The Delaware-Oslo ACL Cohort treatment algorithm yields superior outcomes to usual care 9–12 years after anterior cruciate ligament reconstruction

Abstract

Purpose:

Patient-reported outcomes were compared between participants who followed the treatment algorithm of the Delaware-Oslo ACL Cohort, consisting of progressive preoperative and postoperative rehabilitation, patient education, clinical testing and shared decision-making about treatment choice, and those who followed usual care 9–12 years after anterior cruciate ligament reconstruction (ACLR).

Methods:

Participants with primary ACLR were included from the Norwegian arm of the Delaware-Oslo ACL Cohort and the Norwegian Knee Ligament Registry (usual care). The Knee Injury and Osteoarthritis Outcome Score (KOOS) subscale scores and the International Knee Documentation Committee-Subjective Knee Form (IKDC-SKF) scores were compared. KOOS scores for the usual care group were converted to IKDC-SKF scores with recently published validated crosswalk. The percentages of participants with scores above predefined thresholds for patient acceptable symptom state (PASS) were also calculated.

Results:

Eighty of 100 (80%) participants from the Delaware-Oslo ACL Cohort and 1588 of 3248 (49%) from the usual care group participated in the follow-up. Participants from the Delaware-Oslo ACL Cohort had higher KOOS subscale (p < 0.001) and IKDC-SKF scores (p < 0.001), and a higher percentage reached PASS (84%–96% vs. 62%–76%, p ≤ 0.002) for KOOS Pain, symptoms, activities of daily living and sports compared to the usual care group. No significant differences were found for KOOS quality of life scores (not significant [n.s.]) or PASS percentages (80% vs. 74%, n.s.).

Conclusion:

Participants with ACLR who followed the Delaware-Oslo ACL Cohort treatment algorithm had reduced knee symptoms, superior function and higher percentages of satisfactory outcomes than participants who followed usual care.

Level of Evidence:

Level II.

INTRODUCTION

Rupture of the anterior cruciate ligament (ACL) is common in the young, active population [1]. While anterior cruciate ligament reconstruction (ACLR) is often recommended to restore knee laxity and improve knee function, many patients still experience poorer knee function and impaired quality of life (QOL) in the long term [2, 3]. There is a need to assess strategies that can lead to improved long-term outcomes after ACLR.

The Delaware-Oslo ACL Cohort treatment algorithm incorporates progressive active preoperative and postoperative rehabilitation, patient education, clinical testing and shared treatment decision-making [4–6]. This clinical treatment algorithm has demonstrated promising short- and medium-term results, with superior knee-related patient-reported outcomes two years after ACLR compared to usual care in Norway and the United States [7, 8]. Further research is required to determine whether this treatment algorithm contributes to successful long-term outcomes.

Whether treatment after ACLR is successful [9] depends on patients’ satisfaction with their outcomes [10, 11]. Patient satisfaction with current knee status often seems more clinically valuable than the reported outcome scores [10]. The patient-acceptable symptoms state (PASS) is a simple and clinically relevant measure that refers to a condition that a patient would deem satisfactory [12]. Clinicians, researchers and policymakers can use valid PASS thresholds for a given patient-reported outcome measure to better understand the clinical meaningfulness of the scores and to identify what constitutes a satisfactory outcome [10, 11].

Therefore, the aim was to compare patient-reported outcomes between participants in the Norwegian arm who had followed the Delaware-Oslo ACL Cohort treatment algorithm and participants in the Norwegian Knee Ligament Registry (NKLR) who had received usual care 9–12 years after ACLR. The scores and the percentages of participants who scored above predefined PASS thresholds were compared. The hypothesis was that participants in the Norwegian arm who had followed the Delaware-Oslo ACL Cohort treatment algorithm would have superior patient-reported outcomes compared to participants who had received usual care long-term after ACLR.

MATERIALS AND METHODS

This is a prospective cohort study of participants who followed either the Delaware-Oslo ACL Cohort treatment algorithm or usual care (NKLR) after ACLR. Approval from the Regional Committee for Medical Research Ethics was obtained (reference number 2018/433), and patients provided informed consent before data collection. The Delaware-Oslo ACL Cohort is a binational prospective cohort study where participants were recruited by the University of Delaware and the Norwegian Sports Medicine Clinic (Nimi) in Oslo (Norwegian arm) [5]. Since participants from the usual care group (NKLR) were from Norway, only the Norwegian arm of the Delaware-Oslo ACL Cohort was included in the present study. The cohort receiving usual care consisted of participants from the NKLR, which prospectively records data on all patients undergoing cruciate ligament surgery in Norway since 2004 [13, 14].

Participants from both cohorts had to be enrolled between 2007 and 2011, be 13–40 years, and have undergone primary unilateral ACLR within 2 years of injury. Exclusion criteria were concomitant knee ligament surgery or cartilage surgery, concomitant posterior cruciate and lateral collateral knee ligament injury, posterolateral corner injuries, previous cruciate ligament injury to the contralateral knee or previous surgery to the index knee. Participants with an ACL injury unrelated to sports activity or an ACL graft other than patella or hamstring tendon autografts were also excluded. Records of participants from the Delaware-Oslo ACL Cohort were removed from the NKLR to avoid duplication.

Rehabilitation of the Delaware-Oslo ACL Cohort

All participants from the Delaware-Oslo ACL Cohort underwent a 5-week (10 sessions) rehabilitation programme as described by Eitzen et al. [4]. The goal was to maximise muscle strength and neuromuscular control before any further decisions about treatment were made. The programme comprised progressive neuromuscular exercises, heavy resistance strength training and plyometrics after initial swelling and range of motion impairments were resolved (mean 62 days after injury). The participants also received education and performed clinical tests before they chose their treatment in a shared decision-making process with their orthopaedic surgeons and physical therapists [5]. Participants chose either ACLR or rehabilitation alone.

Postoperative rehabilitation consisted of three phases and was individually tailored based on concomitant injuries, graft type and clinical knee status [15]. The first phase (acute postoperative phase) aimed to resolve swelling, optimise range of motion and limit atrophy. The second phase aimed to achieve ≥80% of muscle strength and hop performance compared to the contralateral leg. The milestones of the third phase (return-to-sport phase) were to achieve 90% muscle strength and hop performance and a progressive transition to sport. It comprised heavy resistance strength training and increasingly demanding plyometric and sport-specific exercises. Further details about the treatment algorithm have been described in previous works [6, 15].

Outcome measures

Participants from the Delaware-Oslo ACL Cohort and the NKLR completed the Knee Injury and Osteoarthritis Outcome Score (KOOS) questionnaire 9–12 years after ACLR [16]. The KOOS assesses knee-related symptoms and function and consists of five subscales scored separately on a 0 (worst) to 100 (best) scale: pain, other symptoms, activities of daily living (ADL), function in sport and recreation (Sports) and knee-related QOL. Missing items in any subscale were handled as described in the KOOS scoring instructions from 2012. The KOOS shows good internal consistency, test–retest reliability, construct validity and responsiveness for individuals with ACL injury or reconstruction [17, 18]. Participants from the Delaware-Oslo ACL Cohort also completed the International Knee Documentation Committee-Subjective Knee Form (IKDC-SKF) questionnaire [19]. To estimate the postoperative IKDC-SKF scores for the NKLR, a validated crosswalk that converts group-based KOOS subscales into IKDC-SKF scores was used [20]. The IKDC-SKF measures symptoms, function and sports activity in patients with a variety of knee pain, such as ligament, meniscus and cartilage damage, osteoarthritis and patellofemoral pain, and results in a single score ranging from 0 (worst) to 100 (best). The IKDC-SKF shows good test-retest reliability, construct validity and responsiveness in individuals after ACL injury [21].

PASS thresholds for KOOS and IKDC-SKF scores were used to assess the percentage of patient-perceived satisfactory outcomes. These thresholds were validated and specifically established for participants 10 years after ACL injury [22] and derived from the dichotomous yes/no answers to the following PASS question [12]: ‘Considering your knee function, do you feel that your current state is satisfactory? With knee function, you should take into account all activities during your daily life, sport and recreational activities, your level of pain and other symptoms, and also your knee-related QOL.

Statistical analysis

Consistent with previous work [8], the a priori sample size estimation to detect a between-group difference for KOOS scores larger than 10 points [23] with a standard deviation (SD) of 20, α-level of 0.05 and 80% power, showed that 64 participants in each cohort were needed.

Descriptive variables included sex, age, body mass index, time from injury to surgery, graft type, cartilage and meniscus injury presence and treatment at ACLR. All the variables were summarised with descriptive statistics such as the mean, SDs, or frequency (percentage). After distributions were checked for severe skewness using Q–Q plots, group differences were assessed with χ² or Fisher’s exact tests for nominal variables and t-tests for continuous data [24].

To assess patterns of missingness, participants with missing and complete KOOS responses (noncompleters and completers) were compared in terms of sex, age, body mass index, time from injury to surgery, presence of meniscus or cartilage injuries, graft type and preoperative KOOS subscales.

Student t-tests were used to compare KOOS and IKDC-SKF scores of the Delaware-Oslo ACL Cohort and the NKLR. The percentages of participants that exceeded PASS thresholds were calculated for the KOOS and the IKDC-SKF scores. Validated PASS thresholds for KOOS subscales are 85.4 points in Pain, 76.5 in Symptoms, 93.8 points in ADL, 71.6 points in Sports, 59.0 points in QOL and for IKDC-SKF 76.2 points [22]. Differences in percentages between the two cohorts were assessed using the χ² test. Statistical analyses were performed in R. Version 4.2.3 (R Core Team).

RESULTS

Eighty of 100 participants (80%) from the Norwegian arm of the Delaware-Oslo ACL Cohort completed the KOOS 9–12 years after ACLR (Figure 1). From the NKLR database, 1588 of 3248 (49%) participants had completed the KOOS 9–12 years after ACLR and were eligible based on the inclusion criteria (Figure 1).

FIGURE 1.

Study flowchart.

Noncompleters were more often male (difference in percentages, 10%), younger (mean difference, 1.3 years), and had lower preoperative KOOS subscale scores (mean differences, 1–3 points) than completers (Supporting Information material).

The mean time from surgery to follow-up was 10.5 years (min–max: 9.7–12.0) for the Delaware-Oslo ACL Cohort and 10.0 years (min–max: 8.9–12.1) for the usual care group. The time from injury to surgery was significantly shorter, and medial meniscus injuries were less often resected and more often left untreated in the Delaware-Oslo ACL Cohort (Table 1), where 54 (67.5%) were operated on within the first 6 months of ACL injury compared to 809 (51%) participants in the usual care group. There were no statistically significant differences in sex, age, body mass index, graft type and presence of concomitant cartilage injuries and meniscus injuries between cohorts. The distribution of the scores in the Delaware-Oslo ACL Cohort and the usual care group is illustrated in Figure 2. The Delaware-Oslo ACL Cohort had statistically significantly higher scores for KOOS Pain, Symptoms, ADL, Sports and IKDC-SKF (p < 0.001) (Table 2).

TABLE 1.

Descriptive characteristics of participants from the Delaware-Oslo ACL Cohort and the NKLR (usual care).

Descriptive characteristics	Delaware-Oslo ACL Cohort (N = 80)a	NKLR (N = 1588)a	p-Valuesb
Sex			n.s.
Men	38 (48%)	784 (49%)
Women	43 (52%)	804 (51%)
Age at surgery (years)	24.6 (7.2)	24.6 (7.9)	n.s.
Body mass index (kg/m2)	23.7 (3.3)	24.3 (3.5)	n.s.
Months from injury to surgery	6.1 (3.9)	7.3 (5.1)	<0.01*
ACL graft			0.06
Bone-patella tendon-bone	27 (34%)	388 (24%)
Hamstring	53 (66%)	1200 (76%)
Cartilage injuries	11 (14%)	235 (15%)	n.s.
Medial meniscus injuries	28 (35%)	460 (29%)	n.s.
Medial meniscus treatment
Resection	3 (4%)	222 (14%)	<0.001*
Suture/fixation	15 (19%)	164 (10%)	n.s.
Trephination	0 (0%)	13 (<1%)	n.s.
Untreated	10 (12.5%)	61 (4%)	<0.01*
Lateral meniscus injuries	19 (24%)	449 (28%)	n.s.
Lateral meniscus treatment			n.s.
Resection	15 (19%)	289 (18%)
Suture/fixation	0 (0%)	57 (3.5%)
Trephination	0 (0%)	13 (<1%)
Untreated	4 (5%)	90 (6%)

Note: Body mass index was measured at surgery in the registry and at screening 8.3 (±15.2) months before surgery in the Delaware-Oslo ACL Cohort. Twelve meniscus injury sides could not be classified, and two lateral meniscus sutures were accompanied with partial resection in the NKLR.

Abbreviations: ACL, anterior cruciate ligament; NKLR, Norwegian Knee Ligament Register; n.s., not significant.

^an (%) or mean (standard deviation).

^bPearson’s χ² test; Welch two sample t-test; Fisher’s exact test.

^*p < 0.05.

FIGURE 2.

KOOS and IKDC-SKF outcome scores of the Delaware-Oslo ACL Cohort (in dark blue) or NKLR (usual care; in light blue) around the PASS thresholds (dotted lines). ACL, anterior cruciate ligament; ADL, activities of daily living; IKDC-SKF, International Knee Documentation Committee Subjective Knee Form; KOOS, Knee Osteoarthritis Outcome Score; NKLR, Norwegian Knee Ligament Register; PASS, patient acceptable symptom state; QOL, quality of life.

TABLE 2.

KOOS and IKDC-SKF scores for participants from the Delaware-Oslo ACL Cohort and those from the NKLR (usual care).

Scores (0–100)	Delaware-Oslo ACL Cohort (N = 80)a	NKLR (N = 1588)a	Differenceb	95% CIb	p-Valuesb
KOOS Pain	94.4 (6.9)	88.5 (14.6)	5.9	4.0, 7.8	<0.001*
KOOS Symptoms	91.2 (10.1)	84.0 (16.3)	6.8	4.2, 9.4	<0.001*
KOOS ADL	98.8 (2.9)	94.2 (11.0)	4.1	3.0, 5.2	<0.001*
KOOS Sports	85.9 (15.5)	73.5 (25.6)	12.0	8.0, 16	<0.001*
KOOS QOL	76.1 (19.9)	72.7 (23.0)	3.9	−0.8, 8.7	n.s.
IKDC-SKF	87.9 (11.0)	79.1 (19.2)	8.4	5.5, 11	<0.001*

Abbreviations: ACL, anterior cruciate ligament; ADL, activities of daily living; CI, confidence interval; IKDC-SKF, International Knee Documentation Committee Subjective Knee Form; KOOS, Knee Osteoarthritis Outcome Score; NKLR, Norwegian Knee Ligament Register; n.s., not significant; QOL, quality of life

^aMean (standard deviation).

^bWelch two sample t-test; positive values indicate higher scores in the Delaware-Oslo ACL Cohort.

^*p < 0.05.

The differences in percentages (pp) in PASS were 19 pp for KOOS Pain, 17 pp for KOOS Symptoms, 20 pp for KOOS ADL, 22 pp for KOOS Sports, 6 pp for KOOS QOL and 20 pp for IKDC-SKF (Figure 3). The PASS percentages in the Delaware-Oslo ACL Cohort were statistically significantly higher for KOOS Pain (p < 0.001), Symptoms (p = 0.002), ADL (p < 0.001), Sports (p < 0.001) subscales and IKDC-SKF (p = 0.001) compared to usual care.

FIGURE 3.

Percentages of participants with PASS including 95% confidence intervals of participants in the Delaware-Oslo ACL Cohort (in dark blue) and the NKLR (usual care; in light blue). ACL, anterior cruciate ligament; ADL, activities of daily living; IKDC-SKF, International Knee Documentation Committee Subjective Knee Form; KOOS, Knee Osteoarthritis Outcome Score; NKLR, Norwegian Knee Ligament Register; n.s., not significant; PASS, patient acceptable symptom state; QOL, quality of life. *p < 0.05.

DISCUSSION

The most important findings of the present study demonstrate reduced knee symptoms and superior function, and higher percentages of satisfactory outcomes in the Delaware-Oslo ACL Cohort, favouring the treatment algorithm over usual care.

Clinical relevance of differences in knee function and QOL

Among the five KOOS subscales, KOOS Sports and QOL are the most responsive and capture the most important concerns for ACL-injured participants [22, 25, 26]. The superior KOOS Sports scores in the Delaware-Oslo ACL Cohort likely reached clinical relevance given the difference in percentages of participants achieving satisfactory outcomes compared to usual care (22 pp). The same was observed for IKDC-SKF as judged by the difference in satisfactory outcomes (20 pp). Conversely, for KOOS QOL, the difference in satisfactory outcomes was minor (6 pp). These findings demonstrate that participants who followed the Delaware-Oslo ACL Cohort treatment algorithm benefitted from superior knee function. In contrast, knee-related QOL (e.g., knee confidence, lifestyle modification) was similar to the usual care group.

Key components of the treatment algorithm

At the time of ACLR, the participants from the Delaware-Oslo ACL Cohort had resolved initial acute impairments and completed a 5-week rehabilitation programme after injury. This approach aligns with current evidence and recommendations for ACL rehabilitation, which encourage preoperative rehabilitation to optimise postoperative outcomes [27, 28]. The rehabilitation programme included progressive neuromuscular exercises, heavy resistance strength training and plyometrics [4]. The programme aimed to regain at least 90% quadriceps and hamstring strength and hop performance before the decision to undergo ACLR [29, 30]. Importantly, this programme has a low risk of adverse effects [4] and did not delay surgery. The programme improved knee function early after ACL injury, a result that was based on participants who were also included in the present study [4]. Since poor preoperative knee function is a well-known risk factor for inferior 10-year outcomes, the superior long-term effects in the Delaware-Oslo ACL Cohort may be at least partially attributed to their knee status after the 5-week rehabilitation programme and before ACLR [31, 32].

Previous reported findings from the Norwegian arm of the Delaware-Oslo ACL Cohort showed that the participants had superior knee function preoperatively (after progressive rehabilitation) and at 2 years after ACLR compared to usual care in Norway [8]. The KOOS scores in the present study were comparable to those at 2 years, which explains the similar differences (<10 points) between cohorts at both follow-up time points. These findings illustrate that participants who followed a progressive rehabilitation programme experience superior knee function before ACLR and can still benefit from it at moderate and long term after ACLR.

The superior outcomes of the Delaware-Oslo ACL Cohort treatment algorithm can also be attributed to key components beyond the rehabilitation programme. Participants benefitted from early management and impairment resolution [4]. Thorough education and close follow-up through regular clinical testing were also essential components of the treatment algorithm. Those components may have prevented meniscus injuries from worsening and requiring meniscectomy, especially in patients who had recurrent instability episodes [33]. As meniscectomies tend to worsen outcomes after ACLR, the fewer medial meniscectomies in the participants from the Delaware-Oslo ACL Cohort might have contributed to their superior results [34, 35]. Education and close follow-up can further motivate and empower participants to incorporate rehabilitation principles into their practice and maintain their level of knee function.

A limitation of this study is that the study design does not allow us to draw firm conclusions about causality [36], a characteristic of all cohort studies. In addition, loss to follow-up is a common concern in registries and might reflect selection bias [37]. The response rate of 49% corresponds with the 10-year patient-reported follow-ups in the Scandinavian knee ligament registries [32, 38]. Noncompleters at follow-up were more often male and younger, which is consistent with previous reports [32, 39]. Although the pattern of missingness may skew the results in both directions [40, 41], the risk of selection bias can be estimated to be minor as differences in preoperative KOOS scores were trivial (mean differences <3 points).

The major strength of this study is the robustness of the outcome data from the usual care group due to the large sample size. This study focused on the largest patient group that sustains ACL injuries and receives ACLR, namely young and active individuals. The usual care group also reflects outcomes after usual care in Scandinavia, as characteristics were similar to those in the Swedish registry [37]. Another strength of our study is the use of validated PASS thresholds developed for participants 10 years after ACL injury [22].

The Delaware-Oslo ACL Cohort treatment algorithm shows promising results for optimising long-term outcomes following ACLR, including higher percentages of satisfactory knee function than what is observed following usual care. Therefore, clinicians should implement this treatment algorithm into their practice, involving progressive preoperative and postoperative rehabilitation, patient education, regular clinical testing and shared decision-making about treatment choice.

CONCLUSION

Participants who followed the Delaware-Oslo ACL Cohort treatment algorithm showed reduced knee symptoms, superior function and more satisfactory patient-reported outcomes 9–12 years after ACLR compared to participants who received usual care. The treatment algorithm consisting of progressive preoperative and postoperative rehabilitation, patient education, regular clinical testing and shared decision-making about treatment choice holds the potential to optimise long-term outcomes after ACLR.

Abbreviations:

ACLR

anterior cruciate ligament reconstruction

ADL

activities of daily living

IKDC-SKF

International Knee Documentation Committee-Subjective Knee Form

KOOS

Knee Injury and Osteoarthritis Outcome Score

NKLR

Norwegian Knee Ligament Registry

PASS

patient acceptable symptom state

QOL

quality of life

Sports

function in sport and recreation