- Email: [email protected]
An Assessment of 2 Objective Measurements of Web Space Position
SCIENTIFIC ARTICLE
An Assessment of 2 Objective Measurements of Web Space Position Michael A. Tonkin, MD, MBBS, Ee Ming Chew, MBBS, James P. Ledgard, MBBS, Ahmed A. Al-Sultan, MS, Belinda J. Smith, BSc, Richard D. Lawson, MBBS
Purpose To describe 2 simple objective clinical methods of measuring the web position between fingers and to determine their intra-observer and inter-observer reliabilities. Methods Two observers examined the second, third, and fourth web spaces on both hands of 30 adult healthy volunteers. The web index measured the web height as a relative ratio to constant anatomical landmarks on both fingers subtending the web. The dorsal web index took reference from the distance between the metacarpophalangeal and proximal interphalangeal joints, whereas the palmar web index was measured in relation to the distance between the most proximal basal digital and proximal interphalangeal joint creases. The intraclass correlation coefficient was used to determine intra-observer and inter-observer reliability. Results Intraclass correlation coefficient values for intra-observer and inter-observer reliability were greater than 0.80, indicating excellent agreement. There was no statistically significant difference between the dorsal or palmar measurement methods in terms of reliability. Conclusions The dorsal or palmar measurement method may be reliably used in healthy adults to establish a web index that describes the web position. The palmar method is considered easier to perform. (J Hand Surg Am. 2015;40(3):456e461. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Diagnostic III. Key words Web position, syndactyly, web creep.
S
a failure of developmental separation of adjacent digital rays and is classified as a malformation of the hand plate in the extended version of the Oberg, Manske, and Tonkin classification.1 Web creep is a common complication after surgical release of syndactyly and many authors have used its YNDACTYLY RESULTS FROM
From the Department of Hand Surgery and Peripheral Nerve Surgery, Royal North Shore Hospital, Children’s Hospital at Westmead, University of Sydney, Sydney, Australia; and the Department of Hand Surgery, Singapore General Hospital, Singapore. Received for publication August 28, 2014; accepted in revised form November 10, 2014. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Michael A. Tonkin, MD, MBBS, Department of Hand Surgery and Peripheral Nerve Surgery, Royal North Shore Hospital, St Leonards 2065, St Leonards, NSW 2065, Australia; e-mail: [email protected]. 0363-5023/15/4003-0006$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2014.11.017
456
r
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
incidence to compare different surgical techniques. Clinical and radiological methods have been deployed to measure change in the position of the web, which can then be used to quantify the severity of web creep. Some techniques describe measurements of web positions with the fingers in abduction.2e5 However, the extent to which the web is abducted and maintained during measurement is poorly described. Variations in the angle of abduction will affect the web position and produce varied results. Dorsal measurement techniques usually require determination of the levels of the metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints and assessment of a web point, which may be difficult to determine in some hands. Palmar measurements usually demand a consistent estimation of palmar creases.2,3 Radiological methods require exposure to radiation and incur financial costs.4,5 Many methods have not undergone rigorous assessment to determine their reliability and reproducibility.6e8
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
457
FIGURE 2: Dorsal method. A represents the distance between the MCP joint and the web point. B represents the distance between the MCP and PIP joints. FIGURE 1: Dorsal method. The examiner places the fingers in maximum passive abduction without force. A perpendicular line is drawn from the midpoint of the web to meet the finger axis at the web point.
abduction angle was documented. We marked the midpoint of the web and drew perpendicular lines from the midpoint of the web to the axial line of each finger. The point at which the lines intersected was termed the web point. The MCP and PIP joints were flexed passively to 90 each, allowing a vernier caliper to be applied squarely on the dorsum of the proximal phalanx (Fig. 2). We measured the distances between the MCP joint and the web point (A) and between the MCP and PIP joints (B) to the nearest millimeter. We calculated the web index as the mean of the ratios (A / B) of each of the 2 adjacent fingers. For example, the dorsal web index for the second web ¼ 0.5 (A / B of index finger þ A / B of middle finger). If web creep developed, the web index was expected to be higher than the baseline measurement and closer to a ratio of one as the web extended toward the PIP joint. If the web extended distal to the PIP joint, distance A had to be measured with the PIP joint extended and the web index became greater than one.
This article details 2 simple, objective, clinical methods of measuring the normal web position and demonstrates their intra-observer and inter-observer reliabilities. MATERIALS AND METHODS We undertook the assessment in adults because the process of obtaining permission to examine minors and the logistics of performing multiple assessments in an appropriate number of children were difficult to satisfy. Ethics approval was obtained from the hospital’s human research ethics committee. We used 2 methods: dorsal and palmar. Dorsal measurement technique The hand was placed with the palm on a flat, firm surface. The web being measured was placed by the examiner in maximal passive abduction without force, with all other fingers relaxed and maintained in this position with the assistance of the subject’s opposite hand. The axes of the fingers subtending the web were then drawn. These lines connected the midpoints of the MCP and PIP joints of these fingers (Fig. 1). The J Hand Surg Am.
Palmar measurement technique The hand was placed with the dorsum on a flat, firm surface. The web being measured was fully and passively abducted by the examiner as for the dorsal method without force. All other fingers were relaxed. We drew axial lines on the palmar aspect of the fingers subtending the web (Fig. 3). We documented the r
Vol. 40, March 2015
458
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
abduction angle, marked the midpoint of the web, and drew perpendicular lines from the midpoint of the web to the axial lines. These lines intersected at the web point. The distance (A) between the most proximal PIP joint crease and the web point and the distance (B) between the most proximal PIP joint crease and the most proximal basal digital crease were measured with a ruler to the nearest millimeter. We calculated the palmar web index as the mean of the ratios (A / B) of the fingers forming the web. For example, the palmar web index for the second web ¼ 0.5 (A / B of index finger þ A / B of middle finger). Many normal web indices had a web index of greater than one, as shown in Figure 3. As web creep occurred, the palmar web index decreased and approached zero. A web position beyond the PIP joint created a negative index. Intra-observer and inter-observer reliability Two observers assessed the web indices of the second, third, and fourth web spaces on both hands of 30 healthy volunteers using a single ruler and vernier calipers. Each assessor independently performed the markings and measurements, separating both intra-observer and inter-observer assessments by a minimum of one week. For the intra-observer error, one observer performed 2 sets of dorsal measurements and 1 set of palmar measurements. The second observer performed 2 sets of palmar measurements and 1 set of dorsal measurements. Dorsal and palmar web indices were calculated from these measurements. We examined reliability with the intraclass correlation coefficient (ICC). This value can range from 0 to 1.9 A value of 0 implies no more than chance agreement between observations, whereas an ICC of 1 indicates perfect agreement. Coefficients are often interpreted according to the guidelines described by Landis and Koch9: Values less than 0.01 indicate poor agreement; 0.01 to 0.20, slight agreement; 0.21 to 0.40, fair agreement; 0.41 to 0.60, moderate agreement; 0.61 to 0.80; substantial agreement; and more than 0.80, excellent agreement. Intra-observer reliability for the dorsal technique was examined with the ICC by comparing both sets of dorsal web indices measured by the first observer. Intra-observer reliability for the palmar technique was determined by comparing both sets of palmar web indices measured by the second observer. We determined inter-observer reliability for the dorsal technique by comparing the mean of the dorsal web indices measured by the first observer with those measured by the second observer. Inter-observer reliability for the palmar technique was calculated by J Hand Surg Am.
FIGURE 3: Palmar method. A represents the distance between the most proximal crease of the PIP joint crease and the web point. B represents the distance between the most proximal crease of the PIP joint and the most proximal basal digital crease.
comparing the palmar web indices measured by the first observer with the mean of the palmar web indices measured by the second observer. We decided to use the mean of 2 indices of one observer after determining that intra-observer reliability was excellent. We obtained advice from our hospital’s research statistician regarding the application of these assessments. RESULTS We measured 11 men and 19 women (180 webs) to obtain web indices (Table 1). Mean age was 37 years (range, 22e66 y). Table 2 lists the ICC values for intra-observer and inter-observer reliability. All values were greater than 0.80, indicating excellent agreement. The 95% confidence intervals of the ICC for the dorsal and palmar techniques overlapped, meaning there was no statistically significant difference of reliability between methods. DISCUSSION Web creep is a common complication of syndactyly release, involving 3% to 22% of patients.10,11 The incidence of web creep has been used to compare different surgical techniques. However, the postoperative web appearance has been imprecisely r
Vol. 40, March 2015
459
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
TABLE 1.
Average and SD of Web Indices for Each Web, as Measured by 2 Assessors Dorsal
Left second Left third Left fourth Right second Right third Right fourth
Palmar
Assessor 1 (Average of 2 Measurements)
Assessor 2
Assessor 1
Assessor 2 (Average of 2 Measurements)
Average
0.45
0.46
1.09
1.06
SD
0.07
0.06
0.07
0.05
Average
0.51
0.51
1.02
0.98
SD
0.05
0.06
0.05
0.03
Average
0.43
0.46
1.12
1.06
SD
0.08
0.08
0.07
0.05
Average
0.46
0.47
1.07
1.05
SD
0.06
0.06
0.06
0.05
Average
0.51
0.52
1.02
0.98
SD
0.06
0.06
0.04
0.04
Average
0.45
0.47
1.10
1.05
SD
0.06
0.07
0.08
0.06
adult hands relative to those in children. Although their findings were statistically significant, we question whether the mild differences they documented are clinically relevant. For example, the difference in third web space ratio between children aged 5 to 6 years and adults altered from 1.81 to 1.88, a difference of less than 2 mm in the adult hand web position. These results indicate no substantial alteration in web position with growth. The next parameter to be considered is the establishment of a reliable method for determining reference levels from which a web ratio or web index is calculated. Dorsal clinical measurements taking reference from the MCP or PIP joints must overcome the difficulty of palpation of joint levels in some hands. In palmar measurement techniques, the reference levels are usually the palmar and digital creases. Multiple creases at the base of the digit and interphalangeal joints or the absence of creases may render standardized measurement by crease position difficult. Third, we agree with Shewell et al2 that the web appearance and position depend on the position of the fingers. The angle of abduction also depends on whether the fingers are actively or passively abducted and whether the adjacent webs are relaxed or abducted. An objective and reliable method of measuring web position demands that the degree of finger abduction be standardized so that the point of measurement of web position is consistent. Finally, some researchers have used the center of the web to mark its position and some have used the point at which the web skin meets the finger skin,
TABLE 2. Intraclass Correlation Coefficients for Dorsal and Palmar Web Indices Intra-Observer Reliability (95% confidence interval)
Inter-Observer Reproducibility (95% confidence interval)
Dorsal
0.88 (0.84e0.91)
0.89 (0.85e0.92)
Palmar
0.85 (0.80e0.89)
0.82 (0.77e0.87)
Web Index
described as “adequate,”12 “satisfactory,”11,13 “maintained at an appropriate level,”14 or “no gross recurrence.”15 Others have attempted to quantify web creep more objectively but acknowledged difficulties.2e8 We have described 2 methods of measuring web position in adult hands and have demonstrated the reliability of the 2 methods. A number of factors may affect an objective and reliable assessment of web position and its application to a measurement of postoperative web creep. These include change of web position with growth, identification of consistent reference points against which the web position is compared, change of web position with differing angles of finger abduction, and the anatomical point within the web that is used to establish the web level. Any measurement of web creep must take into account the normal web position and its alteration with growth of the hand. Ratios are useful in this respect. However, Patterson and Nancarrow3 noted a mild alteration in web position with more distal placement in
J Hand Surg Am.
r
Vol. 40, March 2015
460
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
which is less precise. Both change with differing abduction angles. Brown6 and Toledo and Ger7 clinically assessed web height using the normal palmar crease and adjacent normal web as reference points. These methods can be applied only to patients with single syndactyly and normal palmar creases, and assume that the second, third, and fourth web positions relative to the palmar crease are the same. Consistent abduction angles were not used. Shewell et al2 used photographs of adult hands held flat against a glass plate with fingers abducted 10 . This palmar technique measured the web position as a ratio based on the dominant PIP and distal interphalangeal joint creases. Dominant creases were judged where they crossed the edge of the finger. If there were 2 dominant creases, measurements were taken at the midpoint. Patterson and Nancarrow3 subsequently confirmed that this method was applicable to the hands of children. This method allows for easy comparison of serial postoperative measurements but requires repetitive photography. The authors conceded that the measurements were reproducible only if marked and measured by a single observer. This renders the validity of comparison between studies doubtful unless original photographs are available and are reassessed by a single observer. Their measurements of web position also relied on a web point that was judged to meet the skin of the finger. They found that with 10 separate observers who were asked to mark and measure the same pair of hands, individual differences in measurement ranged up to 7 mm, demonstrating a lack of inter-observer reliability; hence, the need for the same single observer for measurements over time. Withey et al8 developed a semi-quantitative grading system that assigned the level of web creep into thirds of the distance between the normal web position and the PIP joint crease. The normal web position was established by observing adjacent webs and those of the opposite hand, which must be normal for validity. The finger abduction angle was not standardized. Although simple to use, this subjective method was not assessed for intra-observer and inter-observer error. Radiological methods project soft and bony tissues in 2 dimensions, which makes measurements of the web space easier with respect to joint levels. The method described by Richterman et al4 had high intra-observer and inter-observer reproducibility. Vekris et al5 also assessed web creep using serial x-rays. However, these methods call into question the need to expose children to multiple doses of radiation. Furthermore, the financial burden for follow-up is increased. There is also no description of standardizing finger abduction angles during the radiological examination. J Hand Surg Am.
In view of the lack of precise descriptions of some existing methods, the limitations of others, and poor evidence of reliability, we assessed 2 clinical methods that measure web position. If web creep occurs, the index changes; this translates to an objective assessment of severity. We overcame the inaccuracy of palpating the MCP and PIP joint levels in the dorsal method by resting the calipers squarely on the dorsum of the proximal phalanx with both joints passively flexed as near to 90 as possible. In effect, this measurement takes reference from the dorsal metacarpal surface and the distal end of the proximal phalanx. The presence of joint stiffness, such as symphalangism in Apert syndrome, precludes the use of this method. In the palmar measurement technique, the most proximal basal digital and PIP joint creases in the axial line were easily identifiable in adults.2 We found this to be the same in simple and complex syndactylies of children. By definition, the former is only an anomaly of skin whereas the latter also has a bony connection distally. This observation was confirmed by the work of others who used similar methods but with less precise measurements, consistency of digit position, and assessment of reproducibility and reliability.2,3,8 We chose to standardize the position of the fingers in maximal passive abduction without excessive force and with relaxation of the adjacent webs. The absolute position of abduction in degrees is not important; rather, consistent placement of the fingers allows the web being measured to be taut. This minimizes the variability of web position with the varying ability of the patient to abduct the fingers actively. In this position of abduction, assessment of a web point may be performed in a more accurate manner by constructing a line from the midpoint of the web to the perpendicular axial line at 90 . The efforts of rigorous standardization yielded correlation coefficients of more than 0.8, indicating excellent intra-observer and inter-observer reliability for both dorsal and palmar techniques. We acknowledge that only one examiner established the ICC for intra-observer error for each technique and that the inter-observer error ICC was established between 2 examiners. Regrettably, the logistics of performing multiple measurements on volunteers returning for multiple clinic visits prevented an increased number of assessments. Nevertheless, either method may be used to establish a web index that describes the web position and quantify its change. We prefer the palmar method because it is easier to perform. It does not require joint flexion or the use of calipers. The web index can then be used to describe the web position clinically at the basal crease level r
Vol. 40, March 2015
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
461
determining the clinical applicability of the method after syndactyly surgery is to use these techniques to measure the web index while the child is under anesthetic at the completion of surgery, and then at appropriate times postoperatively with the assistance of parents and other medical staff, to determine any change in the web index. Figure 4 shows an example of an immediate postoperative web index measurement. This picture also demonstrates the presence of double basal finger and PIP joint creases. We acknowledge that postoperative scarring and angulatory deformity may alter the position of creases and decrease the reliability of reference points for subsequent postoperative measurements. We do not believe this possibility will prevent the ability to use the web index efficiently to monitor web creep, but this can be proven only by future studies. As with most clinical measurements of web creep, these methods are not applicable to complicated syndactylies in which there is an absence of creases, joint stiffness, and angulatory deformity, because it is often impossible to establish reference points for web position. However, we believe that it is appropriate to apply the palmar technique to measure simple and complex syndactyly web position. FIGURE 4: Example of measurement of a postoperative web index for a simple, incomplete syndactyly. The web index measured 0.95. Note the double basal and PIP joint creases in the child’s hand.
REFERENCES 1. Tonkin MA, Tolerton SK, Quick TJ, et al. Classification of congenital anomalies of the hand and upper limb: development and assessment of a new system. J Hand Surg Am. 2013;38(9):1845e1853. 2. Shewell PC, Nancarrow JD, Fatah F. Quantifying interdigital web morphology. J Hand Surg Br. 1992;17(2):198e200. 3. Paterson P, Nancarrow JD. Defining a normal finger web in children. J Hand Surg Br. 1998;23(4):496e498. 4. Richterman IE, DuPree J, Thoder J, Kozin SH. The radiographic analysis of web height. J Hand Surg Am. 1998;23(6):1071e1076. 5. Vekris MD, Lykissas MG, Soucacos PN, Korompilias AV, Beris AE. Congenital syndactyly: outcome of surgical treatment in 131 webs. Tech Hand Up Extrem Surg. 2010;14(1):2e7. 6. Brown PM. Syndactyly—a review and long term results. Hand. 1977;9(1):16e27. 7. Toledo LC, Ger E. Evaluation of the operative treatment of syndactyly. J Hand Surg Am. 1979;4(6):556e564. 8. Withey SJ, Kangesu T, Carver N, Sommerlad BC. The open finger technique for the release of syndactyly. J Hand Surg Br. 2001;26(1):4e7. 9. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159e174. 10. Keret D, Ger E. Evaluation of a uniform operative technique to treat syndactyly. J Hand Surg Am. 1987;12(5 pt 1):727e729. 11. Percival NJ, Sykes PJ. Syndactyly: a review of the factors which influence surgical treatment. J Hand Surg Br. 1989;14(2):196e200. 12. Sharma RK, Tuli P, Makkar SS, Parashar A. End-of-skin grafts in syndactyly release: description of a new flap for web space resurfacing and primary closure of finger defects. Hand (N Y). 2009;4(1):29e34. 13. Nakamura J, Yanagawa H, Kubo E, Endo T. New modified method for the surgical treatment of syndactyly. Ann Plast Surg. 1989;23(6):511e518. 14. Ostrowski DM, Feagin CA, Gould JS. A three-flap web-plasty for release of short congenital syndactyly and dorsal adduction contracture. J Hand Surg Am. 1991;16(4):634e641. 15. Killian JT, Neimkin RJ. Syndactyly reconstruction by a modified Cronin method. South Med J. 1985;78(4):414e418.
(one-third or two-thirds the distance between the basal crease and PIP joint, at or distal to PIP joint crease) in a manner similar to but more precise than that of Withey et al.8 This study has a number of limitations, some of which have been already addressed. Placement and maintenance of the finger position on either side of the web to be measured may be less than consistent. However, we believe our method is more precise than those previously described, allowing a more accurate evaluation of the central web point position than previously obtained. Our methods use identical positioning and reference points no matter the age of the subject. We can see no problem that would preclude their application in the child’s hand, as Shewell et al2 and Paterson and Nancarrow3 found in the transfer of their method from adult to child. We acknowledge that application of these techniques of measurement is more difficult in the hands of children, because cooperation and maintenance of position are more difficult. The increased difficulty should not deter efforts to achieve a more precise and reliable assessment of web position and web creep after syndactyly surgery. Having confirmed the reliability of measurement in normal hands, the next phase in J Hand Surg Am.
r
Vol. 40, March 2015
An Assessment of 2 Objective Measurements of Web Space Position Michael A. Tonkin, MD, MBBS, Ee Ming Chew, MBBS, James P. Ledgard, MBBS, Ahmed A. Al-Sultan, MS, Belinda J. Smith, BSc, Richard D. Lawson, MBBS
Purpose To describe 2 simple objective clinical methods of measuring the web position between fingers and to determine their intra-observer and inter-observer reliabilities. Methods Two observers examined the second, third, and fourth web spaces on both hands of 30 adult healthy volunteers. The web index measured the web height as a relative ratio to constant anatomical landmarks on both fingers subtending the web. The dorsal web index took reference from the distance between the metacarpophalangeal and proximal interphalangeal joints, whereas the palmar web index was measured in relation to the distance between the most proximal basal digital and proximal interphalangeal joint creases. The intraclass correlation coefficient was used to determine intra-observer and inter-observer reliability. Results Intraclass correlation coefficient values for intra-observer and inter-observer reliability were greater than 0.80, indicating excellent agreement. There was no statistically significant difference between the dorsal or palmar measurement methods in terms of reliability. Conclusions The dorsal or palmar measurement method may be reliably used in healthy adults to establish a web index that describes the web position. The palmar method is considered easier to perform. (J Hand Surg Am. 2015;40(3):456e461. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Diagnostic III. Key words Web position, syndactyly, web creep.
S
a failure of developmental separation of adjacent digital rays and is classified as a malformation of the hand plate in the extended version of the Oberg, Manske, and Tonkin classification.1 Web creep is a common complication after surgical release of syndactyly and many authors have used its YNDACTYLY RESULTS FROM
From the Department of Hand Surgery and Peripheral Nerve Surgery, Royal North Shore Hospital, Children’s Hospital at Westmead, University of Sydney, Sydney, Australia; and the Department of Hand Surgery, Singapore General Hospital, Singapore. Received for publication August 28, 2014; accepted in revised form November 10, 2014. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Michael A. Tonkin, MD, MBBS, Department of Hand Surgery and Peripheral Nerve Surgery, Royal North Shore Hospital, St Leonards 2065, St Leonards, NSW 2065, Australia; e-mail: [email protected]. 0363-5023/15/4003-0006$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2014.11.017
456
r
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
incidence to compare different surgical techniques. Clinical and radiological methods have been deployed to measure change in the position of the web, which can then be used to quantify the severity of web creep. Some techniques describe measurements of web positions with the fingers in abduction.2e5 However, the extent to which the web is abducted and maintained during measurement is poorly described. Variations in the angle of abduction will affect the web position and produce varied results. Dorsal measurement techniques usually require determination of the levels of the metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints and assessment of a web point, which may be difficult to determine in some hands. Palmar measurements usually demand a consistent estimation of palmar creases.2,3 Radiological methods require exposure to radiation and incur financial costs.4,5 Many methods have not undergone rigorous assessment to determine their reliability and reproducibility.6e8
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
457
FIGURE 2: Dorsal method. A represents the distance between the MCP joint and the web point. B represents the distance between the MCP and PIP joints. FIGURE 1: Dorsal method. The examiner places the fingers in maximum passive abduction without force. A perpendicular line is drawn from the midpoint of the web to meet the finger axis at the web point.
abduction angle was documented. We marked the midpoint of the web and drew perpendicular lines from the midpoint of the web to the axial line of each finger. The point at which the lines intersected was termed the web point. The MCP and PIP joints were flexed passively to 90 each, allowing a vernier caliper to be applied squarely on the dorsum of the proximal phalanx (Fig. 2). We measured the distances between the MCP joint and the web point (A) and between the MCP and PIP joints (B) to the nearest millimeter. We calculated the web index as the mean of the ratios (A / B) of each of the 2 adjacent fingers. For example, the dorsal web index for the second web ¼ 0.5 (A / B of index finger þ A / B of middle finger). If web creep developed, the web index was expected to be higher than the baseline measurement and closer to a ratio of one as the web extended toward the PIP joint. If the web extended distal to the PIP joint, distance A had to be measured with the PIP joint extended and the web index became greater than one.
This article details 2 simple, objective, clinical methods of measuring the normal web position and demonstrates their intra-observer and inter-observer reliabilities. MATERIALS AND METHODS We undertook the assessment in adults because the process of obtaining permission to examine minors and the logistics of performing multiple assessments in an appropriate number of children were difficult to satisfy. Ethics approval was obtained from the hospital’s human research ethics committee. We used 2 methods: dorsal and palmar. Dorsal measurement technique The hand was placed with the palm on a flat, firm surface. The web being measured was placed by the examiner in maximal passive abduction without force, with all other fingers relaxed and maintained in this position with the assistance of the subject’s opposite hand. The axes of the fingers subtending the web were then drawn. These lines connected the midpoints of the MCP and PIP joints of these fingers (Fig. 1). The J Hand Surg Am.
Palmar measurement technique The hand was placed with the dorsum on a flat, firm surface. The web being measured was fully and passively abducted by the examiner as for the dorsal method without force. All other fingers were relaxed. We drew axial lines on the palmar aspect of the fingers subtending the web (Fig. 3). We documented the r
Vol. 40, March 2015
458
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
abduction angle, marked the midpoint of the web, and drew perpendicular lines from the midpoint of the web to the axial lines. These lines intersected at the web point. The distance (A) between the most proximal PIP joint crease and the web point and the distance (B) between the most proximal PIP joint crease and the most proximal basal digital crease were measured with a ruler to the nearest millimeter. We calculated the palmar web index as the mean of the ratios (A / B) of the fingers forming the web. For example, the palmar web index for the second web ¼ 0.5 (A / B of index finger þ A / B of middle finger). Many normal web indices had a web index of greater than one, as shown in Figure 3. As web creep occurred, the palmar web index decreased and approached zero. A web position beyond the PIP joint created a negative index. Intra-observer and inter-observer reliability Two observers assessed the web indices of the second, third, and fourth web spaces on both hands of 30 healthy volunteers using a single ruler and vernier calipers. Each assessor independently performed the markings and measurements, separating both intra-observer and inter-observer assessments by a minimum of one week. For the intra-observer error, one observer performed 2 sets of dorsal measurements and 1 set of palmar measurements. The second observer performed 2 sets of palmar measurements and 1 set of dorsal measurements. Dorsal and palmar web indices were calculated from these measurements. We examined reliability with the intraclass correlation coefficient (ICC). This value can range from 0 to 1.9 A value of 0 implies no more than chance agreement between observations, whereas an ICC of 1 indicates perfect agreement. Coefficients are often interpreted according to the guidelines described by Landis and Koch9: Values less than 0.01 indicate poor agreement; 0.01 to 0.20, slight agreement; 0.21 to 0.40, fair agreement; 0.41 to 0.60, moderate agreement; 0.61 to 0.80; substantial agreement; and more than 0.80, excellent agreement. Intra-observer reliability for the dorsal technique was examined with the ICC by comparing both sets of dorsal web indices measured by the first observer. Intra-observer reliability for the palmar technique was determined by comparing both sets of palmar web indices measured by the second observer. We determined inter-observer reliability for the dorsal technique by comparing the mean of the dorsal web indices measured by the first observer with those measured by the second observer. Inter-observer reliability for the palmar technique was calculated by J Hand Surg Am.
FIGURE 3: Palmar method. A represents the distance between the most proximal crease of the PIP joint crease and the web point. B represents the distance between the most proximal crease of the PIP joint and the most proximal basal digital crease.
comparing the palmar web indices measured by the first observer with the mean of the palmar web indices measured by the second observer. We decided to use the mean of 2 indices of one observer after determining that intra-observer reliability was excellent. We obtained advice from our hospital’s research statistician regarding the application of these assessments. RESULTS We measured 11 men and 19 women (180 webs) to obtain web indices (Table 1). Mean age was 37 years (range, 22e66 y). Table 2 lists the ICC values for intra-observer and inter-observer reliability. All values were greater than 0.80, indicating excellent agreement. The 95% confidence intervals of the ICC for the dorsal and palmar techniques overlapped, meaning there was no statistically significant difference of reliability between methods. DISCUSSION Web creep is a common complication of syndactyly release, involving 3% to 22% of patients.10,11 The incidence of web creep has been used to compare different surgical techniques. However, the postoperative web appearance has been imprecisely r
Vol. 40, March 2015
459
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
TABLE 1.
Average and SD of Web Indices for Each Web, as Measured by 2 Assessors Dorsal
Left second Left third Left fourth Right second Right third Right fourth
Palmar
Assessor 1 (Average of 2 Measurements)
Assessor 2
Assessor 1
Assessor 2 (Average of 2 Measurements)
Average
0.45
0.46
1.09
1.06
SD
0.07
0.06
0.07
0.05
Average
0.51
0.51
1.02
0.98
SD
0.05
0.06
0.05
0.03
Average
0.43
0.46
1.12
1.06
SD
0.08
0.08
0.07
0.05
Average
0.46
0.47
1.07
1.05
SD
0.06
0.06
0.06
0.05
Average
0.51
0.52
1.02
0.98
SD
0.06
0.06
0.04
0.04
Average
0.45
0.47
1.10
1.05
SD
0.06
0.07
0.08
0.06
adult hands relative to those in children. Although their findings were statistically significant, we question whether the mild differences they documented are clinically relevant. For example, the difference in third web space ratio between children aged 5 to 6 years and adults altered from 1.81 to 1.88, a difference of less than 2 mm in the adult hand web position. These results indicate no substantial alteration in web position with growth. The next parameter to be considered is the establishment of a reliable method for determining reference levels from which a web ratio or web index is calculated. Dorsal clinical measurements taking reference from the MCP or PIP joints must overcome the difficulty of palpation of joint levels in some hands. In palmar measurement techniques, the reference levels are usually the palmar and digital creases. Multiple creases at the base of the digit and interphalangeal joints or the absence of creases may render standardized measurement by crease position difficult. Third, we agree with Shewell et al2 that the web appearance and position depend on the position of the fingers. The angle of abduction also depends on whether the fingers are actively or passively abducted and whether the adjacent webs are relaxed or abducted. An objective and reliable method of measuring web position demands that the degree of finger abduction be standardized so that the point of measurement of web position is consistent. Finally, some researchers have used the center of the web to mark its position and some have used the point at which the web skin meets the finger skin,
TABLE 2. Intraclass Correlation Coefficients for Dorsal and Palmar Web Indices Intra-Observer Reliability (95% confidence interval)
Inter-Observer Reproducibility (95% confidence interval)
Dorsal
0.88 (0.84e0.91)
0.89 (0.85e0.92)
Palmar
0.85 (0.80e0.89)
0.82 (0.77e0.87)
Web Index
described as “adequate,”12 “satisfactory,”11,13 “maintained at an appropriate level,”14 or “no gross recurrence.”15 Others have attempted to quantify web creep more objectively but acknowledged difficulties.2e8 We have described 2 methods of measuring web position in adult hands and have demonstrated the reliability of the 2 methods. A number of factors may affect an objective and reliable assessment of web position and its application to a measurement of postoperative web creep. These include change of web position with growth, identification of consistent reference points against which the web position is compared, change of web position with differing angles of finger abduction, and the anatomical point within the web that is used to establish the web level. Any measurement of web creep must take into account the normal web position and its alteration with growth of the hand. Ratios are useful in this respect. However, Patterson and Nancarrow3 noted a mild alteration in web position with more distal placement in
J Hand Surg Am.
r
Vol. 40, March 2015
460
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
which is less precise. Both change with differing abduction angles. Brown6 and Toledo and Ger7 clinically assessed web height using the normal palmar crease and adjacent normal web as reference points. These methods can be applied only to patients with single syndactyly and normal palmar creases, and assume that the second, third, and fourth web positions relative to the palmar crease are the same. Consistent abduction angles were not used. Shewell et al2 used photographs of adult hands held flat against a glass plate with fingers abducted 10 . This palmar technique measured the web position as a ratio based on the dominant PIP and distal interphalangeal joint creases. Dominant creases were judged where they crossed the edge of the finger. If there were 2 dominant creases, measurements were taken at the midpoint. Patterson and Nancarrow3 subsequently confirmed that this method was applicable to the hands of children. This method allows for easy comparison of serial postoperative measurements but requires repetitive photography. The authors conceded that the measurements were reproducible only if marked and measured by a single observer. This renders the validity of comparison between studies doubtful unless original photographs are available and are reassessed by a single observer. Their measurements of web position also relied on a web point that was judged to meet the skin of the finger. They found that with 10 separate observers who were asked to mark and measure the same pair of hands, individual differences in measurement ranged up to 7 mm, demonstrating a lack of inter-observer reliability; hence, the need for the same single observer for measurements over time. Withey et al8 developed a semi-quantitative grading system that assigned the level of web creep into thirds of the distance between the normal web position and the PIP joint crease. The normal web position was established by observing adjacent webs and those of the opposite hand, which must be normal for validity. The finger abduction angle was not standardized. Although simple to use, this subjective method was not assessed for intra-observer and inter-observer error. Radiological methods project soft and bony tissues in 2 dimensions, which makes measurements of the web space easier with respect to joint levels. The method described by Richterman et al4 had high intra-observer and inter-observer reproducibility. Vekris et al5 also assessed web creep using serial x-rays. However, these methods call into question the need to expose children to multiple doses of radiation. Furthermore, the financial burden for follow-up is increased. There is also no description of standardizing finger abduction angles during the radiological examination. J Hand Surg Am.
In view of the lack of precise descriptions of some existing methods, the limitations of others, and poor evidence of reliability, we assessed 2 clinical methods that measure web position. If web creep occurs, the index changes; this translates to an objective assessment of severity. We overcame the inaccuracy of palpating the MCP and PIP joint levels in the dorsal method by resting the calipers squarely on the dorsum of the proximal phalanx with both joints passively flexed as near to 90 as possible. In effect, this measurement takes reference from the dorsal metacarpal surface and the distal end of the proximal phalanx. The presence of joint stiffness, such as symphalangism in Apert syndrome, precludes the use of this method. In the palmar measurement technique, the most proximal basal digital and PIP joint creases in the axial line were easily identifiable in adults.2 We found this to be the same in simple and complex syndactylies of children. By definition, the former is only an anomaly of skin whereas the latter also has a bony connection distally. This observation was confirmed by the work of others who used similar methods but with less precise measurements, consistency of digit position, and assessment of reproducibility and reliability.2,3,8 We chose to standardize the position of the fingers in maximal passive abduction without excessive force and with relaxation of the adjacent webs. The absolute position of abduction in degrees is not important; rather, consistent placement of the fingers allows the web being measured to be taut. This minimizes the variability of web position with the varying ability of the patient to abduct the fingers actively. In this position of abduction, assessment of a web point may be performed in a more accurate manner by constructing a line from the midpoint of the web to the perpendicular axial line at 90 . The efforts of rigorous standardization yielded correlation coefficients of more than 0.8, indicating excellent intra-observer and inter-observer reliability for both dorsal and palmar techniques. We acknowledge that only one examiner established the ICC for intra-observer error for each technique and that the inter-observer error ICC was established between 2 examiners. Regrettably, the logistics of performing multiple measurements on volunteers returning for multiple clinic visits prevented an increased number of assessments. Nevertheless, either method may be used to establish a web index that describes the web position and quantify its change. We prefer the palmar method because it is easier to perform. It does not require joint flexion or the use of calipers. The web index can then be used to describe the web position clinically at the basal crease level r
Vol. 40, March 2015
OBJECTIVE MEASUREMENTS OF WEB SPACE POSITION
461
determining the clinical applicability of the method after syndactyly surgery is to use these techniques to measure the web index while the child is under anesthetic at the completion of surgery, and then at appropriate times postoperatively with the assistance of parents and other medical staff, to determine any change in the web index. Figure 4 shows an example of an immediate postoperative web index measurement. This picture also demonstrates the presence of double basal finger and PIP joint creases. We acknowledge that postoperative scarring and angulatory deformity may alter the position of creases and decrease the reliability of reference points for subsequent postoperative measurements. We do not believe this possibility will prevent the ability to use the web index efficiently to monitor web creep, but this can be proven only by future studies. As with most clinical measurements of web creep, these methods are not applicable to complicated syndactylies in which there is an absence of creases, joint stiffness, and angulatory deformity, because it is often impossible to establish reference points for web position. However, we believe that it is appropriate to apply the palmar technique to measure simple and complex syndactyly web position. FIGURE 4: Example of measurement of a postoperative web index for a simple, incomplete syndactyly. The web index measured 0.95. Note the double basal and PIP joint creases in the child’s hand.
REFERENCES 1. Tonkin MA, Tolerton SK, Quick TJ, et al. Classification of congenital anomalies of the hand and upper limb: development and assessment of a new system. J Hand Surg Am. 2013;38(9):1845e1853. 2. Shewell PC, Nancarrow JD, Fatah F. Quantifying interdigital web morphology. J Hand Surg Br. 1992;17(2):198e200. 3. Paterson P, Nancarrow JD. Defining a normal finger web in children. J Hand Surg Br. 1998;23(4):496e498. 4. Richterman IE, DuPree J, Thoder J, Kozin SH. The radiographic analysis of web height. J Hand Surg Am. 1998;23(6):1071e1076. 5. Vekris MD, Lykissas MG, Soucacos PN, Korompilias AV, Beris AE. Congenital syndactyly: outcome of surgical treatment in 131 webs. Tech Hand Up Extrem Surg. 2010;14(1):2e7. 6. Brown PM. Syndactyly—a review and long term results. Hand. 1977;9(1):16e27. 7. Toledo LC, Ger E. Evaluation of the operative treatment of syndactyly. J Hand Surg Am. 1979;4(6):556e564. 8. Withey SJ, Kangesu T, Carver N, Sommerlad BC. The open finger technique for the release of syndactyly. J Hand Surg Br. 2001;26(1):4e7. 9. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159e174. 10. Keret D, Ger E. Evaluation of a uniform operative technique to treat syndactyly. J Hand Surg Am. 1987;12(5 pt 1):727e729. 11. Percival NJ, Sykes PJ. Syndactyly: a review of the factors which influence surgical treatment. J Hand Surg Br. 1989;14(2):196e200. 12. Sharma RK, Tuli P, Makkar SS, Parashar A. End-of-skin grafts in syndactyly release: description of a new flap for web space resurfacing and primary closure of finger defects. Hand (N Y). 2009;4(1):29e34. 13. Nakamura J, Yanagawa H, Kubo E, Endo T. New modified method for the surgical treatment of syndactyly. Ann Plast Surg. 1989;23(6):511e518. 14. Ostrowski DM, Feagin CA, Gould JS. A three-flap web-plasty for release of short congenital syndactyly and dorsal adduction contracture. J Hand Surg Am. 1991;16(4):634e641. 15. Killian JT, Neimkin RJ. Syndactyly reconstruction by a modified Cronin method. South Med J. 1985;78(4):414e418.
(one-third or two-thirds the distance between the basal crease and PIP joint, at or distal to PIP joint crease) in a manner similar to but more precise than that of Withey et al.8 This study has a number of limitations, some of which have been already addressed. Placement and maintenance of the finger position on either side of the web to be measured may be less than consistent. However, we believe our method is more precise than those previously described, allowing a more accurate evaluation of the central web point position than previously obtained. Our methods use identical positioning and reference points no matter the age of the subject. We can see no problem that would preclude their application in the child’s hand, as Shewell et al2 and Paterson and Nancarrow3 found in the transfer of their method from adult to child. We acknowledge that application of these techniques of measurement is more difficult in the hands of children, because cooperation and maintenance of position are more difficult. The increased difficulty should not deter efforts to achieve a more precise and reliable assessment of web position and web creep after syndactyly surgery. Having confirmed the reliability of measurement in normal hands, the next phase in J Hand Surg Am.
r
Vol. 40, March 2015