Background: Low-intensity pulsed ultrasound (LIPUS) is frequently used to enhance or to accelerate fracture-healing, but its clinical role and effectiveness as a treatment modality remain uncertain. We performed a systematic review and meta-analysis of randomized controlled trials to determine the efficiency of LIPUS on bone-healing and/or fracture union, as well as on functional recovery.
Methods: The databases of PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, CINAHL, Web of Science, and Embase were searched for trials concerning LIPUS stimulation and bone-healing or fracture repair, in any language, published from the inception of the database to January 2, 2015. Eligible studies were randomized controlled trials that enrolled patients with any type of fracture, delayed union, or nonunion and randomly assigned them to LIPUS treatment or a control group. Two reviewers independently agreed on eligibility, assessed methodological quality, and extracted outcome data. All relevant outcomes were pooled, and a meta-analysis was performed.
Results: Twenty-four unique randomized trials were selected for analysis after the search of all databases and the inclusion of one trial by the senior author. Time to radiographic fracture union was the most common primary outcome measure evaluated. After pooling the data concerning time to radiographic healing in the combined patient population (n = 429), LIPUS treatment resulted in a mean reduction in healing time of 39.8 days (95% confidence interval, 17.7 to 62.0 days; I2 = 94%). The most reduction in time to radiographic union by LIPUS treatment was seen in fractures with a long natural healing tendency. Three trials evaluating the time to return to work or active duty, as a surrogate for functional recovery, were unable to demonstrate a beneficial effect of LIPUS (n = 179). Evidence from two high-quality trials implied that LIPUS enhances fracture-healing through increased bone formation in cases of delayed and/or impaired bone-healing. The prevention of delayed union or nonunion by LIPUS treatment could not be demonstrated.
Conclusions: LIPUS treatment effectively reduces the time to radiographic fracture union, but this does not directly result in a beneficial effect of accelerated functional recovery or the prevention of delayed union or nonunion. The increase in bone formation as a result of LIPUS treatment may provide a valuable tool in fracture repair, but it does not always lead to healing. Future studies should focus on reporting of a combination of subjective signs of clinical healing, functional recovery, and radiographic union to determine the effectiveness of LIPUS treatment in clinical fracture-healing.
Level of Evidence: Therapeutic Level II. See Instructions for Authors for a complete description of levels of evidence.
Fracture-healing is a unique and typically successful process; nevertheless, 5% to 10% of fractures show a delay in healing after initial operative or nonoperative treatment1. Impairment of fracture-healing leads to a delay in union or may even result in nonunion. Impaired fracture-healing has an important socioeconomic impact2.
Enhancement and acceleration of fracture-healing to prevent a delay in union or to treat impaired fracture-healing form a multibillion-dollar industry3. Low-intensity pulsed ultrasound (LIPUS) forms the basis for one of the currently available noninvasive, bone-healing enhancement devices. LIPUS affects fracture-healing through high-frequency acoustic pressure waves (30 milliwatts [mW]/cm2) that are delivered transcutaneously to the fracture site4. A randomized controlled trial by Heckman et al. provided important evidence of the effectiveness of LIPUS in clinical, fresh fracture-healing, resulting in the approval of LIPUS by the U.S. Food and Drug Administration (FDA) in 19945.
Since then, numerous studies have evaluated the capacities of LIPUS to stimulate bone-healing and fracture repair6-8. The National Institute for Health and Care Excellence (NICE) has advised health-care professionals in the United Kingdom that LIPUS represents a safe and effective treatment option to reduce the healing time in patients with fractures9. However, the outcomes of a previously performed meta-analysis did not demonstrate a beneficial clinical effect on fracture repair. Only moderate-quality to very low-quality evidence for the acceleration of fracture repair was seen3,10. These findings suggest that future randomized controlled trials should focus on important patient-related outcomes, that is, return to function, to determine the true clinical effectiveness of LIPUS in clinical fracture repair.
LIPUS is frequently used to enhance or to accelerate the fracture-healing process, but its clinical effectiveness as a treatment modality still remains uncertain. To provide better guidance for the clinical use and indication for LIPUS, a comprehensively updated systematic review and meta-analysis of randomized controlled trials was performed. To determine the effectiveness of LIPUS treatment on bone-healing, our primary aim was to evaluate the acceleration of radiographic fracture union, clinical healing, and functional recovery. Secondary outcome measures were the prevention of delayed union and nonunion and the increase of bone formation associated with LIPUS treatment.
Materials and Methods
The literature search that was performed for this review was limited to published original randomized or quasi-randomized controlled trials concerning LIPUS treatment of adult patients with all types of fractures, delayed unions or nonunions, and osteotomies. In quasi-randomized trials, the allocation of patients to a treatment is not strictly random but is based on date of birth, hospital record number, or alternation in order of inclusion. Expert opinions, systematic reviews, and narrative review articles were excluded. There was no language restriction; all relevant articles were translated.
Included articles were limited to those concerning adults (skeletally mature patients) with any type of fracture, delayed union or nonunion, or osteotomy who were subjected to LIPUS treatment (1 to 50 mW/cm2) or allocated to the control group. Articles were excluded if patients had metabolic and/or pathological bone disease. Delayed union was defined as no union for three months, and nonunion was defined as no union for a period of nine months or no progression of healing at six months following the fracture.
Comparisons and Outcome Measures Evaluated
Articles were included if LIPUS treatment was compared with either a sham treatment or untreated controls for patients with fresh fractures and/or osteotomies, those undergoing distraction osteogenesis, and those with delayed unions or nonunions. Subgroup analysis was performed according to five clinical categories: (1) nonoperatively treated fresh fractures and/or osteotomies, (2) operatively treated fresh fractures and/or osteotomies, (3) distraction osteogenesis, (4) nonoperatively treated delayed unions and/or nonunions, and (5) operatively treated delayed unions and/or nonunions. Data were sought for radiographic union (defined as the bridging of at least three cortices2), clinical healing (defined as the absence of pain on axial loading or manual palpation of the fracture site2,6), functional recovery (defined as the return to the pre-injury level of activity, that is, return to active duty or work and sports), bone formation (an increase in bone mineral density or bone volume being used as a surrogate for the acceleration of bone formation and/or healing), prevention of delayed union and/or nonunion, and LIPUS-related complications. If possible, the values of subsequent primary and secondary outcome measures were pooled, and a meta-analysis was performed by using the inverse variance method to combine the natural logarithms of the ratio of the mean value for the LIPUS-treated group to the mean value for the sham-treated or placebo control group. In trials in which the investigator and an independent radiologist assessed radiographic healing, only the data from the radiologist were used for further analysis.
Search Strategy, Trial Selection, and Data Abstraction
The following databases were searched from inception to January 2, 2015, to identify potentially relevant studies: PubMed/MEDLINE, Cochrane Central Register of Controlled Trials, CINAHL, Web of Science, and Embase. The following MEDLINE/PubMed search phrase was used: ((randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR clinical trials as topic[mesh:noexp] OR randomly[tiab] OR trial[ti]) NOT (animals[mh] NOT (animals[mh] AND humans[mh]))) AND ((((((((fracture-healing)) OR (osseous callus)) OR (bone remod*)) OR (bone fractures)) OR (fracture*)) OR (bone stress)) AND (ultrasonic therapy OR ultrasonography OR low-intensity ultrasound OR low intensity ultrasound)). The search was performed independently by two reviewers (S.R. and M.P.J.v.d.B.). The bibliographies of retrieved publications and other relevant articles were manually checked to identify all studies meeting the inclusion criteria that were missed by the electronic search. The International Standard Randomised Controlled Trial Number (ISRCTN), ClinicalTrials.gov, and World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) trial registers were assessed to identify currently ongoing trials.
Trial selection was performed by reviewing titles and abstracts to identify relevant articles for our meta-analysis. The full article was retrieved when the title, keywords, and/or abstract revealed insufficient information to determine whether inclusion would be appropriate. All identified studies were independently assessed for inclusion, according to the above-mentioned criteria, by two reviewers11. Disagreement was resolved by discussion, with arbitration by a third reviewer when differences remained.
Articles were not blinded for author, affiliation, or source. In cases in which the eligibility of a study was unclear, the authors of the study were contacted to provide additional information. Data for meta-analyses were extracted from the included studies by one reviewer with use of a pre-piloted data-extraction tool. Extraction was verified by a second reviewer. Disagreements were resolved by consensus or, if necessary, by third-party adjudication.
Explicit criteria were used when making judgments. Determination of the quality of evidence in this systematic review required an assessment of the validity of the results presented in the individual studies. Thus, the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach was used to assess the quality of the collected evidence for each separate outcome measure12.
Data were entered in a database (Review Manager 5.1; Cochrane), and analyses were performed with use of SPSS software (IBM). The results of comparable studies were pooled with use of random-effects or fixed-effects models when appropriate13. In the presence of heterogeneity, a random-effects meta-analysis weights the studies relatively more equally than a fixed-effect meta-analysis. Individual and pooled statistics were reported as relative risks with 95% confidence intervals (95% CIs) for dichotomous outcomes11. When different scales were used, standardized mean differences and 95% CIs were used to assess continuous outcomes measurements. Sensitivity analyses were conducted (when possible) according to the above-mentioned clinical categories and according to the quality of randomization of the included studies. Heterogeneity was examined with use of the chi-square test and the I2 statistic, which indicates the percentage of variability among studies that is due to true differences between studies (heterogeneity) rather than sampling error (chance). The present systematic review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines14. An I2 value of >50% was considered to represent substantial heterogeneity13.
The combined search of all databases resulted in the identification of a total of 816 eligible articles (Fig. 1). In total, sixty-seven articles were retrieved in full-text format, including ten meta-analyses and nine review articles. One additional article was identified through a review of a previously published meta-analysis3. One trial that was included in a previously published meta-analysis on LIPUS was not included in the present review because the ultrasound signal that had been used (1 W/cm2) was not considered to be low intensity10,15. Twenty-nine potentially relevant articles were identified. One research group described their results in four articles16-19, representing only two separate trials with similar study protocol and study design. The early outcomes on lateral malleolar fractures in the first trial were assessed in one article16, and the late outcomes for the same study population were assessed in a second article17. The outcomes of the second trial were initially published in the authors’ own language in a non-PubMed-indexed local trauma journal18. Emami et al. reported on the clinical outcomes of intramedullary nailing of tibial fractures20 as well as on the effect of serum on bone formation and resorption markers21. Cook et al.22 performed a subanalysis of smoking behavior as studied in two previously published trials2,6. The recently published trial by Busse et al. was a pilot study that was performed only to assess the feasibility of conducting a larger trial to reevaluate the use of LIPUS for the treatment of tibial fractures (Trial to Re-evaluate Ultrasound in the Treatment of Tibial Fractures [TRUST]); however, the recorded outcomes related to LIPUS treatment were not published23. We included one additional trial as well as the clinical outcomes of a previously published trial; both of these trials formed thesis chapters by two of the authors of the present study (S.R. and P.A.N.)24,25. The additional trial evaluated the effect of LIPUS on internally fixed closing-wedge metaphyseal osteotomies around the knee and unfixed osteotomies of the fibula that were performed as part of a high tibial osteotomy24. The clinical outcomes of the previously published trial form an addition to the histological findings of delayed unions of unfixed fibular osteotomies treated with LIPUS26,27. In total, twenty-four unique randomized controlled trials were included in the meta-analysis. For comparison reasons, the procedures in the trial by Nolte et al. were divided into two subgroups; the first subgroup consisted of unfixed fibular osteotomies (referred to in the tables and figures as Nolte 2002a), and the second subgroup consisted of internally fixed osteotomies (referred to in the tables and figures as Nolte 2002b) (Table I)24.
Consultation of trial registers did not identify any ongoing controlled trials evaluating the effect of LIPUS on bone-healing or fracture-healing. Three registered trials had ended and were awaiting publication of outcomes: (1) Ultrasound as Adjunct Therapy for Increasing Fusion Success After Lumbar Surgery (EXO-SPINE, NCT00744861), (2) Effects of Low-Intensity Pulsed Ultrasound (LIPUS) Treatment in Post-Operative Scaphoid Delayed Unions or Nonunions (JPRN-UMIN000017303), and (3) Pulsed Ultrasound to Speed Up Healing After Intramedullary Nailing of Tibia Fractures (ISRCTN90844675). These three trials that were awaiting publication as well as the unpublished outcomes of LIPUS treatment in the TRUST study could therefore constitute a risk of publication bias in the current meta-analysis.
The risk-of-bias summary of the included trials is shown in Figure 2. In eight trials, the patients were randomly allocated to the treatment arm, but no sham treatment was given in the control group28-35. In all other trials involving the use of sham treatment, the placebo devices were identical to the active LIPUS devices, except in the trial by Leung et al.36. Patients were analyzed according to the treatment to which they had been randomly allocated, and no crossover was reported. Eight trials included patients who were lost to follow-up; the lost patients were excluded from outcome calculations (i.e., excluded from both the numerator and denominator).
All trials involved the use of LIPUS and provided a peak pressure of 30 mW/cm2. A daily single twenty-minute treatment was administered on an outpatient basis, except in one trial that involved the use of a twice-daily regime29 and one trial that involved the use of only fifteen minutes of treatment31. The total duration of treatment varied among the trials and was determined on the basis of the presence of a radiographically healed fracture, the ability to remove the external fixator, or a preset time frame, which ranged from four weeks to five months (Table I). No complications or adverse events were attributed to the LIPUS treatment, and most studies demonstrated good to excellent compliance with LIPUS treatment.
Bone Union (Radiographic Outcomes)
Time to radiographic fracture union was the primary outcome measure in ten trials (Table I). Outcomes were shown as a proportional prevalence of union at several time points (ranging from one to six weeks), rather than as a time-to-event analysis. The trial by Strauss et al. was published as a structured abstract only29. Contacting the author did not result in sufficient information to allow pooling of these data. Pooling of data on the time to radiographic union showed that LIPUS reduced (p < 0.001) the mean time to radiographic union by 39.8 days (95% CI, 17.7 to 62.0 days; I2 = 94%; heterogeneity, p < 0.00001) (Fig. 3). In this heterogeneous study population, the greatest effect of LIPUS treatment was seen in fractures with a prolonged natural healing tendency, that is, unfixed fibular osteotomies24 and complex fractures of the tibia5,36. The exclusion of lower-quality trials30,32,36 resulted in an additional decrease in time to radiographic healing of three days (mean, 43.0 days [95% CI, 13.3 to 73.0 days]; I2 = 95%; heterogeneity, p < 0.00001). Subgroup analysis showed that patients with operatively treated fresh fractures and/or osteotomies did not significantly benefit (p = 0.07) from LIPUS treatment, in contrast to those with fresh fractures and/or impaired fracture-healing (Fig. 3). The GRADE description for the effect of LIPUS on radiographic fracture union provided low-quality evidence on the acceleration of fracture union by LIPUS treatment (Table II). In two trials (n = 52), the dichotomous outcome measure of “healed” or “failed” was used, and no difference was reported between the study groups on the basis of radiographs or multidetector computed tomography (CT)16,17,19.
Three trials (n = 76) evaluated the effect of LIPUS treatment on time to radiographic union after tibial distraction osteogenesis and bone transportation28,34,35. External fixation was removed after radiographic consolidation of the distraction gap had occurred. The outcomes were expressed as the time to union (in days) divided by the length of the distraction gap, resulting in a distraction consolidation index. Pooling of the data on the distraction consolidation index revealed a mean decrease of 16.5 days/cm (95% CI, 10.6 to 22.4 days/cm; p < 0.001) as result of LIPUS treatment (I2 = 0%; heterogeneity, p = 0.68) (Fig. 4). The investigators reported random allocation of patients to LIPUS treatment or placebo treatment, but they did not clarify how randomization was achieved, nor did they explicitly report the blinding of the outcome assessors; therefore, low-quality evidence indicated that LIPUS accelerates callus maturation and/or bone formation in distraction osteogenesis and bone transportation of the tibia.
In most trials, a form of clinical healing assessment was performed. Pooling of data from trials that evaluated time to clinical healing as a separate outcome (Fig. 5) showed a mean reduction of 14.2 days (95% CI, 1.9 to 26.5 days; I2 = 96%; heterogeneity, p < 0.00001) in time to healing, providing low-quality evidence for the acceleration of clinical fracture-healing by LIPUS (Table II). The study by Gan et al. was excluded from pooling because the authors were not able to provide us with the standard deviations, and the provided data did not allow for recalculation of the standard deviations37. Additional analysis with the exclusion of low-quality trials31,36 failed to demonstrate an enhancing effect of LIPUS on time to clinical healing (mean, 9.2 days [95% CI, –5.5 to 23.8 days]; I2 = 97%; heterogeneity, p < 0.000001).
Only five trials evaluated patient-important outcomes, and data on functional recovery did not demonstrate a beneficial effect of LIPUS19,37-40. Pooling of data from three trials in which time to return to work or active duty was used as a surrogate for functional recovery demonstrated similar findings for LIPUS-treated patients and sham-treated controls (Fig. 6). Therefore, evidence of moderate quality showed that LIPUS treatment had no beneficial effect in terms of the acceleration of functional recovery (Table II).
Several trials focused on LIPUS-increased bone mineral density or bone volume as a surrogate for the acceleration of bone formation and/or healing by LIPUS (Table I). In patients with complex tibial fractures, LIPUS significantly increased bone mineral density at six, fifteen, eighteen, and twenty-one weeks of treatment (p < 0.05), accompanied by increased plasma bone-specific alkaline phosphatase activity in the six-week to twenty-seven-week follow-up period36. However, the increase in serum bone markers by LIPUS was not confirmed in a study evaluating reamed intramedullary nailing of the tibia21. Bone mineral density measurement did not demonstrate a difference between LIPUS-treated patients and untreated controls with anatomically reduced and internally fixed lateral malleolar fractures; interestingly, it also demonstrated no difference between bone mineral density measurements at the time of the surgical procedure and those at the time of the twelve-week postoperative evaluation16,19. Liu et al. found an increase in the gray-level ratio on radiographs as a surrogate for increased bone mineral density as a result of LIPUS treatment, but the described technique had serious limitations31.
LIPUS treatment for sixteen weeks resulted in increased bone formation in patients with delayed unions of the tibial shaft as demonstrated by a significant increase (p = 0.002) in bone mineral density of 34% (95% CI, 14% to 57%)41. Biopsy specimens from the sites of delayed unions of the fibula indicated increased bone formation after two to three months of LIPUS treatment25. LIPUS significantly increased bone volume by 33% (p = 0.023) at the fracture ends. Both of these trials provided moderate-quality evidence that LIPUS enhances bone-healing through increased bone formation in patients with delayed and/or impaired fracture-healing25,41.
Furthermore, contradictory outcomes were observed in trials assessing distraction osteogenesis. LIPUS was associated with the acceleration of callus maturation over a four-week consolidation period in patients with opening-wedge high tibial osteotomies33, whereas in studies involving LIPUS-treated mandibles and sham-treated controls, histological and microradiographic analysis of biopsy specimens showed similar types and amounts of bone formation toward the center of the distraction gap42,43.
Prevention of Delayed Union, Nonunion, and Malunion
The occurrence of a delay in union or nonunion (twelve LIPUS-treated fractures and twenty fractures in the control group) was only observed in seven studies20,24,28,29,34,36,38. Pooling did not show a risk reduction for impaired healing by LIPUS (relative risk, 0.70 [95% CI, 0.31 to 1.58]; I2 = 14%; heterogeneity, p = 0.33). A decreased loss of reduction or correction as a result of LIPUS treatment was reported in two trials. Fresh radial fractures that were treated with active ultrasound had a mean total reduction in loss (and standard deviation) of only 20% ± 6% compared with the controls, which had a mean total reduction in loss of 43% ± 8% (p < 0.01)6. Furthermore, a decreased loss of correction (p = 0.046) was reported for the first distal metatarsal articular angle in patients who had undergone chevron osteotomies and were treated with LIPUS, from 7.4° (range, 0° to 18°) at the time of operation to 7.9° (range, 0° to 22°) at the time of follow-up, compared with a decrease in the controls from 5.2° (range, 0° to 14°) at the time of operation to 10.6° (range, 0° to 18°) at the time of follow-up40.
This systematic review and meta-analysis is an extensive update of previously published meta-analyses3,10, focusing on radiographic union and patient-important outcomes to determine the effectiveness of LIPUS treatment (30 mW/cm2) in clinical practice. Radiographic evaluation of time to union of fresh fractures and delayed unions and nonunions was most frequently reported (ten trials, n = 429). Pooling showed a mean reduction of forty days as a result of LIPUS treatment and provided low-quality evidence that LIPUS accelerates fracture union. The most reduction in time to radiographic union by LIPUS was observed in fractures with a long natural healing tendency (i.e., unfixed fibular osteotomies24, complex fractures of the tibia)2,36. In patients with operatively treated fresh fractures (i.e., patients managed with reamed intramedullary nailing20 and those with closing-wedge high tibial osteotomies24), LIPUS did not accelerate fracture union. In most of the operatively treated cases, the natural healing tendency was already good or could even be stimulated by reaming or compression of the fracture. Although fracture-healing was accelerated by LIPUS, a decline in the number of delayed unions or nonunions could not be demonstrated (n = 890, with 2.7% in the LIPUS treatment group compared with 4.5% in the control group). These findings may imply that LIPUS is capable of accelerating, but not initiating, fracture-healing.
Clinical healing was evaluated by pooling six trials and functional recovery was evaluated by pooling three trials. LIPUS treatment provided a significant reduction in the time to clinical healing, but it did not lead to a reduction in the time to functional recovery. In addition, pooling only high-quality trials (n = 4) could not demonstrate a reduction in the time to clinical fracture-healing as a result of LIPUS treatment. The authors of a previous systematic review and meta-analysis of randomized trials stated the need for large trials of high methodological quality, focusing on patient-important outcomes such as quality of life and return to function, to determine the role of LIPUS in fracture treatment3. Eight trials have been published since that time, and only two trials evaluated patient-important outcomes such as time to return to work or sports. In general, the clinical goal for patients with healing fractures is the early return to the pre-injury functional level. Fracture union is the key factor in the restoration of function, but fracture union does not guarantee a full restoration to the pre-injury functional level. Fracture characteristics, the initial treatment modality, and comorbidities also may play a substantial role in time to functional recovery or return to the pre-injury functional level44. In determining the effectiveness of LIPUS on fracture-healing, one may argue what it is best to investigate: the process that LIPUS primarily affects (bone formation) or the process that LIPUS secondarily influences (functional recovery).
Contradictory results were observed in trials evaluating the effect of LIPUS on bone mineral density. The methodological quality and trial design did not allow for overall pooling of these outcomes. However, two high-quality trials suggested that LIPUS enhanced fracture-healing through increased bone formation in patients with delayed and/or impaired fracture-healing25,41. Increased bone formation and acceleration of callus maturation may result in advanced fracture-bridging and increased fracture stability, thereby allowing for earlier functional usage.
The present systematic review and meta-analysis had limitations; specifically, the overall quality varied among the studies that were included, and some trials had serious methodological limitations. The primary outcome measures among the trials were diverse, and the difference in fracture location and type resulted in substantial heterogeneity and, therefore, predominantly low-quality evidence. Some form of publication bias was present because of the inability to incorporate the outcomes of registered trials awaiting publication and the potential for unreported trials with negative results. None of the trials that were pooled for functional recovery could also be included in the analysis of time to radiographic healing. Ideally, any future high-quality randomized controlled trial should evaluate subjective signs of clinical healing, functional recovery, and radiographic union to determine the effectiveness of LIPUS and its clinical role in bone-healing. In conclusion, low-quality evidence showed that LIPUS treatment accelerated time to clinical and radiographic union in fresh fractures and delayed unions and nonunions, but acceleration of fracture union did not lead to accelerated functional recovery. As ≥90% of all fractures already heal spontaneously within three months after initial surgical or nonsurgical treatment2, fractures that are prone to a long time to healing and/or impaired fracture-healing could particularly benefit from the bone-healing capacity of LIPUS. Future research should therefore focus on fractures with a prolonged natural healing tendency by evaluating subjective signs of clinical healing and functional recovery, as well as radiographic union, to determine the clinical role of LIPUS treatment in fracture-healing.
The authors thank M.H. Vriends (Medical Library, Academic Medical Centre, Amsterdam, the Netherlands) for her assistance with the literature search.
Source of Funding: No external funding was received for this study.
Investigation performed at the Department of Orthopaedic Surgery, Spaarne Hospital, Hoofddorp, the Netherlands
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated