Background: Medial plica syndrome is the most common symptomatic plica. The size and shape of the plica have an important impact on impingement on the femoral condyle and hence, symptoms. The validity of the classification systems of such injuries is essential for prospective studies. The study was designed to assess the reproducibility and reliability of Iino and Jee classification systems of medial plica syndrome. The agreement among multiple surgeons for medial plica syndrome has not been established before.

Methods: We validated both classification systems independently from the original authors at our institution. Arthroscopic videos from 30 patients were reviewed by 6 consultant surgeons, 6 registrars and 6 house officers. Intra- and inter-observer reliability and reproducibility were assessed. Each observer scored the videos on two separate occasions and classified the medial plica according to its type (A, B, C and D) for Iino classification and (1, 2, 3 and 4) for Jee classification system.

Results: The results were subjected to weighted kappa analysis. Intra-observer agreements were 0.76 for consultants, 0.64 for registrars and 0.60 for house officers for Iino classification system. They were 0.81 for consultants, 0.75 for registrars and 0.71 for house officers for Jee classification system. Total unanimity (18 observers assigned same grade for medial plica) was achieved in 23% (7 of 30) with Iino system and 40%(12 of 30) for Jee system. Inter-observer agreement was 0.63 for the first reading and 0.68 for the second reading for Iino system. They were 0.72 for the first reading and 0.80 for the second reading for Jee system. Validity analysis showed a kappa value of 0.78 (substantial agreement).

Conclusions: Jee classification system showed a better inter and intra-observer agreement compared to Iino’s system. Consultants had a better intra-observer agreement compared to their registrars. We have shown that substantial agreement can be found between individuals with no specialist training. Such reliability is crucial for multi-centre clinical research studies involving arthroscopic knee surgery. Jee’s classification system did not consider femoral condyle impingment which has an important effect on symptoms. Both systems should be used in prospective studies to evaluate the state of the medial plica.

Background: Avascular necrosis is a well recognised complication of displaced proximal humeral fractures irrespective of conservative and operative treatment. The reported rate of AVN with open reduction and internal fixation varies from 22 to 40%. The aim of our study was to look at the functional outcome and the incidence of AVN with operative treatment using locking plate with a minimum 3 year follow up.

Materials and methods: We retrospectively reviewed a consecutive series of 50 patients with displaced proximal humerus fractures treated with ORIF from 2002 to 2004. All patients were operated by the two senior authors employing anterior deltopectoral approach, indirect reduction, secure suture repair of the tuberosities and fixation with locking plate. The minimum follow up was 3 years. There were 9 two-part, 19 three-part and 22 four-part fractures. Their average age was 63 years. All patients were assessed objectively with Constant score and subjectively with Oxford questionnaire by an independent observer. Fracture healing and complications were recorded.

Results: 47 patients were available for follow-up. All fractures united. The average Constant score was 84. Their mean Oxford score was 16. There was no infection or metal work failure. One patient fractured below the plate after a fall but went onto uneventful union. 4 of 47 patients (8.5%) developed avascular necrosis. Three were four-part fracture and one was two-part fracture. Three patients underwent hemiarthroplasty of shoulder with good functional recovery. One patient declined further operative intervention due to low level of symptoms.

Conclusion: Indirect reduction and secure fixation of the tuberosities onto the humeral head with a locking plate is a reliable technique of treating displaced proximal humeral fractures. Our experience of avascular necrosis in only 8.5% of these fractures is much lower than any reported series after open surgery.