If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Reid W. Draeger, MD, Department of Orthopaedics, University of North Carolina School of Medicine, 3102 Bioinformatics Building, CB #7055, Chapel Hill, NC 27599-7055.
Using a cadaveric model simulating clinical situations experienced during open reduction and internal fixation of proximal phalangeal fractures, the aim of this study was to evaluate the relationship between level of training and the rates of short, long, and ideal screw length selection based on depth gauge use without fluoroscopy assistance.
Methods
A dorsal approach to the proximal phalanx was performed on the index, middle, and ring fingers of 4 cadaveric specimens, and 3 drill holes were placed in each phalanx. Volunteers at different levels of training then measured the drill holes with a depth gauge and selected appropriate screw sizes. The rates of short, long, and ideal screw selection were compared between groups based on level of training. Ideal screws were defined as a screw that reached the volar cortex but did not protrude more than 1 mm beyond it.
Results
Eighteen participants including 3 hand fellowship–trained attending physicians participated for a total of 648 selected screws. The overall rate of ideal screw selection was lower than expected at 49.2%. There was not a statistically significant relationship between rate of ideal screw selection and higher levels of training. Attending surgeons were less likely to place short screws and screws protruding 2 mm or more beyond the volar cortex
Conclusions
Overall, the rate of ideal screw selection was lower than expected. The most experienced surgeons were less likely to place short and excessively long screws.
Clinical relevance
Based on the low rate of ideal screws, the authors recommend against overreliance on depth gauging alone when placing screws during surgery. The low-rate ideal screw length selection highlights the potential for future research and development of more accurate technologies to be used in screw selection.
However, the screw tip should not protrude excessively beyond the cortical surface because this may cause symptoms owing to irritation of nearby structures. These competing interests must be weighed against one another based on the anatomical location of the screw being placed.
There is a transversely oriented concavity of the volar surface of the proximal phalanges in the hand, which is most prominent at the proximal portion of the bone (Fig. 1).
Given the intimate relationship between the flexor tendons and the volar cortical surface, this concavity is clinically important when placing a dorsal-to-volar screw during treatment of a proximal phalanx fracture and emphasizes the importance of skill in accurately measuring depth.
Figure 1Volar concavity. Axial cut from micro–computed tomography scan demonstrates the transversely oriented concavity of the volar surface of a proximal phalanx.
Although multiple factors are used to determine appropriate screw length during surgery—including depth gauge measurements, fluoroscopic imaging, tactile feedback with screw placement, and knowledge of surgical anatomy and manufacturer’s screw specifications—accurate use of a depth gauge plays a critical role in this determination. Obtaining and interpreting information from a depth gauge can be difficult. Recent literature has demonstrated improved performance of surgical skills with higher levels of training utilizing simulated environments.
Surgeons may select a screw length based on depth gauging alone, without fluoroscopic guidance. The benefit of this approach may be faster operative time as well as decreased radiation exposure. The potential disadvantage to this approach is an increased rate of short or long screws, as well as the possibility of placing screws that need to be replaced. Prior biomechanical literature has demonstrated a decrease in pull-out strength of self-tapping cortical screws with multiple reinsertions.
To the authors’ knowledge, the ability of surgeons to select ideal screw lengths based on depth gauge skills alone without fluoroscopic assistance has not been investigated.
We used a cadaveric model to simulate proximal phalangeal open reduction and internal fixation (ORIF) to further evaluate appropriate screw selection using a depth gauge without fluoroscopic confirmation. The aims of this study were to evaluate the relationship between level of training and rate of ideal screw length selection based on depth gauge measurements in isolation. We hypothesized that increased level of training would lead to increased accuracy of screw selection.
Materials and Methods
The study protocol was reviewed by our institutional review board and was given exempt status.
Four fresh-frozen cadaveric upper extremity specimens were obtained through a standard cadaveric specimen procurement process. These 4 specimens came from 4 separate donors, 3 male and 1 female, with an average age of 71 years, and no known history of underlying pathology of or surgery to the hand. A dorsal approach to the proximal phalanx as described by Pratt
was performed on the index, middle, and ring fingers of each specimen utilizing a straight longitudinal skin incision. The extensor tendons were divided in line with the skin incision. Using a 1.4-mm drill bit (Stryker Variax Hand Plating System; Stryker Corporation, Kalamazoo, MI), 3 midline dorsal-to-volar holes were drilled in each bone. The most proximal and distal holes were drilled at the level of the proximal and distal metaphyseal-diaphyseal junction. The starting point for the middle hole was placed at the midpoint between the proximal and the distal holes (Fig. 2). Nine holes were drilled in 4 specimens for a total of 36 holes.
Figure 2Location of holes drilled in this study. Dorsal-to-volar holes were drilled at the proximal and middle metadiaphyseal junctions and the midpoint between them.
Orthopedic surgery residents and hand fellow-ship–trained attending physicians then voluntarily participated in the study. Using a standard 1.2-mm-diameter depth gauge (Stryker Variax Hand Plating System; Stryker Corporation, Kalamazoo, MI), each participant was asked to use a depth gauge on each drill hole and then report the screw length they would use for the respective drill hole, with options consisting of integral increments of 1 mm (and without consideration of plate variables). No screws were actually placed by the participants during this portion of the study. No image intensifier was made available for use during the measuring portion of the study.
All proximal phalanges were then stripped of all soft tissue. The drill holes were then measured using a digital caliper (Mitutoyo Absolute Series 500 digital caliper; Mitutoyo Corporation, Kawasaki, Kanagawa, Japan) as demonstrated in Figure 3.
Figure 3Digital calipers were outfitted with conical tips, calibrated, and used to measure the thickness of the phalangeal shaft at each drill hole to determine the ideal screw length for each site. The caliper points registered at the midline of the phalangeal shaft on the cortical surface immediately adjacent to the drill hole.
In theory, an ideal screw is the smallest available size in which screw threads obtain purchase in the volar cortex. Prior biomechanical data have demonstrated that screws that extend past the volar cortical surface have higher pull-out strength than screws that do not extend past the surface and that there is no marginal increase in pull-out strength for screws that extend beyond 1 mm past the volar surface of the bone.
For the purposes of this study, we assumed fully threaded screws were available in 1-mm integral increments. We assumed the threaded portion of the screw was equal to the length of the screw, without correction for head height or nonthreaded tips, 2 variables that vary among manufacturers. We defined an ideal screw as one in which the tip of the screw reached the surface of the volar cortex but did not protrude more than 1.0 mm beyond it (Fig. 4).
Figure 4Concept of ideal relative screw tip position. Axial and sagittal cuts from micro–computed tomography scan of the proximal phalanx. For the purposes of this study, an ideal screw was considered one in which the tip of the selected screw reached the volar cortical surface but did not protrude more than 1 mm (as indicated by the green lines) beyond the volar cortical surface, thereby minimizing irritation to the flexor tendons.
To characterize each selected screw as ideal or nonideal, the selected size was compared with the depth of the holes as measured by the digital caliper. This relationship characterizes the location of the screw tip relative to the volar surface of the phalanx. This calculated relationship was referred to as the relative screw tip position, with negative values indicating that the tip of the screw remained contained within the phalanx, and positive values indicating protrusion of the screw tip beyond the volar surface (Fig. 4). This formula is summarized as
Ideal screws in this study were screws with relative screw tip position between 0 and 1.0 mm. Screws outside of this range were considered nonideal screws and were further subdivided into short screws (with a negative relative screw tip position) and long screws (with a relative screw tip position > 1.0 mm).
Statistical analysis
Each screw selected by each participant for each drill hole was classified as ideal, short, or long. The rates of ideal, short, and long screws for each group—attending-level physicians, upper-level (postgraduate year [PGY] 4,5) and lower-level (PGY1,2,3) residents—were then compared using Pearson chi-square analysis, except in instances when a cell in the contingency table was less than 5, in which case the Fischer exact test was used. For all statistical comparisons, P less than .05 was considered significant.
We performed a power analysis assuming alpha of 0.05 and power of 80%. Assuming a rate of ideal screw selection of 75% in the attending group, with 3 attendings and 15 residents participating, our experiment would allow us to detect a difference in rate of 14%.
Results
Eighteen volunteers selected screw sizes for 36 holes resulting in a total of 648 screw sizes chosen. Three hand fellowship–trained attending physicians, 9 junior-level residents (PGY levels 1, 2, and 3), and 6 senior-level residents (PGY levels 4 and 5) participated. Figure 5 demonstrates a scatterplot summarizing the relative screw tip position of each screw size selected.
Figure 5Screw tip position relative to the volar cortex for each screw selected in the study by level of training. Positive screw tip positions indicate protrusion of the screw tip beyond the volar cortex, whereas negative relative screw tip positions indicate the screw tip resides within the phalanx and does not protrude beyond the volar cortex. The solid black line indicates the value 0, representing the volar surface. Data points between the solid black line and the solid green line indicate screws with tips within the ideal range of 0 to 1.0 mm beyond the volar cortical surface.
Ideal screws were those in which the tip of the screw reached the surface of the volar cortex but did not protrude more than 1.0 mm beyond it. The overall rate of ideal screws was 49.2% (319 of 648) (Table 1). The rate of ideal screws by attending physicians was 54.6% (59 of 108), upper-level residents was 50.9% (110 of 216), and lower-level residents was 46.3% (150 of 324). There was not a statistically significant relationship between rate of ideal screws with increased level of training (Fig. 6).
Table 1Rate of Ideal Screw Selection by Level of Training
Figure 6Rate of ideal screws by level of training. Ideal screws were those in which the tip of the screw reached the surface of the volar cortex but did not protrude more than 1.0 mm beyond it. Error bars indicate 95% confidence intervals.
Short screws were those that failed to reach the volar cortical surface. The overall rate of short screws was 9.3% (60 of 648). Attending physicians were significantly less likely to select a short screw than resident physicians (3.7% [4 of 108] vs 10.3% [56 of 540]; P < .05). Breaking the resident groups into upper- and lower-level groups, the rate of short screws was decreased in the attending group compared with the lower-level resident group (3.7% [4 of 108] vs 12% [39 of 324]; P < .05) but not significantly decreased compared with the upper-level resident group (3.7% [4 of 08] vs 7.9% [17 of 216]; P = .24). There was no statistically significant difference in rate of short screws between the upper- and the lower-level resident groups (7.9% [17 of 216] vs 12.0% [39 of 324]; P = .79). Table 1 and Figure 7 summarize rates of short screws by level of training.
Figure 7Rate of short screws by level of training. Attending physicians were significantly less likely to select a short screw than resident physicians (4 of 108 [3.7%] vs 10.3% [56 of 40]; P < .05) as denoted by the *. Error bars indicate 95% confidence intervals.
Long screws were those which protruded more than 1 mm beyond the volar surface. The overall rate of long screws was 41.5% (269 of 648). There was no difference between groups in the rate of long screws. Figure 8 summarizes rates of long screws by level of training.
Figure 8Rate of long and long ≥ 2-mm screws by level of training. There was no difference in rate of long screws (relative screw tip position > 1 mm) by level of training. In the subgroup analysis, attending physicians were significantly less likely to place a screw that protruded 2 mm or more (long ≥ 2 mm) beyond the volar cortical surface as denoted by the *. Error bars indicate 95% confidence intervals.
Clinically, screws that protrude more beyond the volar cortex would be more likely to lead to clinically relevant pathology. Therefore, we performed a subanalysis of long screws evaluating screws protruding 2 mm or more beyond the volar cortex (long ≥ 2 mm). The overall rate of long of 2 mm or greater was 13.9% (90 of 648), with rates of attendings versus upper-level versus lower-level residents 5.6% (6 of 108), 13.0% (28 of 216) and 17.3% (56 of 324), respectively. The rate of long 2 mm or greater was significantly lower in the attending surgeon group versus the residents (5.6% [6 of 108] vs 15.6% [84 of 108]; P < .05) as demonstrated in Figure 8.
Discussion
Screw selection is a challenging skill. Multiple key pieces of information should be integrated to select an ideal screw. Knowledge of local surgical anatomy as well as knowledge of screw design are important variables with which surgeons should be familiar. During surgery, the ability to use a depth gauge and interpret the measurements as well as utilization of and interpreting information relating to fluoroscopy are vital skills in selecting appropriately sized screws. Finally, in addition to being a tool to assist in screw selection, fluoroscopy can also be a tool to assess appropriateness of a screw after it is placed. Assessment of tactile feedback when placing the screw is yet another checking mechanism to ensure appropriate length when a screw is placed.
The current experiment evaluated subjects’ ability to pick appropriate screw lengths for a specific bone (proximal phalanx) using a depth gauge without other modalities typically available during surgery (including fluoroscopy/tactile feedback while placing the screw). The study controlled for screw design variability (making the assumption that fully threaded screws were available to participants in integral increments). Therefore, the information that went into subjects’ screw selection was the information provided by the depth gauge and the subjects’ baseline knowledge of the anatomy proximal phalanx.
The key finding from the study is the overall low rate of ideal screw placement. Our overall rate of ideal screws was lower than what we anticipated in our power analysis, even among our most experienced group consisting of fellowship-trained attending physicians. Given the overall lower than expected rate of ideal screws, the study may have been underpowered to detect a difference between the attending and the resident groups. In the current study, increased level of training was associated with significantly lower rates of short and long 2 mm or greater screws; however, there was no difference in rate of ideal screws between groups. It is possible that excessively long screws placed dorsal to volar and resultant hardware tendinopathy of the flexor tendons may account for some of the postoperative pain and stiffness associated with ORIF of proximal phalanx fractures (Fig. 9).
Other anatomical areas in which excessively long screws may have a high rate of clinically adverse effects include any articular surface and volar-to-dorsal screws for ORIF of distal radius fractures. Whether or not tactile and visual feedback from actual screw placement or fluoroscopic imaging would have increased the accuracy of screws selected was not investigated in the current study, and future research may determine whether or not these adjuncts would increase the rate of ideal screw placement, However, based on the overall low rate of ideal screw length, the authors would advocate against overreliance on depth gauging alone.
Figure 9Qualitative demonstration of A short, B ideal, and C long screws in a proximal phalanx prior to soft tissue stripping. A Short screw, in which the tip of the screw does not reach the volar cortical surface. B Ideal screw, in which the tip of the screw reaches the volar cortical surface but protrudes < 1.0 mm. C Long screw. In this example, the tip of the screw protrudes > 1.5 mm from the volar cortical surface. * bone; t, tendon.
Furthermore, this study highlights some of the limitations of current depth gauging technology that may benefit from future research and development. Technologies that facilitate improved accuracy of screw placement may help minimize the need to switch out screws, which may be advantageous from a biomechanical standpoint.
There is variability between manufacturers in head height and terminal thread-to-tip distance in self-tapping screws used for ORIF of small bones. For the purposes of this study, we made the assumption that the reported length of the screws selected was equal to the threaded shank of the screw and the screws were threaded to the tip. In true clinical situations, an understanding of manufacturer-reported screw length versus functional screw length is important to ensure optimal screw selection. For example, in screws utilized in our institution, manufacturer reported screw length is 0.8 mm greater than the threaded shank length because head height is incorporated into manufacturer-reported length (Fig. 10). Failing to appreciate this manufacturer-related variability may make a surgeon more likely to place a biomechanically inferior screw.
Figure 10Example of discrepancy between manufacturer-reported length and functional length of a screw. In this example, the 0.8-mm head height is incorporated into the manufacturer-reported screw length. Failure to take this into account may cause a surgeon to place an excessively short, mechanically inferior screw.
This study has several limitations. There is a paucity of biomechanical data defining when a screw is of ideal or adequate mechanical strength. Therefore, challenges arise in defining an ideal screw length that takes into account both biomechanical integrity as well as minimizing irritation of nearby soft tissue. In this study, we defined an ideal screw as the smallest screw available in which the threads of the screw wound have reached the surface of the far cortex. The authors believe this was the most methodologically sound manner in which to define the ideal screw length. Future studies to determine cortical purchase of a small-bone screw necessary to achieve adequate mechanical strength are necessary to correctly define true ideal screw lengths for small-bone screws. In clinical practice, there may be screws that obtain adequate biomechanical stability without putting the patient at risk for soft tissue irritation. For example, in patients or portions of bone with thicker volar cortices, adequate mechanical stability may be obtained even if the screw tip does not protrude beyond the volar surface of the bone.
Thus, the rate of ideal screws in this study likely underestimates the rate of clinically acceptable screws. Furthermore, in the operative theater, an image intensifier is available to assess appropriateness of a selected screw. In addition, when a screw is placed during surgery, an increase in resistance in the terminal rotations of the screw provides confirmation that a screw is of an appropriate length to obtain purchase in the volar cortex. This can be helpful in elucidating whether a screw is of adequate length and potentially minimize the selection of short screws. The measurers in this study did not have either of these checking mechanisms at their disposal, and therefore, a higher rate of excessively long and short screws may have been selected than would be expected clinically. Finally, given the concave surface of the volar cortex of proximal phalanx, the measurement provided by the depth gauge may vary depending on which part of the concavity the foot of the depth gauge engaged, which may have had an effect on the screw selection for each hole by the participants.
Acknowledgments
Funding for the cadaveric specimens used in the study was from intradepartmental research funds at the research institution. The Stryker Variax Hand System screws and accessories (including depth gauge devices) were provided free of cost by Stryker Corporation, Kalamazoo, MI. The representatives from the company did not play a role in designing the study nor any role in data analysis nor manuscript preparation.
30683-4&id=gr1.jpg){kind=link}
30683-4&id=gr2.jpg){kind=link}
30683-4&id=gr3.jpg){kind=link}
30683-4&id=gr4.jpg){kind=link}
30683-4&id=gr5.jpg){kind=link}
30683-4&id=gr6.jpg){kind=link}
30683-4&id=gr7.jpg){kind=link}
30683-4&id=gr8.jpg){kind=link}
30683-4&id=gr9.jpg){kind=link}
30683-4&id=gr10.jpg){kind=link}