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: Aviram M. Giladi, MD, MS, c/o Anne Mattson, 3333 North Calvert St, The Curtis National Hand Center, MedStar Union Memorial Hospital, Baltimore, MD.
Whether low bone mineral density affects loss of reduction for distal radius fractures (DRFs) managed without surgery is unknown. Our purpose was to understand how bone mineral density, based on second metacarpal cortical percentage (2MCP) measurement, affects DRF healing after nonsurgical treatment.
Methods
We retrospectively reviewed 304 patients from 2 health systems with DRFs treated without surgery. The AO classification, 2MCP (<50% indicating osteoporosis), and fracture stability based on Lafontaine criteria were determined from prereduction radiographs. Radial inclination, radial height, volar tilt, ulnar variance, and intra-articular stepoff were measured on initial and 6-week final follow-up radiographs and compared. Bivariate analysis was used to evaluate the association between Lafontaine criteria or 2MCP and changes in radiographic parameters. Radiographic parameters with significant associations in bivariate analysis were evaluated in multivariable models adjusted for age, sex, initial radiographic parameters, reduction status, and AO fracture type.
Results
Across all patients, after 6 weeks of nonsurgical treatment, ulnar variance (shortening of the radius) increased by an average of 1.4 mm. Bivariate analysis showed that lower 2MCP and unstable fractures per Lafontaine criteria were each significantly associated with an increase in ulnar variance (P < .05). In adjusted multivariable models, having both 2MCP less than 50% and an unstable fracture together was associated with an additional 1.2-mm increase in ulnar variance (P < .05).
Conclusions
A 2MCP in the osteoporosis range and unstable fractures by Lafontaine criteria were each associated with a significant increase in ulnar variance after nonsurgical treatment for DRFs. Patients with unstable fractures and 2MCP less than 50% are likely to have an additional increase of greater than 1 mm in ulnar variance at the end of nonsurgical fracture treatment than patients with similar injuries, but without these features. Using initial radiographs to identify patients with low bone mineral density that may be at risk for more substantial loss of reduction can assist with decision making for managing DRFs.
The Journal of Hand Surgery will contain at least 2 clinically relevant articles selected by the editor to be offered for CME in each issue. For CME credit, the participant must read the articles in print or online and correctly answer all related questions through an online examination. The questions on the test are designed to make the reader think and will occasionally require the reader to go back and scrutinize the article for details.
The JHS CME Activity fee of $15.00 includes the exam questions/answers only and does not include access to the JHS articles referenced.
Statement of Need: This CME activity was developed by the JHS editors as a convenient education tool to help increase or affirm reader’s knowledge. The overall goal of the activity is for participants to evaluate the appropriateness of clinical data and apply it to their practice and the provision of patient care.
Accreditation: The American Society for Surgery of the Hand (ASSH) is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians.
AMA PRA Credit Designation: The ASSH designates this Journal-Based CME activity for a maximum of 1.00 AMA PRA Category 1 Credits™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.
ASSH Disclaimer: The material presented in this CME activity is made available by the ASSH for educational purposes only. This material is not intended to represent the only methods or the best procedures appropriate for the medical situation(s) discussed, but rather it is intended to present an approach, view, statement, or opinion of the authors that may be helpful, or of interest, to other practitioners. Examinees agree to participate in this medical education activity, sponsored by the ASSH, with full knowledge and awareness that they waive any claim they may have against the ASSH for reliance on any information presented. The approval of the US Food and Drug Administration (FDA) is required for procedures and drugs that are considered experimental. Instrumentation systems discussed or reviewed during this educational activity may not yet have received FDA approval.
ASSH Disclosure Policy: As a provider accredited by the ACCME, the ASSH must ensure balance, independence, objectivity, and scientific rigor in all its activities.
Disclosures for this Article
Editors
Ryan Calfee, MD, MSc, has no relevant conflicts of interest to disclose.
Authors
All authors of this journal-based CME activity have no relevant conflicts of interest to disclose. In the printed or PDF version of this article, author affiliations can be found at the bottom of the first page.
Planners
Ryan Calfee, MD, MSc, has no relevant conflicts of interest to disclose. The editorial and education staff involved with this journal-based CME activity has no relevant conflicts of interest to disclose.
Learning Objectives
Upon completion of this CME activity, the reader will understand:
•
The impact of osteoporosis on patients with distal radius fractures.
•
Evaluation of osteoporosis in association with distal radius fractures.
•
Understand the prevalence of osteoporosis in the setting of a distal radius fracture.
Deadline: Each examination purchased in 2021 must be completed by January 31, 2022, to be eligible for CME. A certificate will be issued upon completion of the activity. Estimated time to complete each JHS CME activity is up to one hour.
A prospective randomized trial comparing nonoperative treatment with volar locking plate fixation for displaced and unstable distal radial fractures in patients sixty-five years of age and older.
Nonoperative treatment compared with volar locking plate fixation for dorsally displaced distal radial fractures in the elderly: a randomized controlled trial.
Non-operative treatment of displaced distal radius fractures leads to acceptable functional outcomes, however at the expense of 40% subsequent surgeries.
found that patients with at least 3 of the following 5 criteria were at risk for loss of reduction: initial dorsal tilt greater than 20°, dorsal comminution, intra-articular extension of fractures, concomitant ulna fracture, and age greater than 60 years. However, Nesbitt et al
evaluated the Lafontaine criteria and found that age was the only significant factor in predicting secondary displacement. Other authors have more recently proposed specific formulas to anticipate fracture instability.
None of these methods have been validated and predicting loss of reduction during nonsurgical management remains difficult.
Osteoporosis is another potentially important factor for DRF stability. Investigators studying fractures treated with open reduction and internal fixation have found no differences in radiographic outcomes between patients with and without osteoporosis.
However, the effect of osteoporosis on radiographic outcomes in patients undergoing nonsurgical treatment is not well studied. Given the frequent nature of DRF and that osteoporosis is increasingly prevalent,
identified a method of reliably evaluating global bone mineral density by measuring the second metacarpal cortical percentage (2MCP) on hand or wrist radiographs. The 2MCP positively correlated with total bone mineral density from Dual Energy X-ray Absorptiometry (DEXA) scans, and a 2MCP threshold of 50% had excellent sensitivity and specificity for differentiating patients with osteoporosis. The primary goal of this study was to evaluate the effect of low bone mineral density on loss of reduction for DRFs treated with closed reduction and immobilization. Using 2MCP as an indicator of bone mineral density, we hypothesized that 2MCP at the time of fracture would predict change in radiographic measurements. The secondary goal of the study was to further evaluate whether Lafontaine criteria would predict change in radiographic measurements.
Materials and Methods
Study sample
After obtaining institutional review board approval, medical records were reviewed for patients treated at 2 health care systems from January 2014 to May 2019 with a diagnosis of DRF. Inclusion criteria were individuals with (1) a DRF treated with immobilization with or without a closed reduction and (2) digital wrist radiographs including a complete posteroanterior and lateral view at the time of injury and at a minimum of 6 weeks after injury. Patients who were younger than 18 years old, had multiple fractures (excluding distal ulna fractures), or underwent surgical treatment with placement of any hardware prior to obtaining 6-week follow-up radiographs were excluded from the study. The preliminary search revealed 2,508 patients with a diagnosis of DRF treated at our 2 hospital systems. Of these, 741 underwent surgery within 6 weeks of injury and 1,463 had additional fractures or did not have adequate postreduction or 6-week radiographs, including a substantial majority with nondisplaced or minimally displaced fractures. Ultimately, 304 patients were included in the study (Fig. 1).
Figure 1Study flow diagram. ICD-10, International Classification of Diseases, 10th Revision.
Five primary measurements were made by 2 experienced providers—a fellowship-trained hand surgeon (J.G.) and a research fellow with a medical degree (I.Y.) who was trained by fellowship-trained hand surgeons (A.G. and K.M.)—on each radiograph using previously described techniques
with digital software (Centricity PACS): volar tilt, radial inclination, radial height, ulnar variance, and intra-articular displacement. Any discrepancy of more than 0.1 mm was rereviewed with a fellowship-trained hand surgeon (A.G.). In addition, measurement fidelity checks were randomly performed on over 10% of the patients by this fellowship-trained hand surgeon (A.G.) to confirm consistency and accuracy. The measurements were made on initial radiographs for all patients, postreduction radiographs for those who underwent closed reduction, and again on 6-week radiographs for all patients. On initial radiographs, the presence of dorsal comminution, intra-articular fractures, or a concomitant ulnar styloid fracture was also recorded. If a closed reduction was performed but no postreduction radiographs were available for review, the patient was excluded from the study. Change from initial to final radiographs was calculated. Of note, for DRF, an increase in ulnar variance occurs due to radial shortening in comparison with the stable distal ulna. However, we refer to this parameter as an increase in ulnar variance to remain consistent with prior literature and common usage, and to avoid confusion with our descriptions of radial height and radial inclination.
Primary independent variable
Fracture instability was our primary independent variable. A fracture was considered unstable according to Lafontaine et al
if 3 or more of the following 5 criteria were met: initial dorsal tilt greater than 20°, dorsal comminution, intra-articular fracture, ulnar styloid fracture, and age greater than 60 years. These radiographic parameters were measured on injury radiographs. A dichotomous variable was created accordingly as stable or unstable. We chose to assess Lafontaine criteria as a single entity rather than evaluate the 5 components separately because it has mostly been studied and clinically utilized as a dichotomous variable.
The covariates included 2MCP, sex, race, reduction status (whether patient underwent a closed reduction or not), initial radiographic measures (from injury radiographs if no reduction was performed, or from immediate postreduction radiographs if reduction was performed), fracture type, and Charlson Comorbidity Index (CCI) score, which we calculated after reviewing each patient’s medical history. The 2MCP was calculated on a true posteroanterior view of the hand from either a hand or a wrist digital radiograph as described by Schreiber et al.
The 2MCP measurement was made on the initial injury radiographs prior to formal measurements of the 6-week radiographs. Utilizing 2MCP as an indicator of baseline bone mineral density at the time of fracture, it was categorized into a dichotomous variable (>50% vs <50%). Age was used as a continuous measure. Race was categorized as white, African American, Hispanic, and other. Each fracture was classified into 9 categories according to the AO/OTA (Orthopaedic Trauma Association) classification system. Reduction status was a dichotomous variable depending on whether or not closed reduction was performed. Initial radiographic parameters were used as a continuous measure. The CCI score was calculated for each patient and evaluated as a continuous variable.
Statistical methods
Means ± SDs and frequencies were calculated and summarized (Table 1). We compared differences between initial and final follow-up radiographic parameters using 2-sided paired sample t tests. Separate bivariate linear regression models were used to evaluate the association of Lafontaine criteria for instability (“instability”) and each of the final follow-up radiographic parameters. For ulnar variance, which was the only radiographic parameter that had significant bivariate association with instability, we conducted multivariable linear regression analysis adjusting for 2MCP, sex, reduction status, initial radiographic measurement, CCI score, and AO/OTA fracture type. Because 2MCP level may be an important effect modifier of the relationship between radiographic outcomes and instability, we added an interaction term in our model to assess whether the associations varied by 2MCP level (low or high). This interaction term allows an evaluation of whether differences in bone mineral density (2MCP) alter the impact that instability has on radiographic outcomes. In addition, variance inflation factors were examined to ensure that the primary covariates of interest showed no collinearity according to the 5 rule
(Appendix A; available on the Journal’s Web site at www.jhandsurg.org). Based on regression diagnostics and clinical relevance, final follow-up ulnar variance was chosen as the main outcome measure. Significance was set at .05.
Table 1Summary of Descriptive Statistics (n = 304)
Patients had an average age of 60 ± 18 years old (range, 18–96 years). The majority were female (238 patients; 78.0%). Based on Lafontaine criteria, 158 patients (51.9%) had fractures that were determined to be stable and 146 patients (48.1%) had unstable fractures. One hundred thirty patients (42.7%) had a 2MCP less than 50% and were classified as osteoporotic. Other demographics and care details are in Table 1.
When comparing measurements from initial radiographs with those of final follow-up radiographs, there was a statistically significant change in all radiographic parameters across the entire cohort (Table 2). Bivariate linear regression models found a significant association between unstable fractures and ulnar variance at final follow-up (1.2 mm [95% confidence interval, 0.58–1.85; P < .05]). In addition, 2MCP less than 50%, CCI score, fractures undergoing reduction, initial ulnar variance, as well as AO/OTA type C2 and C3 fractures were all significantly associated with an increase in ulnar variance (Appendix B, Unadjusted Model data, p all < .05; available on the Journal’s Web site at www.jhandsurg.org). None of the other radiographic parameters had significant findings on bivariate analyses; therefore, we only evaluated ulnar variance in subsequent models.
Table 2Changes in Radiographic Outcomes From Postreduction to ≥6 Wk Follow-Up
We then developed a multivariable linear regression model using final ulnar variance as the main outcome measure and Lafontaine instability as our primary independent variable. The results of our final model are shown in Table 3 [Adjusted Model F(14,289) = 23.74; P < .05; R2 = 0.53]. Adjusting for other covariates, the interaction term between instability and 2MCP level was significantly associated with the outcome, showing that, at final follow-up, patients with unstable fractures and less than 50% 2MCP had an additional 1.2-mm increase in ulnar variance (95% confidence interval, 0.31–2.17; P < .05). The CCI score, initial ulnar variance, reduction status, and C3 fracture type also remained significantly associated with an increase in ulnar variance (p all < .05).
Table 3Adjusted b Estimates for Multivariable Linear Regression Model, With Final Ulnar Variance as Outcome Measure
A separate multivariable model was tested to explore the association of unstable fractures and 2MCP with change in ulnar variance from initial to final follow-up radiographs as the outcome variable, rather than evaluating initial variance and final variance as separate variables. In analyzing the interaction between 2MCP and instability, the impact on ulnar variance was essentially the same; however, overall model performance was worse, potentially due to removal of initial ulnar variance as a separate variable (Appendix C; available on the Journal’s Web site at www.jhandsurg.org). In addition, we ran another separate model without including reduction status as a variable, to avoid markedly overcontrolling for instability, and found that the overall results were essentially unchanged and the R2 decreased slightly (not shown).
Discussion
Osteoporosis-related fractures, including DRFs, represent an increasing proportion of health care expenditure in the United States.
Many of these fractures are treated without surgical management; however, substantial loss of reduction and associated persistent symptoms can frequently lead to subsequent surgical treatment and higher overall costs.
Non-operative treatment of displaced distal radius fractures leads to acceptable functional outcomes, however at the expense of 40% subsequent surgeries.
Understanding which patients are at greater risk of progression to malunion may help guide initial treatment. This study sought to use predicted instability by Lafontaine criteria alongside a simple method of evaluating bone mineral density with 2MCP to determine the effect of poor bone health on radiographic outcomes of distal radius fractures treated without surgery. Our final multivariable model identified that, for patients with bone mineral density in the osteoporotic range as well as an unstable fracture, ulnar variance increased more than an additional 1 mm beyond the changes observed for other patients. However, there was no significant association with volar tilt, radial inclination, radial height, or articular stepoff.
Across all patients, we saw a small, statistically significant change in all radiographic parameters with unclear clinical relevance. Although change in radiographic parameters for DRFs treated with surgery is rarely reported, our results were similar to those reported by Fitzpatrick et al
evaluated previous studies and determined clinically meaningful change at 5° decrease in volar tilt, 3° decrease in radial inclination, 2 mm loss of radial height, and 1 mm increase in ulnar variance. The overall mean change in ulnar variance across all patients in our study was 1.4 mm. Our patients with low 2MCP and unstable fractures had an additional 1.2-mm increase in ulnar variance beyond this. Although the exact amount of change in ulnar variance is difficult to predict, especially considering potential variability in arm position for x-rays (both in our study and in the clinical experience with injured patients), surgeons might take into account this additive predicted change in ulnar variance from initial radiographs when counseling these patients about nonsurgical and surgical care. The impact on clinical and functional outcomes, especially considering that low 2MCP did not have a significant impact on the other radiographic parameters, is important for further study.
The results of this study validate Lafontaine criteria of instability in bivariate and multivariable analyses with respect to ulnar variance. Patients with fractures that met at least 3 of 5 criteria were more likely to have a greater increase in ulnar variance. Nesbitt et al
evaluated Lafontaine criteria in 50 patients who underwent closed reduction and orthosis wear and analyzed each criterion alone and in combination with each other. Their study found that age was the only factor significantly associated with loss of reduction. However, LaMartina et al
found in univariate analysis that Lafontaine criteria were significantly associated with a worse final position for ulnar variance, radial height, and radial inclination, but not dorsal tilt. Accordingly, our study provides further evidence that Lafontaine criteria provides valuable information to surgeons and their patients about the likely radiographic progression of distal radius alignment with nonsurgical management. This was strengthened by our finding that the interaction between instability and low 2MCP predicted the greatest increase in ulnar variance. Whether this is because patients with lower 2MCP have worse initial fractures (in ways not captured by instability criteria or AO/OTA classification), more instability or likelihood of collapse during healing, or other contributory factors is unclear and requires further study. However, this finding may provide insight as to why some patients deemed radiographically unstable by Lafontaine criteria do not progress to notable malunion while others do. Our findings indicate that a more robust indicator of fracture stability may be needed, one that considers additional elements beyond those included by Lafontaine et al.
Compared with previous studies that reported no effect of osteoporosis on radiographic outcomes after internal fixation of DRFs, we found that bone mineral density can be associated with loss of reduction for those with unstable fractures treated without surgery.
but also may provide additional information on prognosis of DRFs treated without surgery, especially when findings that predict instability are also present. Clayton et al
studied the association between bone mineral density and severity of distal radius fractures in 137 patients, in which 92 patients underwent nonsurgical management. Their outcomes included early instability, malunion, and late carpal malalignment. They found that patients with lower t scores on DEXA scan had a higher likelihood of early instability, malunion, and late carpal malalignment. However, they only used predetermined thresholds for dorsal angulation and ulnar variance and did their measurements on manual radiographs, as opposed to the change in ulnar variance during the course of treatment that we evaluated on digital radiographs. Although those authors did not measure radial height, radial inclination, or intra-articular stepoff, our results are consistent with their other findings
in that the patients with lower bone mineral density tended to have more radiographic displacement of their unstable DRFs.
Our study has several limitations. First, patient demographics and radiographs were reviewed retrospectively and without strictly controlling arm positioning and specific technique for the imaging. Second, only radiographic outcomes and not clinical outcomes were evaluated. In fact, the American Academy of Orthopaedic Surgeons’ Clinical Practice Guidelines
on DRFs states, “We are unable to recommend for or against operative treatment for patients over age 55 with distal radius fractures” largely owing to inconsistency between radiographic findings and clinical findings especially in older patients. Patients may be able to achieve good clinical outcomes despite poor radiographic outcomes, especially in elderly and low-demand individuals,
found that ulnar variance was 1 of the 2 most important radiographic parameters for obtaining good functional outcomes in patients treated with internal fixation. Third, we only included patients who had documented nonsurgical management across 6 weeks. As such, those patients lost to follow-up were excluded. Many of those excluded had nondisplaced or minimally displaced fractures, and providers did not ask these patients to obtain radiographs or even follow-up at 6 weeks or later. Also, patients with more severely displaced fractures or worsening collapse who may have undergone surgical treatment were excluded. However, knowing that many patients progress through seemingly appropriate nonsurgical treatment and yet heal in a malunited position makes nonsurgical cohorts like ours important to study more thoroughly. Fourth, the changes in radiographic parameters are small and measurement errors, as well as measurement inconsistency owing to technique variations, may play a role in the significance of our findings. Garon and Kleinman
found high intraobserver and interobserver reliability for measuring ulnar variance as we did (intraclass correlation coefficient of 0.92 for intraobserver reliability and 0.87 for interobserver reliability), but what impact positioning had on the radiographs is a notable unknown in our study. We did not control for care location in our analyses, and we only included 2 sites, potentially having an impact on our results or limiting generizability, especially if reduction techniques or postinjury care vary elsewhere. Finally, we did not directly measure bone mineral density, but rather used 2MCP as a validated indicator; how 2MCP measurements are best made and/or used in regular clinical care is the subject of future work.
In our study, we found that having 2MCP in the osteoporosis range and an unstable distal radius fracture by Lafontaine criteria was associated with an additional statistically significant increase in ulnar variance at the time of fracture healing. Patients with unstable fractures and a 2MCP less than 50% are likely to have an additional increase of 1 mm or more in ulnar variance compared with patients with otherwise similar injuries. Whether low bone mineral density led to greater fracture instability not directly measured by Lafontaine et al,
or had an impact on healing of the fracture after it occurred, remains unclear. However, although further clinical and outcomes studies are needed, the relationship between Lafontaine instability and low 2MCP could potentially factor into patient and surgeon decision-making for fractures that are on the border between surgical or nonsurgical management at the time of initial evaluation.
Acknowledgments
Institutional review board approval was obtained before conducting this research study.
The study was funded by a Raymond M. Curtis Research Foundation Grant at The Curtis National Hand Center, MedStar Union Memorial Hospital, Baltimore, MD.
Appendix A
Appendix AStatistics Evaluating for Multicollinearity
Variable
VIF
Tolerance
Lafontaine criteria
2.03
0.493
2MCP50
2.25
0.443
Lafontaine’s criteria—2MCP50
3.31
0.302
Reduction status
1.2
0.835
Initial ulnar variance
1.18
0.844
Gender
1.07
0.933
CCI score
1.32
0.754
AO/OTA 2
10.07
0.099
AO/OTA 3
3.56
0.281
AO/OTA 4
6.75
0.148
AO/OTA 5
2.3
0.434
AO/OTA 6
1.75
0.571
AO/OTA 7
8.53
0.117
AO/OTA 8
6.43
0.155
AO/OTA 9
3.12
0.320
Mean VIF: 3.66
2MCP50, second metacarpal cortical percentage 50% or below (indicating likely osteoporosis);
Appendix CAdjusted B Estimates for Multivariable Linear Regression With Change in Ulnar Variance as the Primary Outcome Measure F(14,289) = 6.05; P < .05; R2 = 0.19
A prospective randomized trial comparing nonoperative treatment with volar locking plate fixation for displaced and unstable distal radial fractures in patients sixty-five years of age and older.
Nonoperative treatment compared with volar locking plate fixation for dorsally displaced distal radial fractures in the elderly: a randomized controlled trial.
Non-operative treatment of displaced distal radius fractures leads to acceptable functional outcomes, however at the expense of 40% subsequent surgeries.
00087-3&id=gr1.jpg){kind=link}