An Independent Learning Method for Orthopaedic Surgeons Performing Shoulder Ultrasound to Identify Full-Thickness Tears of the Rotator Cuff

Murphy, Richard James MA, MBChB; Daines, Michael Todd MD; Carr, Andrew Jonathan MA, ChM, FRCS, FMedSci; Rees, Jonathan Lloyd MBBS, FRCS(Eng), MD, FRCS(Tr&Orth)

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.K.00706
Scientific Articles
Abstract

Background: There is an evolving interest in shoulder ultrasound performed by orthopaedic surgeons as part of routine clinical assessment of the rotator cuff in a so-called one-stop clinic. This study investigated the accuracy of ultrasound assessment of rotator cuff integrity performed by orthopaedic surgeons without prior experience of ultrasound who were following our proposed learning protocol.

Methods: We studied four surgeons without previous experience with shoulder ultrasound and monitored their ability to evaluate rotator cuff integrity using ultrasound compared with findings at arthroscopy. The surgeons attended a formal training course and were taught a protocol to identify and size full-thickness tears of the rotator cuff. The surgeons performed preoperative scans on the day that patients underwent shoulder arthroscopy. This allowed the surgeons to receive same-day feedback with comparison of arthroscopic images and ultrasound images.

Results: One hundred and fifty-nine shoulders were scanned by the surgeons in the study. In the initial training period, surgeons who performed >100 scans demonstrated a sensitivity of 94% and a specificity of 88% (a positive predictive value of 79% and a negative predictive value of 97%) for the identification of a full-thickness tear and agreed with intraoperative sizing of the defect in 84% of the scans. In the later training period, the predictive values showed a sensitivity of 90% and a specificity of 97% (a positive predictive value of 95% and a negative predictive value of 94%) for the identification of a full-thickness tear and agreement with intraoperative sizing for 95% of the scans.

Conclusions: The predictive values obtained in this study for the evaluation of rotator cuff integrity were comparable with published results from experienced radiologists. This study demonstrates the capacity of our proposed learning protocol to train surgeons without previous ultrasound experience to reliably evaluate rotator cuff integrity using ultrasound within fifty to 100 scans.

Level of Evidence: Diagnostic Level I. See Instructions for Authors for a complete description of levels of evidence.

Author Information

1Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, Botnar Research Centre, Nuffield Orthopaedic Centre, Oxford, OX3 7LD, UK. E-mail address for R.J. Murphy: richard.murphy@ndorms.ox.ac.uk

Article Outline

Rotator cuff tendinopathy is the most common cause of shoulder pain and results in more elective surgical interventions to the shoulder than any other pathological condition1-3. In the treatment of rotator cuff tendinopathy, information regarding rotator cuff integrity is valuable when one is making decisions about surgical intervention. Routine clinical assessment in an outpatient setting may require imaging of the rotator cuff to evaluate the integrity of the cuff tendons. The presence of a full-thickness rotator cuff tear can influence the decision to undertake surgery and the type of procedure that may be performed.

Ultrasound is commonly used to image the shoulder and is reliable and accurate in identifying full-thickness tears of the rotator cuff compared with surgical findings as a gold standard and when performed by an experienced radiologist4-9. Comparison of preoperative ultrasound and magnetic resonance imaging has shown the two modalities to have equal predictive capabilities in identifying and quantifying defects of the rotator cuff10-13.

Advances in ultrasound technology have led to the development of affordable, high resolution, portable scanners that present the opportunity to perform ultrasound evaluations outside the radiology department. Consequently, a number of clinical specialties, including cardiology, anesthesiology, rheumatology, and shoulder surgery, use this versatile, dynamic, and real-time imaging modality. Studies of orthopaedic surgeons with substantial ultrasound experience have shown that their assessment of the rotator cuff is comparable with that done by musculoskeletal radiologists14-16. These clinicians use ultrasound within the boundaries of their own area of expertise to perform a specific evaluation of the rotator cuff tendons as part of their clinical management.

The expansion in ultrasound performed by clinicians who are not radiologists has led to debate over training and level of competence. The American Institute of Ultrasound in Medicine, the Royal College of Radiologists (United Kingdom), and the European Federation of Societies for Ultrasound in Medicine and Biology have each published training guidelines for medical practitioners who are not radiologists to direct development of proficiency in musculoskeletal ultrasound17-19. These publications suggest that a clinician must perform a minimum of 150 to 300 scans under the supervision of a qualified musculoskeletal ultrasonographer or radiologist to develop proficiency in musculoskeletal ultrasound. Each of the guidelines encompasses musculoskeletal ultrasound and does not offer a training solution for practitioners wishing to utilize ultrasound for a focused and anatomically limited purpose within their area of specialty and in a setting that does not require constant supervision from a trainer. These learning protocols rely on the expertise of the trainer as a best practice standard against which the trainee can be compared and ultimately deemed competent.

The aim of this study was to investigate the accuracy of ultrasound assessment of rotator cuff integrity performed by orthopaedic surgeons without prior experience of ultrasound who followed our proposed learning protocol.

Back to Top | Article Outline

Materials and Methods

Four surgeons, including one specialist shoulder surgeon, one shoulder fellow, one third-year orthopaedic resident, and one first-year orthopaedic resident, took part in the study, providing a spectrum of clinical experience. None of the surgeons had prior experience of performing shoulder ultrasound. Prior to participant recruitment, each surgeon completed a formal two-day training course in shoulder ultrasound instructed by a faculty of experienced musculoskeletal radiologists. The course included theory and practical classes with development and evaluation of technical skills and diagnosis with use of a variety of ultrasound machines and scanning performed on volunteers with normal and abnormal shoulders.

Following approval from the local research ethics committee, participants were prospectively recruited between June 2009 and December 2010 from patients scheduled to have arthroscopic subacromial decompression with or without rotator cuff repair. These patients presented with subacromial impingement pain diagnosed by the history and findings on physical examination, including painful arc and positive Hawkins and Jobe tests; all had failed conservative therapy in the form of subacromial corticosteroid injections and physiotherapy. Each participant underwent ultrasound scanning of the affected shoulder performed on the day of surgery. When it was possible, study participants were scanned by all of the surgeons involved in the study. Each surgeon performed the scan independently, participants were not permitted to divulge any prior knowledge of their condition to the surgeons performing the scans, and discussion between surgeons was not permitted until all ultrasound scans had been completed and reported. To maintain a high standard of patient care, the specialist shoulder surgeon involved in the study had previously examined the participants who were on his operating list when they were seen in his outpatient clinic, but the other three surgeons in the study performed their ultrasound scans without any prior knowledge of the participants and without performing a physical examination.

The first fifty scans performed by each surgeon were considered to represent the initial training period, and the second fifty scans performed by each surgeon were considered to represent the later training period.

The ultrasound scans were carried out with use of a portable scanner (Voluson i; GE Healthcare) with a 4.7 to 13-MHz small-parts linear-array probe (SP10-16-RS; 34.5 mm in length). The machine settings were optimized for shoulder ultrasound prior to the start of the study with help from an industry product and applications specialist. These settings were used by each of the surgeons in the study. A standardized shoulder ultrasound scanning protocol was used to cover the key areas of potential rotator cuff pathology on the basis of published techniques20,21. Patients were seated or slightly reclined for the scans. With the arm in a neutral position (hand resting on the ipsilateral thigh, with the palm upward), the lesser tuberosity, subscapularis, and bicipital groove were assessed in axial and sagittal planes. The arm was moved into external rotation to improve visualization of the subscapularis tendon. The arm was then moved into a position of extension (the hand placed on the iliac crest) to draw the rotator cuff tendons out from beneath the acromion. The supraspinatus and infraspinatus tendons were assessed in longitudinal (oblique-coronal) and cross-sectional (oblique-sagittal) planes. Finally, the posterior fibers of the infraspinatus and teres minor were visualized with the hand resting on the contralateral shoulder. Defects within the tendons were identified as previously described by Teefey et al.5. The surgeons saved ultrasound images to demonstrate the condition of each of the rotator cuff tendons and completed a report form that detailed the findings of the scan. Measurements of the width (anteroposterior measurement) of any full-thickness tears identified were determined; those measuring ≤3 cm were reported to be small to medium sized, and those measuring >3 cm were reported to be large to massive. An operative record of arthroscopic findings was retained for each patient, and the surgeons who had performed the scan of the patient then received same-day feedback after completion of the operation. Tear size was measured intraoperatively in the subacromial space with use of a calibrated probe in the anterolateral portal and was reported with use of the same measurements as for the ultrasound analysis. Using this same-day feedback method, the surgeons were able to recall the ultrasound scans they had performed earlier in the day and derive maximum learning benefit from the surgical findings. For further reinforcement and when ultrasound reporting was incorrect, the saved images from both ultrasound and arthroscopy were reviewed. Surgeon-performed ultrasound scans and reports and the operative record were retained for each participant. Any preoperative ultrasound reports that had been completed by the local musculoskeletal radiology department were also collected for comparison. The results were analyzed for the presence or absence of a full-thickness rotator cuff tear and the anteroposterior size of the tears, features that have the most substantial impact on our surgical management and decision-making.

We used this observational review of the learning curve associated with this method for training in shoulder ultrasound to suggest a pragmatic and sensible learning program for surgeons, which is summarized in the Appendix.

Back to Top | Article Outline
Statistical Methods

The results were tabulated to compare ultrasound findings with surgical findings for each of the surgeons. Predictive values (sensitivity, specificity, the positive predictive value, the negative predictive value, and accuracy) for the identification of full-thickness tears of the rotator cuff were calculated from these data. Ninety-five percent confidence intervals were calculated for each of the proportions with use of a normal approximation method. Partial-thickness tears and intact rotator cuffs were both considered to be intact. Chance-corrected agreement between the results from each of the surgeons and the local musculoskeletal radiology department was determined by calculating kappa values. The surgeons’ results were combined to demonstrate the performance of the group as a whole during two time periods: (1) the initial training period, represented by the first fifty ultrasound scans performed by each surgeon and (2) the later training period, represented by the second fifty ultrasound scans performed by the two surgeons who each completed 100 scans.

The reported measurements of tear size as small to medium or large to massive from ultrasound and arthroscopy were cross-tabulated to illustrate the proportion of imaging and surgical reports that yielded the same results. The proportion of matching reports was calculated as a percentage with confidence intervals as described above.

Back to Top | Article Outline
Source of Funding

The study was funded by the National Institute for Health Research, National Health Service.

Back to Top | Article Outline

Results

Three hundred and three ultrasound scans were performed on 159 shoulders from 156 patients during the study period. Two of the surgeons, the specialist shoulder surgeon and the first-year orthopaedic resident, scanned 101 shoulders each; the shoulder fellow scanned fifty shoulders; and the third-year orthopaedic resident scanned fifty-one shoulders. As such, all four surgeons’ scans contributed to the analysis of the initial training period of the group, but only the specialist and the first-year resident completed sufficient scans to define the later training period. For 119 shoulders, a preoperative radiology department ultrasound report was available for comparison. Each surgeon performed his ultrasound scans with similar frequency: the specialist shoulder surgeon and first-year orthopaedic resident scanned patients over the full duration of the study, with the shoulder fellow involved for the first half of the study and the third-year orthopaedic resident for the second half.

Operative findings showed that seventy-four shoulders (47%) had an intact rotator cuff, thirty-two (20%) had a partial-thickness tear, and fifty-three (33%) had a full-thickness tear. All fifty-three full-thickness tears included a component in the supraspinatus tendon; thirty-six of these were <3 cm in width, involving the supraspinatus tendon with or without extension into the infraspinatus tendon; seventeen were >3 cm, involving both the supraspinatus and infraspinatus tendons; and four had a concomitant full-thickness tear of the subscapularis tendon.

Table I details the outcome of each surgeon’s ultrasound scans matched with intraoperative findings. Table II details the predictive values for these data allowing comparison of performance between individuals.

Sixty-three shoulders were scanned by both the specialist shoulder surgeon and the first-year orthopaedic resident. They agreed about the presence or absence of a full-thickness rotator cuff tear in 86% of the shoulders and showed a kappa value of 0.71 (standard error = 0.09), suggesting substantial agreement among the clinicians. The agreement among these clinicians and the radiology department was equally high, with the first-year orthopaedic resident agreeing with radiology about the findings in 85% of eighty-two shoulders that were scanned by both the resident and a radiologist (kappa value = 0.70, standard error = 0.08) and the specialist shoulder surgeon agreeing with radiology about the findings in 83% of seventy-eight shoulders that were scanned by both the specialist and a radiologist (kappa value = 0.67, standard error = 0.08).

Table III shows ultrasound sizing of correctly identified full-thickness rotator cuff tears from all four surgeons matched with intraoperative sizing for the two training periods. Ultrasound and intraoperative sizing of the full-thickness tears as small to medium or large to massive showed agreement of 84% (95% confidence interval [CI], 71% to 97%) for the initial training period of the first fifty scans and agreement of 95% (95% CI, 88% to 100%) for the later training period of the second fifty scans for the surgeons who completed 100 scans each.

For the 303 ultrasound scans performed by the surgeons in the study, there was the potential for 108 scans to identify a full-thickness tear. Eleven scans failed to identify a full-thickness tear; ten of these were reported as partial-thickness tears and only one was reported as an intact rotator cuff. All of the missed tears were small to medium sized. Fourteen scans incorrectly identified a full-thickness tear; nine of these had partial-thickness tears at arthroscopy.

Back to Top | Article Outline

Discussion

The use of shoulder ultrasound by orthopaedic surgeons as an adjunct to clinical history and physical examination can provide important information at point of access for the patient and allows the provision of a so-called one-stop clinic, saving time and hospital visits, reducing demand on radiology departments, and potentially offering a cost saving.

This study aimed to bridge the gap between the uninitiated orthopaedic surgeon and those with many years of ultrasound experience in an effort to establish some evidence for a suitable minimum training period for individuals wishing to follow this independent learning method.

Current training guidelines in the use of musculoskeletal ultrasound for clinicians who are not radiologists require the learning clinician to be supervised directly by a qualified musculoskeletal ultrasonographer or radiologist for at least 150 to 300 scans17-19. The need for a supervising radiologist or sonographer to act as the gold standard within the confines of the radiology department is rather limiting. The substantial time investment required by the learning clinicians and supervising radiologists or sonographers to complete this training renders these protocols unfeasible for clinicians who are not radiologists.

Our learning method utilizes arthroscopy as a gold standard for the assessment of rotator cuff integrity. Combining this gold standard with same-day comparison of ultrasound and arthroscopic findings provides a learning method whereby surgeons can hone their ultrasound technique and image interpretation quickly and accurately by using a definitive outcome measure against which they can judge their findings.

The predictive values for ultrasound identification of a full-thickness tear attained by the surgeons using our learning method from their initial training period of their first fifty scans show a high level of proficiency even at this stage. The values attained during the later training period of the second fifty scans are comparable with published data for surgeons experienced in shoulder ultrasound and for musculoskeletal radiologists10,14-16. These data are detailed for comparison in Table IV.

Compared with the predictive values for the identification of full-thickness tears demonstrated by the local musculoskeletal radiology department (sensitivity of 93%, specificity of 85%, positive predictive value of 76%, negative predictive value of 96%, and accuracy of 87%), the results attained in the study show that the surgeons demonstrated a similar level of ability to identify full-thickness tears within the initial training period of their first fifty scans using the learning method.

The ultrasound measurement of tear size demonstrated by the surgeons was also comparable with the results seen from the local musculoskeletal radiology department. When categorizing the tears into clinically useful groups, as small to medium (≤3 cm) or large to massive (>3 cm), the agreement between preoperative ultrasound measurements and intraoperative sizing increased from 84% in the initial training period to 95% in the later training period. Due to the small sample size of tears identified, statistical analysis of the difference between these values was not valid. Results from the local musculoskeletal radiology department demonstrated agreement for the same measurements in 71% of the shoulders, which was again equaled by the surgeons within the initial training period. It is important to note, however, that preoperative radiology ultrasound scans were not performed on the day of surgery. The time delay between the radiology scan and the day of surgery may have allowed for tears to increase in size and impact the level of agreement between the radiology reports and intraoperative findings. This limitation does not detract from the high level of agreement between the surgeons’ ultrasound measurements and intraoperative findings.

Eleven scans failed to identify a full-thickness rotator cuff tear and fourteen scans mistakenly identified partially torn or intact tendons as containing a full-thickness tear. These mistakes may reflect the inexperience of the user, limitations of the imaging modality in terms of technological capability, or even patient variables such as anatomical differences and body habitus. Nonetheless, the rate and type of mistakes made are comparable with those noted in recent published results of ultrasound reporting compared with surgical findings by experienced users of shoulder ultrasound4-7,9,10,14-16.

Perhaps of most importance, the study showed a high negative predictive value for the detection of full-thickness rotator cuff tears for all of the participants. As a consequence, a negative scan performed in the acute or chronic setting can reassure the assessing surgeon that the presence of a full-thickness rotator cuff tear is highly unlikely and therefore the potential need for early repair is not indicated.

The learning protocol used in this study offers an independent method of learning for surgeons that may also be used for continued performance monitoring and self-assessment after completion of training to maintain a high standard of care. When using ultrasound in an outpatient setting, a surgeon can retain an image log and report findings as part of the medical record; predictive values can be reviewed for the patients who progress to surgery.

With any clinical investigation that relies on subjective evaluation and technical ability, there are limitations that may impact its evaluation. The study was limited by the nature of our patient group; arthroscopic diagnosis was the chosen outcome measure for comparison and, as such, the patients included in the study had already failed conservative therapy and may have had a higher prevalence of structural rotator cuff defects than a cross section of patients seen in a shoulder clinic. This may represent an example of spectrum bias as the prevalence of a condition within a study population may differ from the target population and affect the apparent predictive values obtained by a diagnostic test used in that population. Table IV details the prevalence of full-thickness tears in this study population and that of other recent publications, showing this study to have a comparable or lower prevalence of tears than the other investigations. This finding could lead to this study having a falsely lower sensitivity for the detection of full-thickness tears compared with the other publications.

During the patients’ clinic attendance, at least two months prior to surgery, the specialist shoulder surgeon in the study had previously performed a physical examination on the patients he scanned and, in some cases, would have seen a prior radiology report. These issues may have led to bias in his ultrasound reporting, although a substantial time had passed between prior clinical assessment and the day of ultrasound scanning for all shoulders, and we did not see a difference in performance between the specialist and the remainder of the group who had not examined the patients or had any prior knowledge of their history.

The ultrasound technique in this study was not a detailed musculoskeletal ultrasound scan, as may be performed by a radiologist, but aimed to establish the most important aspect of image analysis in rotator cuff pathology: the presence or absence of a full-thickness tear. Including this fast, affordable, and immediate imaging method as an adjunct to clinical evaluation does not eradicate the need for further musculoskeletal imaging but may help to screen patients effectively prior to more advanced imaging methods in some cases.

The study showed that surgeons could develop proficiency in ultrasound assessment of rotator cuff integrity in a relatively short period of time by using an independent training method and produce results that are comparable with published data for experienced musculoskeletal radiologists. It was not within the scope of this study, however, to establish a minimum level of proficiency required for surgeons to perform ultrasound as part of clinical practice or the number and frequency of ultrasound scans that need to be completed after the initial training period to maintain that level of proficiency.

In conclusion, the study demonstrates that orthopaedic surgeons with varying levels of clinical experience can become competent in shoulder ultrasound in a relatively short period of time by following an independent learning program. Importantly, this method of learning offers the opportunity for shoulder surgeons to use their everyday surgical practice to audit and improve their ultrasound assessment of the rotator cuff. Combined with the growing affordability and quality of portable ultrasound machines, this learning method facilitates independent training and ongoing validation for surgeon-performed ultrasound.

Back to Top | Article Outline

Appendix Cited Here...

A figure showing the shoulder ultrasound training program for surgeons is available with the online version of this article as a data supplement at jbjs.org.

Investigation performed at the Oxford National Institute for Health Research Biomedical Research Unit, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, United Kingdom

Back to Top | Article Outline

References

1. Ostör AJ Richards CA Prevost AT Speed CA Hazleman BL. Diagnosis and relation to general health of shoulder disorders presenting to primary care. Rheumatology (Oxford). 2005 Jun;44(6):800–5. Epub 2005 Mar 15.
2. Chard MD Hazleman R Hazleman BL King RH Reiss BB. Shoulder disorders in the elderly: a community survey. Arthritis Rheum. 1991 Jun;34(6):766–9.
3. Vecchio P Kavanagh R Hazleman BL King RH. Shoulder pain in a community-based rheumatology clinic. Br J Rheumatol. 1995 May;34(5):440–2.
4. Mack LA Matsen FA 3rd Kilcoyne RF Davies PK Sickler ME. US evaluation of the rotator cuff. Radiology. 1985 Oct;157(1):205–9.
5. Teefey SA Hasan SA Middleton WD Patel M Wright RW Yamaguchi K. Ultrasonography of the rotator cuff. A comparison of ultrasonographic and arthroscopic findings in one hundred consecutive cases. J Bone Joint Surg Am. 2000 Apr;82(4):498–504.
6. Brenneke SL Morgan CJ. Evaluation of ultrasonography as a diagnostic technique in the assessment of rotator cuff tendon tears. Am J Sports Med. 1992 May-Jun;20(3):287–9.
7. Wiener SN Seitz WH Jr. Sonography of the shoulder in patients with tears of the rotator cuff: accuracy and value for selecting surgical options. AJR Am J Roentgenol. 1993 Jan;160(1):103–7; discussion 109-10.
8. Hedtmann A Fett H. [Ultrasonography of the shoulder in subacromial syndromes with disorders and injuries of the rotator cuff]. Orthopade. 1995 Nov;24(6):498–508. German.
9. Ottenheijm RP Jansen MJ Staal JB van den Bruel A Weijers RE de Bie RA Dinant GJ. Accuracy of diagnostic ultrasound in patients with suspected subacromial disorders: a systematic review and meta-analysis. Arch Phys Med Rehabil. 2010 Oct;91(10):1616–25.
10. Teefey SA Rubin DA Middleton WD Hildebolt CF Leibold RA Yamaguchi K. Detection and quantification of rotator cuff tears. Comparison of ultrasonographic, magnetic resonance imaging, and arthroscopic findings in seventy-one consecutive cases. J Bone Joint Surg Am. 2004 Apr;86(4):708–16.
11. Bryant L Shnier R Bryant C Murrell GA. A comparison of clinical estimation, ultrasonography, magnetic resonance imaging, and arthroscopy in determining the size of rotator cuff tears. J Shoulder Elbow Surg. 2002 May-Jun;11(3):219–24.
12. de Jesus JO Parker L Frangos AJ Nazarian LN. Accuracy of MRI, MR arthrography, and ultrasound in the diagnosis of rotator cuff tears: a meta-analysis. AJR Am J Roentgenol. 2009 Jun;192(6):1701–7.
13. Fotiadou AN Vlychou M Papadopoulos P Karataglis DS Palladas P Fezoulidis IV. Ultrasonography of symptomatic rotator cuff tears compared with MR imaging and surgery. Eur J Radiol. 2008 Oct;68(1):174–9. Epub 2007 Dec 21.
14. Al-Shawi A Badge R Bunker T. The detection of full thickness rotator cuff tears using ultrasound. J Bone Joint Surg Br. 2008 Jul;90(7):889–92.
15. Ziegler DW. The use of in-office, orthopaedist-performed ultrasound of the shoulder to evaluate and manage rotator cuff disorders. J Shoulder Elbow Surg. 2004 May-Jun;13(3):291–7.
16. Jeyam M Funk L Harris J. Are shoulder surgeons any good at diagnosing rotator cuff tears using ultrasound?: A comparative analysis of surgeon vs radiologist. Int J Shoulder Surg. 2008 Jan;2(1):4–6.
17. American Institute of Ultrasound in Medicine. Official statement: training guidelines for the performance of musculoskeletal ultrasound examinations. 2009 Nov 14. http://www.aium.org/resources/statements.aspx. Accessed 2012 Aug 15.
18. Board of the Faculty of Clinical Radiology. Ultrasound training recommendations for medical and surgical specialties. The Royal College of Radiologists; 2005. p 1–58.
19. Minimum training requirements for the practice of Medical Ultrasound in Europe. Ultraschall Med. 2010 Aug;31(4):426-7. Epub 2010 Aug 19.
20. Allen GM. Shoulder ultrasound imaging-integrating anatomy, biomechanics and disease processes. Eur J Radiol. 2008 Oct;68(1):137–46. Epub 2008 Apr 21.
21. Mack LA Nyberg DA Matsen FA 3rd. Sonographic evaluation of the rotator cuff. Radiol Clin North Am. 1988 Jan;26(1):161–77.

Disclosure: One or more 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 an 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 2013 by The Journal of Bone and Joint Surgery, Incorporated