Assessment of the Diagnostic Accuracy of Ultrasound (USG) for the Detection of Rotator Cuff Tear with Respect to Magnetic Resonance Imaging (MRI) IRIA

Reason for choosing this research question: Previously published studies are done among western population and no Indian data exists in this regard. Most of the studies do not mention the transducer frequency.

Aim and Objectives:  To assess the diagnostic accuracy of ultrasound for the detection of rotator cuff tear with respect to Magnetic Resonance Imaging (MRI)  To quantify sensitivity, specificity, positive and negative predictive values for the same  To assess common risk factor for rotator cuff tears among Indian population

Normal Anatomy: SubscapularisSupraspinatus Infraspinatus Teres Minor Wikipedia

Literature Review: More than 50% of individuals more than 80 years of age show rotator cuff tears (1) Cause restriction in activities of daily living (2) No published Indian data Sensitivity and specificity of ultrasound to diagnose rotator cuff tear with respect of MRI Sr. No.StudySensitivitySpecificity 1Misamore %60.00% 4Nicoletti %80.00% 2Martin-Hervas %67.00% 3Wallny %82.00% 5De Candia %100.00% 6Teefey %67.00% 7Iannotti %80.00% 8Sipola %45.00%

Sample size calculation for a new diagnostic tool: Expected sensitivity / specificity of new test 80% Precision of 10% Alpha error 5% (confidence interval of 95%) Minimum sample ~ 35 (MRI positive cases) Software used for sample size calculation: nMaster 2.0 Note: Specificity taken as 80% based on the study done by Iannotti et al which showed that ultrasound has as specificity of 80% to diagnose any rotator cuff tear.(3)

Study population recruitment: Inclusion criteria:  Shoulder pain, both acute and chronic  Stiffness of shoulder  Restriction in activities of daily living  Trauma to shoulder Exclusion criteria:  Refusal for the ultrasound study  Female subjects without a chaperone  Post-operative cases  Subjects unable to cooperate due to pain  Patients presenting for evaluation of tumours / malignancies The study was approved by the institutional review board

Design of data collection:  Demographic data and questions related to risk factors were collected prospectively  Available lab investigations were checked retrospectively Personnel:  Ultrasound operator: PG resident doing the study  MRI reporting doctor: Musculoskeletal radiologist Minimizing bias:  The index test (ultrasound) were interpreted independently of the reference standard (MRI)

Equipment: USG machine: Siemens ACUSON S2000 MRI scanner: Philips Achieva 3.0T MRI 14 MHz ultrasound transducer used for the study

Ultrasound criteria for rotator cuff pathology (4): Partial thickness tendon tear: Characterised by a well-defined hypoechoic or anechoic abnormality that disrupt the tendon fibres  Interstitial tear: does not extend to bursal or articular surface  Articular tear: extends to articular surface  Bursal tear: extends to bursal surface Full thickness tear: Characterised by a well-defined hypoechoic or anechoic abnormality that disrupts the hyperechoic tendon fibres and extend from the articular to bursal surface of the tendon.

Study Protocol: Ultrasound: Biceps was assessed with elbow in 90 degree flexion Subscapularis was assessed with the arm in external rotation and elbow at 90 degree flexion Abduction of arm done to assess for impingement Supraspinatus and infraspinatus assessed after keeping the arm in internal rotation such that the dorsal of the hand touched the back of the subject MRI:  Proton density weighted (PDW) axial SPAIR (fat suppressed)  T2W SPAIR coronal  T2W SPAIR sagittal  PDW coronal  PDW sagittal  T1W axial

Data Collection: Epidemiological data: Age, gender, handedness, occupation, history of smoking and trauma Patient record for diabetes, hypertension, cholesterol and serum Vit. D levels Ultrasound data: Tendon morphology, tear (partial vs full thickness), synovial thickening and bursal fluid MRI data: Similar to ultrasound

Results - Epidemiological data: Gender Age Average age of the subjects was 39.6 years (+/ years) Average age for men was 38.3 years (+/ years) Average age for women was 44.4 years (+/- 12 years) Age >45 years – significant association for supraspinatus tendinopathic changes / tear (Pearson Chi-Square test)

Right shoulder was evaluated in 45 subjects Left shoulder was evaluated in 25 subjects 80% did light work and 20% heavy work 38% gave history of smoking 50% gave history of previous trauma 10 % had diabetes 17% had hypertension * 29% had elevated serum cholesterol level In 18 subjects Vit. D levels were assessed, only 3 had normal level * – significant association for supraspinatus tendinopathic changes / tear (Pearson Chi-Square test)

Diagnosis of rotator cuff tendon tear: Ultrasound vs MRI USG vs MRI Supraspi natus Infraspin atus Subsca pularis Any tendon tear Sensitivity 84%57%38% 73.08% Specificity 78.79%93.65%83.87% 88.39% PPV 89%50%23% 67.86% NPV 74.29%95.16%91.23% 90.73% Accuracy 81.43%90.00%78.50% 84.54% PPV: Positive predictive value NPV: Negative predictive value Statistics: -True positive and negative and false positive and negative observations were tabulated using 2x2 tables -Sensitivity, specificity, PPV, NNV and diagnostic accuracy calculated -Pearson Chi Square test used to assess for significant associations

Misdiagnosis: Misdiagnosed as no tear Partial thickness tear of right supraspinatus

Complete tear of supra and Infraspinatus Misdiagnosed as no tear Supraspinatus tear Infraspinatus tear Bursal thickening misdiagnosed as normal cuff muscle

Conclusion: Ultrasound is a sensitive and cost effective modality for the diagnosis of rotator cuff injuries Sensitivity for the diagnosis of supraspinatus tendon tears 84% Age (>45 yrs.) and presence of hypertension had significant association for the presence of supraspinatus tendinopathy / tear Results of our study as compared to published data: Sr. No.StudySensitivitySpecificity 1 Misamore %60.00% 4 Nicoletti %80.00% 2 Martin-Hervas %67.00% 3 Wallny %82.00% 5 De Candia %100.00% 6 Teefey %67.00% 7 Iannotti %80.00% 8 Sipola %45.00% Our Study %88.00%

Scope for improvement: More practice in performing shoulder ultrasound before starting the study would have reduced misdiagnosis Higher transducer frequency (>15 MHz) may have improved the diagnostic yield Thank you

References: 1. Tashjian RZ. Epidemiology, natural history, and indications for treatment of rotator cuff tears. Clin Sports Med Oct;31(4):589– Nakajima D, Yamamoto A, Kobayashi T, Osawa T, Shitara H, Ichinose T, et al. The effects of rotator cuff tears, including shoulders without pain, on activities of daily living in the general population. J Orthop Sci Mar;17(2):136–40. 3.Iannotti JP, Ciccone J, Buss DD, Visotsky JL, Mascha E, Cotman K, et al. Accuracy of office-based ultrasonography of the shoulder for the diagnosis of rotator cuff tears. J Bone Joint Surg Am Jun;87(6):1305– Jacobson JA. Fundamentals of Musculoskeletal ultrasound. 2nd ed. Elsevier; p.