTICLE
Accuracy of Visual Estimates of Partial Flexor Tendon Lacerations Tamir Pritsch, MD,* Corrine Wong, MD,† Douglas M. Sammer, MD†
Purpose To determine whether hand surgeons could accurately and consistently estimate the size of partial flexor tendon lacerations. Materials and methods Thirty-two partial flexor tendon lacerations were made in the flexor digitorum profundus tendons of a fresh-frozen cadaveric hand. Four hand surgeons and 5 residents estimated the size of the lacerations. Estimates were repeated 3 days later. Magnified images of the laceration cross-section were used to calculate the true size of each laceration. Inter- and intrarater reliability were calculated using the intraclass correlation coefficient. Accuracy was measured with the mean bias error and the mean absolute error. Results Interrater and intrarater reliabilities were both high. There was a high level of consistency for both surgeons and residents. In terms of accuracy, there was a 3% bias toward underestimation. The mean absolute error was 11%. There was no statistically significant difference between the accuracy of attending hand surgeons and that of residents. Participants were less accurate when estimating lacerations close to a 60% laceration threshold for surgical repair (lacerations in the 50%e70% range). For lacerations within this range, an incorrect management decision would have been made 17% of the time, compared with 7% of the time for lacerations outside that range. Conclusions The accuracy and reliability of surgeon estimates of partial flexor tendon laceration size were high for surgeons and residents. Accuracy was lower for lacerations close to the threshold for repair. Clinical relevance Visual estimation is acceptable for evaluating partial flexor tendon lacerations, but it may be less reliable for lacerations near the threshold for repair. Therefore, surgeons should be cautious when deciding whether or not to repair partial lacerations in the borderline range. (J Hand Surg Am. 2015;40(12):2421e2426. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Key words Flexor, tendon, laceration, estimation, partial.
From the *Division of Hand Surgery, Department of Orthopedic Surgery, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel; and the †Department of Plastic Surgery, University of Texas Southwestern Medical School, Dallas, TX. Received for publication January 26, 2015; accepted in revised form September 8, 2015. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Douglas M. Sammer, MD, University of Texas Southwestern Medical School, 1801 Inwood Rd., Dallas, TX 75390; e-mail: Douglas.Sammer@ UTSouthwestern.edu. 0363-5023/15/4012-0015$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2015.09.009
R
EPAIRING A MINOR PARTIAL FLEXOR tendon laceration can weaken the tensile strength of the tendon and can increase bulk resulting in poor gliding and increased work of flexion.1e3 Therefore, only substantial lacerations that weaken the tendon to the point that it is at risk for rupture (or that cause catching on the tendon sheath) should be repaired.4e8 The recommended threshold for repairing a partial flexor tendon laceration varies. Many authors advocate repair if the laceration involves more than 50% to 60% of the tendon’s
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
2421
2422
ESTIMATION OF PARTIAL TENDON LACERATIONS
cross-sectional area,9e13 but recommendations range from 30% to 95%.14e18 The purpose of this study was to determine whether hand surgeons could accurately and consistently estimate the size of partial flexor tendon lacerations. MATERIALS AND METHODS Partial flexor tendon laceration model One fresh-frozen hand was obtained from the institution’s willed body program and allowed to thaw for 24 hours. Bruner incisions were made over volar aspect of the index, middle, ring, and little fingers to expose the flexor tendon sheaths from the distal phalanges into the midpalm. The flexor tendon sheaths were divided longitudinally, exposing the flexor tendons. Using a scalpel, 8 partial flexor tendon lacerations of varying sizes were made in the flexor digitorum profundus tendons of each finger from zone I through zone III, creating a total of 32 partial lacerations. The lacerations were oriented transversely to the long axis of the tendon. A random number was used to generate 32 percentage values, ranging between 1% and 99%. An effort was made to create lacerations that approximated the depth assigned by the random number generator. In this way, 32 partial flexor tendon lacerations were created with a wide range of laceration depths. The true depth of each laceration was unknown at the time of estimation and was precisely measured later in the experiment. Estimation of laceration size Four hand surgeons and 5 plastic surgery residents (postgraduate year 2, n ¼ 2; 3, n ¼ 2; 4, n ¼ 1) estimated the size (percent involvement of cross-sectional surface area) of the lacerations under loupe magnification and with the use of standard surgical instruments including calipers and ruler, although none of the participants ended up using the calipers. Participants were invited to repeat the size estimates in order to evaluate intrarater reliability. The second session was performed 3 days after the first session and the tendon lacerations were evaluated in reverse order to reduce recall bias. All aspects of the study were approved by our institutional review board. Measurement of true laceration size The flexor digitorum profundus tendons containing the lacerations were then removed from the hand and fixed in formalin (10% neutral buffered formalin; 24 hours). The laceration surfaces were painted with black ink (black permanent ink; Newell Rubbermaid Inc., Freeport, IL). Ink was used rather than a histological stain in J Hand Surg Am.
r
FIGURE 1: Magnified photograph of the tendon cross-section through the partial laceration. The inked surface represents the partial laceration.
order to avoid stain penetration outside the zone of laceration. Next, using a scalpel, the tendon was completely divided through each laceration. This converted each partial laceration into a complete laceration with an inked surface representing the surface area of the partial laceration (Fig. 1). Magnified electronic images of each inked tendon cross-section were obtained, and digital image processing software (Photoshop; Adobe Systems Inc., San Jose, CA) was used to calculate the percentage of the total crosssectional surface area of each partial tendon laceration. Statistical methods Inter- and intrarater reliability were measured by calculating the intraclass correlation coefficient (ICC, 2,1) for absolute agreement.19 Interrater reliability was calculated for all participants and for surgeon and resident subgroups separately. Intrarater reliability was calculated for all participants separately. An ICC less than 0.40 was considered poor, an ICC between 0.40 and 0.59 was considered fair, an ICC between 0.60 and 0.74 was considered good, and an ICC greater than 0.74 was considered excellent.19,20 Accuracy was measured by calculating the mean bias error (MBE) and the mean absolute error (MAE). Accuracy was measured for all participants and for the surgeon and resident subgroups separately. Comparisons of MBE and MAE were performed using the Student t test, and P values less than .05 were considered statistically significant. Finally, the estimation accuracy for lacerations that were close to the 60% threshold for surgical repair (50%e70% true size range) were analyzed separately and compared with lacerations that fell outside this range. This comparison was performed in order to Vol. 40, December 2015
ESTIMATION OF PARTIAL TENDON LACERATIONS
2423
FIGURE 2: A scatter plot of estimated laceration size versus true laceration size. The black line represents a linear regression. The red line represents perfect accuracy, where the estimated and true lacerations sizes are equal. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
determine whether estimation accuracy is different for lacerations that are close to the threshold for surgical repair compared with lacerations that were obviously above or below the threshold and what effect that may have on surgical decision making. RESULTS Nine participants (5 residents and 4 attending hand surgeons) completed the initial estimation session. Eight of the study participants (5 of the residents and 3 of the attending hand surgeons) completed the second estimation session. Reliability The overall interrater reliability was excellent (ICC, 0.88). When the attending hand surgeon and resident subgroups were analyzed separately, the interrater reliability remained excellent for both attending hand surgeons (ICC, 0.88) and residents (ICC, 0.87). The intrarater reliability of all participants, including both attending hand surgeons and residents, fell within the excellent range with ICCs greater than 0.74 and was 0.91 for all 3 attending hand surgeons and ranged from 0.78 to 0.91 for the 5 residents. J Hand Surg Am.
r
Accuracy Figure 2 demonstrates size estimates plotted against true laceration size. The MBE of all participants was e3%, demonstrating a slight bias toward underestimation (Fig. 3). The MAE was 11% (Fig. 4). The MBE and MAE were not significantly different for attending surgeons compared with residents (MBE e1.9 vs e3.7, P ¼ .85; MAE 10.8 vs 11.0, P ¼ .74). The MAE for lacerations in the 50% to 70% size range, which were close to the 60% threshold for surgical repair, was 13%, compared with an MAE of 10% for lacerations outside this range (P ¼ .015) (Fig. 5). For lacerations in the 50% to 70% size range, an incorrect management decision would have been made 17% of the time owing size estimation error. For lacerations outside the 50% to 70% range, an error in management would have been made 7% of the time. DISCUSSION Small partial flexor tendon lacerations do not require repair, and in fact, repairing a small laceration may weaken the tendon or increase its bulk. Conversely, large partial lacerations should be repaired to avoid Vol. 40, December 2015
2424
ESTIMATION OF PARTIAL TENDON LACERATIONS
FIGURE 3: Graph represents estimation error plotted against true laceration size. The red line represents zero estimation error. The black line represents the mean bias error of e3%, demonstrating a slight bias for underestimation of laceration size. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
FIGURE 4: Graph represents absolute estimation error plotted against true laceration size. The black line represents the mean absolute estimation error of 11%. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
J Hand Surg Am.
r
Vol. 40, December 2015
ESTIMATION OF PARTIAL TENDON LACERATIONS
2425
FIGURE 5: Graph demonstrates the mean absolute estimation error of lacerations close to the threshold for surgery (range, 50%e70%) and the mean absolute estimation error for lacerations outside that range. CI, confidence interval.
tendon rupture. The exact size of partial laceration that should be repaired is unknown with recommendations ranging from 30% to 95%.9e18 For the purposes of this study, involvement of 60% of the tendon cross-sectional surface area was considered the threshold for surgical repair. Most surgeons (89%) estimate the size of partial lacerations in the operating room by visual inspection without the use of a ruler or caliper.10 Even with the aid of calipers, calculating the surface area of flexor tendons can be inaccurate.12 Because most surgeons perform a visual estimation of partial tendon lacerations, it is important to know whether surgeons are capable of performing this task well. Although studies of visual perception demonstrate that estimates of surface area are more difficult to make than estimates of angles, length, and size,21,22 this study showed a high level of accuracy, interrater reliability, and intrarater reliability. Studies of visual perception further demonstrate a difference between experienced and novice estimators,23 but we did not detect a difference between attending surgeons and residents. It is unclear why laceration size was more difficult to estimate when lacerations were near the 60% threshold for surgical repair. A similar inaccuracy in measuring midrange partial lacerations has been noted in a study in which calipers were used to measure J Hand Surg Am.
r
tendon diameters and laceration size.12 Although the exact threshold for surgical repair is not clear in the literature, our results suggest that increased care should be taken when estimating the size of partial tendon lacerations in the range of 50% to 70%. If a more exact threshold for surgical repair is defined in the future, it would become important to develop an even more accurate and precise method of measuring laceration size. One limitation of the study is the fact that the flexor tendons were widely exposed, which did not represent a typical clinical scenario. The wide exposure of the flexor tendons in this experimental model could conceivably have improved the accuracy of estimation over that which would occur clinically. Another limitation is the fact that the lacerations in this study were all sharp and perpendicular to the axis of the tendon. Oblique or irregular lacerations may reduce the accuracy of size estimation. These issues would benefit from further study. REFERENCES 1. Ollinger H, Wray RC, Weeks PM. Effects of suture on tensile strength gain of partially and completely severed tendons. Surg Forum. 1975;26:63e64. 2. Bishop AT, Cooney WP, Wood MB. Treatment of partial flexor tendon lacerations: the effect of tenorrhaphy and early protected mobilization. J Trauma. 1986;26(4):301e312.
Vol. 40, December 2015
2426
ESTIMATION OF PARTIAL TENDON LACERATIONS
3. Grewal R, Saw SS, Bastidas JA, Fischer KJ, Sotereanos DG. Passive and active rehabilitation for partial lacerations of the canine flexor digitorum profundus tendon in zone II. J Hand Surg Am. 1999;24(4): 743e750. 4. Seki Y, Kuroda H. Locking finger due to a partial laceration of the flexor digitorum superficialis tendon: a case report. Hand Surg. 2014;19(3):437e439. 5. Frewin PR, Scheker LR. Triggering secondary to an untreated partially-cut flexor tendon. J Hand Surg Br. 1989;14(4):419e421. 6. Fujiwara M. A case of trigger finger following partial laceration of flexor digitorum superficialis and review of the literature. Arch Orthop Trauma Surg. 2005;125(6):430e432. 7. Janecki DJ. Triggering of the finger caused by flexor-tendon laceration. J Bone Joint Surg Am. 1976;58(8):1174e1175. 8. Tohyama M, Tsujio T, Yanagida I. Trigger finger caused by an old partial flexor tendon laceration: a case report. Hand Surg. 2005;10(1): 105e108. 9. McGeorge DD, Stilwell JH. Partial flexor tendon injuries: to repair or not. J Hand Surg Br. 1992;17(2):176e177. 10. McCarthy DM, Boardman ND III, Tramaglini DM, Sotereanos DG, Herndon JH. Clinical management of partially lacerated digital flexor tendons: a survey of hand surgeons. J Hand Surg Am. 1995;20(2): 273e275. 11. Schlenker JD, Lister GD, Kleinert HE. Three complications of untreated partial laceration of flexor tendon—entrapment, rupture, and triggering. J Hand Surg Am. 1981;6(4):392e398. 12. McCarthy DM, Tramaglini DM, Chan SS, Schmidt CC, Sotereanos DG, Herndon JH. Effect of partial laceration on the
J Hand Surg Am.
r
13.
14.
15. 16. 17. 18.
19. 20. 21. 22. 23.
structural properties of the canine FDP tendon: an in vitro study. J Hand Surg Am. 1995;20(5):795e800. Chow SP, Yu OD. An experimental study on incompletely cut chicken tendons—a comparison of two methods of management. J Hand Surg Br. 1984;9(2):121e125. Haddad R, Scherman P, Peltz T, Nicklin S, Walsh WR. A biomechanical assessment of repair versus nonrepair of sheep flexor tendons lacerated to 75 percent. J Hand Surg Am. 2010;35(4): 546e551. Reynolds B, Wray RC, Weeks PM. Should an incompletely severed tendon be sutured? Plast Reconstr Surg. 1976;57(1):36e38. Wray RC, Weeks PM. Treatment of partial lacerations. Hand (N Y). 1980;12(2):163e166. Dobyns RC, Cooney WC, Wood MB. Effect of partial lacerations on canine flexor tendons. Minn Med. 1982;65(1):27e32. McMaster PE. Tendon and muscle ruptures: clinical and experimental studies on the cause and locations of subcutaneous rupture. J Bone Joint Surg Am. 1933;15(3):705e722. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86(2):420e428. Chilton NW, Fleiss JL. Design and analysis of plaque and gingivitis clinical trials. J Clin Periodontol. 1986;13(5):400e410. Hollands JG, Spence I. The discrimination of graphical elements. Appl Cognit Psychol. 2001;15(4):413e431. Cleveland WS. The Elements of Graphing Data. Pacific Grove, CA: Wadsworth; 1985. Lewandowsky S, Spence I. Discriminating strata in scatterplots. J Am Statistic Assoc. 1989;407(84):682e688.
Vol. 40, December 2015
Accuracy of Visual Estimates of Partial Flexor Tendon Lacerations Tamir Pritsch, MD,* Corrine Wong, MD,† Douglas M. Sammer, MD†
Purpose To determine whether hand surgeons could accurately and consistently estimate the size of partial flexor tendon lacerations. Materials and methods Thirty-two partial flexor tendon lacerations were made in the flexor digitorum profundus tendons of a fresh-frozen cadaveric hand. Four hand surgeons and 5 residents estimated the size of the lacerations. Estimates were repeated 3 days later. Magnified images of the laceration cross-section were used to calculate the true size of each laceration. Inter- and intrarater reliability were calculated using the intraclass correlation coefficient. Accuracy was measured with the mean bias error and the mean absolute error. Results Interrater and intrarater reliabilities were both high. There was a high level of consistency for both surgeons and residents. In terms of accuracy, there was a 3% bias toward underestimation. The mean absolute error was 11%. There was no statistically significant difference between the accuracy of attending hand surgeons and that of residents. Participants were less accurate when estimating lacerations close to a 60% laceration threshold for surgical repair (lacerations in the 50%e70% range). For lacerations within this range, an incorrect management decision would have been made 17% of the time, compared with 7% of the time for lacerations outside that range. Conclusions The accuracy and reliability of surgeon estimates of partial flexor tendon laceration size were high for surgeons and residents. Accuracy was lower for lacerations close to the threshold for repair. Clinical relevance Visual estimation is acceptable for evaluating partial flexor tendon lacerations, but it may be less reliable for lacerations near the threshold for repair. Therefore, surgeons should be cautious when deciding whether or not to repair partial lacerations in the borderline range. (J Hand Surg Am. 2015;40(12):2421e2426. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Key words Flexor, tendon, laceration, estimation, partial.
From the *Division of Hand Surgery, Department of Orthopedic Surgery, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel; and the †Department of Plastic Surgery, University of Texas Southwestern Medical School, Dallas, TX. Received for publication January 26, 2015; accepted in revised form September 8, 2015. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Douglas M. Sammer, MD, University of Texas Southwestern Medical School, 1801 Inwood Rd., Dallas, TX 75390; e-mail: Douglas.Sammer@ UTSouthwestern.edu. 0363-5023/15/4012-0015$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2015.09.009
R
EPAIRING A MINOR PARTIAL FLEXOR tendon laceration can weaken the tensile strength of the tendon and can increase bulk resulting in poor gliding and increased work of flexion.1e3 Therefore, only substantial lacerations that weaken the tendon to the point that it is at risk for rupture (or that cause catching on the tendon sheath) should be repaired.4e8 The recommended threshold for repairing a partial flexor tendon laceration varies. Many authors advocate repair if the laceration involves more than 50% to 60% of the tendon’s
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
2421
2422
ESTIMATION OF PARTIAL TENDON LACERATIONS
cross-sectional area,9e13 but recommendations range from 30% to 95%.14e18 The purpose of this study was to determine whether hand surgeons could accurately and consistently estimate the size of partial flexor tendon lacerations. MATERIALS AND METHODS Partial flexor tendon laceration model One fresh-frozen hand was obtained from the institution’s willed body program and allowed to thaw for 24 hours. Bruner incisions were made over volar aspect of the index, middle, ring, and little fingers to expose the flexor tendon sheaths from the distal phalanges into the midpalm. The flexor tendon sheaths were divided longitudinally, exposing the flexor tendons. Using a scalpel, 8 partial flexor tendon lacerations of varying sizes were made in the flexor digitorum profundus tendons of each finger from zone I through zone III, creating a total of 32 partial lacerations. The lacerations were oriented transversely to the long axis of the tendon. A random number was used to generate 32 percentage values, ranging between 1% and 99%. An effort was made to create lacerations that approximated the depth assigned by the random number generator. In this way, 32 partial flexor tendon lacerations were created with a wide range of laceration depths. The true depth of each laceration was unknown at the time of estimation and was precisely measured later in the experiment. Estimation of laceration size Four hand surgeons and 5 plastic surgery residents (postgraduate year 2, n ¼ 2; 3, n ¼ 2; 4, n ¼ 1) estimated the size (percent involvement of cross-sectional surface area) of the lacerations under loupe magnification and with the use of standard surgical instruments including calipers and ruler, although none of the participants ended up using the calipers. Participants were invited to repeat the size estimates in order to evaluate intrarater reliability. The second session was performed 3 days after the first session and the tendon lacerations were evaluated in reverse order to reduce recall bias. All aspects of the study were approved by our institutional review board. Measurement of true laceration size The flexor digitorum profundus tendons containing the lacerations were then removed from the hand and fixed in formalin (10% neutral buffered formalin; 24 hours). The laceration surfaces were painted with black ink (black permanent ink; Newell Rubbermaid Inc., Freeport, IL). Ink was used rather than a histological stain in J Hand Surg Am.
r
FIGURE 1: Magnified photograph of the tendon cross-section through the partial laceration. The inked surface represents the partial laceration.
order to avoid stain penetration outside the zone of laceration. Next, using a scalpel, the tendon was completely divided through each laceration. This converted each partial laceration into a complete laceration with an inked surface representing the surface area of the partial laceration (Fig. 1). Magnified electronic images of each inked tendon cross-section were obtained, and digital image processing software (Photoshop; Adobe Systems Inc., San Jose, CA) was used to calculate the percentage of the total crosssectional surface area of each partial tendon laceration. Statistical methods Inter- and intrarater reliability were measured by calculating the intraclass correlation coefficient (ICC, 2,1) for absolute agreement.19 Interrater reliability was calculated for all participants and for surgeon and resident subgroups separately. Intrarater reliability was calculated for all participants separately. An ICC less than 0.40 was considered poor, an ICC between 0.40 and 0.59 was considered fair, an ICC between 0.60 and 0.74 was considered good, and an ICC greater than 0.74 was considered excellent.19,20 Accuracy was measured by calculating the mean bias error (MBE) and the mean absolute error (MAE). Accuracy was measured for all participants and for the surgeon and resident subgroups separately. Comparisons of MBE and MAE were performed using the Student t test, and P values less than .05 were considered statistically significant. Finally, the estimation accuracy for lacerations that were close to the 60% threshold for surgical repair (50%e70% true size range) were analyzed separately and compared with lacerations that fell outside this range. This comparison was performed in order to Vol. 40, December 2015
ESTIMATION OF PARTIAL TENDON LACERATIONS
2423
FIGURE 2: A scatter plot of estimated laceration size versus true laceration size. The black line represents a linear regression. The red line represents perfect accuracy, where the estimated and true lacerations sizes are equal. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
determine whether estimation accuracy is different for lacerations that are close to the threshold for surgical repair compared with lacerations that were obviously above or below the threshold and what effect that may have on surgical decision making. RESULTS Nine participants (5 residents and 4 attending hand surgeons) completed the initial estimation session. Eight of the study participants (5 of the residents and 3 of the attending hand surgeons) completed the second estimation session. Reliability The overall interrater reliability was excellent (ICC, 0.88). When the attending hand surgeon and resident subgroups were analyzed separately, the interrater reliability remained excellent for both attending hand surgeons (ICC, 0.88) and residents (ICC, 0.87). The intrarater reliability of all participants, including both attending hand surgeons and residents, fell within the excellent range with ICCs greater than 0.74 and was 0.91 for all 3 attending hand surgeons and ranged from 0.78 to 0.91 for the 5 residents. J Hand Surg Am.
r
Accuracy Figure 2 demonstrates size estimates plotted against true laceration size. The MBE of all participants was e3%, demonstrating a slight bias toward underestimation (Fig. 3). The MAE was 11% (Fig. 4). The MBE and MAE were not significantly different for attending surgeons compared with residents (MBE e1.9 vs e3.7, P ¼ .85; MAE 10.8 vs 11.0, P ¼ .74). The MAE for lacerations in the 50% to 70% size range, which were close to the 60% threshold for surgical repair, was 13%, compared with an MAE of 10% for lacerations outside this range (P ¼ .015) (Fig. 5). For lacerations in the 50% to 70% size range, an incorrect management decision would have been made 17% of the time owing size estimation error. For lacerations outside the 50% to 70% range, an error in management would have been made 7% of the time. DISCUSSION Small partial flexor tendon lacerations do not require repair, and in fact, repairing a small laceration may weaken the tendon or increase its bulk. Conversely, large partial lacerations should be repaired to avoid Vol. 40, December 2015
2424
ESTIMATION OF PARTIAL TENDON LACERATIONS
FIGURE 3: Graph represents estimation error plotted against true laceration size. The red line represents zero estimation error. The black line represents the mean bias error of e3%, demonstrating a slight bias for underestimation of laceration size. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
FIGURE 4: Graph represents absolute estimation error plotted against true laceration size. The black line represents the mean absolute estimation error of 11%. Open circles represent a single estimation value and shaded circles indicate that more than 1 participant made the same estimation value.
J Hand Surg Am.
r
Vol. 40, December 2015
ESTIMATION OF PARTIAL TENDON LACERATIONS
2425
FIGURE 5: Graph demonstrates the mean absolute estimation error of lacerations close to the threshold for surgery (range, 50%e70%) and the mean absolute estimation error for lacerations outside that range. CI, confidence interval.
tendon rupture. The exact size of partial laceration that should be repaired is unknown with recommendations ranging from 30% to 95%.9e18 For the purposes of this study, involvement of 60% of the tendon cross-sectional surface area was considered the threshold for surgical repair. Most surgeons (89%) estimate the size of partial lacerations in the operating room by visual inspection without the use of a ruler or caliper.10 Even with the aid of calipers, calculating the surface area of flexor tendons can be inaccurate.12 Because most surgeons perform a visual estimation of partial tendon lacerations, it is important to know whether surgeons are capable of performing this task well. Although studies of visual perception demonstrate that estimates of surface area are more difficult to make than estimates of angles, length, and size,21,22 this study showed a high level of accuracy, interrater reliability, and intrarater reliability. Studies of visual perception further demonstrate a difference between experienced and novice estimators,23 but we did not detect a difference between attending surgeons and residents. It is unclear why laceration size was more difficult to estimate when lacerations were near the 60% threshold for surgical repair. A similar inaccuracy in measuring midrange partial lacerations has been noted in a study in which calipers were used to measure J Hand Surg Am.
r
tendon diameters and laceration size.12 Although the exact threshold for surgical repair is not clear in the literature, our results suggest that increased care should be taken when estimating the size of partial tendon lacerations in the range of 50% to 70%. If a more exact threshold for surgical repair is defined in the future, it would become important to develop an even more accurate and precise method of measuring laceration size. One limitation of the study is the fact that the flexor tendons were widely exposed, which did not represent a typical clinical scenario. The wide exposure of the flexor tendons in this experimental model could conceivably have improved the accuracy of estimation over that which would occur clinically. Another limitation is the fact that the lacerations in this study were all sharp and perpendicular to the axis of the tendon. Oblique or irregular lacerations may reduce the accuracy of size estimation. These issues would benefit from further study. REFERENCES 1. Ollinger H, Wray RC, Weeks PM. Effects of suture on tensile strength gain of partially and completely severed tendons. Surg Forum. 1975;26:63e64. 2. Bishop AT, Cooney WP, Wood MB. Treatment of partial flexor tendon lacerations: the effect of tenorrhaphy and early protected mobilization. J Trauma. 1986;26(4):301e312.
Vol. 40, December 2015
2426
ESTIMATION OF PARTIAL TENDON LACERATIONS
3. Grewal R, Saw SS, Bastidas JA, Fischer KJ, Sotereanos DG. Passive and active rehabilitation for partial lacerations of the canine flexor digitorum profundus tendon in zone II. J Hand Surg Am. 1999;24(4): 743e750. 4. Seki Y, Kuroda H. Locking finger due to a partial laceration of the flexor digitorum superficialis tendon: a case report. Hand Surg. 2014;19(3):437e439. 5. Frewin PR, Scheker LR. Triggering secondary to an untreated partially-cut flexor tendon. J Hand Surg Br. 1989;14(4):419e421. 6. Fujiwara M. A case of trigger finger following partial laceration of flexor digitorum superficialis and review of the literature. Arch Orthop Trauma Surg. 2005;125(6):430e432. 7. Janecki DJ. Triggering of the finger caused by flexor-tendon laceration. J Bone Joint Surg Am. 1976;58(8):1174e1175. 8. Tohyama M, Tsujio T, Yanagida I. Trigger finger caused by an old partial flexor tendon laceration: a case report. Hand Surg. 2005;10(1): 105e108. 9. McGeorge DD, Stilwell JH. Partial flexor tendon injuries: to repair or not. J Hand Surg Br. 1992;17(2):176e177. 10. McCarthy DM, Boardman ND III, Tramaglini DM, Sotereanos DG, Herndon JH. Clinical management of partially lacerated digital flexor tendons: a survey of hand surgeons. J Hand Surg Am. 1995;20(2): 273e275. 11. Schlenker JD, Lister GD, Kleinert HE. Three complications of untreated partial laceration of flexor tendon—entrapment, rupture, and triggering. J Hand Surg Am. 1981;6(4):392e398. 12. McCarthy DM, Tramaglini DM, Chan SS, Schmidt CC, Sotereanos DG, Herndon JH. Effect of partial laceration on the
J Hand Surg Am.
r
13.
14.
15. 16. 17. 18.
19. 20. 21. 22. 23.
structural properties of the canine FDP tendon: an in vitro study. J Hand Surg Am. 1995;20(5):795e800. Chow SP, Yu OD. An experimental study on incompletely cut chicken tendons—a comparison of two methods of management. J Hand Surg Br. 1984;9(2):121e125. Haddad R, Scherman P, Peltz T, Nicklin S, Walsh WR. A biomechanical assessment of repair versus nonrepair of sheep flexor tendons lacerated to 75 percent. J Hand Surg Am. 2010;35(4): 546e551. Reynolds B, Wray RC, Weeks PM. Should an incompletely severed tendon be sutured? Plast Reconstr Surg. 1976;57(1):36e38. Wray RC, Weeks PM. Treatment of partial lacerations. Hand (N Y). 1980;12(2):163e166. Dobyns RC, Cooney WC, Wood MB. Effect of partial lacerations on canine flexor tendons. Minn Med. 1982;65(1):27e32. McMaster PE. Tendon and muscle ruptures: clinical and experimental studies on the cause and locations of subcutaneous rupture. J Bone Joint Surg Am. 1933;15(3):705e722. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86(2):420e428. Chilton NW, Fleiss JL. Design and analysis of plaque and gingivitis clinical trials. J Clin Periodontol. 1986;13(5):400e410. Hollands JG, Spence I. The discrimination of graphical elements. Appl Cognit Psychol. 2001;15(4):413e431. Cleveland WS. The Elements of Graphing Data. Pacific Grove, CA: Wadsworth; 1985. Lewandowsky S, Spence I. Discriminating strata in scatterplots. J Am Statistic Assoc. 1989;407(84):682e688.
Vol. 40, December 2015