Dysglycemia in the post-operative period has been associated with increased rates of infection following total joint arthroplasty. Additionally, patients without clinical diagnosis of diabetes with hemoglobin A1c% values between 5.7% and 6.4% have been shown to be dysglycemic in the peri-operative period. This study examines the influence of post-operative blood glucose management on rates of re-operation for infection in patients undergoing total joint arthroplasty in diabetic, and clinically non-diabetic patients.

We performed a retrospective review of prospectively collected data, for all primary, elective total hip and total knee arthroplasties performed at The Ottawa Hospital between April 2010 and October 2017. Kaplan-Meier survivorship, and mutivarient regression analysis were used to determine predictors of infection based on pre-operative diabetes status, HbA1c%, and post-operative insulin management.

4159 joints met inclusion criteria. Patients with HbA1c 5.7–6.4% had lower rate of revision for infection if they received post-operative blood glucose management (0.80%), vs without glucose management (1.20%). Kaplan-Meier survivorship analysis showed this difference to be insignificant (p=0.23). Patients with DM1 had statistically worse survivorship when compared to other groups (p=0.010).

Patients with undiagnosed perioperative dysglycemia may be at a greater predisposition for developing infection, requiring re-operation, following total joint arthroplasty. This study may be underpowered due to overall low rates of revision. More rigorous peri-operative glucose management strategies may be required for patients, who otherwise receive no glycemic management outside of hospital. This is more pertinent with trends towards decreasing post-operative lengths-of-stay.

The effectiveness of total hip replacement as a surgical intervention has revolutionized the care of degenerative conditions of the hip joint. However, the surgeon is still left with important decisions in regards to how best deliver that care with choice of surgical approach being one of them especially in regards to the short-term clinical outcome. It is however unclear if a particular surgical approach offers a long-term advantage. This study aims to determine the influence of the three main surgical approaches to the hip on patient reported outcomes and quality of life after 5 years post-surgery.

We extracted from our prospective database all the patients who underwent a Total Hip Replacement surgery for osteoarthritis or osteonecrosis between 2008 and 2012 by an anterior, posterior or lateral approach. All the pre-operative and post-operative HOOS (Hip disability and Osteoarthritis Outcome Score) and WOMAC (Western Ontario and McMaster Universities Osteoarthritis Index) scores were noted. Analysis of covariance (ANCOVAs) were used to study the relationship between amount of change in HOOS and WOMAC subscales (dependant variables) and approach used, by also including confounding factors of age, gender, ASA (American Society of Anaesthesiologists) score, Charnley score and Body Mass Index. A total of 1895 patients underwent a primary total hip arthroplasty during the considered period. Among them, 367 had pre-operative and ≥5 years post operative PROM scores (19.47%)

The mean follow-up for the study cohort was 5.3 years (range 5 to 7 years) with, 277 at 5 years, 63 at 6 years, and 27 at 7 years. In the posterior approach group we had 138 patients (37.60%), 104 in the lateral approach (28.34%) and 125 in the anterior approach (34.06%). There were no significant differences between the 3 groups concerning the Charnley classification, BMI, Gender, ASA score, side and pre-operative functional scores. We did not observe any significant difference in the amount of change in HOOS and WOMAC subscales between the 3 groups. There were no differences either in the post-operative scores in ultimate value.

Our monocentric observational study shows that these three approaches provide predictable and comparable outcomes on HRQL and PROMs at long-term follow-up both in terms of final outcome but also in percent improvement. This study has several limitations. We excluded patients who underwent revision surgery leaving the unanswered question of how choice of surgical approach could lead to different revision rates, which have an impact on the functional outcomes. Moreover, even if we controlled for the most important confounders by a multivariate analysis model, there is still some involved cofounders, which could potentially lead to a bias such as smoking, socio-economical status or femoral head diameter. But we do not have any reason to think that these parameters could be unequally distributed between the three groups. Finally, our study cohort represents of 19.47% of the complete cohort. The fact that not all patients have PROM's was pre-determined as eight years ago we instituted that only 1 in 5 patients that returned their pre-operative questionnaire would get their PROM's at follow-up. Despite this, our statistical power was sufficient.