|Home | About | Journals | Submit | Contact Us | Français|
The wish to return to level I pivoting sports is a major indication of ACL-reconstruction. Patients usually return to pivoting sports between 6 months and 1 year postoperatively, but no matched study has so far examined 1-year return to sport rates in nonoperatively and operatively treated ACL-injured patients.
ACL-injured patients following a nonoperative treatment course, including recommendation of activity modification, will have lower return to pivoting sport rates than operatively treated patients 1 year after baseline testing/surgery, when matched by preinjury sports activity, age and sex.
Pair-matched cohort study
Sixty-nine nonoperatively treated ACL-injured patients were pair-matched with 69 operatively treated patients (n=138), based on specific preinjury sport, age and sex. Nonoperatively treated patients were recommended not to return to level I sports. Patients were defined as nonoperatively or operatively treated according to their status at follow-up. The baseline and follow-up testing included registration of sports participation, KT-1000 measurements, 4 hop tests, and patient-reported outcome measures. McNemar’s test and paired t-tests or Wilcoxon’s test were used to compare outcomes of nonoperatively and operatively treated patients.
No significant baseline differences were found. At 12.9±1.2 months (mean ± standard deviation) after baseline testing (nonoperative) and 12.7±1.2 months after surgery (operative), there was no significant difference in overall return to sport rates (nonoperative: 68.1 %, operative: 68.1 %, p=1.000), or in return to level I sport rates (nonoperative: 54.8 %, operative: 61.9 %, p=0.664). Nonoperatively treated patients who participated in level I sports prior to injury had a significantly lower return to sport rate (54.8 %) than nonoperatively treated patients who participated in level II sports (88.9 %, p=0.003). The nonoperatively treated patients had significantly higher knee joint laxity, but significantly better hop test limb symmetry indexes, KOS-ADLS scores, and IKDC 2000 scores. None of the functional differences were larger than the smallest detectable difference.
ACL-injured patients following a nonoperative treatment course, including recommendations of activity modifications, and operatively treated patients did not have significantly different rates of returning to pivoting sports after 1 year in this pair-matched cohort study. Clinicians should be aware of a potentially high level of noncompliance to recommendations of activity modifications. While these results show that it is possible for nonoperatively treated patients to return to sport after rehabilitation, future follow-ups are needed to examine whether these patients maintain sports participation over time, and what long-term consequences they may suffer regarding subsequent injuries and knee osteoarthritis.
Injuries to the anterior cruciate ligament (ACL) are common in pivoting sports such as soccer, team handball, basketball, alpine skiing and tennis.8 After an ACL injury, patients have shown increased knee joint laxity,12 altered neuromuscular strategies,36,39 proprioceptive deficits6,10 and a loss of muscle strength.14,43 For a substantial number of patients, the injury and resulting loss of dynamic knee stability lead to a lower activity level.3,31 Additionally, psychological and social factors may also influence whether or not a patient returns to sport after injury.13,38
The main treatment options after an ACL injury are nonoperative and operative management. These two courses of treatment differ not only with regard to the surgical intervention, but also the content and length of rehabilitation, and recommendations for future sports activity. It is frequently advocated that patients who aim at returning to pivoting sports should undergo operative management with ACL reconstruction.9,25,26,28 The rationale behind this treatment algorithm is that an ACL reconstruction will improve passive knee stability and limit the risk of subsequent injuries, in particular meniscus and cartilage injuries, in patients who subject their knee to high loads.7,37 On the other hand, patients who are willing to modify their activity level may not need the additional passive knee stability provided by an ACL reconstruction, and can therefore be considered candidates for nonoperative treatment. Because the patient’s motivation for resuming pivoting sports is a major indication for surgery, it should be expected that existing cohort studies reveal a lower return to sport rate in nonoperatively treated patients. Furthermore, nonoperatively treated patients are recommended not to return to high demand pivoting sports.18,20,27,41 Still, several studies do not show a difference in return to sport rates between nonoperatively and operatively treated patients.18,29,42 This may in part be explained by a difference in preinjury activity level, as nonoperatively treated patients have been shown to participate in less demanding sports activities than operatively treated patients.11,15,18 Thus, it is possible that even if the return to sport rates are comparable, the nonoperatively treated patients return to less challenging sports activities.
In a recent meta-analysis, Ardern et al.3 found that the mean time between ACL-reconstruction and resumption of sport is 7.3 months. However, the timing of return to sport varies between institutions and individual patients. Both nonoperatively and operatively treated patients are expected to have finished rehabilitation and attempted to return to sports activities within 1 year.1,5,8,20,35,45 To our knowledge, no previous study has reported 1-year return to sport rates for nonoperatively and operatively treated patients with identical preinjury activity levels.
Therefore, the purpose of this study was to compare the 1-year return to sport rates in patients following a nonoperative treatment course, including recommendations of activity modifications, and individually matched operatively treated patients, who all participated in pivoting sports prior to injury. We hypothesized that nonoperatively treated patients would have lower return to pivoting sport rates than operatively treated patients, when matched by specific preinjury sport, sex and age.
This study is based on data from a prospective cohort study on ACL-injured subjects. Patients were consecutively included from the Musculoskeletal and Sport Medicine Clinic Hjelp24 NIMI in Oslo, Norway. To be included in the study, patients had to have sustained a unilateral ACL-rupture within the last 6 months (confirmed by MRI and ≥3 mm side-to-side difference with a KT-1000 arthrometer),12 be between 13 and 60 years of age, and participate in pivoting sports (Table 1) at least twice a week prior to injury. Pivoting sports was defined as level I or level II sports according to the activity level classification by Hefti et al.,21 modified to European sport activities.16,30 Sports with frequent pivoting (e.g. soccer, team handball, floorball, basketball) are classified as level I sports. Level II sports are pivoting sports with less frequent pivoting than level I sports (e.g. racket sports, alpine skiing, snowboarding, gymnastics and aerobics). Patients that participated in level III sports (e.g. running, cross-country skiing and weight lifting) were not included in this study. The exclusion criteria were bilateral injuries, previous knee injuries, or symptomatic concomitant injuries. Out of a total of 233 patients, 208 (89.2 %) attended the follow-up 1 year after baseline testing (nonoperatively treated patients) or surgery (operatively treated patients) (Figure 1). Nonoperatively treated patients with available 1-year follow-up data (n=88) were eligible for pair-matching with operatively treated patients (n=120), based on specific preinjury sport, sex and age (+/− 3 years). The study was carried out in accordance with the Declaration of Helsinki. All subjects signed a written consent prior to inclusion, and the study was approved by the Regional Ethical Committee for Eastern Norway.
At our institution (Hjelp24 NIMI), all subjects undergo active rehabilitation before the decision of nonoperative or operative treatment is made. After resolving initial impairments, the rehabilitation program consists of heavy resistance strength training, neuromuscular training, and plyometric exercises.16 During this time, the patients are informed about both nonoperative and operative treatment. The decision of nonoperative or operative treatment is made by the responsible orthopaedic surgeon, in close communication with the patient and treating physical therapist. A wish to return to level I sports (Table 1), dynamic instability, young age (but skeletally mature), and a patient’s preference for surgery are factors that positively influence a surgical decision. Following a nonoperative treatment course, the patients undergo continued active rehabilitation as needed. Nonoperatively treated patients at our institution typically undergo 3-4 months of rehabilitation after initial impairments have been resolved. Patients are discharged from rehabilitation when they meet specific criteria for quadriceps and hamstrings strength (limb symmetry index [LSI] > 90 %) and 4 single-legged hop tests (LSI > 90 % for all 4 hop tests).
Reconstructive surgery is performed with either a hamstring tendon or a bone-patellar tendon-bone (BPTB) autograft. In the early postoperative phase (0-6 weeks), the aim of the rehabilitation is to eliminate effusion, regain full range of motion and minimize muscular atrophy. From 2 to 6 months postoperatively, the aim of the rehabilitation is to regain adequate muscle strength and dynamic knee stability. The rehabilitation aims from 6 to 12 months postoperatively are to normalize muscle strength and dynamic knee stability, and to prepare the patient for return to sport with sport-specific training. Rehabilitation protocols are adjusted individually by the responsible physical therapist.
The return to sport criteria at our institution are at least 90 % hamstring and quadriceps strength, and limb symmetry indexes of at least 90 % on 4 hop tests.33 Operatively treated patients are recommended not to return to level II or lower sports earlier than 6 months postoperatively, and to level I sports no earlier than 9 months postoperatively. Nonoperatively treated patients may return to level II or lower sports as soon as the return to sports criteria are fulfilled; however, they are recommended to refrain from returning to return to level I sports without reconstructive surgery.
All patients went through baseline functional testing after inclusion. The same tests were then performed 1 year after baseline testing (nonoperatively treated patients) or 1 year after surgery (operatively treated patients). The initial KT-1000 measurements were included in the baseline data. Before testing, all patients performed a 10-minute standardized warm-up on a stationary bicycle. Thereafter, 4 hop tests (the single hop for distance, the crossover hop for distance, the triple hop for distance, and the 6-meter timed hop)33 were performed. These hop tests have previously shown to be reliable and valid for the ACL-injured population, and have a reported minimal detectable difference of 7-13 LSI.34 For the first 3 hop tests, the hop distance in centimeters was measured with a standard tape measure. For the 6-meter timed hop test, a stop watch was used to record the time to the nearest 100th of a second. All patients performed 1 practice trial followed by 2 recorded test trials on each leg, and the uninvolved leg was always tested first. After the hop tests, the patients completed the Knee Outcome Survey Activities of Daily Living (KOS-ADLS)24 and the International Knee Documentation Committee Subjective Knee Form (IKDC 2000).22,23 Both questionnaires are scored from 0 (worst) to 100 (best), and have established reliability, validity and responsiveness in patients with knee injuries.22-24 The KOS-ADLS has a reported standard error of measurement of 3.2,24 which corresponds to a minimal detectable difference of 8.9.44 The minimal detectable difference of the IKDC 2000 has been estimated to 12.8.22,23 The patients also completed a self-reported numeric global rating scale (GRS) of knee function from 0 (cannot do daily activities) to 100 (preinjury knee function), and reported what specific types of sports activity they participated in prior to injury and present.
Single-legged hop test limb symmetry indexes (LSI) were calculated as the longest distance hopped on the involved leg divided by the longest distance hopped on the uninvolved leg × 100. For the 6-meter timed hop test, the LSI was calculated as the fastest time hopped on the uninvolved leg divided by the fastest time hopped on the involved leg × 100.
Preinjury sports activity was defined as the patient’s self-reported main sports activity prior to injury. The patient was registered as having returned to sport if their sports activity at follow-up matched their preinjury main sports activity. Patients that returned to other sports than their preinjury main sports activity were thus classified as not having returned to sport.
Comparisons of return to sport rates between nonoperatively treated patients and operatively treated patients were conducted using the the Mantel-Haenszel estimate of the risk ratio, which takes pairing into account. The analyses were also stratified by preinjury activity level, defined by the preinjury main sport activity of the patient. P-values for comparisons of all other nominal variables were obtained using McNemar’s chi-squared test. For continuous variables, differences between nonoperatively treated patients and operatively treated patients were analyzed using paired t-tests for normally distributed variables, and Wilcoxon’s signed rank test for variables that were not normally distributed. The level of significance was set to .05 for all analyses. All analyses were conducted using SPSS v. 17.0 for Windows (SPSS Inc., Chicago, IL).
From the 88 available patients following a nonoperative treatment course, including recommendations of activity modifications, 69 pairs were formed. A total of 138 patients were thus included in the final analyses. The reasons for not being able to match nonoperatively treated patients were: No operatively treated patients participating in the same preinjury sport (n=12; 57.1 %), not matching age (n = 4; 19.0 %), and a combination of these two factors (n = 3; 14.3 %). Of the 69 operatively treated patients, 53 (76.8 %) had undergone reconstructive surgery with a hamstring autograft, and 16 (23.2 %) with a BPTB autograft. The mean time from injury to surgery was 5.5 ± 2.3 (SD) months.
In both the nonoperatively and operatively treated group, 42 patients (60.9 %) participated in level I sports and 27 patients (39.1 %) in level II sports prior to injury. The most frequent preinjury main sports were soccer (20 pairs, 29 %), alpine skiing/snowboarding (17 pairs, 24.6 %), and team handball (12 pairs, 17.4 %). There were no significant differences in descriptive variables, concomitant injuries or any functional outcomes at baseline (Table 2).
Nonoperatively treated patients returned for follow-up testing 12.9 ± 1.2 months (mean ± SD) after baseline testing, and the operatively treated patients were tested 12.7 ± 1.2 months after surgery (p=0.477). At the follow-up, 47 out of 69 nonoperatively treated patients had returned to sport (68.1 %). The return to sport rate of the operatively treated patients was also 68.1 % (RRMH = 1.00 [0.61-1.63], p=1.000). There were no significant differences between nonoperatively and operatively treated patients in return to level I or level II sports (Table 3). Out of the 42 patients in each group that participated in level I sports prior to injury, 23 (54.8 %) of the nonoperatively treated patients returned to sport, while 26 (61.9 %) of the operatively treated patients returned to sport (p=0.664). Out of the 27 patients in each group that participated in level II sports prior to injury, the nonoperative return to sport rate was 88.9 %, compared to an operative return to sport rate of 77.8 % (p=0.508).
Within nonoperatively treated patients only, patients that participated in level I sports had significantly lower return to sport rates (54.8 %) than patients that participated in level II sports (88.9 %, p=0.003). There was no significant difference in return to sport between operatively treated level I and level II athletes (return to level I sports: 61.9 %, return to level II sports: 77.8 %, p=0.195).
The KT-1000 measurements revealed that nonoperatively treated patients had significantly higher knee joint laxity (Table 3). The nonoperatively treated patients had significantly higher limb symmetry indexes for all 4 hop tests, KOS-ADLS scores, and IKDC 2000 scores (all p<0.05). There was no significant difference between nonoperatively treated patients and operatively treated patients for the GRS for knee function.
At the time of reconstruction, injuries to the medial meniscus were recorded in 20 of the operatively treated patients (29.0 %). Eleven (55 %) of these were treated with partial resection, 7 (35 %) were sutured, and 2 (10 %) were left untreated. Fourteen (20.3 %) of the operatively treated patients had a lateral meniscus injury, whereof 10 (71.4 %) were treated with partial resection, 1 was sutured (7.1 %), and 3 (21.4 %) were left untreated. No surgical procedures were performed at any point during this study in the nonoperatively treated group.
Based on the medical records, there was no significant difference between nonoperatively and operatively treated patients in the number of complications, new injuries or exacerbated concomitant injuries at follow-up (Nonoperative: 9, Operative: 10, p=0.708). Nor was there any significant difference in the number of patients that incurred complications, new injuries or exacerbated concomitant injuries between the two groups (Nonoperative: 6, Operative: 10, p=0.332). In the nonoperatively treated patients, 3 (4.3 %) symptomatic medial meniscus injuries, 4 (5.8 %) symptomatic lateral meniscus injuries, 1 (1.4 %) lateral cartilage injury, and 1 (1.4 %) contralateral ACL rupture were recorded at follow-up. In the operatively treated patients, 1 (1.4 %) symptomatic medial meniscus injury, 4 (5.8 %) symptomatic lateral meniscus injuries, 2 (2.9 %) symptomatic medial cartilage injuries, 1 (1.4 %) rerupture of the ACL, 1 (1.4 %) patella subluxation and 1 (1.4 %) surgically treated Cyclops lesion were recorded.
The main finding of this study was that patients following a nonoperative treatment course, including recommendations of activity modifications, and operatively treated patients, matched by specific preinjury sport, sex and age, did not have significantly different return to pivoting sport rates after 1 year. Our hypothesis was therefore not supported.
To our knowledge, this is the first matched study to compare return to pivoting sport rates for nonoperatively and operatively treated ACL-injured patients at a time when most patients are expected to have returned to sport (1 year). Furthermore, it is so far the largest matched study in this field, and the only study to match patients by their specific preinjury sport. A difference in preinjury sports would likely represent a major confounding factor in cohort studies that compare return to sport rates. As operatively treated patients have shown to participate in sports that may entail a higher degree of knee loading,11,15,18 returning to sport would also pose higher demands on dynamic knee stability for this group. By matching the patients by their specific preinjury sport, each pair of patients in this study thus had to achieve the same level of dynamic knee stability in order to return to sport. Patients that followed a nonoperative treatment course, including recommendations of activity modifications, had a return to sport rate of 68 %, suggesting that the majority of nonoperatively treated patients were able to return to pivoting sports. The return to sport rate in this study is comparable to our research group’s previously published 1-year results from a prospective cohort,29 yet higher than in several other previous studies.2,4,40 Both nonoperatively and operatively treated patients in the current study underwent active rehabilitation with emphasis on heavy resistance strength training, neuromuscular training, and plyometrics.16 Furthermore, patients were not discharged from rehabilitation until they met specific functional criteria. This must be taken into consideration when comparing our results to other cohorts.
At the follow-up, nonoperatively treated patients had significantly higher anterior knee laxity as measured with the KT-1000, but also significantly higher scores on all 4 hop tests, the KOS-ADLS, and the IKDC 2000. However, none of the differences in functional outcomes were larger than the previously reported minimal detectable difference of these tests.22-24,34 These small differences, although statistically significant, are thus probably of no clinical relevance. In line with our results, two other matched studies did not find significant differences in functional outcomes or activity level between nonoperatively and operatively treated patients.27,41 However, these studies had a follow-up of 10 and 15 years. Moreover, the only published RCT on rehabilitation plus early ACL reconstruction versus rehabilitation plus optional delayed ACL reconstruction showed no significant differences in self-reported outcome measures or activity level after 2 years.19
ACL-injured patients that are treated at our institution are recommended not to participate in level I sports without reconstructive surgery. In spite of this, our results surprisingly showed that 55 % of the nonoperatively treated level I athletes returned to sport (Table 3). These results indicate a high level of noncompliance to the activity modifications recommended to reduce the risk of give-way episodes. Institutions and clinicians that recommend activity modifications in combination with nonoperative treatment should therefore be cautious of the risk that patients might not follow their recommendations. While these results show that it is possible for nonoperatively treated patients to return to sport after rehabilitation, future follow-ups are needed to examine whether these patients maintain sports participation over time, and what long-term consequences they may suffer regarding subsequent injuries and knee osteoarthritis.
The low rate of subsequent injuries and exacerbated concomitant injuries observed in this study can likely be attributed to patients having returned to sport only months before the follow-up. Furthermore, only symptomatic injuries were recorded. While both nonoperatively and operatively treated patients underwent MRI prior to inclusion, there was no systematic follow-up MRI in this study. Future studies are therefore needed to examine a possible difference in the incidence of further injury to the knee (in particular meniscus injuries, cartilage injuries, reruptures and give-way episodes resulting in late reconstructive surgery) between nonoperatively and operatively treated patients. While the prevalence of meniscectomy in high level athletes 10 years after an ACL injury has been reported to be as high as 80 % and 68 % for nonoperatively and operatively treated patients, respectively,27 it is crucial that future studies also record the athletic exposure of the patients. Previous reports on subsequent meniscus/cartilage injuries in ACL-injured patients describe that activity modifications were recommended to the nonoperatively treated patients.18,27,41 However, noncompliance to these recommendations places the nonoperatively treated patients at considerably higher risk of further knee injury than suggested by the treatment guidelines. The results from this study clearly highlight the importance of recording athletic exposure in studies on subsequent injuries, as the frequency of injuries depends on the proportion returning to sport and the duration of sports participation.32
Nonoperatively treated patients who participated in level I sports prior to injury had a significantly lower return to sports rate (55 %) than patients who participated in level II sports prior to injury (89 %). There was no significant difference in return to sport rates between operatively treated patients with different preinjury activity levels. As the nonoperatively treated patients in this study were advised not to return to level I sports, this finding was expected. So far, few studies have reported separate return to sport rates for different types of sports. At a 6-year follow-up, Fink et al.17 reported that nonoperatively and operatively treated patients had slightly more comparable sports participation in low risk sports than in high risk sports. However, neither the aforementioned study,17 nor our study, was statistically powered to address this issue. Future studies are therefore needed to establish whether there is a significant interaction between preinjury sports type and return to sport in nonoperatively and operatively treated patients, even within patients that participate in pivoting sports prior to injury.
A disadvantage of any nonrandomized study on nonoperative and operative treatment is the inability to control for the factors that led to surgery. In this patient population, factors that were given weight in the surgical decision-making included a wish to return to level I sports, dynamic instability, age, and the preferred treatment of the patient. The two groups were not significantly different in terms of age, additional injuries or any baseline functional outcome measure. However, we did not record the patients’ intention to return to level I sports, or which treatment they preferred. It is not unlikely that the nonoperatively and operatively treated patients differed in terms of these two factors, and both the presence of self-selection and a potential difference in the motivation for returning to sport may have introduced a confounding bias that could not be eliminated by the matching process. As ACL-injuries typically occur in an athletically active population,8 return to sports participation after ACL-injury is an important measure. In this study, the patients were classified as having returned to sport if they reported participation in their preinjury main sport at the 1-year follow-up. Still, it is unknown if the patients had returned to their previous level of play, or level of performance. This definition also excludes patients that chose to quit their preinjury main sport, but were active in other sports. Lastly, we did not register the reasons why patients had not returned to sport. A systematic review recently reported fear of reinjury to be a more dominant reason for not returning to sport after ACL reconstruction than problems with the reconstructed knee.3 A more detailed registration of athletic participation might therefore be needed to disclose other differences between nonoperatively and operatively treated patients.
There was no significant difference in return to pivoting sport rates between ACL-injured patients following a nonoperative treatment course, including recommendations of activity modifications, and operatively treated patients in this pair-matched cohort study. Although nonoperatively treated patients were recommended not to return to level I sports, 55 % of these patients returned to sport. Clinicians should therefore be aware of a potentially high level of noncompliance to recommendations of activity modifications.
We thank physical therapists Annika Storevold, Ida Svege, Espen Selboskar, Karin Rydevik, and Marte Lund with their assistance in data collection for this study. We also acknowledge the Musculoskeletal and Sport Medicine Clinic, NIMI (www.nimi.no), for supporting the Norwegian Research Center for Active Rehabilitation (NAR; www.active-rehab.no) with rehabilitation facilities and research staff.