Evaluating the predictive power for second anterior cruciate ligament injury of patient-reported outcomes after anterior cruciate ligament reconstruction

An anterior cruciate ligament (ACL) injury is a common injury in the active population, with an annual incidence of 68.6 per 100,000 person-years [19]. An ACL injury can be career-ending, especially for patients active at lower sports levels [20] and is commonly treated with ACL reconstruction.

Nearly two thirds of patients treated with ACL reconstruction report improvements in patient-reported outcome (PRO) measures during the recovery period after ACL reconstruction, but improvements in knee function and psychological status may not be interdependent [14]. While patient-reported outcomes do not differ across graft types, normal knee kinematics is often not fully restored after ACL reconstruction [18]. Bone-patella-tendon-bone (BPTB) and quadriceps tendon (QT) autografts have shown lower second ACL injury rates compared to hamstring tendon (HT) or allografts [18]. Rehabilitation should not only focus on functional recovery but also address mental recovery, psychological factors, and negative psychological responses during rehabilitation [3]. Mental recovery, alongside individualized rehabilitation, strength and neuromuscular control and sport-specific training, are the pillars of a comprehensive, personalized approach to rehabilitation which aims to improve outcomes and reduce the risk of second ACL injury [8].

In the realm of sports medicine research, ACL injuries are among the most extensively studied, and several national registries prospectively collect patient data to monitor patient's well-being and the impact of injury or treatment on the individual [4, 6, 10].

Standardized PROs have been established to gauge the perceived impact of treatments or injuries. PROs can be longitudinally compared for the same patient or contrasted with a cohort suffering from similar conditions to measure changes over time [2]. As of 2018, there were 24 different PROs available to evaluate patients from an ACL injury [7]. Among these PROs, the International Knee Documentation Committee (IKDC) form was the most commonly used, along with the Knee Injury and Osteoarthritis Outcome Score (KOOS) [7]. The ACL-Return to Sport after Injury Scale (ACL-RSI) was identified as the PRO with the highest methodological quality to evaluate patients with an ACL injury [7].

Mental recovery, individualized rehabilitation, strength and neuromuscular control, and sport-specific training are pillars of a comprehensive approach aimed at reducing the risk of second ACL injury. As such, PROs fill an important function, as PROs can be used to monitor the mental recovery of patients. Answers to PROs have been linked to a second ACL injury upon return to sport (RTS) [11, 15]; however, both high and low results on PROs have shown an association with second ACL injury [5]. Ideally, for each PRO, one cut-off could be used to identify whether patients ‘pass’ (patients can be cleared to RTS since the risk for second injury is low) or ‘fail’ (patients should be advised against RTS since the risk for second injury is high) the PRO included in RTS testing.

Our research team recently conducted a study [16], where PRO results were analyzed for 641 patients who underwent ACL reconstruction and returned to knee-strenuous sports, rated at level 6 or higher on the Tegner Activity Scale, which includes sports like snowboarding, floorball or baseball. Patients' PRO scores at the time in which patients rated to have returned to knee-strenuous sport were extracted from a rehabilitation-specific registry. In this registry, patients are asked to answer PROs and perform muscle function tests at regular intervals after ACL reconstruction.

The PROs examined included the KOOS subscales for Sports & Recreation and Quality of Life (QoL), the ACL-RSI and the Knee Self-Efficacy Scale (K-SES₁₈) [1]. Out of the included patients, 10% (n = 64) experienced a second ACL injury within 2 years of returning to sport. Cut-off values for each PRO that best distinguished between patients who suffered a second ACL injury and those who did not were determined. To assess the ability of these PROs to differentiate between the two groups, a receiver operating characteristics (ROC) analysis was performed. ROC analysis is a graphical method used to display the discriminatory accuracy of markers to distinguish between patients affected by an outcome, that is, a second ACL injury within 2 years after RTS, and patients not affected by the outcome of interest.

Results revealed that the cut-offs for the KOOS subscales Sports and QoL, the ACL-RSI and the K-SES₁₈ could not effectively distinguish between patients who experienced a second ACL injury and patients who did not. In other words, these cut-offs showed no discriminative capacity, as indicated in Table 1. In summary, approximately half of the patients who suffered a second ACL injury scored above the calculated cut-off for each of the PROs, while the other half scored below, as depicted in Figure 1. Moreover, when a patient scored above a calculated cut-off, there was approximately a 50% chance of correctly predicting whether the patient would go on to experience a second ACL injury. This suggests that the cut-offs had no better predictive ability than a coin flip, which is insufficient for clinicians who treat patients after ACL reconstruction.

Data for the publication were extracted from a rehabilitation-specific registry, which poses some limitations. For instance, no information about the trauma that leads to the ACL injury is registered.

One plausible explanation for the lack of predictive value could be the questionable psychometric properties exhibited by most PROs used in patients who suffer an ACL injury. The most essential psychometric property for a questionnaire is content validity [9], which refers to what the questionnaire actually measures. Unfortunately, content validity is notably lacking or questionable for most questionnaires used in patients who suffer from ACL injuries. Consequently, it is inappropriate for clinicians to use PROs to evaluate patients after an ACL injury because we do not know what PROs are actually evaluating [7]. Another possible reason for the lack of predictive value is that these PROs were not originally designed to forecast second ACL injuries. Additionally, psychological variables are subject to high variability. Individuals might be very confident about their ability to run on uneven ground but lose all confidence after attempting the activity. PROs often encompass various constructs, such as fear, confidence, QoL or self-efficacy, all categorized as ‘psychological aspects’. Therefore, the cross-sectional use of PROs does not reflect the variability in the outcome measured, where psychological response to a physical task could vary from very positive to very negative while performing the physical task itself.

One important point is that different activity levels and different age groups exhibit diverse responses to ACL injuries, reconstructions, and risk of second ACL injury [12, 23]. High-level athletes face intense physical demands and mental pressures, compared with recreational athletes. Further, older and less active patients may prioritize QoL over peak performance. Understanding these nuances is essential for personalized rehabilitation and evaluation. More active or younger patients might have different scores on PROs that are more predictive of a second ACL injury. For instance, elite athletes might be more susceptible to fear of re-injury, making scores on fear-related PROs potentially predictive of a second ACL injury. Conversely, non-elite athletes or older patients might prioritize QoL, with scores on QoL-related PROs being more predictive for these groups. This variability underscores the need for tailored rehabilitation programmes and evaluations that consider individual psychological and physical profiles. It is crucial to address both physical and mental aspects of recovery to improve outcomes and reduce the risk of second ACL injury, especially considering the varying needs and goals across different patient demographics. In fact, younger athletes, especially adolescents, show a dramatic increase in ACL reconstructions and recurrent injuries, suggesting a need for improved prevention strategies [21]. Additionally, the variability in return-to-sport rates and the durability of high-level sports careers post-ACL reconstruction highlight the need for tailored rehabilitation programmes [21]. Only a fraction of athletes returns to their pre-injury performance levels, with significant differences across sports. For instance, alpine skiing may be more forgiving than sports involving cutting and pivoting, such as basketball and football. This perspective underscores the necessity for individualized treatment and evaluation plans that consider the unique demands of each sport and athlete [21]. More comprehensive, age and sport-specific studies are essential to enhance injury prevention, treatment, and rehabilitation strategies, ultimately improving outcomes for all athletes.

To evaluate patients when designing a clinical trial or registry, it is crucial to select patient-centred outcome measures. These measures assess outcomes that are important to patients and allow patients to perceive the treatment as effective [2]. Interview studies conducted on patients with ACL injuries have reported that patients experience fear about undergoing the entire process of surgery and rehabilitation again [17], as well as fear of re-injury [13, 22]. However, none of the most widely used PROs in patients with ACL injuries, such as the KOOS, IKDC and ACL-RSI, address the fear of re-injury. Therefore, clinicians are using PROs to measure constructs of questionable importance to patients. As shown in our study [16], the PROs demonstrate no ability to distinguish between patients who experience a second ACL injury and those who do not. More work is needed from clinicians and the research community to develop, or adapt, PROs to reflect and measure factors important to patients. Clinically, the results from our study suggest that the analyzed PROs should not be used to predict a second ACL injury, but should be employed to assess what the PROs are intended to measure at the time of responding to the questionnaire. The questionnaires analyzed, including the KOOS subscales Sports and QoL, the ACL-RSI, and the K-SES₁₈, could serve as foundational material for discussions with patients. During patient-practitioner discussions, clinicians should remember that PRO scores do not possess the ability to predict future events and should be used cautiously in decision-making, such as RTS or discharge from rehabilitation.

Instead of delving into the underlying causes behind the lack of discriminant validity in the PROs used to address second ACL injuries, research efforts should focus on adapting and enhancing the content validity of PROs for outcomes that are genuinely important to patients themselves. As highlighted by Zsidai et al. [24], ‘the identification and universal acceptance of a PRO measure that demonstrates strong evidence for content, reliability, validity and responsiveness should be embraced as the primary outcome measure for all ACL clinical trials and knee ligament registries’. Importantly, the development of such a PRO, whether newly created or an improvement of an existing one, must incorporate patient perspectives to ensure high content validity for outcomes that truly matter to patients.

Funding information is not available.

The authors declare no conflict of interest.

The ethics statement is not available.