To

Surgical

Tech

nique

1

SURGICAL TECHNIQUE

Computer-Assisted Navigation of Volar Percutaneous

Scaphoid Placement

Eric Walsh, MD, Joseph J. Crisco, PhD, Scott W.Wolfe, MD

Purpose To investigate a computer-assisted technique for retrograde insertion of a percutaneousscaphoid screw and compare insertion time, accuracy, and radiation exposure to the traditionaltechnique. We hypothesize that computer-assisted navigation of volar percutaneous scaphoidscrew placement would improve accuracy, require less time, and diminish radiation exposurewhen compared to the traditional technique.

Methods Ten matched pairs of cadaveric wrists were randomized to computer-assisted versustraditional volar percutaneous scaphoid screw placement. Time of the overall procedure, set-uptime, time for ideal guide wire placement, and radiation time were recorded. Number of K-wireattempts was also recorded. Finally, accuracy of planned screw axis and actual screw axis werecompared. Student’s t-tests and rank sums were used to determine whether the differences inoutcome variables between computer-assisted and traditional techniques was significant, with analpha level of 0.05.

Results Although the overall time of the 2 procedures and the set-up time were not differentbetween the 2 groups, the time for placement of the K-wire was halved in the computer-assistedpercutaneous scaphoid fixation group, and the number of K-wire attempts needed for accuratescrew placement approached clinical significance. Although the radiation exposures for theindividual components of set-up time and final check time were not different, the radiationexposures for global time of the procedure, K-wire placement, and screw placement wereclinically significant.

Conclusions Computer-assisted navigation of volar percutaneous scaphoid screw placement takesno more time that traditional methods and significantly reduces the amount of radiation exposureto the patient. Although not statistically significant, the technique reduced the number ofincorrect passes of the K-wire, requiring a single attempt in 4 of the 5 specimens. (JHand Surg 2009;34A:1722–1728. Copyright © 2009 by the American Society forSurgery of the Hand. All rights reserved.)

Key words Navigation, percutaneous, scaphoid.

(nhj

HE ADVANCEMENT OF modern surgical tech-niques has allowed surgeons to treat patientswith arthroscopic, minimally invasive, mini-

pen, and now computer-assisted orthopedic surgery

From the Orthopedic Group, Inc, Pawtucket, RI; Brown Medical School, Rhode Island Hospital, Provi-dence, RI: Weill Medical College of Cornell University, Hospital for Special Surgery, New York, NY.

Received for publication April 6, 2008; accepted in revised form August 15, 2009.

The authors acknowledge Joshua Richards, MD, who contributed to the design of the study; EvanLeventhal, MS, for his work on part of the statistical calculations; Christopher Plaskos, PhD, ResearchDirector, Praxim Med Division, Hospital for Special Surgery, who helped with the computer naviga-tion, methods, and materials; and Carrine Granchi, MSc, Praxim Med Division, Hospital for Special

Surgery, who helped with the computer navigation, methods, and materials.

722 � © ASSH � Published by Elsevier, Inc. All rights reserved.

CAOS). Hand surgeons frequently use the first 3 tech-iques with success. The CAOS surgical techniques,owever, have been reserved for the treatment of largeoints like hips and knees. The technology used in

esearch support was donated by Praxim (Praxim, Med Division, Grenoble, France). The implantsere Acutrak 2 headless screws donated by Acumed (Acumed LLC, Hillsboro, OR) in support of our

tudy. No other benefits were received.

orrespondingauthor:EricWalsh,MD,HandandUpperExtremitySurgery,OrthopedicGroup, Inc.,88 Pawtucket Avenue, Pawtucket, RI 02860; e-mail: EFWalshMD@yahoo.com.

363-5023/09/34A09-0023$36.00/0oi:10.1016/j.jhsa.2009.08.009

Rws

C5

0d

es.

COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION 1723

SurgicalTechnique

CAOS has allowed reconstructive orthopedic surgeons

to provide greater accuracy in total joint replacement.1

We have developed a modification of CAOS technol-ogy known as computer-assisted percutaneous scaphoidfixation (CAPS). The CAPS technology has been de-veloped specifically to increase the accuracy of smallbone and joint surgery, allowing a computer to provideaccurate calculation and guidance for retrograde percu-taneous scaphoid screw placement.

Percutaneous fixation has been developed as a viablealternative to open surgery without surgical wounds andthe associated ligamentous disruption.2–12 The accuracyof percutaneous fracture fixation, however, is limited bythe experience of the surgeon and the imaging tech-niques available in the operating room. Percutaneousfixation of the scaphoid is difficult because of the smallsize and complex shape of the bone. In addition, themajority of the scaphoid is covered by articular carti-lage, making the placement of a percutaneous screwwithin the confines of the subchondral bone critical.

Recent studies have looked closely at 2 aspects ofpercutaneous scaphoid screw fixation: optimizing theposition of the screw within the scaphoid and confirm-ing complete containment of the screw within the bone.These studies describe various approaches, includingdorsal, volar, and trans-trapezial, with or without radio-graphic, computed tomography, or arthroscopic assistedtechniques.13–15 Even with these modern imaging mo-dalities, a seasoned hand surgeon can struggle to accu-rately place a percutaneous screw safely within thescaphoid.

This proof of concept study examines the accuracy andtime involved with a new technique of percutaneousscaphoid screw placement using computer-assisted navi-gation. We hypothesized that computer-assisted scaphoidscrew placement would improve the accuracy, minimize

FIGURE 1: The 2 compared set-ups for the procedures. Twoscaphoid screw placement and the B computer-assisted techniqu

the time and the number of K-wire attempts, and diminish

JHS �Vol A, No

the surgeon and patient’s radiation exposure when com-pared to the traditional percutaneous technique.

MATERIALS AND METHODSThis study was approved by the institutional reviewboard of the research division of Hospital for SpecialSurgery. Ten cadaver wrists were randomized to 1 of 2surgical procedures. Five were randomized to comput-er-assisted volar percutaneous scaphoid screw place-ment (Acutrak screw, Acumed LLC, Hillsboro, OR)and 5 were randomized to traditional volar percutane-ous screw placement.

All surgical procedures were performed by a singlesurgeon. For each surgical procedure, the followingoutcome variables were measured: time of overall pro-cedure global time; set-up time; number of K-wireattempts; time for ideal guide wire placement; radiationtime for overall procedure global, set-up time, K-wireplacement, screw placement, and final check; and pre-cision of the actual screw axis to the planned target axis.

Student’s t-tests and rank sums were used to deter-mine whether the differences in outcome variables be-tween computer-assisted and traditional techniques wassignificant, with an alpha level of 0.05.

Surgical protocol for traditional volar percutaneous screwplacement

The traditional technique we use is that described byHaddad and Goddard.8 Each cadaveric wrist was sus-pended in skeletal traction with a Steinmann pinthrough the olecranon and the thumb suspended in afinger trap (Fig. 1). Ten pounds of skeletal traction wasused. The wrist was then placed into the field of view ofa mini C-arm (OEC GE Medical Systems, Inc., Law-rence, MA). The fluoroscope was then adjusted to en-sure that adequate images of the scaphoid could be

res of the setup for both the A traditional volar percutaneous

pictu

obtained, and the time for this set-up recorded and

vember

1724 COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION

Surgical

Tech

nique

defined as the time for procedure—set-up time. A K-wire was then placed with a standard K-wire driverthrough a volar percutaneous incision at the tip of thescaphoid tubercle. No wires were placed across thetrapezium nor was any part of the trapezium removed tohelp position the K-wire. The starting point was deter-mined via the fluoroscope to correspond to a point 2 to4 mm ulnar and dorsal to the scaphoid tubercle.

When the guide wire was positioned and confirmedon multiple fluoroscope images, the time was againrecorded and defined as time for procedure—time forK-wire placement. The number of K-wire attempts toachieve this ideal placement was also recorded. Theguide wire length from the visible subchondral boneedges proximally to distally within the scaphoid wasmeasured with the screw measuring device and con-firmed with another guide wire held at the entry pointand marked with a hemostat. We subtracted 4 to 5 mm(as is generally accepted in practice) from this measure-ment to calculate the screw length to ensure againstarticular protrusion. The wire was then drilled throughthe scaphoid and into the distal radius to hold it in placefor the drilling. The wire was over-drilled with a can-nulated reamer and the starter drill used for the proxi-mal cortex. The fluoroscope was used to determineproper trajectory and depth. A cannulated screw wasthreaded over the guide wire. Before the wire wasremoved, fluoroscopic images were taken to confirmposition of the screw in multiple planes. The wire wasremoved and the overall time for the procedure—global was recorded.

The radiation time was recorded for the entire pro-cedure global and also for the set-up time, placement ofthe K-wire, screw placement, and final check (Fig. 4).

Surgical protocol for computer-assisted volar percutaneousscrew placement

The only difference in the set-up between the CAPStechnique and the traditional technique involved thefabrication and application of a custom-molded, ther-moplastic short-arm wrist splint to the cadaveric wrists.The splint was tightly affixed with straps to the under-lying hand and wrist and equipped with immobile re-flective markers for communication with the 2 camerasof the navigation device (Praxim, Med Division, Grenoble,France) (Fig. 1). When the splint was in place and thenavigation cameras were confirmed to be appropri-ately positioned, the timer was started for the proce-dure. Four to 9 fluoroscopic images were taken in 20°increments from lateral to posteroanterior, while thewrist position was captured by the navigation system.

This allowed the surgeon to choose and compare

JHS �Vol A, No

images that best aided in ensuring that the screwtrajectory was safely contained within the scaphoid.When the images were completed, the computer reg-istered the fluoroscopic field within the navigationenvironment. The fluoroscope was then removedfrom the surgical field to allow the surgeon to per-form the percutaneous scaphoid procedure unhin-dered by the C-arm (Fig. 2). Each of the surgicalinstruments was instrumented with reflective markersso that a virtual K-wire was projected onto eachsaved fluoroscopic image simultaneously. The sur-geon then aligned the guide wire in all views toidentify and lock in an ideal trajectory within thescaphoid. The ideal trajectory was then digitally pro-jected onto each image, and the surgeon prepared todrill. In most cases, the guide wire was placed using4 images simultaneously, and the others were used toconfirm that the trajectory never violated a corticalsurface. During drilling, the navigation device in-stantly enabled the surgeon to verify that the actualwire was proceeding along the ideal trajectory path inall fluoroscopic views. When the wire was in place,the length of the screw could be determined from10-mm increments on the K-wire trajectory, and theappropriate-length screw was advanced into thescaphoid. The CAPS system provided 10-mm incre-mental markers on the trajectory path (yellow dots inFig. 2). The length of the screw was calculated di-rectly from the trajectory markers and confirmedwith the traditional measuring device for the Acutraksystem. We again subtracted 4 to 5 mm from themeasurement to ensure complete intraosseous screwplacement beneath the articular surface of the bone.Final position was checked with the fluoroscope, andboth the surgical time and the radiation exposurewere recorded. The procedure times and radiationtimes are presented in Figures 3 and 4. Subsequently,the actual screw axis was compared to the targetedideal axis, and variance was calculated.

When all 10 cannulated screws had been placed intothe cadaveric scaphoids, the scaphoids were surgicallyremoved and visually inspected.

RESULTSThe mean global time of the procedure for the tradi-tional method was 44.6 min (SD � 18.94 min) com-pared to 50.6 min (SD � 14.38 min) for the CAPSmethod. There was no statistical difference betweenthese times (p � .58). The set-up time of the procedurewas 10.8 min (SD � 8.47 min) for the traditional and11.0 min (SD � 4.3 min) for the CAPS method; again

there was no statistical difference between these times

vember

COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION 1725

SurgicalTechnique

FIGURE 2: The Praxim image viewer. This picture shows the surgeon using 9 captured images that were used to place the volar scaphoid

screw. The blue line is the actual computer-navigated guide wire, and the dotted yellow line is the projected trajectory of the wire and screw.

FIGURE 3: Three different times for the procedure: time for global procedure; time for setup; time for K-wire placement. All

times were not statistically significant.

JHS �Vol A, November

1726 COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION

Surgical

Tech

nique

(p � .96). Finally, the time for K-wire placement was15 min (SD � 8.15 min) for the traditional method and7.8 min (SD � 6.98 min) for the CAPS method, which,given the small number of specimens tested, did notreach statistical significance (p � .17) (Fig. 3.).

The number of K-wire attempts was 2.4 (SD � 1.14)for the traditional method and 1.2 (SD � 0.45) for theCAPS technique, which approached statistical signifi-cance (p � .06).

The radiation exposure for the traditional methodversus CAPS is documented in Figure 4 and Table 1.The p values were significant for the global radiationexposure time (p � .02), the placement of the K-wire(p � .05) and the screw placement (p � �0.001). Theradiation time for the set-up (p � .71) and the finalcheck were not statistically significant (p � .71).

The angle between the projected and planned axis of

TABLE 1. Time (seconds) of Radiation Exposurefor the Traditional Method vs CAPS

Traditional CAPS

Global radiation 258.20 � 142.70* 72.20 � 7.20*

Setup 36.40 � 17.71 32.4 � 15.01

Placement of K-wire 160.00 � 133.98* 22.60 � 8.71*

Screw placement 45.80 � 7.26* 3.40 � 4.67*

Final check 16.00 � 10.27 13.80 � 7.69

*Indicates statistical difference.

FIGURE 4: This graph compares the radiation exposure for thand final check were not statistically significant; however, theplacement, and screw placement were statistically significant.

the scaphoid screw and the actual screw placement for

JHS �Vol A, No

the computer-assisted navigation technique was calcu-lated. The planned versus actual angle was 1.64° SD �0.56°.

Finally, on gross inspection, 1 cortical violation ofthe proximal scaphoid articular surface occurred usingthe traditional method—the screw was positioned tooradially. In addition, 1 cortical violation of scaphoradialarticulation occurred using the navigation method—thescrew was in ideal axis position but initially was ad-vanced too far and was then repositioned before the endof the procedure. All computer-assisted screw positionswere in the center of the scaphoids.

DISCUSSIONAlthough percutaneous techniques show encouragingresults and might have advantages over conservativeand open treatments, they are not without complication.Bushnell et al. looked retrospectively at 24 patients whohad percutaneous screw fixation using cannulatedscrews.16 They cite a 29% complication rate, includingnonunion, screw malposition, hardware problems, andpostoperative fracture of a proximal pole.

Several studies have been done to refine the tech-nique of volar percutaneous placement. In 2001, Mena-pace et al. performed a cadaver study and determined a“safe zone” for screw placement and an entry point fora guide wire on the scaphoid tuberosity using anthro-pometric analysis of radiographs and computed tomog-raphy scans.17 Kamineni et al. performed volar percu-taneous scaphoid screw fixation in 32 cadaveric wrists

rious parts of the procedure. Radiation exposure for the setupiation exposure for the global time for the procedure, K-wire

e varad

and reported measurements between the starting point

vember

COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION 1727

SurgicalTechnique

and vital structures.10 They determined that the super-ficial palmar branch of the radial artery might be at risk,and they recommend dissection to the scaphotrapezialjoint under direct visualization. Levitz et al. used quan-titative computed topographic analysis to find the opti-mal screw insertion point through a volar percutaneousapproach.13 They determined that a volar approach isfrequently hindered by the trapezium and a radial start-ing point or a transtrapezial approach to avoid theconcave volar surface of the scaphoid and the dorsalulnar cartilaginous surface.

Whether or not a dorsal approach confers technicalor safety advantages has not been conclusively deter-mined. Chan et al. compared volar and dorsal ap-proaches to determine which approach would allowmore central placement of a screw.18 They found thatthe dorsal–proximal approach allowed a more centralplacement in the distal pole of the scaphoid.

From a mechanical perspective, central placement ofthe screw in the proximal scaphoid pole has demon-strated advantages. Trumble et al. demonstrated thatcentral screw placement significantly decreases time tounion, and placement within the central third will alsoeliminate the chance of violating a cortical surface.19

McCallister et al. determined that central placement ofa screw in the proximal fragment had superior stiffnessand load to displacement compared with those placedeccentrically.15 Dodds et al. demonstrated that longerheadless screws (placed 2 mm from the each cortex) inthe central axis of the scaphoid provided greater me-chanical stability than short screws (placed 4 mm fromeach cortex), thereby increasing the number of screwthreads on either side of the fracture.20

Our study looks at a novel method to accurately andpredictably place a guide wire and a cannulated scaph-oid screw percutaneously within the safe zone of ascaphoid. Although the overall time of the 2 proceduresand the set-up time were not different between the 2groups, the time for placement of the K-wire washalved in the CAPS group, and the number of K-wireattempts needed for accurate screw placement ap-proached significance (p � .06). The scaphoid idealaxis is narrow, and the ideal starting point is quite small.Every time a K-wire is placed and removed from thescaphoid using the traditional technique makes subse-quent attempts more difficult and time consuming.Most important, each computer-assisted screw wasplaced precisely in the central third of the scaphoidproximal pole, whereas 1 of the 5 traditional scaphoidscrews was outside the central axis and breached thedorsal-radial articular surface. The clinical relevance of

the screw position is inherently important for prevent-

JHS �Vol A, No

ing malposition and decreasing the risk of leaving theoperating room with a screw that violates a corticalsurface. The cost of a single errant screw placement onthe viability of the neighboring articular surfaces cannotbe directly calculated, but it is a preventable error usingcomputer-assisted techniques.

Not surprising, there was a clinically significant re-duction in the radiation exposures for the global time ofthe procedure (p � .02), placement of the K-wire (p �.05), and placement of the screw (p � .001). Althoughthe radiation exposures for the individual componentsof set-up time and final check time were not different,the radiation exposures for the global time of the pro-cedure, K-wire placement, and screw placement wereclinically significant (p � .02). The radiation exposureoverall was reduced by 28%. Athwal et al. showed thatmini C-arms have significantly reduced the exposure tosurgeons working with these devices during treatmentof distal radius fractures.21 It should be appreciated,however, that the surgeon’s fingers are often directlywithin the beam when performing percutaneous scaph-oid surgery using traditional methods, and Athwal et al.cautioned strongly against this practice. Moreover, thepatient is always and necessarily exposed during sur-gery and, given the opinion by several scientists thatthere is no level of acceptable radiation, any methods bywhich the amount of radiation can be limited to thepatient and operating room staff should be consideredbeneficial.21

Although the scaphoid is small and asymmetric, theaccuracy of the navigation system was identified to bewithin 2° of the planned trajectory (1.64°). This allowedfor precise planning of the scaphoid screw axis to lie inthe center of the proximal pole, so as to optimize boththe position and the length of the scaphoid screw. Thesystem eliminated the repeated and cumbersome posi-tioning of the fluoroscopy unit while checking K-wireand screw position, minimized radiation exposure to thesurgeon and patient, and virtually eliminated malposi-tioned K-wire tracks in the bone. Screw length could bemeasured directly and accurately off the ideal trajectorypath, eliminating a potential source of error.

Because we limited this pilot study to 10 cadav-eric wrists, the small size of our study group mighthave limited our ability to define differences inother parameters that approached statistical signif-icance, such as the number and timing of K-wireattempts. The power of our comparison of K-wireattempts was only 0.4, indicating the likelihood ofa Beta error; that is, the inability to demonstrate asignificant difference due to the small number of

specimens, when a difference likely exists.

vember

1728 COMPUTER NAVIGATION-ASSISTED SCHAPHOID FIXATION

Surgical

Tech

nique

As with other applications, computer-navigated sys-tems require careful setup to ensure optimal accuracy.Proper positioning of the sensors on the wrist splint wascrucial to the system’s detection of the diodes in differ-ent wrist positions. It was also important that the cus-tom splint fit securely on each cadaveric wrist to pre-vent motion within the splint and introduction of error.Our accuracy measurements and low standard errorsindicated that the splint fabrication and fit achieved ourgoals. There certainly will initially be an increased costof using a custom splint and the CAPS method. Webelieve, however, that the increased accuracy and safetyof the CAPS technique would offset the serious com-plications and associated potential costs of subsequenttreatments required from a malpositioned screw.

This proof of concept study of the CAPS techniquefor percutaneous scaphoid surgery demonstrates that asurgeon can accurately and quickly place a guide wireand screw without the constraints of a fluoroscope andsignificantly reduce the radiation exposure to the patientand the surgeon. The high accuracy of the technique,the ability to rapidly and consistently place the screw inthe central third of the scaphoid proximal pole, and thereduction in the number of K-wire attempts are attrac-tive advantages of the technique, particularly for sur-geons who are relatively inexperienced in the nuancesof percutaneous scaphoid screw placement. Theremight be several other potential applications of thistechnology in surgery of the upper limb.

REFERENCES1. Parratte S, Argenson JN. Validation and usefulness of a computer-

assisted cup-positioning system in total hip arthroplasty. A prospec-tive, randomized, controlled study. J. Bone Joint Surg 2007;89A:494–499.

2. Wozasek GE, Moser KD. Percutaneous screw fixation for fracturesof the scaphoid. J Bone Joint Surg 1991;73B:138–142.

3. Rettig ME, Raskin KB. Retrograde compression screw fixation ofacute proximal pole scaphoid fractures. J Hand Surg 1999;24A:1206–1210.

4. Rettig ME, Kozin SH, Cooney WP. Open reduction and internal

fixation of acute displaced scaphoid waist fractures. J Hand Surg2001;26A:271–276.

JHS �Vol A, No

5. Haisman JM, Rohde RS, Weiland AJ, American Academy of Or-thopaedic Surgeons. Acute fractures of the scaphoid. J Bone JointSurg 2006;88A:2750–2758.

6. Adolfsson L, Lindau T, Arner M. Acutrak screw fixation versus castimmobilization for undisplaced scaphoid waist fractures. J HandSurg 2001;26B:192–195.

7. Inoue G, Shionoya K. Herbert screw fixation by limited access foracute fractures of the scaphoid. J Bone Joint Surg 1997;79B:418–421.

8. Haddad FS, Goddard NJ. Acute percutaneous scaphoid fixation. Apilot study. J Bone Joint Surg 1998;80B:95–99.

9. Yip HS, Wu WC, Chang RY, So TY. Percutaneous cannulated screwfixation of acute scaphoid waist fractures. J Hand Surg 2002;27B:42–46.

10. Kamineni S, Lavy CB. Percutaneous fixation of scaphoid fractures.An anatomical study. J Hand Surg 1999;24B:85–88.

11. Bedi A, Jebson PLJ, Hayden R, Jacobson J, Martus JE. Internalfixation of acute, nondisplaced scaphoid waist fractures via a limiteddorsal approach: an assessment of radiographic and functional out-comes. J Hand Surg 2007;32A:326–333.

12. Bond CD, Shin AY, McBride MT, Dao KD. Percutaneous screwfixation or cast immobilization for nondisplaced scaphoid fractures.J Bone Joint Surg 2001;83:1674–1681.

13. Levitz S, Ring D. Retrograde (volar) scaphoid screw insertion—aquantitative computed tomographic analysis. J Hand Surg 2005;30A:543–548.

14. Slade JF, Geissler WB, Gutow AP, Merrell GA. percutaneous inter-nal fixation of selected scaphoid nonunions with an arthroscopicallyassisted dorsal approach. J Bone Joint Surg 2003;85A:20–32.

15. McCallister WV, Knight J, Kaliappan R, Trumble TE. Central place-ment of the screw in simulated fractures of the scaphoid waist. Abiomechanical study. J Bone Joint Surg 2003;85A:72–77.

16. Bushnell BD, McWilliams AD, Messer TM. Complication in dorsalpercutaneous cannulated screw fixation of nondisplaced scaphoidwaist fractures. J Hand Surg 2007;32A:827–833.

17. Manapace KA, Larabee L, Arnoczky SP, Neginhal VS, Dass AG,Ross LM. Anatomic placement of the Herbert-Whipple screw inscaphoid fractures: a cadaveric study. J Hand Surg 2001;26A:883–892.

18. Chan KW, McAdams TR. Central screw placement in percutaneousscrew scaphoid fixation: a cadaveric comparison of proximal anddistal techniques. J Hand Surg 2004;29A:74–79.

19. McCallister WV, Knight J, Kaliappan R, Trumble TE. Central place-ment of the screw in simulated fractures of the scaphoid waist:a biomechanical study. J Bone Joint Surg 2003;85A:72–77.

20. Dodds SD, Panjabi MM, Slade JF. Screw fixation of scaphoidfractures: a biomechanical assessment of screw length and screwaugmentation. J Hand Surg 2006;31A:405–413.

21. Athwal GS, Bueno RA, Wolfe SW. Radiation exposure in handsurgery: mini versus standard C-arm. J Hand Surg 2005;30A:1310–

1316.

vember