Clinical decision support tools are software that match the input characteristics of an individual patient to an established knowledge base to create patient-specific assessments that support and better inform individualized healthcare decisions. Clinical decision support tools can facilitate better evidence-based care and offer the potential for improved treatment quality and selection, shared decision making, while also standardizing patient expectations. Predict+ is a novel, clinical decision support tool that leverages clinical data from the Exactech Equinoxe shoulder clinical outcomes database, which is composed of >11,000 shoulder arthroplasty patients using one specific implant type from more than 30 different clinical sites using standardized forms. Predict+ utilizes multiple coordinated and locked supervised machine learning algorithms to make patient-specific predictions of 7 outcome measures at multiple postoperative timepoints (from 3 months to 7 years after surgery) using as few as 19 preoperative inputs. Predict+ algorithms predictive accuracy for the 7 clinical outcome measures for each of aTSA and rTSA were quantified using the mean absolute error and the area under the receiver operating curve (AUROC).Introduction
Methods
Reverse total shoulder arthroplasty (RTSA) has a proven track record as an effective treatment for a variety of rotator cuff deficient conditions. However, glenoid erosion associated with the arthritic component of these conditions can present a challenge for the shoulder arthroplasty surgeon. Options for treatment of glenoid wear include partial reaming with incomplete baseplate seating, bony augmentation using structural or impaction grafting techniques, and augmented baseplates. Augmented components have the advantage of accommodating glenoid deformity with a durable material and also ream less subchondral bone; both of which may offer an advantage over traditional bone grafting. Biomechanical and early clinical studies of augmented glenoid baseplates suggest they are a reasonable treatment option, though posteriorly augmented baseplates have shown better performance than superiorly augmented implants. However, there are no mid- or late-term studies comparing augmented baseplates to bone grafting or partial reaming. We present a live surgical demonstration of RTSA for a patient with advanced glenoid erosion being treated with an augmented glenoid baseplate that can be dialed in the direction of any deformity (superior, posterior, etc.). This versatility allows the surgeon to place the augment in any direction and is not confined to the traditional concepts of glenoid wear in a single vector. Clearly, longer term follow up studies are needed to determine the ultimate effectiveness of these devices in treating glenoid deformity in RTSA.
As reverse total shoulder arthroplasty (RTSA) systems expand with longer durations in vivo, so does the concern and potential complications of wear, debris and osteolysis. Despite some other profound attempts, no wear testing method has stood out to compare implants across systems and labs. The main reasons may have been the diverse sources of forces and motions used in testing, widely different wear amounts which resulted and the general lack of dedicated shoulder simulators. To add a dedicated shoulder simulator to hip and knee simulators would burden the resources of any testing lab. In this study we propose a shoulder wear test method which addresses the above. Harnessing the wealth of force-motion data from telemetrized shoulder implants from the Bergman's group in Berlin, we synthesized their results to devise a wholistic multi-axes simulation regime for reverse shoulders. The alignment and motions of the humeral cup and the glenosphere were kept anatomically correct (relative to each other) and yielded a physiologically realistic wear-inducing articulation. However, we opted for a very unusual installation/orientation of the whole implant system to allow a twelve station AMTI (hip) simulator to be adapted for this study. The shoulder constructs were aligned with novel fixtures such that the machine's vertical compressive force mimicked the average forces of the shoulder found from the in vivo telemetry data in magnitude and nominal direction. Aligned thus, a patient with a shoulder installed would neither stand, nor lie down, but be oriented in a composite angle relative the simulator original axes. Each anatomic shoulder motion would be achieved by unique computed combinations of the three simulator motion actuators, none of which would be aligned anatomically for the shoulder on its own. The maximum ranges of cyclic shoulder motion achieved with the constraints of the simulator were 38°–79° of forward elevation repeated in two separate (15°and 45°) elevation planes. The change of elevation plane inherently involved abduction-adduction motion, and simultaneously also involved variation of internal-external rotation within a 57° range. Each elevation rise (twice per cycle) was also accompanied by a sinusoidally rising and falling compressive load in the range 50N–1700N. The test method was tested (!) by simulating for 2.5 million of the above (double-elevation) cycles and gravimetrically measuring wear of twelve 36 mm size RTSA systems. We compared six systems having vitamin E-infused highly cross-linked polyethylene bearings (100 kGy radiation) to six controls with a medium cross-linked polyethylene of half the radiation dose. Significant wear resulted for the control bearing material (average 17.9 ± 0.851 mg/MC) which was no less than many hips and knees. Multiply (and statistically significantly, p < 0.001) less average wear (3.42 ± 0.22 mg/MC) resulted for the highly cross linked bearings. The above demonstrated the effectiveness of the test method. Significant wear resulted under physiologically realistic cyclic motion and forces with strong discrimination between two systems whose bearing materials were known to be different in resilience to wear. Using novel fixtures and unusual orientation to utilize a standard commercially available joint simulator promises efficacy of the test method and utility across different labs.
