Aim. This retrospective study evaluated the outcome of treatment for unhealed fracture-related infections (FRI). Methods. We identified a consecutive, single-centre cohort of patients having treatment for an FRI Consensus confirmed FRI. All fractures were unhealed at the time of treatment. Patients were followed up for at least one year. Successful outcome was a healed fracture without recurrent infection. Lack of union, persistent infection and/or unplanned reoperation defined failure. Results. Demographics: 183 patients (184 FRIs) with mean age 52.1 years (range 17-96) were treated and followed up for a mean of 2.8 years (range 1-9.4). Mean duration of FRI was 1.1 years with 65 (35.5 %) presenting within 6 months of injury. 118 patients had established infected non-union. FRI was most frequent in the tibia (74), femur (48) and humerus (24). 171 patients were BACH Complex. 75.5% of FRIs were culture positive, with Staph. aureus being the most frequent organism. Polymicrobial infection and Gram negative cultures were common (25.5% and 33.6%). Treatment: 98.3% of surgeries were performed in one stage with just 3 planned 2-stage procedures (2 endoprosthetic replacements and 1 free fibular flap). No bone graft was used in any surgery and all wounds were closed at first operation. 48 cases (26%) required flap coverage (29 free flaps and 19 local flaps). Local antibiotics were used in 124 cases (67.4%) of primary surgeries. All patients had sampling, debridement, systemic antibiotics and wound closure. 40 (21.7%) had DAIR, 31 (16.8%) had new internal fixation and 105 (57.1%) had external fixation (including 79 Ilizarov fixators). Outcomes: After primary surgery, 84.6% of all patients were infection-free and 77.2% had united. After further surgery, 98.8% were infection-free and 98.1% had united. External fixation techniques achieved infection eradication in 89.1% compared to 71.7% with any internal fixation (p=0.005). Primary internal fixation achieved union in 81.7% compared to 74.3% with external fixation (p=0.27). Secondary surgery after external fixation was mainly docking site fixation. Conclusion. Unhealed FRIs present a difficult challenge for treatment. This large series demonstrated that single-stage treatment, without bone grafting, gave acceptable results with few reoperations. Primary external fixation gave more certainty of infection eradication but required more reoperations to secure union. However, this difference in reoperation was not statistically significant. We strongly advocate managing these patients with a multidisciplinary team which can treat all aspects of the condition

Abstract. Aim. The aim of this study was to present the results of treatment of displaced lateral clavicle fractures by an arthroscopically inserted tightrope device (‘Dogbone’, Arthrex). Methods. We performed a retrospective series of our patients treated with this technique between 2015 and 2019. Patients were identified using the ‘CRS Millennium’ software package and operation notes/clinic letters were analysed. We performed an Oxford Shoulder Score (OSS) on all the patients at final follow-up. Our electronic ‘PACS’ system was used to evaluate union in the post-operative radiographs. Results. We treated 26 patients with displaced lateral clavicle fractures between 2015 and 2019. There were 4 patients who were treated with a ‘dogbone’ and supplementary plate fixation and the remaining 22 were treated with a ‘dogbone’ alone. Radiological union was seen in 22 (84%) patients. The mean Oxford Shoulder Score (OSS) was 46. Apart from one patient who required removal of the superior endobutton and knot under local anaesthetic there was no other secondary surgery. There were no cases of infection, nerve injury or frozen shoulder. Conclusions. Arthroscopic ‘dogbone’ treatment of lateral clavicle fractures is a safe, cosmetically friendly technique with promising high rates of fracture union and return to normal function. We recommend its use over the more conventional treatment of a hook plate

Distal femoral physeal fractures can cause of growth distrurbance which frequently requires further surgical intervention. The aim of this study was to determine if tibial tuberosity ossification at the time of injury can predict further surgery in patients who have sustained a physeal fracture of the distal femur. We retrospectively investigated all patients who had operative treatment for a distal femoral physeal fracture at a paediatric level one trauma center over a 17 year period. Logistic regression analysis was performed investigating associations between the need for further surgery to treat growth disturbance and tibial tuberosity ossification, age, Salter Harris grade, mode of fixation or mechanism of injury. 74 patients met the inclusion criteria. There were 57 boys (77%) and 17 girls (23%). The average age at time of injury was 13.1 years (range 2.-17.1 years). Following fixation, 30 patients (41%) underwent further surgery to treat growth disturbance. Absence of tibial tuberosity fusion to the metaphysis was significantly associated with need for further surgery (p = < 0 .001). Odds of requiring secondary surgery after tibial tuberosity fusion to metaphysis posteriorly (compared with not fused) were 0.12, 95% CI (0.04, 0.34). The estimate of effect of tibial tuberosity ossification on reoperation rates did not vary when adjusted for gender, mechanism, fixation and Salter Harris grade. When accounting for age, the odds of further operation if the tibial tuberosity is fused to the metaphysis posteriorly (compared with not fused) were 0.28, 95% CI (0.08, 0.94). Tibial tuberosity ossification stage at time of injury is a predictor of further surgery to treat growth disturbance in paediatric distal femoral fractures. Children with distal femoral physeal fractures whose tibial tuberosity was not fused to the metaphysis posteriorly were 8.3 times more likely to require further surgery

Whether or not to resurface the patella in total knee arthroplasty (TKA) remains controversial. Several methods of dealing with the patella exist: ALWAYS resurface; NEVER resurface; SOMETIMES resurface. There is good reason to consider selective patellar resurfacing. First, in an age of partial knee arthroplasty we have become more tuned in to analyzing patterns of arthritis. In TKA there is a high percentage of patients who do not have significant patellar cartilage wear or anterolateral knee pain. These patients may be candidates for leaving the patella unresurfaced in TKA. Arno et al found that 42% of patients had no significant patellar arthritis at the time of TKA. Roberts et al found that only 15% of patients should undergo patella resurfacing based on the presence of exposed bone on the patella; the other 85% could be considered suitable for leaving the patella unresurfaced. Second, despite a cumulative incidence of less than 5–10%, problems related to patellar resurfacing account for perhaps the most catastrophic complications encountered, with treatments that have limited success. These complications include fracture, avascular necrosis, extensor mechanism disruption, and anterior knee pain. Third, it is a fallacy to think that anterior knee pain (AKP) does not exist despite primary patellar resurfacing in TKA. Meftah (Ranawat) et al found that AKP persists in 30% of patients and new AKP develops in 10% of patients after TKA with patellar resurfacing. Barrack et al found that with patellar resurfacing the incidence of AKP is 28% in patients without preop AKP and 9% in those with preop AKP. They also found that without patellar resurfacing the incidence of new AKP was 14% and persistent AKP was 23%. Fourth, only roughly 44–64% of patients who undergo secondary patellar resurfacing for AKP after TKA with an unresurfaced patella actually get relief of their pain, suggesting that there is some other etiology of anterior knee pain. Residual component malalignment, boxy femoral components, PF overstuffing, referred pain or asymmetric resurfacing may explain ongoing pain. Finally, the data in well-designed studies show that selective patellar resurfacing can produce similar outcomes with and without resurfacing, particularly in those without significant patellar arthritis. In multiple studies, higher rates of secondary surgery occur when the patella is left unresurfaced in primary TKA, but this is for “pain” without clear etiology. On the other hand secondary surgery is rarely performed in TKA with patellar resurfacing for “pain” only, despite its high incidence. The quality of patellar cartilage at the time of primary TKA should be considered, as that may be the best indicator of whether a knee will do well without patellar resurfacing (that is, selective patellar resurfacing may be a better idea than never resurfacing the patella). While patellar resurfacing remains controversial in modern TKA, excellent outcomes are achievable with, and without, primary patellar resurfacing. Selectively leaving the patella unresurfaced when there is limited patellar arthritis may not only be highly effective, but it may also limit the incidence of secondary resurfacing that may occur with more substantial patellar arthritis while also minimizing the risk of some of the devastating complications that can occur due to patellar resurfacing in TKA

Aims

Excision of chronic osteomyelitic bone creates a dead space which must be managed to avoid early recurrence of infection. Systemic antibiotics cannot penetrate this space in high concentrations, so local treatment has become an attractive adjunct to surgery. The aim of this study was to present the mid- to long-term results of local treatment with gentamicin in a bioabsorbable ceramic carrier.

Aim. Simultaneous use of Ilizarov techniques with transfer of free muscle flaps is not current standard practice. This may be due to concerns about duration of surgery, clearance of infection, potential flap failure or coordination of surgical teams. We investigated this combined technique in a consecutive series of complex tibial infections. Method. A single centre, consecutive series of 45 patients (mean age 48 years; range 19–85) were treated with a single stage operation to apply an Ilizarov frame for bone reconstruction and a free muscle flap for soft-tissue cover. All patients had a segmental bone defect in the tibia, after excision of infected bone and soft-tissue defects which could not be closed directly or with local flaps. We recorded comorbidities, Cierny-Mader and Weber-Cech classification, the Ilizarov method used, flap type, follow-up duration, time to union and complications. Results. 26 patients had osteomyelitis and 19 had infected non-union. Staphylococci were cultured in 25 cases and 17 had polymicrobial infections. Ilizarov monofocal compression was used in 14, monofocal distraction in 15, bifocal compression/distraction in 8, and bone transport in 8. 8/45 had an additional ankle fusion, 7/45 had an angular deformity corrected at the same time and 24 also had local antibiotic carriers inserted. Median time in frame was 5 months (3–14). 38 gracilis, 7 latissimus dorsi and 1 rectus abdominus flaps were used. One flap failed within 48 hours and was revised (flap failure rate 2.17%). There were no later flap complications. Flaps were not affected by distraction or bone transport. Mean follow-up was 23 months (10–89). 44/45 (97.8%) achieved bony union. Recurrence of infection occurred in 3 patients (6.7%). Secondary surgery was required to secure union with good alignment in 8 patients (17.8%; docking site surgery in 6, IM nailing in 2) and in 3 patients for infection recurrence. All were infection free at final follow-up. Conclusions. Simultaneous Ilizarov reconstruction with free muscle flap transfer is safe and effective in treating segmental infected tibial defects, and is not associated with an increased flap failure rate. It shortens overall time spent in treatment, with fewer operations per patient. However, initial theatre time is long and a committed multidisciplinary team is required to achieve good results

Radial head fractures with fragment displacement should be reduced and fixed, when classified as Mason II type injuries. We describe a method of arthroscopic fixation which is performed as a day case trauma surgery, and compare the results with a more traditional fixation approach, in a case controlled manner. We prospectively reviewed six Mason II radial head fractures which were treated using an arthroscopic reduction and fixation technique. The technique allows the fracture to be mobilised, reduced, and anatomically fixed using headless screws. All arthroscopic surgeries were conducted as day-cases. We retrospectively collected age and sex matched cases of open reduction and fixation of Mason II fractures using headless screws. The arthroscopic cases required less analgesia, shorter hospital admissions, and had fewer complications. The averaged final range of follow-up, at 1 year post-operation was 15 to 140 degrees in the arthroscopic group and 35 to 120 degrees in the open group. The Mayo Elbow Performance Score was 95/100 and 90/100 respectively. No acute complications were noted in the arthroscopic group, and a radial nerve neuropraxia [n=1], superficial wound infection [n=1], and loose screw [n=1]. Two patients of the arthroscopic group required secondary motion gaining operations [n=1 arthroscopic anterior capsulectomy for a fixed flexion contracture of 35 degrees, and n=1 loss of supination requiring and arthroscopic radial scar excision]. Three patients in the open group required secondary surgery [n=2 arthroscopic anterior capsulectomy for fixed flexion deformities, and n=1 arthroscopic anterior capsulectomy for fixed flexion deformities, and n=1 arthroscopic radial head excision for prominent screws, loss of forearm rotation, and radiocapitellar arthrosis pain]. The technique of arthroscopic fixation of Mason II radial head fractures appears to be valid, with respect to anatomical restoration of the fracture, minimal hospital admission, reduction in analgesia requirement, fewer complications, and a decreased need for secondary surgery

End-stage ankle arthritis (ESAA) is a debilitating disease that does not affect all individuals equally. Gender differences have been identified in patients with end-stage hip and knee arthritis and have stimulated research to explain these findings. The present study was undertaken to examine if gender has a significant effect on pre-operative disability and post-operative outcomes in patients with ESAA. Patients undergoing ankle arthrodesis (AA) or total ankle replacement (TAR) with minimum 2-year follow-up were identified in the Canadian Orthopaedic Foot and Ankle Society prospective ankle reconstruction database. Demographic data, revision data, patient satisfaction questionnaires, and outcome data using the Ankle Osteoarthritis Scale (AOS) and Short-form 36 (SF-36) health survey were collected. TAR: 384 patients were included, with 198 females and 186 males. Patient BMI, comorbidities, and duration of follow-up were similar between groups. Males were slightly older at the time of surgery (65.1 vs 62.4 years, p=0.01)). The most common etiology was post-traumatic arthritis for both genders, however females had a higher rate of rheumatoid arthritis (17% vs 5%, p=0.001). Implant types included STAR, Hintegra, and Mobility, and were similar between groups. Preoperatively females had higher rates of pain and disability, demonstrated by lower SF-36 physical component scores (PCS) (31.0 vs 34.5, p<0.001), and higher AOS pain (54.7 vs 51.1, p=0.05) and AOS disability scores (66.5 vs 59.6, p<0.001). Postoperatively, both groups had significant improvement in PCS, AOS pain, and AOS disability scores. Females, however, continued to demonstrate lower PCS scores (38.3 vs 41.9, p<0.001) and higher AOS disability (31.0 vs 25.8, p=0.02) than males. Regression analysis found that preoperative PCS, gender, age, and arthritis etiology all had a significant impact on postoperative PCS scores, with preoperative PCS scores having the largest impact. Preoperative AOS pain and disability scores had the largest impact on postoperative AOS pain and disability scores, respectively. Gender had no significant impact on AOS pain and disability scores postoperatively. Patient satisfaction was similar between males and females postoperatively. Secondary surgery was performed in 13.6% of females and 16.1% of males. Five males and five females underwent revision to arthrodesis. In patients with ESAA, females tend to have higher pre-operative levels of pain and disability compared to males, which persists post-operatively. This is consistent with the hip and knee arthroplasty literature. This finding may be due to females undergoing surgery at more advanced disease states, arthritis etiology, referral bias, or treatment bias. Both males and females have significant and similar degrees of improvement in pain and disability scores after TAR, and reoperation rates and patient satisfaction rates are similar despite the apparent disparity in outcomes

Background. Glenoid component aseptic loosening is the most common source of total shoulder arthroplasty (TSA) revision. In an attempt to strengthen cemented glenoid component fixation, divergent pegged glenoids were designed. Divergent peg creation was intended to increase cement purchase and provide resistance to component rocking. Methods. Thirty-four patients who underwent divergent peg TSA had data collected prospectively. The data from these patients was retrospectively reviewed, primarily for radiographic evidence of glenoid component loosening. The endpoint was defined as the need for revision secondary to glenoid loosening. Secondary outcome measures such as SPADI (shoulder pain and disability index), active forward elevation, abduction, internal rotation, and external rotation were also collected. Data was obtained preoperatively and at the following postoperative intervals: 3 months, 6 months, and yearly. The last available postoperative radiographs were also reviewed and graded on a modified Franklin glenoid lucency scale described by Lazarus et al. Results. The mean follow-up was 5.6 years ± 2.44 (range, 2–10 years). Of the 34 patients in the study, 20 patients had radiographs available for review. Radiographs showed complete lucency around one or less pegs in 14/20 patients (grade 0–2). Six out of twenty glenoids had complete radiolucency around two or more pegs with gross loosening seen in one shoulder (grade 3–5). No secondary surgery was performed in any patients. Significant improvements were seen in active elevation 30.2 degrees ±44.3 (p=0.043) at the last follow up. Improvement was also seen in internal rotation of 4.3 levels ± 4.9 (p=0.016) at the last follow-up. The mean postoperative SPADI score decreased by 48.1 points (p=0.039). Conclusions. Patients with divergent pegged glenoid TSA showed significant improvement in postoperative shoulder elevation, internal rotation, and SPADI scores at final follow-up. No patients underwent revision. However, there were a significant number of patients with radiolucency seen around 2 or more pegs

Introduction and aims. We present a series of patients who have had secondary reconstruction of war injuries to the upper and lower limbs, sustained during the Iraq and Afghanistan conflicts. Material and Methods. All patients were seen at the combined Peripheral Nerve Injuries Clinic at the Defence Medical Centre for Rehabilitation, Headley Court. All surgery was performed at Odstock Hospital. Procedures include scar excision and neurolysis (all patients), release of scar contractures, tenolysis, tendon transfers, revision nerve grafts, excision of neuroma, and soft tissue reconstruction using pedicled or free flaps. Results. 24 patients have been treated at the time of submission. We have used 13 free flaps (1 free groin flap, 9 anterolateral thigh, 3 parascapular, with 4 as through-flow flaps) and 1 pedicled groin flap, with no flap losses. There were 6 amputation stump revisions (1 above elbow, 5 below knee). The majority (n=23) have had nerve recovery distal to the level of injury following revision surgery. Conclusions. Nerve repairs recover following neurolysis (and revision nerve graft if necessary) with provision of good soft tissue cover. Release of scar contractures with flap cover allows healing of chronic wounds and permits mobilisation of joints. Thin fascio-cutaneous flaps provide good contour and can be elevated more easily than skin grafted muscle flaps for secondary surgery. Free or regional flaps are preferable to local flaps in high energy-transfer military wounds. Immediate complex reconstruction is not always appropriate in multiply-injured patients

Purpose. Limited information is available regarding the functional outcomes of radial head fractures managed with open reduction and internal fixation (ORIF). The purpose of this study was to determine the functional outcomes of radial head fractures treated with ORIF. Method. Fifty-two patients, with a mean age of 4412 years, who were treated with radial head ORIF were evaluated at a mean of 4.42.4 years. Thirty were isolated radial head fractures (Group A), 13 (Group B) were associated with a complex fracture-dislocation (terrible triad variants), and 5 (Group C) were associated with a proximal ulnar fracture (Monteggia/trans-olecranon variants). Fourty-four were partial articular fractures and 8 were complete articular fractures. Outcomes were assessed with physical and radiographic examination, and validated self-reported questionnaires. Results. The average PREE score (Patient Rated Elbow Evaluation) for Groups A, B, and C were 7.613.1, 12.313.4, and 108.5, respectively. The average MEPI (Mayo Elbow Performance Score) for Groups A, B, and C were 8913, 8511, and 918, respectively. For Groups A, B, and C respectively, the prevalence of radiographic radiocapitellar arthritis was 30%, 46%, and 20%. The average flexion/extension arc for Groups A, B, and C were 7 to 132, 6 to 134, and 10 to 132 respectively. Secondary surgery was performed in 17% of cases, most commonly for decreased motion. Three comminuted fractures failed ORIF and required conversion to radial head arthroplasty. Conclusion. Patients with radial head fractures, including those associated with complex fracture-dislocations, can achieve excellent functional outcomes with low self reported pain and disability when treated with ORIF, despite radiographic evidence of mild post-traumatic arthritis

We report on the first clinical cases of the Arthrex Ankle Syndesmosis TightRope (winner of 2003 BOA Technological Achievement Award and 2004 Cutlers' Prize), which has recently been licensed for use where classically a syndesmosis screw would be used. Twelve patients with Weber C ankle fractures treated with Arthrex TightRope syndesmosis fixation have a minimum of six months follow-up. The syndesmosis was fixed with the ankle in plantarflexion to aid reduction. Patient demographics, including fracture classification and mode of injury were obtained. Parameters measured at follow-up included ankle range of motion, maintenance of reduction and fibular length, and AOFAS ankle outcome score. The patient cohort showed a typical bimodal distribution of age. Age over 65 years was associated with a poorer outcome. Five patients had ankle fracture-dislocations, which was a factor for a poorer outcome. Nine patients had fibular plate fixation in addition to syndesmosis fixation, whilst three patients with Maisonneuve injuries had syndesmosis fixation only. There were no major complications, loss of reduction, wound problems, implant loosening or osteolysis. Ankle dorsiflexion was not restricted and mean total ankle range of motion was comparable to the uninjured side. No patient required secondary surgery for any reason, including hardware removal. Arthrex TightRope fixation is a simple, safe and effective method of ankle syndesmosis fixation, which allows physiological micromotion. Fixation in plantarflexion provides optimum syndesmosis compression for reduction, and does not compromise ankle range of motion. The Arthrex Ankle Syndesmosis TightRope may become the treatment of choice in Weber C ankle fractures

Introduction. In recent years, there has been a growing interest in bioresorbable metals. Orthopaedic components made from these materials do not require removal by secondary surgery, and offer superior load bearing capability compared to the existing biodegradable polymers. Research on bioresorbable metals have largely focused on alloys based on a subset of the Mg-Zn-Ca ternary system [1, 2], which are pre-existing elements inside the human body. Cytocompatibility assessments of these alloys have reported no signs of inflammation or adverse cellular reactions [2-4]. Rather than designing for longevity, bioresorbable metals rely on their tendency to corrode in a controlled manner. Hence, controlling their corrosion rates is of utmost importance. In the present work, we have explored the effect of compositional variation on the properties of the Mg-Zn-Ca amorphous metals. Subsequent characterisations are performed to assess their suitability as a bioresorbable material. Materials and Methods. A mixture of pure elements and master alloys, namely magnesium, zinc, calcium, and Mg-Ca master alloy, were melted in an induction furnace, followed by injection casting to produce the amorphous metallic samples. Pure magnesium (crystalline) was also used in the subsequent characterisation tests for comparison. The thermophysical properties of the as-cast amorphous metals were characterized using x-ray diffraction (XRD) and differential scanning calorimetry (DSC). The biocorrosion performance was assessed by a combination of immersion, potentiodynamic polarisation (PDP) and hydrogen evolution studies. These tests were conducted in cell media, with a sodium bicarbonate buffer, at 37°C and pH 7.4 in a humidified CO. 2. atmosphere. Results and Discussion. A range of amorphous metal compositions, from Mg-rich to Ca-rich, were successfully produced. XRD confirmed that the alloys were amorphous. Subsequent characterisation tests revealed that minor alterations in composition were not detrimental to thermophysical properties; however, the critical casting size and corrosion rates were much more sensitive to alloy chemistry. In comparison, the Mg-rich alloys have superior corrosion resistance, whereas the Ca-rich alloys have improved thermophysical properties, thereby allowing them to undertake more complex thermoplastic forming processes. Conclusion. We have successfully produced amorphous metals with a range of corrosion resistance and thermophysical properties. The combination of biocompatible elements, superior corrosion resistance and reduced hydrogen evolution, make these amorphous metals more suitable for use as bioresorbable orthopaedic components than their crystalline counterparts. Acknowledgements. The authors would like to thank the Australian Research Council (ARC) for partial funding of this work via the ARC Centre of Excellence for Design in Light Metals (CE0561574)

Introduction:. Two fixed bearing options exist for tibial resurfacing when performing unicompartmental knee arthroplasty (UKA). Inlay components are polyethylene-only implants inserted into a carved pocket on the tibial surface, relying upon the subchondral bone to support the implant. Onlay components have a metal base plate and are placed on top of a flat tibial cut, supported by a rim of cortical bone. To our knowledge, there is no published report that compares the clinical outcomes of these two implants using a robotically controlled surgical technique. We performed a retrospective review of a single surgeon's experience with Inlay versus Onlay components, using a robotic-guided protocol. Methods:. All surgeries were performed using the same planning software and robotic guidance for execution of the surgical plan (Mako Surgical, Fort Lauderdale, FL). The senior surgeon's prospective database was reviewed to identify patients with 1) medial-sided UKA and 2) at least two years of clinical follow up. Eighty-six patients met these inclusion/exclusion criteria: 41 Inlays and 45 Onlays. Five patients underwent a secondary or revision procedure during the follow up period and were considered separately. Our primary outcome was the WOMAC score, subcategorized by the Pain, Stiffness, and Function sub-scores. The secondary outcome was need for secondary surgery. Continuous variables were analyzed using the two-tailed Student's t-test; categorical variables were analyzed using Fisher's exact test. Results:. Average follow up was 2.7 years and 2.3 years in the Inlay and Onlay groups, respectively. The post-op WOMAC Pain score was 3.1 for Inlays and 1.6 for Onlays (p = 0.03). The post-op Stiffness score was 1.8 for Inlays and 1.4 for Onlays (p = 0.19). The post-op Function score was 10.3 for Inlays and 6.7 for Onlays (p = 0.12). We identified a subgroup of 51 patients (23 Inlay, 28 Onlay) for whom there was both pre- and post-op WOMAC data available. There were no differences in the pre-op Pain, Stiffness, or Function scores between groups. In this subgroup, the Pain score improved from 8.3 to 4.0 for Inlays, versus an improvement from 9.2 to 1.7 for Onlays (p = 0.01). The Stiffness score improved from 3.9 to 2.2 for Inlays, versus an improvement from 4.3 to 1.5 for Onlays (p = 0.08). The Function score improved from 27.5 to 12.5 for Inlays, versus an improvement from 32.1 to 7.3 for Onlays (p = 0.03). When all 86 patients in the study were considered, 4/41 Inlays (9.8%) underwent a secondary procedure during the two year follow up (two conversions to total knee replacement [TKR]; one conversion to an Onlay component; one internal fixation of a subchondral compression fracture). In the Onlay group, 1/45 patients (2.2%) underwent a secondary procedure, a conversion to TKR (p = 0.20). Discussion:. Onlay UKR demonstrated improved pain relief compared to inlay UKR at two years of follow up. There was a trend toward improved function and fewer secondary surgical procedures with the onlay device. We advise surgeons performing UKA to preferentially use Onlay implants in order to maximize clinical benefit