Thermal Debridement in the Management of Low Grade Scapholunate Ligament Tears.

Low grade scapholunate interosseous ligament (SLIL) tears are often managed with mechanical or thermal arthroscopic debridement, although this remains controversial. This study aimed to assess the short-term outcomes of thermal debride- ment of low-grade SLIL tears. Patients with low grade SLIL tears who underwent arthroscopic thermal debridement between 2010 and 2017 were identified and divided into two groups: isolated thermal debridement and concomitant pro- cedures. Patient reported outcomes, wrist range of motion, grip strength, return to work, and baseline activities were evaluated. Twenty-seven patients underwent isolated thermal debridement and 20 underwent concomitant procedures. Pain significantly improved in both groups. Grip strength significantly improved in the concomitant procedure group. There was no significant change in wrist range of motion in either group. Most patients returned to baseline activities. Arthroscopic thermal debridement provides good outcomes in patients with low grade SLIL tears both in isolation and in association with other injuries.

Postoperative Inpatient Conversions Following Ambulatory Orthopedic Surgery.

BACKGROUND: Despite the recent trend toward outpatient orthopedic surgical procedures, there are patients who incur unanticipated conversions to inpatient status either immediately following ambulatory surgery or within days to weeks via presentation to the emergency department. The purpose of this study was to examine characteristics, co-morbidities, and causes of admissions in these populations in order to help identify factors for which strategies can be developed in order to minimize unanticipated admissions and medical costs. METHODS: Using a major academic medical center's bill-ing department's database, 95 outpatients were identified who were immediately converted into inpatient status and another 84 outpatients who were admitted within 30 days of surgery. The reasons for admission, length of procedure, length of admission, ASA score, comorbidities, and insur-ance type were assessed. RESULTS: For the patients who were converted to inpa-tient status postoperatively, pain accounted for 57% of conversions. Hypertension was the most commonly seen comorbidity (44%). In patients admitted within 30 days of ambulatory surgery, infection (25%) was the most common cause of admission. Smoking (46.4%) represented the most common comorbidity in this cohort. CONCLUSIONS: The majority of immediate inpatient con-versions were due to pain, emphasizing the need to optimize perioperative analgesia and provide impactful patient education regarding postoperative pain expectations. For patients admitted within 30 days of surgery, infection represented the majority of readmissions, and smoking was the most common comorbidity. While, some infections may be unavoidable, this stresses the importance of medical and social factor optimization prior to surgery. Addressing these factors leading to unanticipated admissions can have a profound effect on health care expenditures and patient outcomes.

Tourniquet Use for Short Hand Surgery Procedures Done Under Local Anesthesia Without Epinephrine.

PURPOSE: Wide-awake local anesthesia no tourniquet (WALANT) is an increasingly popular surgical technique. However, owing to surgeon preference, patient factors, or hospital guidelines, it may not be feasible to inject patients with solutions containing epinephrine the recommended 25 minutes prior to incision. The purpose of this study was to assess pain and patient experience after short hand surgeries done under local anesthesia using a tourniquet rather than epinephrine for hemostasis. METHODS: Ninety-six consecutive patients undergoing short hand procedures using only local anesthesia and a tourniquet (LA-T) were assessed before and after surgery. A high arm pneumatic tourniquet was used in 73 patients and a forearm pneumatic tourniquet was used in 23. All patients received a local, unbuffered plain lidocaine injection. No patients received sedation. Pain related to local anesthesia, pneumatic tourniquet, and the procedure was assessed using a visual analog scale (VAS). Patient experience was assessed using a study-specific questionnaire based on previous WALANT studies. Tourniquet times were recorded. RESULTS: Mean pain related to anesthetic injection was rated 3.9 out of 10. Mean tourniquet related pain was 2.9 out of 10 for high arm pneumatic tourniquets and 2.3 out of 10 for forearm pneumatic tourniquets. Patients rated their experience with LA-T favorably and 95 of 96 patients (99%) reported that they would choose LA-T again for an equivalent procedure. Mean tourniquet time was 9.6 minutes and only 1 patient had a tourniquet inflated for more than 20 minutes. Tourniquet times less than 10 minutes were associated with less pain than tourniquet times greater than 10 minutes (P < .05); however, both groups reported the tourniquet to be on average less painful than the local anesthetic injection. CONCLUSION: Short wide-awake procedures using a tourniquet are feasible and well accepted. Local anesthetic injection was reported to be more painful than pneumatic tourniquet use. Tourniquets for short wide-awake procedures can be used in settings in which preprocedure epinephrine injections are logistically difficult or based on surgeon preference. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic IV.

Assessment of Pronator Quadratus Repair Integrity Using Dynamic Ultrasonography Following Volar Plate Fixation for Distal Radius Fractures.

Background: Previous work evaluating the pronator quadratus (PQ) muscle following volar plate fixation (VPF) of distal radius fractures (DRF) suggests that PQ repair often fails in the postoperative period. The purpose of this investigation was to assess PQ repair integrity following VPF of DRF using dynamic musculoskeletal ultrasonography. Methods: Twenty adult patients who underwent VPF of DRF with repair of the PQ with a minimum follow-up of 3 months underwent bilateral dynamic wrist ultrasonography. The integrity of the PQ repair, wrist range of motion (ROM) and strength, and functional outcome scores were assessed. Results: Mean patient age at the time of surgery was 59 ± 14 years, and 50% underwent VPF of their dominant wrist. Patients were evaluated at a mean 9 ± 4 months after VPF. All patients had an intact PQ repair. The volar plate was completely covered by the PQ in 55% of patients and was associated with a larger PQ when compared to patients with an incompletely covered volar plate (P = .026). The flexor pollicis longus tendon was in contact with the volar plate in 20% of patients, with those patients demonstrating a trend toward significantly increased wrist flexion (P = .053). No difference in ROM, strength, or outcome scores was noted among wrists with completely or incompletely covered volar plates. Conclusions: The PQ demonstrates substantial durability after repair following VPF. Wrist ROM, strength, and functional outcomes are similar in wrists in which the volar plate is completely or incompletely covered by the repaired PQ.

Bone Graft Options in Upper-Extremity Surgery.

Bone grafting in the upper extremity is an important consideration in patients with injuries or conditions resulting in missing bone stock. A variety of indications can necessitate bone grafting in the upper extremity, including fractures with acute bone loss, nonunions, malunions, bony lesions, and bone loss after osteomyelitis. Selecting the appropriate bone graft option for the specific consideration is important to ensure optimal patient outcomes. Considerations such as donor site morbidity and the amount and characteristics of bone graft needed all weigh in the decision making regarding which type of bone graft to use. This article reviews the options available for bone grafting in the upper extremity.

Scaphoid Nonunions.

Scaphoid nonunions are challenging injuries to manage and the optimal treatment algorithm continues to be debated. Most scaphoid fractures heal when appropriately treated; however, when nonunions occur, they require acute treatment to prevent future complications like scaphoid nonunion advanced collapse. Acute nonunion treatment technique depends on nonunion location, vascular status of the proximal pole, fracture malalignment, and pre-existing evidence of arthrosis. Bone grafting and vascular grafts are common in nonunion management. Chronic nonunions that have progressed to scaphoid nonunion advanced collapse often require a salvage procedure such as four corner fusions, proximal row carpectomy, or wrist fusion. Herein, we review the current literature regarding scaphoid nonunions with regards to their anatomy, natural history, classification, diagnostic imaging, surgical management, and clinical outcomes.

Flexor Tendon Injuries.

Flexor tendon injuries of the hand are uncommon, and they are among the most challenging orthopaedic injuries to manage. Proper management is essential to ensure optimal outcomes. Consistent, successful management of flexor tendon injuries relies on understanding the anatomy, characteristics and repair of tendons in the different zones, potential complications, rehabilitation protocols, recent advances in treatment, and future directions, including tissue engineering and biologic modification of the repair site.

Musculoskeletal Injuries in Yoga.

While yoga has been widely studied for its benefits to many health conditions, little research has been performed on the nature of musculoskeletal injuries occurring during yoga practice. Yoga is considered to be generally safe, however, injury can occur in nearly any part of the body-especially the neck, shoulders, lumbar spine, hamstrings, and knees. As broad interest in yoga grows, so will the number of patients presenting with yoga-related injuries. In this literature review, the prevalence, types of injuries, forms of yoga related with injury, specific poses (asanas) associated with injury, and preventive measures are discussed in order to familiarize practitioners with yoga-related injuries.

Scaphoid Healing Required for Unrestricted Activity: A Biomechanical Cadaver Model.

PURPOSE: To determine if scaphoid fractures with bridging bone of 50% of their width treated with a centrally placed screw will restore biomechanical integrity equivalent to that of the intact scaphoid. METHODS: Twenty-four fresh cadaver scaphoids were used. Six were left intact to serve as the control group. Six were osteotomized 50% of their width and made up the osteotomy without screw group. Six were included in the 50% osteotomy plus compression screw group. The remaining 6 were to be treated with an osteotomy of 25% or 75% with a screw, based upon the results of the 50% osteotomy with screw group. Biomechanical testing was performed using an Instron testing machine, with a load applied to the scaphoid's distal pole. Load to failure and stiffness were measured. RESULTS: Intact scaphoids had an average load to failure of 610.0 N. The average load to failure of the 50% osteotomy group without a screw was 272.0 N and with a screw was 666.3 N. There was no significant difference in load to failure between the 50% osteotomy plus screw and the intact scaphoid. The 75% osteotomy plus screw was found to have a load to failure of 174.0 N, significantly lower than the intact scaphoid. The 50% osteotomy plus screw had a significantly higher stiffness than the intact scaphoid control. CONCLUSIONS: A 50% intact scaphoid with a centrally placed screw showed similar load to failure and significantly higher stiffness than the intact scaphoid when tested in cantilever bending. CLINICAL RELEVANCE: This study demonstrates that patients with scaphoid waist fractures who undergo surgery with a compression screw may be able to return to unrestricted activity with 50% partial healing.

Management of Pisotriquetral Instability.

Pisotriquetral instability is an often-overlooked condition that can lead to ulnar-sided wrist pain and dysfunction. Various case series and biomechanical studies have been published regarding the diagnosis and treatment of this condition. We review current methods for examining, diagnosing, and treating pisotriquetral instability.

The Relationships Between Surface Measurements and Underlying Tendon Autograft Length for Upper Extremity Reconstructive Surgery.

PURPOSE: The availability of tendon grafts is an important consideration for successful upper extremity reconstructive surgery, including flexor or extensor tendon reconstructions, tendon transfers, and ligament reconstructions. Graft selection is based on availability, expendability, ease of harvest, and length. Given variations in patient height and extremity length, existing average values may provide suboptimal insight into actual tendon lengths available. The purpose of this study is, therefore, to pursue a method of estimating available donor tendon lengths based on easily measured anatomical surface landmarks. METHODS: Thirty cadaveric upper and lower extremity limbs were dissected and the length of commonly harvested tendon grafts including the palmaris longus, extensor indicis proprius, extensor digiti minimi, plantaris, and second long toe extensor was measured. Surface forearm length (from finger tip to cubital fossa) and surface fibular length (from lateral malleolus to fibular head) were also measured. Correlations between surface measurements and underlying tendon lengths were analyzed, and linear models were generated that predicted tendon length as a function of surface measurements. RESULTS: Surface measurements were correlated with underlying tendon length (R = 0.46 - 0.66). Linear models could predict tendon lengths based on surface measurements. A ratio of donor tendon length compared with the limb segment measured was established for each tendon and can be applied to estimate donor tendon length. For the upper extremity tendons, the multipliers for the palmaris longus, extensor indicis proprius, and extensor digiti minimi were 0.51, 0.20, and 0.18, respectively. Lower extremity tendon ratios for the plantaris and extensor digitorum longus were 0.69 and 0.60, respectively. CONCLUSIONS: Although length of available donor tendon can be a limiting variable at the time of surgery, surgeons may be better able to estimate underlying tendon lengths using easily obtained superficial measurements. CLINICAL RELEVANCE: Information obtained from these cadaveric measurements may aid in preoperative planning in hand and upper extremity surgery.

Hand Society and Matching Program Web Sites Provide Poor Access to Information Regarding Hand Surgery Fellowship.

The Internet is a common resource for applicants of hand surgery fellowships, however, the quality and accessibility of fellowship online information is unknown. The objectives of this study were to evaluate the accessibility of hand surgery fellowship Web sites and to assess the quality of information provided via program Web sites. Hand fellowship Web site accessibility was evaluated by reviewing the American Society for Surgery of the Hand (ASSH) on November 16, 2014 and the National Resident Matching Program (NRMP) fellowship directories on February 12, 2015, and performing an independent Google search on November 25, 2014. Accessible Web sites were then assessed for quality of the presented information. A total of 81 programs were identified with the ASSH directory featuring direct links to 32% of program Web sites and the NRMP directory directly linking to 0%. A Google search yielded direct links to 86% of program Web sites. The quality of presented information varied greatly among the 72 accessible Web sites. Program description (100%), fellowship application requirements (97%), program contact email address (85%), and research requirements (75%) were the most commonly presented components of fellowship information. Hand fellowship program Web sites can be accessed from the ASSH directory and, to a lesser extent, the NRMP directory. However, a Google search is the most reliable method to access online fellowship information. Of assessable programs, all featured a program description though the quality of the remaining information was variable. Hand surgery fellowship applicants may face some difficulties when attempting to gather program information online. Future efforts should focus on improving the accessibility and content quality on hand surgery fellowship program Web sites.

Anatomic Locking Plate Fixation for Scaphoid Nonunion.

Nonunion can occur relatively frequently after scaphoid fracture and appears to be associated with severity of injury. There have been a number of techniques described for bone grafting with or without screw fixation to facilitate fracture healing. However, even with operative fixation of scaphoid fractures with bone grafting nonunion or malunion rates of 5 to 10 percent are still reported. This is the first report of an anatomic locking plate for scaphoid fracture repair in a 25-year-old right hand dominant healthy male.

Does Brachial Plexus Blockade Result in Improved Pain Scores After Distal Radius Fracture Fixation? A Randomized Trial.

BACKGROUND: Distal radius fractures are very common injuries and surgical treatment for them can be painful. Achieving early pain control may help improve patient satisfaction and improve functional outcomes. Little is known about which anesthesia technique (general anesthesia versus brachial plexus blockade) is most beneficial for pain control after distal radius fixation which could significantly affect patients' postoperative course and experience. QUESTIONS/PURPOSES: We asked: (1) Did patients receiving general anesthesia or brachial plexus blockade have worse pain scores at 2, 12, and 24 hours after surgery? (2) Was there a difference in operative suite time between patients who had general anesthesia or brachial plexus blockade, and was there a difference in recovery room time? (3) Did patients receiving general anesthesia or brachial plexus blockade have higher narcotic use after surgery? (4) Do patients receiving general anesthesia or brachial plexus blockade have higher functional assessment scores after distal radius fracture repair at 6 weeks and 12 weeks after surgery? METHODS: A randomized controlled study was performed between February, 2013 and April, 2014 at a multicenter metropolitan tertiary-care referral center. Patients who presented with acute closed distal radius fractures (Orthopaedic Trauma Association 23A-C) were potentially eligible for inclusion. During the study period, 40 patients with closed, displaced, and unstable distal radius fractures were identified as meeting inclusion criteria and offered enrollment and randomization. Three patients (7.5%), all with concomitant injuries, declined to participate at the time of randomization as did one additional patient (2.5%) who chose not to participate, leaving a final sample of 36 participants. There were no dropouts after randomization, and analyses were performed according to an intention-to-treat model. Patients were randomly assigned to one of two groups, general anesthesia or brachial plexus blockade, and among the 36 patients included, 18 were randomized to each group. Medications administered in the postanesthesia care unit were recorded. Patients were discharged receiving oxycodone and acetaminophen 5/325 mg for pain control, and VAS forms were provided. Patients were called at predetermined intervals postoperatively (2 hours, 4 hours, 6 hours, 12 hours, 24 hours, 48 hours, and 72 hours) to gather pain scores, using the VAS, and to document the doses of analgesics consumed. In addition, patients had regular followups at 2 weeks, 6 weeks, and 12 weeks. Pain scores were again recorded using the VAS at these visits. RESULTS: Patients who received general anesthesia had worse pain scores at 2 hours postoperatively (general anesthesia 6.7 ± 2.3 vs brachial plexus blockade 1.4 ± 2.3; mean difference, 5.381; 95% CI, 3.850-6.913; p < 0.001); whereas reported pain was worse for patients who received a brachial plexus blockade at 12 hours (general anesthesia 3.8 ± 1.9 vs brachial plexus blockade 6.3 ± 2.4; mean difference, -2.535; 95% CI, -4.028 to -1.040; p = 0.002) and 24 hours (general anesthesia 3.8 ± 2.2 vs brachial plexus blockade 5.3 ± 2.5; mean difference, -1.492; 95% CI, -3.105 to 0.120; p = 0.031).There was no difference in operative suite time (general anesthesia 119 ± 16 minutes vs brachial plexus blockade 125 ± 23 minutes; p = 0.432), but time in the recovery room was greater for patients who received general anesthesia (284 ± 137 minutes vs 197 ± 90; p = 0.0398). Patients who received general anesthesia consumed more fentanyl (64 µg ± 93 µg vs 6.9 µg ± 14 µg; p < 0.001) and morphine (2.9 µg ± 3.6 µg vs 0.0 µg; p < 0.001) than patients who received brachial plexus blockade. Functional outcome scores did not differ at 6 weeks (data, with mean and SD for both groups, and p value) or 12 weeks postoperatively (data, with mean and SD for both groups, and p value). CONCLUSIONS: Brachial plexus blockade pain control during the immediate perioperative period was not significantly different from that of general anesthesia in patients undergoing operative fixation of distal radius fractures. However, patients who received a brachial plexus blockade experienced an increase in pain between 12 to 24 hours after surgery. Acknowledging "rebound pain" after the use of regional anesthesia coupled with patient counseling regarding early narcotic administration may allow patients to have more effective postoperative pain control. It is important to have a conversation with patients preoperatively about what to expect regarding rebound pain, postoperative pain control, and to advise them about being aggressive with taking pain medication before the waning of regional anesthesia to keep one step ahead in their pain control management. LEVEL OF EVIDENCE: Level 1, therapeutic study.

Factors Affecting Hand Surgeon Operating Room Turnover Time.

Background: The purpose of this study was to determine the factors that affect hand surgeon operating room (OR) turnover time. We hypothesized that surgeon presence in the OR, decreased American Society of Anesthesiologists (ASA) class, smaller case type, and earlier case time, as well as other factors, decreased OR turnover time. Methods: A total of 685 hand surgery cases performed by 5 attending hand surgeons between September 2013 and December 2014 were identified. Turnover time, patient comorbidities (ASA class), surgeon, prior OR surgical procedure, current OR surgical procedure, location of the surgery (ambulatory surgical center [ASC] vs orthopedic specialty hospital [OSH]), time of surgery, and order of OR cases were recorded. The effect of surgeon routine variables, OR case factors, and patient health status on OR turnover was analyzed. Results: Turnover time was significantly shorter in cases where the surgeon remained in the OR during turnover (27.5 minutes vs 30.4 minutes) and when the surgeon incentivized OR staff (24 minutes vs 29 minutes). The ASC was found to have shorter turnover times than the OSH (27.9 minutes vs 36.4 minutes). In addition, ASA class, type of prior OR procedure, type of current OR procedure, and case order all significantly affected turnover time. Comparison of OR turnover time among the 5 surgeons revealed a statistically significant difference at the OSH but not at the ASC. Conclusion: OR turnover time is significantly affected by surgeon routine, location of surgery, patient ASA class, procedure type, and case order. Interestingly, the effect of hand surgeon routine on OR turnover time may be amplified at an academic OSH versus an ASC.

"Hand surgeons probably don't starve": Patient's perceptions of physician reimbursements for performing an open carpal tunnel release.

BACKGROUND: The purpose of this study is to evaluate patient's perceptions of physician reimbursement for the most commonly performed surgery on the hand, a carpal tunnel release (CTR). METHODS: Anonymous physician reimbursement surveys were given to patients and non-patients in the waiting rooms of orthopaedic hand physicians' offices and certified hand therapist's offices. The survey consisted of 13 questions. Respondents were asked (1) what they thought a surgeon should be paid to perform a carpal tunnel release, (2) to estimate how much Medicare reimburses the surgeon, and (3) about how health care dollars should be divided among the surgeon, the anesthesiologist, and the hospital or surgery center. Descriptive subject data included age, gender, income, educational background, and insurance type. RESULTS: Patients thought that hand surgeons should receive $5030 for performing a CTR and the percentage of health care funds should be distributed primarily to the hand surgeon (56 %), followed by the anesthesiologist (23 %) and then the hospital/surgery center (21 %). They estimated that Medicare reimburses the hand surgeon $2685 for a CTR. Most patients (86 %) stated that Medicare reimbursement was "lower" or "much lower" than what it should be. CONCLUSION: Respondents believed that hand surgeons should be reimbursed greater than 12 times the Medicare reimbursement rate of approximately $412 and that the physicians (surgeons and anesthesiologist) should command most of the health care funds allocated to this treatment. This study highlights the discrepancy between patient's perceptions and actual physician reimbursement as it relates to federal health care. Efforts should be made to educate patients on this discrepancy.

The Management of Domestic Animal Bites to the Hand.

Hand bites from domestic animals are extremely common. Though many may initially appear benign, it is important for treating physicians to be aware of the factors that place patients sustaining animal bites at additional risk for infection. As clinicians, we must be able to efficiently diagnose and treat these patients properly to avoid the morbidity that animal bites can provoke. The current paper reviews the evaluation and management of domestic animal bites to the hand.

Impact of Joint Position and Joint Morphology on Assessment of Thumb Metacarpophalangeal Joint Radial Collateral Ligament Integrity.

PURPOSE: A 2-part biomechanical study was constructed to test the hypothesis that coronal morphology of the thumb metacarpophalangeal joint impacts the assessment of instability in the context of radial collateral ligament (RCL) injury. METHODS: Fourteen cadaveric thumbs were disarticulated at the carpometacarpal joint. Four observers measured the radius of curvature of the metacarpal (MC) heads. In a custom jig, a micrometer was used to measure the RCL length as each thumb was put through a flexion and/or extension arc under a 200 g ulnar deviation load. Strain was calculated at maximal hyperextension, 0°, 15°, 30°, 45°, and maximal flexion. Radial instability was measured with a goniometer under 45 N stress. The RCL was then divided and measurements were repeated. Analysis of variance and Pearson correlation metrics were used. RESULTS: The RCL strain notably increased from 0° to 30° and 45° of flexion. With an intact RCL, the radial deviation was 15° at 0° of flexion, 18° at 15°, 17° at 30°, 16° at 45°, and 14° at maximal flexion. With a divided RCL, instability was greatest at 30° of flexion with 31° of deviation. The mean radius of curvature of the MC head was 19 ± 4 mm. Radial instability was inversely correlated with the radius of curvature to a considerable degree only in divided RCL specimens, and only at 0° and 15° of flexion. CONCLUSIONS: The RCL contributes most to the radial stability of the joint at flexion positions greater than 30°. The results suggest that flatter MC heads contribute to stability when the RCL is ruptured and the joint is tested at 0° to 15° of metacarpophalangeal flexion. CLINICAL RELEVANCE: The thumb MC joint should be examined for RCL instability in at least 30° of flexion.

Radiocapitellar joint contact pressures following radial head arthroplasty.

PURPOSE: To determine the radial head arthroplasty length that best replicates the native radiocapitellar contact pressure. METHODS: Eight cadaveric elbows (4 matched pairs) with an average age of 73 years were tested. All specimens were ligamentously stable and without visible cartilage wear. Radiocapitellar contact pressures were digitally analyzed during simulated joint loading at 0°, 45°, and 90° of elbow flexion and neutral rotation in the intact specimens and after ligament-preserving radial head arthroplasty at -2 mm, 0 mm, and +2 mm of the native length. The results were analyzed using 1-way analysis of variance and post hoc Tukey pairwise comparison tests. RESULTS: Paired analysis demonstrated significantly decreased mean contact pressures when comparing the native versus the minus 2 groups. Significantly decreased maximum contact pressures were also noted between the native and the minus 2 groups. Examining the mean contact pressures showed no significant difference between the native and the zero group and the native and the plus 2 groups. As for the maximum contact pressures, there was also no significant difference between the native and the zero group and the native and the plus 2 group. CONCLUSIONS: Up to 2 mm of overlengthening may be tolerated under simulated loading conditions without significantly increasing contact pressures of the radiocapitellar joint. Surgeons can use this knowledge along with radiographic parameters and intraoperative examination of elbow stability to gauge the appropriate size of the radial head implant to be used in order to decrease the risk of overstuffing the joint and minimizing radiocapitellar chondral wear. TYPE OF STUDY/LEVEL OF EVIDENCE: Prognostic IV.