However, in the United States, unlike in Europe and Asia, sonography is not
considered a first-line imaging modality for shoulder pain.7,10 In addition,
the usefulness of shoulder sonography is widely considered to be operator
dependent, with the radiologist’s experience being the primary factor in this
imaging modality’s effectiveness.11–13
Several studies have found little agreement in sonographic results between less- and more-experienced operators, even when the operators were evaluating full-thickness tears.13,14 Other studies have found good to excellent reliability of sonography for the diagnosis of full-thickness rotator cuff tears but less satisfactory detection sensitivity for partialthickness tears.1,4,11,14 In a meta-analysis of 65 studies, de Jesus et al2 found no differences between MRI and sonography in sensitivity or specificity for detection of full or partial rotator cuff tears.
Given these variable results, many non-European radiologists may believe that musculoskeletal sonography is a difficult technique to learn and implement and may thusexclude it from clinical practice in patients with shoulder pain. However, with the current economic climate, we wished to challenge both this hypothesis and the current dependency on MRI for the diagnosis of shoulder disorders. In this study, we assessed the effect of implementing an open-ended, comprehensive training program on the diagnostic accuracy of shoulder sonographic interpretation in a clinical practice.
Several studies have found little agreement in sonographic results between less- and more-experienced operators, even when the operators were evaluating full-thickness tears.13,14 Other studies have found good to excellent reliability of sonography for the diagnosis of full-thickness rotator cuff tears but less satisfactory detection sensitivity for partialthickness tears.1,4,11,14 In a meta-analysis of 65 studies, de Jesus et al2 found no differences between MRI and sonography in sensitivity or specificity for detection of full or partial rotator cuff tears.
Given these variable results, many non-European radiologists may believe that musculoskeletal sonography is a difficult technique to learn and implement and may thusexclude it from clinical practice in patients with shoulder pain. However, with the current economic climate, we wished to challenge both this hypothesis and the current dependency on MRI for the diagnosis of shoulder disorders. In this study, we assessed the effect of implementing an open-ended, comprehensive training program on the diagnostic accuracy of shoulder sonographic interpretation in a clinical practice.
Discussion
Shoulder sonography is becoming increasingly popular for the diagnosis of rotator cuff tears due in part to its lower cost, accessibility, and results that are similar to those obtained with MRI.1–4 However, sonography is not the first-line imaging test for shoulder pain in the United States, likely because operator experience is thought to contribute to variable diagnostic accuracy and reproducibility.2,11–14,16,17 This lack of preference for shoulder sonography is the case despite the American College of Radiology appropriateness criteria rating of sonography as 8 or 9 (usually appropriate) for patients older than 35 years with shoulder pain and suspected rotator cuff tears/impingement.16 In addition, the American College of Radiology appropriateness criteria rate sonography as 5 (maybe appropriate) should MRI (9, usually appropriate) be contraindicated for patients with ersistent pain.
In contrast, the American College of Radiology appropriateness score for sonography increases to 8 of 9 (usually appropriate) for evaluation of the postoperative cuff or in patients older than 35 years with suspected impingement.
In one study, the interobserver concordance for the diagnosis of full- and partial-thickness rotator cuff tears on independent examinations was found to be high (92%) between 2 operators with more than 5 years of shoulder sonography experience.16 Another study found that agreement between an experienced musculoskeletal radiologist and a general radiologist with no experience in shoulder sonography was 98% for full-thickness rotator cuff tears and 90% for partial-thickness tears.11 These studies demonstrated good inter-rater measurement reproducibility; however, in the second study, the sensitivity, specificity, and accuracy for the detection of full-thickness rotator cuff tears relative to surgery as a reference standard were 3% to 4% lower for the general radiologist than for the experienced musculoskeletal radiologist.11 Another study evaluated the learning curves for 2 orthopedic surgeons using office-based sonographic examinations to detect full-thickness supraspinatus tears previously diagnosed with MRI.12 In this study, at least 100 shoulder sonographic examinations were required to enable each surgeon to detect full-thickness tears, with diagnostic accuracy of 67 of 72 (93%) and 92 of 95 (97%), respectively, in the second round of 100 examinations. The variability in reported operator accuracy for rotator cuff disorders other than for full-thickness tears2,11–14,16,17 may have led to a lack of confidence in sonographically based diagnoses. Nevertheless, agreement and accuracy for the diagnosis of full-thickness tears are high.1,2,11,13,14
....
Shoulder sonography is becoming increasingly popular for the diagnosis of rotator cuff tears due in part to its lower cost, accessibility, and results that are similar to those obtained with MRI.1–4 However, sonography is not the first-line imaging test for shoulder pain in the United States, likely because operator experience is thought to contribute to variable diagnostic accuracy and reproducibility.2,11–14,16,17 This lack of preference for shoulder sonography is the case despite the American College of Radiology appropriateness criteria rating of sonography as 8 or 9 (usually appropriate) for patients older than 35 years with shoulder pain and suspected rotator cuff tears/impingement.16 In addition, the American College of Radiology appropriateness criteria rate sonography as 5 (maybe appropriate) should MRI (9, usually appropriate) be contraindicated for patients with ersistent pain.
In contrast, the American College of Radiology appropriateness score for sonography increases to 8 of 9 (usually appropriate) for evaluation of the postoperative cuff or in patients older than 35 years with suspected impingement.
In one study, the interobserver concordance for the diagnosis of full- and partial-thickness rotator cuff tears on independent examinations was found to be high (92%) between 2 operators with more than 5 years of shoulder sonography experience.16 Another study found that agreement between an experienced musculoskeletal radiologist and a general radiologist with no experience in shoulder sonography was 98% for full-thickness rotator cuff tears and 90% for partial-thickness tears.11 These studies demonstrated good inter-rater measurement reproducibility; however, in the second study, the sensitivity, specificity, and accuracy for the detection of full-thickness rotator cuff tears relative to surgery as a reference standard were 3% to 4% lower for the general radiologist than for the experienced musculoskeletal radiologist.11 Another study evaluated the learning curves for 2 orthopedic surgeons using office-based sonographic examinations to detect full-thickness supraspinatus tears previously diagnosed with MRI.12 In this study, at least 100 shoulder sonographic examinations were required to enable each surgeon to detect full-thickness tears, with diagnostic accuracy of 67 of 72 (93%) and 92 of 95 (97%), respectively, in the second round of 100 examinations. The variability in reported operator accuracy for rotator cuff disorders other than for full-thickness tears2,11–14,16,17 may have led to a lack of confidence in sonographically based diagnoses. Nevertheless, agreement and accuracy for the diagnosis of full-thickness tears are high.1,2,11,13,14
....
Because the study was retrospective,
varying standards of patient care may have been used, as well as
nonstandardized radiology and surgical
report language. Standardization of this report nomenclature with prospectively
defined terminology would decrease reporting variability and aid in the comparison
of results. In addition, the range of musculoskeletal sonography experience may
have increased the variability of the study results. We did not separate out the
examinations interpreted by the most experienced sonographer because we believe
that the benefits of acquisition and interpretation standardization as well as
feedback based on surgical correlation also improved the accuracy of this radiologist’s
sonographic work. Finally, because of the retrospective nature of this study,
the patient population was inhomogeneous with regard to referral patterns,
symptoms, and the distribution of tendon tears across groups.
The results of this retrospective study demonstrate that introducing musculoskeletal sonography into a new clinical practice is not only feasible but can be accomplished with high diagnostic accuracy. The use of musculoskeletal sonography may enable a decrease in health care costs by substitution of a diagnostic musculoskeletal sonographic examination for a shoulder MRI examination.7 The use of sonography as a first-line diagnostic imaging modality for shoulder pain is warranted, as evidenced by the European guidelines.10
Furthermore, based on the findings of this study, we believe that the implementation of a systematic quality improvement program, including acquisition protocol standardization and a comprehensive, ongoing educational program for all team members, can improve the diagnostic performance of all aspects of musculoskeletal sonography, not only sonography limited to rotator cuff injuries.
Although operator experience cannot be ruled out as a factor in sonographic interpretation, this study demonstrates that education provided to a group of operators with a wide variety of experience increases the diagnostic sensitivity and accuracy of sonography for detecting full-thickness supraspinatus and infraspinatus tendon tears.
In conclusion, implementation of formal, ongoing training that embraces all team members, standardizes acquisition and interpretation protocols, and provides a forum for continuous quality improvement raises the diagnostic accuracy and sensitivity of shoulder sonography for rotator cuff injuries. Our work supports the potential of musculoskeletal sonography as a first-line imaging modality for shoulder pain when rotator cuff disorders are suspected.7,10 By implementing an open-ended training program for the entire care team, musculoskeletal sonography can be easily and successfully introduced into a new clinical practice with high diagnostic accuracy.
The results of this retrospective study demonstrate that introducing musculoskeletal sonography into a new clinical practice is not only feasible but can be accomplished with high diagnostic accuracy. The use of musculoskeletal sonography may enable a decrease in health care costs by substitution of a diagnostic musculoskeletal sonographic examination for a shoulder MRI examination.7 The use of sonography as a first-line diagnostic imaging modality for shoulder pain is warranted, as evidenced by the European guidelines.10
Furthermore, based on the findings of this study, we believe that the implementation of a systematic quality improvement program, including acquisition protocol standardization and a comprehensive, ongoing educational program for all team members, can improve the diagnostic performance of all aspects of musculoskeletal sonography, not only sonography limited to rotator cuff injuries.
Although operator experience cannot be ruled out as a factor in sonographic interpretation, this study demonstrates that education provided to a group of operators with a wide variety of experience increases the diagnostic sensitivity and accuracy of sonography for detecting full-thickness supraspinatus and infraspinatus tendon tears.
In conclusion, implementation of formal, ongoing training that embraces all team members, standardizes acquisition and interpretation protocols, and provides a forum for continuous quality improvement raises the diagnostic accuracy and sensitivity of shoulder sonography for rotator cuff injuries. Our work supports the potential of musculoskeletal sonography as a first-line imaging modality for shoulder pain when rotator cuff disorders are suspected.7,10 By implementing an open-ended training program for the entire care team, musculoskeletal sonography can be easily and successfully introduced into a new clinical practice with high diagnostic accuracy.
Dedicated Training Program for Shoulder
Sonography, Patricia B. Delzell, MD, Alex Boyle, Erika
Schneider, PhD, J Ultrasound Med 2015; 34:1037–1042
Không có nhận xét nào :
Đăng nhận xét