Abstract
Objectives:
To evaluate cost-effectiveness of anterior cruciate ligament reconstruction (ACLR) combined with lateral extra-articular tenodesis (LET) compared to standard ACLR alone from a healthcare payer perspective in younger patients at high risk of ACLR graft rupture.
Methods:
We conducted a pragmatic, multicenter randomized clinical trial (STABILITY Trial), involving participants aged 14 to 25 years with ACL deficiency who were randomly assigned to undergo either ACLR with or without LET. We evaluated participants at baseline and at 3, 6, 12, and 24 months after ACLR. We measured costs from a healthcare payer perspective within Ontario’s (Canada) publicly funded healthcare system, including direct costs associated with healthcare resource utilization. This included hospital costs, procedure costs, clinician and provider time and billing fees, tests, and medications for individuals with disabilities or aged ≥65 years. We included costs related to the primary ACLR intervention and any additional surgeries due to adverse events (e.g., revision ACLR from graft rupture), which were documented at standard post-ACLR surgical follow-up visits. We measured effectiveness using three outcomes: 1) quality-adjusted life years (QALYs) over 24 months, 2) the Marx Activity Rating Scale at 24 months (adjusted for the baseline score), and 3) the number of participants who returned to a level of sport equal to or greater than their baseline level at 24 months after surgery.
We calculated mean cost and effectiveness and between group differences for each group using generalized linear models and linear regression models, respectively, and applied non-parametric bootstrapping with 1000 replications to improve robustness of 95% confidence intervals (CI) estimates. We also used Chi-squared tests to evaluate between-group differences in return-to-sport. To evaluate cost-effectiveness, we calculated incremental cost-utility and cost-effectiveness ratios (ICUR and ICER). We also completed net benefit regression models to evaluate cost-effectiveness for the 24-month Marx Activity Rating Scale, reported as the incremental net benefit (INB), while controlling for age, sex, height, mass, and baseline Marx score. We used cost-effectiveness planes (CEP) and a cost-effectiveness acceptability curve (CEAC) over a range of willingness-to-pay (WTP) values ($0 to $1,000) to quantify uncertainty and probability of cost-effectiveness. WTP represents the maximum acceptable amount one is willing to pay to achieve one additional unit improvement in effect.
Results:
We randomly assigned a total of 618 participants to either undergo ACLR alone (n=312) or ACLR with LET (n=306). Both groups shared similar baseline characteristics and return-to-sport time (median = 11 months [interquartile range = 8 to 17]) (Table 1). The total number of reoperations was 56 participants (18%) in the ACLR alone group and 46 participants (15%) in the ACLR with LET group.
Comparing overall costs from a healthcare payer perspective, ACLR with LET was slightly more expensive than ACLR alone, with a small mean difference of $145.18 ($95% CI, -128.17 to 418.53) (Table 2). The distinction in QALYs over 24 months between the groups was marginal, with a mean difference of 0.004 QALYs (95% CI, -0.01 to 0.01) (Table 2). ACLR with LET yielded significantly higher 24-month Marx scores, showing a mean difference of 0.72 points (95% CI, 0.07 to 1.37) when adjusted for baseline scores (Table 2). Return-to-sport outcomes did not exhibit statistically significant differences (Table 2).
We did not compute ICUR or ICER for QALYs and return-to-sport outcomes due to lower costs of ACLR alone and negligible differences in effectiveness between groups, indicating the dominance of the ACLR alone intervention (Table 2). Bootstrapped estimates for QALYs on the CEP were spread across all quadrants, with 73% concentrated in the northwest quadrant. This suggests ACLR with LET may not be cost-effective due to higher costs and lower effectiveness (Figure 1A). However, there was a 26% probability indicating ACLR with LET could also be cost-effective (Figure 1A).
The calculated ICER for the 24-month Marx Activity Rating Scale was $201.64 per 1-point increase with ACLR and LET, suggesting it is cost-effective (Table 2). The majority of bootstrapped ICER estimates fell within the northeast and southeast (96%) quadrants on the CEP, indicating ACLR with LET is highly likely to be cost-effective (Figure 1B). For the net benefit regression, the INB became positive at a WTP of $300 for a 1-point improvement in 24-month Marx Activity Rating Scale, indicating ACLR with LET was cost-effective beyond this threshold (Figure 2). According to the CEAC, the probability of ACLR with LET being cost-effective was around 11% at a WTP of $0 per 1-point improvement in the Marx Activity Rating Scale at 24 months and the probability gradually increased as WTP value increased, plateauing at around 88% with a WTP of $1,000 (Figure 2).
Conclusions:
The assessment of cost-effectiveness of ACLR with LET compared to standard ACLR alone provided mixed results. While ACLR with LET was shown to be cost-effective when evaluating the 24-month Marx Activity Rating Scale, there was uncertainty in interpreting cost-effectiveness for QALYs and return to sport, with minimal differences in effect between groups. Therefore, patient selection and individual characteristics should be considered when selecting the most appropriate approach for each patient.
