Abstract
Introduction: Fractures of the fifth metacarpal are a common injury. In children, they are often stable and heal uneventfully, including after closed reduction. Fractures of the metacarpal diaphysis, conversely, tend to be unstable and may require surgical intervention. It is unclear at what point fractures of the fifth metacarpal fracture should be considered unstable. We identify the location at which a fifth metacarpal fracture is at risk for surgical intervention. Methods: A two-year retrospective cohort analysis of all children (age ≤18) with fifth metacarpal fractures presenting to a Pediatric Plastic Surgery clinic was performed. Clinical and radiographical data were acquired from electronic medical records. Fracture location and fracture angulation were measured, descriptive statistics were performed, and multivariable logistic regression was used to identify predictors for surgery. Results: Ninety-eight patients met criteria for review. Mean patient age was 14.0 years (SD 2.2), and 95% were male. Forty-two (44.2%) patients underwent closed reduction, and 12 (12.2%) patients were recommended for surgery. Increased fracture angulation on initial lateral radiograph (OR: 1.08 (1.03-1.14 95% CI), p = 0.006) and more proximal fractures (OR: 0.92 (0.86-0.98 95% CI), p = 0.008) were identified as independent predictors for surgery. Fractures occurring proximal to 70% of metacarpal length resulted in accurate detection of those requiring surgery with 72.4% sensitivity, 75.3% specificity, 95.5% and 27.6% negative and positive predictive value, respectively. Conclusions: In pediatric fifth metacarpal fractures, increased fracture angulation and those occurring proximal to 70% of the metacarpal length were independent predictors for surgery.
Keywords: Fracture, metacarpal, pediatrics, surgery
Résumé
Introduction: Les fractures du cinquième métacarpien sont des blessures fréquentes. Chez l’enfant, ce sont souvent des fractures stables qui guérissent sans problème, même après réduction fermée. En revanche, les fractures de la diaphyse du métacarpe tendent à être instables et peuvent nécessiter une intervention chirurgicale. On ignore à quel point les fractures du cinquième métacarpien doivent être considérées comme instables. Nous identifions l’emplacement auquel une fracture du cinquième métacarpien présente un risque d’intervention chirurgicale. Méthodes: Une analyse de cohorte rétrospective de deux ans a été réalisée avec tous les enfants (âgés de moins de 18 ans) se présentant dans une clinique de chirurgie plastique pédiatrique avec des fractures du cinquième métacarpien. Les données cliniques et radiographiques ont été extraites des dossiers médicaux électroniques. L’emplacement et l’angulation de la fracture ont été mesurés, des statistiques descriptives ont été effectuées et une régression logistique multifactorielle a été utilisée pour identifier les facteurs prédictifs de la chirurgie. Résultats: Quatre-vingt-dix-huit patients satisfaisaient les critères d’inclusion pour l’analyse. L’âge moyen des patients était de 14,4 ans (écart-type = 2,2) et 95% d’entre eux étaient de sexe masculin. Quarante-deux (44,2%) patients ont eu une réduction fermée de la fracture et la chirurgie a été recommandée pour 12 autres patients (12,2%). Un plus grand angle de fracture sur la radiographie initiale de profil (RR : 1,08 [IC à 95% : 1,03 à 1,14], P = 0006) et la nature plus proximale de la fracture (RR : 0,92 [IC à 95% : 0,86 à 0,98], P = 0008) ont été identifiés comme facteurs prédictifs indépendants de la chirurgie. Les fractures survenant en proximal de 70% de la longueur du métacarpe étaient correctement identifiées comme nécessitant une chirurgie avec une sensibilité de 72,4%, une spécificité de 75,3% et une valeur prédictive négative et positive de, respectivement, 95,5% et 27,6%. Conclusions: Dans les fractures pédiatriques du cinquième métacarpien, une augmentation de l’angulation de la fracture et les fractures survenant dans les 70% proximaux de la longueur du métacarpe étaient des facteurs prédictifs indépendants de chirurgie.
Mots-clés: fracture, métacarpien, pédiatrie, chirurgie
Introduction
Fifth metacarpal fractures are the most commonly injured bone in the hand.1,2 In adolescents, metacarpal neck fractures are estimated to account for approximately 28% of all hand fractures. 3 Fractures involving the neck or metaphysis of the metacarpal are termed Boxer's fractures. These fractures are often considered stable and heal uneventfully with immobilization or buddy taping with or without closed reduction.1,2,4,5 A fracture may be deemed stable if a simple orthosis can maintain reduction without causing severe pain or displacement of the fracture. 6 Studies in adults suggest that Boxer's fractures without clinical evidence of rotation, scissoring, or extensor lag can tolerate up to 70 degrees of angulation without requiring surgery or resulting in functional deficits.7,8 In contrast, fractures of the metacarpal shaft or diaphysis tend to be unstable, often requiring surgical fixation. 9 Clinically, however, it is unclear at what point the diaphysis ends and the metaphysis begins in both adults and children. Similarly, the point at which metacarpal fractures should be deemed unstable and considered for surgical intervention remains uncertain.
The purpose of this study was to identify the location at which a pediatric fifth metacarpal fracture may be at increased risk for surgical intervention. Secondary objectives were to identify other risk factors for surgical intervention and summarize the epidemiology of fifth metacarpal fractures in children.
Materials and Methods
A retrospective cohort study was performed from a prospectively maintained database of hand injuries presenting to a tertiary pediatric centre. All patients with a fifth metacarpal fracture between 2017 and 2019 were identified and reviewed. Patients were excluded if they were older than 18 years of age at the time of injury or presented to the Plastic Surgery Clinic more than 21 days post-injury. Fractures of the ‘base’ of the metacarpal were also excluded if identified as such by the treating surgeon or localized to the proximal 30% of the metacarpal when measured.
Available medical records and radiographs were reviewed for all included patients. Patient characteristics relating to age, sex, hand dominance, mechanism of injury and management were recorded. Patients were identified and grouped into one of two cohorts: those requiring or recommended surgery and those treated with immobilization with or without closed reduction. Surgical methods included closed reduction with percutaneous Kirschner-wire fixation or open reduction and internal fixation (ORIF). Indications for surgery were determined by the treating surgeon at the time of presentation based on the presence of persistent or recurrent scissoring, extensor lag, fracture angulation, or displacement despite attempted closed reduction.
All radiographs were reviewed retrospectively for objective measurement using our institution's Picture Archiving and Communication System (PACS) by one reviewer (CD). Thirty percent of all radiographs were independently reviewed by the senior author (KC) to ensure measurement reliability. Fracture angulation was measured on initial posterior-anterior (PA), lateral, and oblique x-rays for all patients when available on PACS. Measurement of fracture angles was based on the method described and validated by Lowdon (Figure 1A). 10 A line was drawn along the longitudinal axis of the medullary canal and a line from the centre of the metacarpal head to the fracture site. Fracture location was determined as the percentage of the total metacarpal length along the metacarpal shaft (Figure 1B).
Figure 1.
(A) Fracture angulation of θ = 37° measured on lateral radiograph 7 ; (B) fracture location of 64% measured as percentage of total metacarpal length (length of x divided by sum of lines x and y) on posterior-anterior radiograph.
Statistical Methods
Descriptive statistics were performed. Patient demographics, fracture characteristics, and fracture management were described using means, proportions, and medians. Univariate analysis of selected characteristics by surgery status were obtained using t-tests or chi-square tests for continuous and categorical factors, respectively. Medians were compared using Mood's test, a nonparametric test to compare medians of two independent samples. 11 Multivariable logistic regressions were used to estimate the effects of type of fracture and fracture location on odds of undergoing surgery. A significance level of 95% (alpha <0.05) was used in all analyses. Inter-rater reliability of radiographic measurements was calculated using intraclass correlation coefficient (ICC) to compare measurements by the two raters, CD and KC.
Results
Patient Demographics
A total of 98 patients met inclusion criteria. The mean age was 14 years (SD 2.2) and 95% of participants were male. The single most common method of presenting injury was an axial load (47%). Patient and clinical characteristics are summarized in Table 1.
Table 1.
Patient Demographics and Clinical Characteristics.
| Variable | Total (n = 98) | Surgery (n = 12) | No surgery (n = 86) | P-value |
|---|---|---|---|---|
| Age; mean (SD), years | 14 (2.2) | 15.2 (1.9) | 13.9 (2.2) | 0.028 |
| Male; n (%) | 93 (95) | 12 (100.0) | 81 (94.2) | NS a |
| Right Hand Dominance; n (%) | 90 (93) | 11 (91.7) | 79 (92.9) | NS |
| Right sided injury; n (%) | 87 (89) | 12 (100) | 75 (87) | NS |
| Mechanism of Injury; n (%) | NS | |||
| Fall during activity | 27 (28) | 2 (16.7) | 25 (29.1) | |
| Punching/axial load | 46 (47) | 5 (41.7) | 41 (47.7) | |
| Other | 25 (25) | 5 (41.6) | 20 (23.2) | |
| Underwent closed reduction; n (%) | 42 (44) | 8 (66.7) | 34 (41.0) | NS |
| Physis, open; n (%) | 60 (63) | 3 (27.3) | 57 (67.1) | 0.018 |
| Concomitant injuries; n (%) | 6 (6) | 0 (0) | 6 (7.1) | NS |
| Fracture angle (PA); mean (SD) | 22.3 (10.1) | 24.4 (12.5) | 22.0 (9.8) | NS |
| Fracture angle (lateral); mean (SD) | 30.9 (14.4) | 45.3 (15.8) | 29.1 (13.2) | 0.004 |
| Fracture location; mean % (SD) | 70.9 (9.7) | 61.7 (10.8) | 72.1 (8.9) | 0.002 |
NS = not significant.
Fracture Angles and Location
ICC of fracture location on PA, and angulation on PA, oblique, and lateral radiographs was 0.89 (95% CI, 0.79 to 0.95), 0.80 (95% CI, 0.63 to 0.93), 0.87 (95% CI, 0.73 to 0.93), and 0.82 (95% CI, 0.63 to 0.91) respectively. Mean fracture angle on lateral radiographs for patients who underwent surgery was 45.3° (SD 15.8) compared to 29.1° (SD 13.2) for non-surgical patients (p = 0.004). Mean fracture location on PA radiographs for surgical patients was at 61.7% (SD 10.8) of the total length compared to a mean of 72.1% (SD 8.9) for non-surgical patients, p = 0.002. Multiple logistic regression showed that an increased fracture angulation on initial lateral radiographs radiograph (OR: 1.08 (1.03-1.14 95% CI), p = 0.006) and more proximal fractures (OR: 0.92 (0.86-0.98 95% CI), p = 0.008) were both predictors of surgical intervention (Table 2). There was not a clear threshold with both high sensitivity and specificity regarding optimal angle on lateral view for prediction of surgery (Figure 2A and Table 3). Fractures occurring proximal to 70% of metacarpal length resulted in accurate prediction of those ultimately requiring surgery with 72.4% sensitivity, 75.3% specificity, 95.5% negative predictive value and 27.6% positive predictive value (Table 4 and Figure 2B).
Table 2.
Logistic Regression Model for Risk of Surgery.
| Beta Estimate | Standard Error | Z value | Pr(>|z|) | OR (95% CI) | |
|---|---|---|---|---|---|
| Intercept | 1.088 | 2.318 | 0.469 | 0.639 | |
| Ratio Of Location | −0.087 | 0.033 | −2.638 | 0.008 | 0.92 (0.86-0.98) |
| Angle measured on Lateral View | 0.073 | 0.026 | 2.756 | 0.006 | 1.08 (1.03-1.14) |
Equation from which we obtained the beta estimates: Logit(Having surgery) = 1.088 −0.087 × ‘Ratio Of Location’ + 0.073 × ‘Angulation measured on Lateral view’
Figure 2.
(A) Receiver operating characteristic curve for (a) angulation on lateral view and (B) fracture location in predicting need for surgery. Area under the curve (AUC): 77% (CI: 63%, 90%); (b): Optimal Cut-Off for Fracture location in Predicting Need for Surgery, Predicted Probabilities and Receiver Operating Characteristic. Area under the curve (AUC): 78% (CI: 63%, 93%).
Table 3.
Predictive Value of Fracture Angulation for Surgical Intervention.
| Threshold (angulation) | Sensitivity | Specificity | NPV | PPV |
|---|---|---|---|---|
| 10 | 100.00 | 3.53 | 100.00 | 11.83 |
| 20 | 100.00 | 28.24 | 100.00 | 15.28 |
| 30 | 100.00 | 49.41 | 100.00 | 20.37 |
| 40 | 45.45 | 75.29 | 91.43 | 19.23 |
| 50 | 36.36 | 97.65 | 92.22 | 66.67 |
| 60 | 18.18 | 98.82 | 90.32 | 66.67 |
| 70 | 9.09 | 98.82 | 89.36 | 50.00 |
| 80 | 0.00 | 100.00 | 88.54 | NaN |
| 90 | 0.00 | 100.00 | 88.54 | NaN |
Positive Predictive Value (PPV), and Negative Predictive Value (NPV); NaN, not a number.
Table 4.
Predictive Value of Fracture Location for Surgical Intervention.
| Threshold (location) | Sensitivity | Specificity | NPV | PPV |
|---|---|---|---|---|
| 40 | 0.00 | 98.82 | 88.42 | 0.00 |
| 50 | 18.18 | 96.47 | 90.11 | 40.00 |
| 60 | 45.45 | 89.41 | 92.68 | 35.71 |
| 70 | 72.73 | 75.29 | 95.52 | 27.59 |
| 80 | 100.00 | 7.06 | 100.00 | 12.22 |
| 90 | 100.00 | 1.18 | 100.00 | 11.58 |
Positive Predictive Value (PPV), and Negative Predictive Value (NPV).
While the terminology used by surgeons at our institution was relatively precise at differentiating fracture location, multiple terms were used to describe the same fracture's location (Figure 3). With the exception of ‘distal metadiaphysis’ and ‘head,’ terminology such as ‘neck’ and ‘subcondylar’ was consistently employed to indicate fractures at 76% (95% CI: 74.8-77.3) and 75% (95% CI: 73.5-76.8) metacarpal length while fractures of the ‘middle third’ and ‘midshaft’ consistently described fractures at 52% (95% CI: 45.0-59.3) and 57% (95% CI: 49.8-63.8) of the metacarpal length. Fractures of the ‘distal third’ consistently referred to fractures at 69% (95% CI: 65.8-71.6) of metacarpal length.
Figure 3.
Surgeon-reported fracture location compared to measured location of fracture with 95% confidence intervals. Size of dot represents the estimates of mean ratios of location and the size corresponds to sample size.
Management and Outcomes
Forty-two (44.2%) of the 95 patients with available data had closed reduction of the metacarpal fracture prior to initial presentation to the Plastic Surgery clinic. Surgery was recommended for 12 (12.2%) patients for indications which included worsening angulation (25%, n = 3), appearance (8.3%, n = 1) and unstable configuration, angulation and/or displacement (66.7%, n = 8). Ten (10.2%) patients elected to undergo surgical management: five were treated with ORIF (50%), and five with closed reduction and Kirshner wires (50%). For the 86 patients who were not surgical candidates or chose to not undergo surgery, there were no issues documented at follow up. The median number of plastic surgery clinic visits was 4.0 (IQR; 3.0, 4.0) for patients who were recommended surgery, and 2.0 (1.0, 2.0) for those who were not recommended surgery. Mean follow up was 3.1 weeks (SD 5.4) with a maximum of 47 weeks. Two patients, one of whom underwent surgery, reported pain at follow up. No patients required secondary surgery. Other complications recorded included concerns with appearance (n = 1), reaction to sutures (n = 1), infection (n = 3), and rash (n = 1).
Discussion
The most common fifth metacarpal fracture presentation in our study cohort was a young male with an axial load injury. This is in agreement with literature on pediatric and adult hand fractures.2,10,12–14 Objective measurement of metacarpal fracture location and angulation at initial presentation identified patients at higher risk for surgical intervention. Our retrospective cohort study of children with fifth metacarpal fractures demonstrates that a fracture location proximal to 70% of the metacarpal length was an independent predictor for surgical intervention. To our knowledge, this is the first report to objectively identify the location at which a fracture may be at higher risk for surgery. Our findings are in agreement with results from Hartley et al, which, despite not identifying the exact location, found that pediatric midshaft metacarpal fractures were associated with an increased need for surgery compared to metacarpal neck fractures. 9 Given the issue of nomenclature, however, the boundaries of where the metacarpal midshaft ends and begins remain debated.
Nomenclature for metacarpal fracture location is subject to some ambiguity. Surgeons at our institution used a variety of terms to describe a fracture's location, but the subjective classification appeared to accurately differentiate between fracture location. Problems in communication or interpretation, however, especially in transfer of care, may result. Clarity of communication with regards to fracture location is important given the potential need for surgical intervention with diaphyseal fractures. Consistent terminology or objective measurement of fracture location is advised.
Our study also found that an increased initial fracture angulation was associated with subsequent surgical intervention. However, the study was unable to define an exact angulation threshold with sufficient sensitivity and specificity to accurately predict need for surgery. While there is no clear consensus on angulation limits for treatment with surgery in the literature, these results do support the systemic review by Dunn et al, which demonstrated that angulation of up to 70° may be acceptable in adults with Boxer's fractures. 4 There is a paucity of evidence, however, to determine whether this conclusion also applies children. 15
While this study included the use of validated methods of measurement for radiographic angulation and location, 10 the current study is underpowered to explore if there is a statistical interaction between location and angle of fracture in predicting need for surgery. Moreover, the study was underpowered to test other variables that demonstrated a significant difference on univariate analysis such as fracture obliquity, physeal status, or patient age. It should be noted that the measured fracture location in our study may be limited by the measurement technique. Because the measurement is only based on the PA x-ray, significant fracture angulation may overestimate the fracture location. The measurement technique presented in this study, however, appears to be simple and reliable. Finally, the decision for surgery is based on individual surgeon judgment and may introduce bias.
Footnotes
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Approval: All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Dr. James J. Wiley CHEO Research Institute Endowment Fund, (grant number 762220101261).
Financial Disclosure Statement: The authors have the following to disclose:
Celina DeBiasio received financial support from the Dr James J. Wiley CHEO Research Institute Endowment Fund.
Description of Individual Author Contributions: CD: collected data, reviewed data including measuring radiographic fracture angles and locations, co-author of manuscript. ES: statistical and data analysis, coauthored and edited manuscript. RJW: statistical and data analysis, coauthored and edited manuscript. ST: co-authored the manuscript and provided edits. KC: reviewed data including measuring radiographic fracture angles and locations, co-authored the manuscript and provided edits.
Institution Where Work Was Conducted: Children's Hospital of Eastern Ontario (CHEO), Ottawa, ON (CD, ES, RJW, ST, KC)
IRB Approval Statement: Children's Hospital of Eastern Ontario (CHEO) REB (Number 19/130×) approval was obtained.
Statement of Informed Consent: As a retrospective review with no individual identifying patient information or data in this report, informed consent was not applicable.
ORCID iDs: Celina DeBiasio https://orcid.org/0000-0002-5960-3821
Kevin Cheung https://orcid.org/0000-0003-2481-5952
References
- 1.Poolman RW, Goslings JC, Lee JB, Statius Muller M, Steller EP, Struijs PAA. Conservative treatment for closed fifth (small finger) metacarpal neck fractures. Cochrane Database Syst Rev. 2005;2005(3):CD003210. http://search.ebscohost.com/login.aspx?direct=true&db=cmedm&AN=16034891&site=ehost-live. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mahabir RC, Kazemi AR, Cannon WG, Courtemanche DJ. Pediatric hand fractures: A review. Pediatr Emerg Care. 2001;17(3):153-156. doi: 10.1097/00006565-200106000-00001. [DOI] [PubMed] [Google Scholar]
- 3.Rajesh A, Basu AK, Vaidhyanath R, Finlay D. Hand fractures: A study of their site and type in childhood. Clin Radiol. 2001;56(8):667-669. doi: 10.1053/crad.2001.0753. [DOI] [PubMed] [Google Scholar]
- 4.Dunn JC, Kusnezov N, Orr JD, Pallis M, Mitchell JS. The boxer’s fracture: Splint immobilization is not necessary. Orthopedics. 2016;39(3):188-192. doi: 10.3928/01477447-20160315-05. [DOI] [PubMed] [Google Scholar]
- 5.Bansal R, Craigen MAC. Fifth metacarpal neck fractures: Is follow-up required? J Hand Surg Eur Vol. 2007;32(1):69-73. http://search.ebscohost.com/login.aspx?direct=true&db=cmedm&AN=17125893&site=ehost-live [DOI] [PubMed] [Google Scholar]
- 6.Cheah AEJ, Yao J. Hand fractures: Indications, the tried and true and new innovations. J Hand Surg Am. 2016;41(6):712-722. doi: 10.1016/j.jhsa.2016.03.007. [DOI] [PubMed] [Google Scholar]
- 7.Statius Muller MG, Poolman RW, van Hoogstraten MJ, Steller EP. Immediate mobilization gives good results in boxer’s fractures with volar angulation up to 70 degrees: A prospective randomized trial comparing immediate mobilization with cast immobilization. Arch Orthop Trauma Surg. 2003;123(10):534-537. http://search.ebscohost.com/login.aspx?direct=true&db=cmedm&AN=14639483&site=ehost-live [DOI] [PubMed] [Google Scholar]
- 8.van Aaken J, Fusetti C, Luchina S, et al. Fifth metacarpal neck fractures treated with soft wrap/buddy taping compared to reduction and casting: Results of a prospective, multicenter, randomized trial. Arch Orthop Trauma Surg. 2016;136(1):135-142. doi: 10.1007/s00402-015-2361-0. [DOI] [PubMed] [Google Scholar]
- 9.Hartley RL, Lam J, Kinlin C, et al. Surgical and nonsurgical pediatric hand fractures. Plast Reconstr Surg - Glob Open. 2020;8(3):e2703. doi: 10.1097/gox.0000000000002703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lowdon IMR. Fractures of the metacarpal neck of the little finger. Injury. 1986;17(3):189-192. doi: 10.1016/0020-1383(86)90332-3. [DOI] [PubMed] [Google Scholar]
- 11.Mangiafico SS. Mood’s median test for two-sample data. https://rcompanion.org/handbook/F_05.html. Published 2016. Accessed October 18, 2021.
- 12.Liu EH, Alqahtani S, Alsaaran RN, Ho ES, Zuker RM, Borschel GH. A prospective study of pediatric hand fractures and review of the literature. Pediatr Emerg Care. 2014;30(5):299-304. doi: 10.1097/PEC.0000000000000118. [DOI] [PubMed] [Google Scholar]
- 13.Davison PG, Boudreau N, Burrows R, Wilson KL, Bezuhly M. Forearm-Based ulnar gutter versus hand-based thermoplastic splint for pediatric metacarpal neck fractures: A blinded, randomized trial. Plast Reconstr Surg. 2016;137(3):908-916. doi: 10.1097/01.prs.0000479974.45051.78. [DOI] [PubMed] [Google Scholar]
- 14.Rw P, Jc G, Lee J MSM, Ep S, Paa S. Conservative treatment for closed fifth metacarpal (small finger) fractures. Orthop Trauma Dir. 2006;4(5):21-28. doi: 10.1055/s-2006-944323. [DOI] [Google Scholar]
- 15.Lee SJ, Merrison H, Williams KA, Vuillermin CB, Bauer AS. Closed reduction and immobilization of pediatric fifth metacarpal neck fractures. HAND. 2022;17(3):416-421. doi: 10.1177/1558944720942890. [DOI] [PMC free article] [PubMed] [Google Scholar]