How Applicable Is Published Minimum Clinically Important Difference Literature to a Single-Center Anterior Cruciate Ligament Reconstruction Population?

1. INTRODUCTION

The field of orthopedic surgery continues to transition towards using patient-reported outcome measures (PROMs) to measure treatment success. Anterior cruciate ligament reconstruction (ACLR) is a highly prevalent procedure that provides high patient satisfaction according to PROMs such as the Knee Injury and Osteoarthritis Outcome Score (KOOS) and the Single Assessment Numerical Evaluation (SANE).^1,2 The SANE score has been correlated to multiple other knee surveys, demonstrating its ability to reflect knee symptoms.^2–4 The KOOS score has been validated for ACLR.⁵

Metrics such as the minimum clinically important difference (MCID) can help define treatment success. MCID is defined as the smallest difference that patients consider significant or perceptible and therefore is used as a goal for clinical improvement. Winterstein et al. published an MCID for the SANE for knee injuries as 19 for a 12-month follow-up in active, healthy female patients.⁶ Spindler et al. concluded a clinically meaningful effect for the KOOS of 8 points based on responsiveness, specifically in patients after ACLR, based on a study by Roos et al.^5,7

To the author’s knowledge, there are no studies comparing patients who reach MCID for SANE and KOOS to patients who do not reach MCID. This comparison has the potential to provide surgeons with the information necessary to improve the number of patients that reach MCID for SANE and KOOS scores after ACLR. The purpose of this study was to identify patients who reached the literature-reported MCID values for KOOS and SANE. The secondary aim was to identify if there were any differences in demographic, injury, and treatment characteristics among patients who reached MCID for KOOS and/or SANE and those who did not achieve any MCID.

2. MATERIAL AND METHODS

3. RESULTS

A total of 322 patients were included [Figure 1]. Baseline and treatment characteristics are listed in Table 1. Most patients were female (62.1%) and had an average age of 28 (range 10- 66). Average BMI was 25.3 (range 15.9 – 42.1). Most patients had an ASA score of 1 (84.1%). A total of 20.5% of patients had a mental health diagnosis of anxiety and/or depression.

Close

Figure 1: CONSORT flow diagram showing cohort selection. CONSORT: Consolidated Standards of Reporting Trials. ACLR: Anterior cruciate ligament reconstruction.

Export to PPT

Average operative time was 108.3 minutes (range 44 – 256 minutes). Most patients received a bone-tendon-bone autograft (49.1%), vs 24% that received a hamstring autograft and 27% that received an allograft. Acute injury was a more common presentation seen in 93.2% percent (n=300) versus 6.8% (n=22) of patients presenting with a chronic injury. There were 143 patients requiring a meniscal procedure (44.4%), with 40 knees undergoing meniscal repair, 194 undergoing meniscal debridement, and 9 undergoing both repair and debridement procedures. The percentage of patients meeting MCID at two years for KOOS and SANE scores is depicted in Figure 2. The proportion of patients who achieved MCID at two years for KOOS was 73.6% (237 patients) and for SANE was 45.7% (147 patients) [Table 2]. The majority of patients achieving SANE MCID achieved KOOS MCID (89.8%), while just 55.7% of patients achieving KOOS MCID achieved SANE MCID. 41.0% of patients achieved MCID for both KOOS and SANE, while 21.7% of patients did not achieve MCID for any scale.

Close

Figure 2: Pie chart demonstrates the percentage of patients that met the MCID for KOOS, SANE, both KOOS and SANE, or didn’t meet any MCID. MCID: Minimum clinically important difference, KOOS: Knee injury and osteoarthritis outcome score, SANE: Single assessment numerical evaluation

Export to PPT

Compared to patients who achieved MCID in at least one PROM, there was a significantly higher proportion of patients who underwent ACLR for chronic injuries among those who did not meet MCID (11.4% vs 5.6%, p=0.004). Graft choice did not statistically differ among patients who achieved MCID and those who did not.

4. DISCUSSION

MCID is a metric that can help better define patient improvement for PRO data. PRO tools such as the SANE score have been correlated to multiple other knee surveys, demonstrating their ability to reflect knee symptoms, and the KOOS score has been validated for ACLR.^2,3,5 MCID is a metric that has the potential to provide more clinically relevant conclusions as opposed to statistical significance and, therefore, may be a better way to report patients’ outcomes.

This study used the MCID of the SANE score calculated by Winterstein et al.⁶ This was calculated for young active female patients, not patients with ACL injuries specifically. Additionally, this was derived from the IKDC MCID calculated by Greco et al. for patients with articular cartilage defects.⁸ This demonstrates multiple inconsistencies in the patient population for calculations of the MCID for the SANE score for patients with ACL tears and reconstructions. This study used the MCID of the KOOS score used by Spindler et al. and determined by Roos et al. as 8.^5,7 This was calculated by Roos et al. by their post-operative ACLR patients demonstrating the KOOS score improves by more than or equal to 8 points at the 6 months mark and less than or equal to 7 points at the 3 months mark.⁵ There are limitations in the calculation, including that it was obtained from a single study’s patient population from the 1990s and an MCID based solely on changes over a certain time period.

Multiple methods of calculating MCID exist, such as the distribution-based method and the anchor-based method. Nwachukwu et al. calculated MCID for IKDC, Lysholm scale, and 12-item short form physical component score and mental component score (SF-12 PCS and MCS) using a distribution-based method.¹ Greco et al. calculated the MCID of IKDC on an anchor-based method, and this was used as an anchor for Winterstein et al. to calculate the MCID of SANE scores.^5,8 McCreary et al. looked at MCID for the patient-rated wrist evaluation, which is a PRO tool used to evaluate patients with distal radius fractures.⁹ They concluded that an anchor question should be included with PRO data collection to calculate MCID. It is possible that MCID may be more useful and accurate if an anchor-based method with an anchor question is used. Other methods have been used. For example, Roos et al. determined an MCID of KOOS based on time points and patient activity levels at certain changes in KOOS score.⁵

McCreary et al. determined that the MCID is altered by many factors, including assessment time points and analytical method.⁹ Greco et al determined that MCID changes with time points of follow-up between 6 and 12 months of follow-up.⁸ Additionally, the method of determining the MCID has a large effect on the calculated value of the MCID. Furthermore, it is likely that diagnosis, treatment, and patient characteristics can alter MCID.

There are several studies that evaluated predictive variables for obtaining MCID in different PRO measures after ACLR. Nwachukwu et al. concluded that higher SF-12 MCS scores predicted MCID achievement on the other knee-specific questionnaires.¹⁰ Kunze et al. published a machine learning algorithm to predict MCID for the IKDC score after ACLR. They found that, based on 6 machine learning algorithms, the femoral tunnel fixation method, range of motion, MCL integrity, BMI, and prior contralateral surgery significantly affected whether patients reached MCID.¹¹ The present study demonstrated that MCID was less likely to be reached in patients with chronic ACL tears than in those with acute tears. It should be noted that several potentially influential factors were not evaluated in this study, including rehabilitation adherence, activity level, and socioeconomic factors, which may influence the likelihood of achieving MCID. (Jones et al., OJSM, 2019).¹²

5. LIMITATIONS

This study has several limitations. First, although the PROM data were collected prospectively, the study design was retrospective and therefore subject to selection bias. Second, the investigation was performed at a single ambulatory surgical center, which may limit the generalizability of the findings to other practice settings, institutions, and patient populations. Third, the MCID thresholds used in this study were derived from prior literature and were not specifically established for the present ACLR cohort. In particular, the SANE MCID was taken from a cohort of young active female patients, which may affect the accuracy of classifying clinically meaningful improvement in patients undergoing ACLR.

6. CONCLUSION

MCID achievement rates differed substantially between KOOS and SANE following ACLR. Patients with chronic ACL injuries were less likely to achieve clinically meaningful improvement, suggesting timing from injury may influence postoperative outcomes.