Importance Multiligament knee injuries are rare, but can lead to significant functional limitations. Surgery has been shown to improve outcomes, however, there remains considerable debate regarding the optimal timing of surgery.
Objective We aimed to determine whether early surgery in the setting of a multiligamentous knee injury was associated with superior functional outcomes when compared with surgery on a delayed basis.
Evidence review A comprehensive literature search of the MEDLINE, EMBASE and PubMed databases was conducted up to March 2018. We identified studies with a sample size greater than 10 that included subjects with an injury to at least two of the four major knee ligaments and compared outcomes between early and delayed surgery. We assessed the methodological quality of each included study using the Newstead-Ottawa Scale.
Findings We identified 11 eligible studies, including a total of 320 patients (195 early and 125 delayed). The mean time to surgery among patients treated early was 11.2 days in comparison to 294.8 days for the delayed group. Early surgery was found to have a statistically significantly higher Lysholm score (p<0.0001) and Meyers rating (p=0.02) when compared with delayed surgery. No statistically significant differences in International Knee Documentation Committee, Tegner Activity Scale, total arc of motion, loss of extension or loss of flexion were demonstrated between early and delayed surgery. Early surgery was noted to have a statistically significantly higher odds of requiring a manipulation under anaesthesia or arthrolysis (p=0.04), however, subsequent subgroup analysis showed no difference between early and delayed surgery when only studies employing an early range of motion protocol were pooled.
Conclusions and relevance Based on the current body of literature, which primarily consists of level IV evidence, early surgery in the setting of multiligament knee injury may provide better functional outcomes without compromising range of motion when using early postoperative mobilisation protocols. Further studies of higher quality are required to corroborate these findings.
Level of evidence Level IV, systematic review.
- knee dislocation
- timing of surgery
Statistics from Altmetric.com
What is already known
Operative treatment of multiligament knee injuries leads to superior functional outcome scores, greater range of motion, fewer contractures, improved ligamentous stability and higher rates of return to sport and employment.
The heterogenous nature of multiligament knee injuries have made them difficult to evaluate with large clinical trials.
What are the new findings
There is a paucity of high-level evidence to provide definitive conclusions and recommendations regarding the optimal timing of surgery for multiligament knee injuries.
Quantitative synthesis of low-quality studies suggests that early surgery may provide superior functional outcomes, without compromising range of motion, provided that early postoperative range of motion is initiated.
Future studies should use a disease-specific outcome measure, such as the Multiligament Quality of Life questionnaire to assess patient-reported outcomes in multiligament knee injuries.
Multiligament knee injuries (MLKI) are rare, however, they can result in significant morbidity.1–4 The substantial force required to cause these injuries is often associated with a knee dislocation or substantial subluxation of the knee resulting in an injury to two or more ligaments.5 6 Its low incidence and heterogeneous injury profile have it made it inherently difficult to evaluate with large clinical trials, resulting in a lack of consensus regarding the most effective management of these complex injuries.5 7–9
While it is well established that operative treatment leads to superior functional outcome scores, greater range of motion, fewer contractures, improved ligamentous stability and higher rates of return to sport and employment, there is still no consensus on the optimal timing of surgery for this challenging injury.7 10–20 Many surgeons have advocated for early surgical treatment to achieve the best result.7 13–16 18 20 21 Proponents of early repair or reconstruction suggest that the first 3 weeks after injury represent a critical time to re-establish anatomic relationships and offer the best chance to restore the central axis of knee motion.22 On the other hand, those who advocate delaying surgery believe that it offers the advantage of obtaining improved preoperative range of motion and provides extra-articular structures the opportunity to heal on their own.7 8 23–26 Others have reported superior outcomes with staged intervention, with repair of the extra-articular injuries acutely, followed by reconstruction of the cruciate ligaments at a later date once knee range of motion improves.10 11 16 19 27
In 2009, Mook et al 28 attempted to clarify this conflicting evidence with a systematic review, which examined over 60 years of published data on the optimal timing of surgery for MLKIs. The authors concluded that while early and delayed surgical intervention may yield equivalent stability outcomes, those who underwent early repair or reconstruction were more likely to require additional interventions for range of motion deficits, even when combined with an early range of motion protocol. However, the results of the systematic review were limited by a paucity of high-level evidence and were less generalisable as only studies with the most severe knee injuries were included. Notably, another systematic review published in the same year concluded that early operative treatment of MLKIs resulted in improved functional and clinical outcomes, which further highlights the heterogeneous nature of the available literature.7 Considering this lack of consensus, the purpose of this study was to provide an update on the optimal timing of surgery for MLKIs by performing a systematic review of the orthopaedic literature to determine whether early surgery results in superior functional outcomes without adversely affecting range of motion.
We conducted a systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.29
Two investigators (US and JS) independently performed a systematic literature search of the electronic databases MEDLINE, EMBASE and PubMed from their inception to March 2018. A keyword search including a combination of the following medical subject headings (MeSH) was performed: knee dislocation, multiligament, reconstructive surgical procedure and repair. Additional articles were detected by searching through the reference lists of eligible studies. The complete search strategy used for EMBASE is presented in online supplementary appendix 1.
Supplementary file 1
Studies were eligible for inclusion if they met the following criteria: (1) knee injury to at least two of the four primary knee ligaments; (2) functional outcome data reported for both the ‘early’ and ‘delayed’ surgery cohorts; and (3) total sample size >10 patients. Early surgery was defined as operative treatment within 4 weeks of injury. There were no age or language restrictions. All review articles, conference abstracts and case reports were excluded.
The titles, abstracts and full text of all retrieved articles were screened independently by two reviewers (US and JS). If any uncertainty was encountered during the screening process, the study was included until full-text review could be performed. Any disagreements at the full-text stage were resolved through discussion with the senior author until consensus was reached.
Two reviewers (US and JS), working independently and in duplicate, extracted all relevant data from eligible studies into a standardised collection form using spreadsheet software. Data collected included: (1) general study information (author, year of publication, level of evidence); (2) study population data (sample size, mean age, gender); (3) injury data (mechanism, type of MLKI based on the Schenck classification30 and mean time of surgery); (4) follow-up data (mean and range); and (5) outcomes reported.
Assessment of risk of bias in included studies
A modified version of the Newstead-Ottawa Scale was used to assess the methodological quality of all included studies.31 32 This rating scale evaluates non-randomised studies on the selection of the study groups, the comparability of the groups and the ascertainment of the outcome of interest. A maximum score of 8 points represents a high-quality study.31 Two independent reviewers assessed the quality of all included studies (US and JS).
Descriptive statistics were calculated with continuous data reported as weighted means with their associated SD and categorical data reported as frequencies with percentages. Pooled analyses were performed using RevMan (Review Manager V.5.3, The Cochrane Collaboration, Oxford, England). Continuous outcomes were reported as weighted mean differences (MD), while dichotomous outcomes were reported as ORs. Pooled summary statistics were calculated using a fixed effects model. Heterogeneity was quantified by the I2 statistic (ie, <25% being low and >75% being high), while the Cochrane X2 test of homogeneity (ie, Q test, p<0.10) was used to test for the presence of heterogeneity.33 For studies that did not report the mean and SD directly, these values were imputed from the median and IQR using well-established statistical techniques.34 An a priori subgroup analysis was performed to assess the effect of early postoperative range of motion on the need for subsequent manipulation under anaesthesia or operative arthrolysis. All tests of significance (two tailed) were performed with an alpha value of 0.05.
The results of our electronic database search and the number of studies excluded at each stage of study selection are depicted in the PRISMA flow diagram (figure 1). The baseline characteristics of all included studies are reported in table 1.
General study characteristics
A total of 320 patients were enrolled in the 11 included studies.15 18–20 35–41 Of note, all the included studies were retrospective cohorts or case series. The mean age of participants across all studies was 31.6 years. The majority of patients were male (86%) and had an injury to at least three of the major knee ligaments (62%). The most common mechanism of injury was a motor vehicle accident (40%) followed by participation in sports (38%). The mean time to surgery among patients treated early was 11.2 days in comparison to 294.8 days for the delayed group. All patients in the early intervention group underwent surgery within 30 days of MLKI. The mean follow-up time across studies was 47 months. The Lysholm15 19 20 35 36 39–41 and overall International Knee Documentation Committee (IKDC) score15 18–20 35–37 39 were the most commonly reported functional outcome measures. Range of motion (eg, total arc of motion, loss of extension and loss of flexion)15 20 35 39 and rates of manipulation under anaesthesia or arthrolysis18 20 35 36 38 39 were also frequently reported.
The details of the methodological quality assessment using the modified Newstead-Ottawa Scale are outlined in online supplementar y appendix 2. The mean score for all included studies was 3.0 (range 1–4), which suggests an overall low-quality rating. Although the majority of studies examined representative cohorts and reported excellent long-term follow-up, the methodological issues inherent to a retrospective study design limited the overall rigour (ie, lack of blinded assessment and adjusting for confounders) of these studies.
Supplementary file 2
The mean Lysholm scores for those who underwent early and delayed surgery were 87.7 and 80.1, respectively. Pooled analyses from eight studies15 19 20 35 36 39–41 demonstrated a statistically significant difference in Lysholm scores between the early and delayed surgery groups favouring early surgery (MD 6.80 (95% CI 4.06 to 9.55); p<0.0001) (figure 2A).
Overall IKDC score
Approximately 69% (88/127) of the early group and 48% (32/66) of the delayed group were found to have a ‘normal’ (IKDC A) or ‘nearly normal’ (IKDC B) knee. A quantitative synthesis of eight studies15 18–20 35–37 39 found no statistically significant difference between the early and delayed surgery groups (OR 1.97 (95% CI 1.00 to 3.88); p=0.05) (figure 2B). When pooling studies with severe injuries (ie, KD III and IV) only, no statistically significant difference in the proportion of overall IKDC scores was noted between the early and delayed surgery groups (4 studies; OR 1.81 (95% CI 0.68 to 4.85); p=0.24).18 20 35 37
Tegner Activity Scale
Pooled estimates from three studies reporting on activity level using the Tegner Activity Scale15 36 39 showed no statistically significant difference in scores between early and delayed surgery (MD 0.42 (95% CI 0.01 to 0.83); p=0.05) (figure 2C).
Meyers Rating Scale
The overall rate of an ‘excellent’ or ‘good’ rating was 89% (25/28) and 56% (9/16) for patients who underwent early and delayed surgery, respectively. The odds of having an ‘excellent’ or ‘good’ rating were approximately 5.5 times higher with early surgery compared with delayed surgery after pooling the results from two studies20 35 reporting the Meyers rating (OR 5.47 (95% CI 1.27 to 23.56); p=0.02) (figure 2D).
Range of motion
Total arc of motion, loss of extension and loss of flexion
There was no statistically significant difference in the total arc of motion (p=0.22), loss of extension (p=0.36) or loss of flexion (p=0.54) between the early (n=72) and delayed (n=35) surgery groups after pooling the estimates from four studies.15 20 35 39
Manipulation under anaesthesia or arthrolysis
The pooled rate of manipulation under anaesthesia or arthrolysis after early and delayed surgery was 17% (17/98) and 2% (1/48), respectively. There was a statistically significant difference in the odds of requiring a manipulation under anaesthesia or arthrolysis between groups, with the odds being 3.9 times higher with early surgery (6 studies; OR 3.91 (95% CI 1.10 to 13.87); p=0.04)18 20 35 36 38 39 (figure 3). However, subgroup analysis demonstrated that when only those studies utilising an early postoperative range of motion protocol were pooled, there was no statistically significant difference in the odds of requiring a manipulation under anaesthesia or arthrolysis (4 studies; OR 3.50 (95% CI 0.72 to 16.95); p=0.12).18 20 36 38 Similarly, no difference in the odds of undergoing subsequent manipulation under anaesthesia or arthrolysis was noted when pooling studies reporting on severe injuries (ie, KD III and IV) alone (4 studies; OR 4.01 (95% CI 0.87 to 18.43); p=0.07).18 20 35 38
The results of the current systematic review suggest that early repair or reconstruction of an MLKI may lead to superior functional outcomes, specifically higher Lysholm scores and Meyers ratings. In addition, a non-significant trend in IKDC scores favouring early surgery was observed, with the majority of patients in the early group (69%) classified as having a ‘normal’ or ‘nearly normal’ knee, compared with less than half of patients (48%) who had delayed surgery. Notably, no difference was observed in the overall range of motion, loss of extension and loss of flexion between the two groups. A statistically significantly higher odds of requiring a manipulation under anaesthesia or operative arthrolysis was noted among patients who underwent early surgery, however, no difference was found between early and delayed surgery when only studies that reported employing an early postoperative range of motion protocol were included.
Our findings seem to diverge from those of Mook et al,28 who found that early surgery yielded significantly more range of motion deficits and additional treatments for joint stiffness (ie, manipulation under anaesthesia or arthrolysis) despite the use of early postoperative mobilisation protocols. In addition, contrary to our findings, Mook et al 28 reported no difference in Lysholm scores when comparing those treated early and on a delayed basis. The fact that Mook and colleagues restricted their systematic review to the study of severe MLKIs (KD III and higher) may account for the differences seen between our results. Another possible explanation may be that our inclusion of more recent studies, utilising contemporary postoperative rehabilitation protocols with an emphasis on early mobilisation, may have led to a true improvement in overall function and range of motion. Interestingly, Levy and colleagues published a systematic review on the management of MLKIs in the same year as Mook et al, and reported findings similar to the current study, with higher mean Lysholm score and higher percentage of excellent and good IKDC scores in patients undergoing early versus delayed surgery.
In 2015, Jiang et al 27 published the results of a systematic review evaluating the ideal time to surgery in patients with KD III (medial or lateral) injuries. They concluded that staged treatment yielded the best clinical results, while no difference was found between the early and delayed surgical groups.27 However, these results were based solely on the proportion of ‘excellent’, ‘good’, ‘fair’ and ‘poor’ IKDC or Lysholm scores. Furthermore, they did not examine the impact of surgical timing on range of motion and the need for additional treatments due to arthrofibrosis. Similar to the findings of the current review, Hohmann et al 42 reported significantly higher Lysholm scores among patients who underwent early surgery, while demonstrating no difference in range of motion between groups. However, Hohmann and colleagues42 did not evaluate the effect of timing of surgery on the development of arthrofibrosis and the need for additional treatments. Finally, Barfield and colleagues43 conducted a systematic review of the literature after 2009 and noted that although there had been a significant increase in the number of publications reporting outcomes after MLKI, there remained a paucity of high-level evidence to provide any conclusive recommendations regarding the timing of surgery.43
This systematic review attempted to provide a quantitative synthesis of all studies comparing early and delayed surgery in patients with an MLKI. However, our results must be interpreted with caution, as they are based on level IV evidence. A lack of high-quality evidence continues to plague the study of MLKIs. The need for large, well-designed, prospective, multicentre randomised trials has been well established,7 43 however, the relative rarity of MLKIs, unique characteristics of each injury pattern and variety of surgical techniques available present significant challenges to organising a randomised study. Moving forward, a prospective multicentre cohort study may be a more plausible study design to adequately evaluate outcomes following MLKIs. The Multicenter Orthopaedic Outcomes Network group have successfully employed this study design to examine outcomes following anterior cruciate ligament reconstruction44 and rotator cuff tears.45
The standard use of Lysholm and IKDC scores has become commonplace in the study of MLKIs despite their lack of validation in the setting of these complex injuries.15 19 20 35 36 39–41 As such, the interpretation of their results, specifically the Lysholm score, can be difficult. For instance, in the current study, the MD in Lysholm scores between the early and delayed surgery groups was 6.8 points, which was found to be statistically significant. However, due to the lack of studies evaluating the validity, reliability and responsiveness of the Lysholm score as an instrument in this population, the minimal clinically important difference remains unknown. As a result, it is difficult to determine whether our results are clinically significant. Therefore, future studies should use a disease-specific patient-reported outcome, such as the Multiligament Quality of Life questionnaire, which is a novel outcome measure that has previously demonstrated excellent validity and reliability in the MLKI population.46
Despite the shortcomings of the existing MLKI literature, this study represents a methodologically rigorous critical appraisal of the current evidence. This systematic review has a number of strengths including the pooling of data across common outcomes from all available studies comparing early and delayed surgery, its generalisability across all levels of severity and minimisation of bias by performing study selection, data extraction and quality assessments in duplicate.
The present systematic review has several limitations. First, the level of evidence and methodological quality of included studies were generally poor as demonstrated by the Newstead-Ottawa Scale score. As such, the results of the current review must be interpreted with caution. Second, the presence of within (eg, cohort sizes, sex) and between (eg, surgical technique, postoperative protocol) study heterogeneity may limit the external validity of our results. Third, the current review did not examine the effect of staged surgery on functional outcomes, as a result, definitive conclusions regarding optimal timing of surgery cannot be made. Lastly, the early and delayed surgery groups may be inherently different from one another (eg, associated injuries, transfer, definition of delayed surgery), thereby introducing potential selection and attrition bias.
Evidence from the current body of literature, although of low quality, suggests that in patients with MLKIs, early surgery may lead to superior functional outcomes with no difference in range of motion when compared with delayed surgery. Although early surgery was associated with a higher odds of requiring a manipulation under anaesthesia or arthrolysis for knee stiffness, subgroup analysis revealed no difference in the need for additional treatments when examining only those studies employing an early postoperative range of motion protocol.
Contributors All of the authors have contributed to the design, data abstraction and preparation of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.