A new approach to differentiate those with anterior cruciate ligament deficient knees into copers and non-copers

1 Introduction

Anterior cruciate ligament (ACL) rupture generally leads to the knee joint instability which in turn may affects individuals’ function during sport and daily activities [1]. Previous studies demonstrated that knee function varies significantly between subjects with ACL rupture [1, 2]. Therefore, patients with ACL injury are usually categorized into two groups, based on the ability to compensate for their injury, as copers or non-copers. Copers are capable of resuming most of their pre-injury activities without considerable functional limitations. These patients are able to maintain stability of their affected knee joint during high level physical activity for many years. Conversely, non-copers experience knee instability even during routine daily activities [3]. Since differentiating copers from non-copers is mandatory to plan a treatment program for ACL deficient (ACLD) individuals, employing a reliable method to subdivide these patients into two separate groups is of utmost importance [4, 5]. Although the decision to prescribe reconstruction surgery depends on many factors [6], the ability to cope with the injury is one of the important predictors. Therefore, being aware of such, not only accelerates reconstruction surgery for non-copers and prevents wasting of time for competitive athletes, it also will prevent unnecessary surgery for potential copers. Furthermore, inability to clearly differentiate copers from non-copers in studies verifying ACLD subjects, may lead to reporting inconsistent movement patterns [7]. Numerous methods have been utilized in the literature to sub-grouping the ACLD subjects as copers or non-copers [4, 9]. Lysholm score, Knee outcome survey sport/ADL, global rating of function, number of giving way, international knee documentation committee (IKDC), quadriceps index and various hopping tests, are methods employed in the literature. However, many suggested methods usually contain hop test which may be a detrimental activity to the ACLD subjects due to the considerable stress imposed on their affected knee [10, 11]. subjects with ACLD knee generally may refuse to perform hop test, possibly to protect their affected knee from being further injured [10]. Furthermore, previous studies showed that the present screening tests may not be promising in categorizing ACLD subjects into copers and non-copers [12]. Utilizing questionnaires is a safe and time conserving tool in assessment and sub-grouping of ACLD patients. The knee injury and osteoarthritis outcome score (KOOS) is a reliable and valid tool which has been widely utilized for ACLD subjects [13]. The aim of the present study was to verify whether various subscales of the KOOS could be considered as possible predictive values in order to differentiate copers from non-copers. Subscales of the KOOS may be grouped together with other criteria, such as global rating of function and number of giving way to hopefully provide a fast, safe and reliable method to discriminate patient who compensate well for their pathology and non-copers.

2 Materials and methods

Thirty-four male participants were recruited from a surgery waiting list of patients whose injury occurred 6 to 12 months earlier. A convenient sample size was utilized in this study. Recruitment period was around 12 months. Inclusion criteria were: (a) non-operated, chronic, right-sided, and complete ACL rupture with or without meniscal injury, diagnosed by MRI and clinical examination; (b) pain less than grade three according to visual analogue scale (0–10) at the time of examination and (c) negative history of any other musculoskeletal and neurological injury. Patients grouped as non-copers if they had at least one of the following criteria:

Experienced two or more episodes of giving way since injury,

Their global rating of function rated less than 60% of that of pre-injury level, and

Acquire less than 60% of that of the intact side in six meters timed hop test,

Otherwise, they were categorized as copers. These criteria were considered as a reference standard to differentiate copers from non-copers and as a dependent variable to evaluate other possible benchmarks for sub-grouping the ACLD individuals. All participants signed an informed consent form before participating in the study. This study was approved by the Ethics Committee of Shiraz University of Medical Science (reference number: CT-91-6336). The KOOS questionnaire was utilized to evaluate functional status in ACLD individuals over the week prior to the testing session (Table 1). This questionnaire includes five subscales, namely: symptom, pain, daily activity, sport/recreation and quality of life, each range between 0–100, with lower score represents more disability. Scoring was performed using established method for the KOOS questionnaire [14]. Each subscale of the KOOS evaluated separately in the present study. Receiver operating characteristics (ROC) analysis is a useful means of evaluating a statistical model which subdivides subjects into two categories, e.g. copers or non-copers. This method was constructed for all subscales of the KOOS questionnaire as independent variables. The cut-off values for differentiating copers from non-copers and maximum sensitivity and specificity of the five subscales of the KOOS were estimated using the ROC curve. This curve is a plot of sensitivity against 1-specificity for different cut-off values. ROC curve also provide a common scale to compare different predictors with different units. The accuracy of ROC curve can be verified by measuring the area under the curve (AUC). A value closer to one indicates that the predictor could reliably distinguish among ACLD subjects with different functional capability. An area of 0.5 represents a worthless test. In other word, the closer the curve to the left and top border, the more accurate the test, the closer the curve to the 45 degree diagonal of the ROC space, the less accurate the test.

3 Results

In the present study, time from injury was in a narrow range, 7.50±2.91 months, resulting in more homogenous ACLD subjects. The participants’ demographic data have been summarized in Table 1. Sub-grouping of the ACLD individuals based on our reference standard demonstrated that 5 (14.7%) patients categorized as copers and 29 patients (85.3%) grouped as non-copers. All subscales of the KOOS demonstrated as a predictor for categorizing ACLD subjects. ROC curve showing the accuracy of the subscales in sub-grouping of the ACLD patients are presented in (Fig. 1A–E). The AUC was 0.83 and the cut-off value was 82.14 for symptom subscale (Fig. 1A). The AUC and the cut-off value were 0.87 and 75.00 for pain subscale respectively, (Fig. 1B). Figure 1C represents ROC curve for daily activity. The AUC was 0.81 and the cut-off value was 72.05. The AUC was 0.91 and the cut-off value was 45.00 for sport/recreation subscale (Fig. 1D). The AUC and the cut-off value were 0.92 and 25.00 for quality of life respectively (Fig. 1E).

4 Discussion

The aim of the present study was to examine whether subscales of the KOOS questionnaire could be used as a predictive value to differentiate copers from non-copers. This questionnaire has been translated and culturally adapted into different languages and is increasingly being utilized for patients undergoing ACL surgery and rehabilitation [15].

The results demonstrated that all subscales of the KOOS questionnaire could be considered as a predictive value in differentiating copers from non-copers. Among these subscales, sport/recreation and quality of life showed the most accuracy with AUC of 0.91 and 0.92, respectively. Pain, symptom and daily activity with AUC of 0.87, 0.83 and 0.81 ranked third to fifth, respectively. This amount of accuracy represents good discriminative ability of the KOOS subscales in differentiating ACLD subjects. In order to offer an effective and timely management of ACLD subjects, it is imperative to know a cut-off value based on which copers can be differentiated from non-copers. These cut-off values are summarized in Table 2. Although hop test is a reliable and valid tool to examine functional status in ACLD subjects and is an impeccable test in differentiating copers from non-copers, it is detrimental to the injured knee of the ACLD subjects. Many subjects with ACLD knee may do not do their best during hopping test in order to prevent further injury by overloading their injured knee [16]. Furthermore, prescribing a safe hop test for ACLD subjects requires some prerequisites such as having pain free knee joint without effusion and a complete range of motion and the ability to hop on the involved side, which are sometimes absent in injured athlete. Decision as to whether a conservative treatment is appropriate for athletes who are in mid-season should usually be made promptly. Considering the aforementioned criteria, the majority of ACLD subjects would be categorized as non-copers, unless they acquire more than 60% of that of uninvolved extremity in hop test. Moreover, evidence supports that the present tests utilized to sub-grouping these patients generally are unable to exactly divide patients into copers and non-copers [12]. This, not only necessitates applying other potential criteria to classify ACLD subjects, it may also introduce a measurement error due to using imperfect standard for comparison. The fact that our subjects were in the chronic stage and there was no contraindication and they were not unwilling to do hop test, we used hop test to differentiate copers from non-copers. In this study, all subscales of the KOOS represented as potential predictive values which may be used in lieu of hop tests in differentiating copers from non-copers. As noted earlier, the AUC for all aforementioned subscales were high (0.81–0.92), denoting high discriminative ability of the KOOS subscales. The results also indicated that five ACLD out of 34 subjects diagnosed as copers and 29 classified as non-copers, denoting that nearly 14 percent of ACLD individuals were copers which is consistent with previous research [8]. There has been some attempts in order to replace hop with another functional test such as lunge test to distinguish copers from non-copers [17]. Applying a faster and safer alternative method like questionnaire, as suggested by this research, would be probably advantageous.

As denoted in the literature, no single tool is capable of determining the functional status in ACLD subjects [4]. Hence, we suggest that predictive values, as presented by subscales of the KOOS, may be used as an adjunct to other discriminative tools in discriminating ACLD subjects with different functional capabilities.

Conflict of interest

None.