With the rising rates, and associated costs, of total knee arthroplasty (TKA), enhanced clarity regarding patient appropriateness for TKA is warranted. Towards addressing this gap, we elucidated in qualitative research that surgeons and osteoarthritis (OA) patients considered TKA need, readiness/willingness, health status, and expectations of TKA most important in determining patient appropriateness for TKA. The current study evaluated the predictive validity of pre-TKA measures of these appropriateness domains for attainment of a good TKA outcome.

This prospective cohort study recruited knee OA patients aged 30+ years referred for TKA at two hip/knee surgery centers in Alberta, Canada. Those receiving primary, unilateral TKA completed questionnaires pre-TKA assessing TKA need (WOMAC-pain, ICOAP-pain, NRS-pain, KOOS-physical function, Perceived Arthritis Coping Efficacy, prior OA treatment), TKA readiness/willingness (Patient Acceptable Symptom State (PASS), willingness to undergo TKA), health status (PHQ-8, BMI, MSK and non-MSK comorbidities), TKA expectations (HSS KR Expectations survey items) and contextual factors (e.g., age, gender, employment status). One-year post-TKA, we assessed for a ‘good outcome’ (yes/no), defined as improved knee symptoms (OARSI-OMERACT responder criteria) AND overall satisfaction with TKA results. Multiple logistic regression, stepwise variable selection, and best possible subsets regression was used to identify the model with the smallest number of independent variables and greatest discriminant validity for our outcome. Receiver Operating Characteristic (ROC) curves were generated to compare the discriminative ability of each appropriateness domain based on the ‘area under the ROC curve’ (AUC). Multivariable robust Poisson regression was used to assess the relationship of the variables to achievement of a good outcome.

f 1,275 TKA recipients, 1,053 (82.6%) had complete data for analyses (mean age 66.9 years [SD 8.8]; 58.6% female). Mean WOMAC pain and KOOS-PS scores were 11.5/20 (SD 3.5) and 52.8/100 (SD 17.1), respectively. 78.1% (95% CI 75.4–80.5%) achieved a good outcome. Stepwise variable selection identified optimal discrimination was achieved with 13 variables. The three best 13-variable models included measures of TKA need (WOMAC pain, KOOS-PS), readiness/willingness (PASS, TKA willingness), health status (PHQ-8, troublesome hips, contralateral knee, low back), TKA expectations (the importance of improved psychological well-being, ability to go up stairs, kneel, and participate in recreational activities as TKA outcomes), and patient age. Model discrimination was fair for TKA need (AUC 0.68, 95% CI 0.63-0.72), TKA readiness/willingness (AUC 0.61, 95% CI 0.57-0.65), health status (AUC 0.59, 95% CI 0.54-0.63) and TKA expectations (AUC 0.58, 95% CI 0.54-0.62), but the model with all appropriateness variables had good discrimination (AUC 0.72, 95% CI 0.685-0.76). The likelihood of achieving a good outcome was significantly higher for those with greater knee pain, disability, unacceptable knee symptoms, definite willingness to undergo TKA, less depression who considered improved ability to perform recreational activities or climb stairs ‘very important’ TKA outcomes, and lower in those who considered it important that TKA improve psychological wellbeing or ability to kneel.

Beyond surgical need (OA symptoms) and health status, assessment of patients’ readiness and willingness to undergo, and their expectations for, TKA, should be incorporated into assessment of patient appropriateness for surgery.

In a cross sectional cohort study three different metal on metal total hip systems were assessed. Two monoblock acetabular designs; the Durom socket (Zimmer, Warsaw, In) and the Birmingham socket (Smith and Nephew, Memphis, TN), and one modular metal on metal total hip system (Pinnacle, Depuy Orthopaedics, Warsaw, In) in patients who have received these implants in the our region. 56 patients were recruited in total. All patients were assessed clinically, radiologically and biochemically. Whole blood Cobalt, Chromium and titanium levels were tested.

The median head size used in the Birmingham group was 52mm (Range 44mm to 56mm), and in the Durom group, 48mm (Range 42 to 54mm). The median head size used in the modular Pinnacle group was 40mm (Range 36–44mm).

The blood metal ions levels in the larger non modular acetabular sockets were significantly raised compared to the Pinnacle group. For Co 1.95 µg/l and 2.70 µg/l in the Durom and Birmingham groups respectively compared to only 0.52 µg/l in the Pinnacle group (P< 0.001). Mean Cr levels were the same for the two monoblock systems, 1.9 µg/l compared to the Pinnacle sockets 1.2 µg/l (P< 0.001).

Our study clearly demonstrates that there is a significant difference in metal ion levels in patients following a monobloc large head arthroplasty system compared to a smaller modular metal on metal hip arthroplasty. The smaller head size appears to produce less metal ions whilst at the same time a 36mm–44mm head size is large enough to increase hip stability and range of movement as well as decreasing the risk of impingement. In our practice we are no longer using this design and the safest strategy, when considering metal on metal bearings, is to use a modular, smaller head system such as the Pinnacle.

Purpose

To determine whether there is a difference in the metal ion levels among three different metal-on-metal total hip systems: two monoblock large heads and one modular metal-on-metal total hip replacement system in patients who received these implants in our region.