Abstract
Purpose
Factors that contribute to early and late re-operation after cruciate reconstruction (CR) have not been evaluated on a population level in a public health system. After surgery patients are at risk for knee stiffness, infection or early graft failure prompting revision. Long-term, ipsilateral revision CR, contralateral CR and potentially even joint replacement may occur. Population research in total joint replacement surgery has demonstrated an inverse relationship between complication/failure rates and surgeon procedural volume. We hypothesized that in Ontario, younger patient age and lower surgeon volume would increase the risk of short and long-term re-operation after CR.
Method
Billing, procedural and diagnostic coding from administrative databases (Ontario Health Insurance Plan, Canadian Institutes of Health Research) were accessed through the Institute for Clinical Evaluative Sciences to develop the cohort of all Ontario residents aged 14 to 60 who underwent anterior or posterior CR from July 1992 to April 2008. Logistic regression analysis was used to calculate the odds ratio for patient (age, gender, comorbidity, income, concurrent knee surgery) and provider (surgeon volume, teaching hospital status) factors for having a surgical washout of the knee, manipulation for stiffness or repeat of the index event within six months. A cox proportional hazards survivorship model was used to calculate the hazard ratio of the same covariates for repeat CR and partial/total knee arthroplasty from inception until end of 2009.
Results
The cohort identified 34,735 CR patients with a median age 28 yrs (IQR 20–36) and 65% male. Re-operation for infection was 0.2% and stiffness 0.5%. The long-term rate of any repeat CR was 7.7% after a mean 4.23.4 years. Female gender (OR=2.8, p<0.0001), overnight hospital stay (OR=2.1, p=0.0005), meniscal repair with CR (OR=1.9, p=0.008) and surgeon volume of 0–12 CR/yr (OR=4.0, p=0.0006), significantly increased the odds of re-operation for stiffness. The odds of re-operation for infection were significantly increased for surgeons performing 0–12 CR/yr (OR=3.8, p=0.007), and for CR performed at a teaching hospital (OR=2.3, p=0.002). Repeat CR was not influenced by surgeon volume at any time-point. Survival analysis demonstrated a long-term repeat CR rate of 13% (HR=1.8, p<0.0001) for age 14–19 yrs compared to the mean cohort age. Late partial or total knee replacement occurred in 0.75% of patients, with increased risk found for patients >30 years (HR=2.5, p=0.002), or who had concurrent surgery for an osteochondral lesion at the index CR (HR=2.3, p=0.001).
Conclusion
Although this data is limited by the ability to distinguish between anterior or posterior and revision or contralateral CR, we have demonstrated that lower volume surgeons have higher complication rates (stiffness, infection) after CR surgery. We have also identified at-risk groups, such as females for stiffness post-CR and osteochondral injury + CR for eventual knee replacement.
