Accelerated Versus Non-accelerated Rehabilitation After Primary Anterior Cruciate Ligament Reconstruction Using Hamstring Autografts: A Systematic Review and Meta-analysis of Comparative Studies

Abstract

Purpose

This study aimed to compare the clinical outcomes between accelerated rehabilitation (AR) and non-accelerated rehabilitation (NR) after anterior cruciate ligament reconstruction (ACLR) using hamstring autografts through a systematic review and meta-analysis.

Methods

To compare the outcomes between AR and NR, we searched Medline, Embase, and the Cochrane Library. Studies comparing their clinical outcomes after ACLR using hamstring autografts were included. AR was characterized by the starting range of motion and weight-bearing within 3 days and return to sports within 6–9 months. A meta-analysis of clinical outcome parameters used in ≥ 3 studies was conducted.

Results

Seven studies were included. The International Knee Documentation Committee subjective score was significantly higher in AR than in NR at the 3- (mean difference [MD], 7.30; 95% confidence interval [CI] 1.55–13.05; P = 0.013) and 6-month follow-ups (MD, 5.64; 95% CI 0.11–11.17; P = 0.046). The side-to-side difference in anterior tibial translation at the final follow-up assessed in four studies was significantly lower in NR than in AR (MD, 0.59; 95% CI 0.12–1.07; P = 0.015). Overall Tunnel widening at the final follow-up assessed in four studies was also smaller in NR than in AR (MD, 0.48; 95% CI 0.00–0.96; P = 0.0479). However, the mean side-to-side difference and overall tunnel widening between them was < 1 mm.

Conclusion

The early subjective clinical outcomes of AR after ACLR using hamstring autografts were superior to those of NR. Although tunnel widening and anterior instability were also significantly larger in AR than in NR, the difference was not clinically relevant.

Introduction

The primary goal of anterior cruciate ligament reconstruction (ACLR) is to maximize knee stability and functional capacity while permitting a safe return to sports [1, 2]. Although surgical aspects are crucial to the success of the treatment, rehabilitation after surgery is equally important [3, 4]. However, currently, there is no consensus on the content and assessment methods of rehabilitation programs [2]. A high variability in the composition and time ranges of rehabilitation components may lead to confusion among patients and therapists [5].

Traditional postoperative restrictions, such as bracing for immobilization, delayed weight-bearing, and limiting early full extension, have all been used in rehabilitation protocols after ACLR to prevent excessive loads on the healing graft [6]. However, Shelbourne and Nitz introduced an accelerated rehabilitation (AR) protocol after ACLR using bone–patellar tendon–bone (BPTB) autografts to overcome many complications, such as prolonged knee stiffness, limitation of complete extension, delay in strength recovery, and anterior knee pain [7]. This rehabilitation protocol permits early range of motion (ROM) and weight-bearing, as well as return to light sporting activities by 2 months and full activity within 4–6 months after reconstruction [7, 8]. Previous studies have reported that AR decreased the incidence of surgical complications and improved knee function in patients who underwent ACLR [9, 10] although it did not affect long-term knee stability [11, 12].

As hamstring autografts have gained popularity in recent years owing to decreased harvest-site morbidity and lower incidence of patellofemoral osteoarthritis [13–16], there has been a concern on whether AR can be applied after ACLR using hamstring autografts. This concern is theoretically based on the slower incorporation rate of soft tissue to the bone tunnel and the increased laxity resulting from excessive graft–tunnel motion [17]. Several clinical studies have evaluated the effectiveness and safety of AR after ACLR using hamstring autografts; however, the results have been controversial [18–21]. Recently, a systematic review reported the clinical outcomes of AR after ACLR using hamstring autografts with the use of a best-evidence synthesis [22]. However, there have been no statistical approach (meta-analysis) to combine the results from previous comparative studies.

Therefore, the purpose of this study was to investigate the effectiveness and safety of AR after ACLR using hamstring autografts through a systematic review and meta-analysis of comparative studies. We hypothesized that AR after ACLR using hamstring autografts would result in early functional recovery and would not cause clinically relevant complications compared with non-accelerated rehabilitation (NR).

Materials and Methods

Literature Search and Data Sources

Multiple comprehensive databases were searched to find studies comparing the clinical outcomes of AR and NR after ACLR using hamstring autografts. This study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [23].

To identify relevant articles for inclusion in this meta-analysis, two different authors separately performed Internet searches in Medline (1950 to June 2019), Embase (1974 to June 2019), and the Cochrane Library (1950 to June 2019) databases on the same day (June 30, 2019). In addition, the search was repeated near the completion of the data analysis (July 20, 2019) to ensure a comprehensive investigation.

The following keywords and medical subject heading terms were included in the searches: (anterior cruciate ligament OR ACL) AND (reconstruction) AND (accelerated OR aggressive) AND (rehabilitation). After the initial electronic search, additional relevant articles were retrieved by searching the bibliographies of all selected full-text articles.

Study Selection

Two reviewers independently confirmed the inclusion of all studies according to the selection criteria. For study selection, the reviewers screened the titles and abstracts of the identified studies and subsequently screened the full text. Table 1 summarizes the inclusion and exclusion criteria. Based on previous studies [22, 24, 25], AR was defined as starting ROM and weight-bearing within 3 days after surgery and return to sports within 6–9 months. If more than one of these criteria (ROM, weight-bearing, and return to sports) was not satisfied, the protocol was defined as NR.

Table 1.

Inclusion and exclusion criteria

Inclusion criteria

Studies comparing the clinical outcomes of AR and NR after primary ACLR

AR characterized by starting ROM and weight-bearing within 3 days and return to sports within 6–9 months

Hamstring autografts

Any surgical methods of primary ACLR (number of graft bundle, tunnel technique, and fixation method)

English language or with a recognizable data table of results

Human subjects

Exclusion criteria

Non-comparative case series

Other than hamstring autografts

Not matched AR/NR protocol

No usable information/data

Review or meta-analysis

Case reports

AR accelerated rehabilitation, NR non-accelerated rehabilitation, ACLR anterior cruciate ligament reconstruction, ROM range of motion

Data Extraction

Two reviewers independently recorded data from each study using a predefined data extraction form; these data included the first author, journal title, year of publication, study design, sample size, mean age, follow-up period, number of graft bundles, femoral tunnel placement technique, graft fixation methods, treatments for meniscal lesions, and values of outcome parameters. Disagreement between the reviewers was resolved by consensus or discussion with a third investigator when consensus was not reached. The primary outcomes of interest were the International Knee Documentation Committee (IKDC) subjective scores at the 3- and 6-month follow-ups. The secondary outcomes of interest included the side-to-side difference (STSD) in anterior tibial translation and tunnel widening of both the femoral and tibial tunnels at the final follow-up.

Study Quality Assessment

Two reviewers independently assessed the risk of bias in the enrolled studies. To assess the risk of bias in these studies, we used the Risk of Bias (RoB) 2 tool for randomized controlled trials and the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) tool for non-randomized studies. The RoB 2 tool prompts judgments regarding biases in five domains: bias arising from the randomization process, derivations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result [26]. The ROBINS-I tool covers seven domains: bias due to confounding, participant selection, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of the reported results [27].

Statistical Analysis

A meta-analysis of the outcome parameters that were used in ≥ 3 studies was conducted (IKDC subjective score, STSD in anterior tibial translation, and tunnel widening). The mean and standard deviation (SD) of these parameters were obtained from each study. In studies with a sample size of ≤ 70 that did not report the SD, we estimated the SD by dividing the range by 4 using the method established by Hozo et al. [28]; in studies that only reported the SD for the total group and did not report the SD and range for each group, we used the SD of the total group in the meta-analysis.

Publication bias could not be assessed in this study. Tests for funnel plot asymmetry are typically performed only when at least ten studies are included in the meta-analysis [29]. As our analysis included only seven studies, tests for asymmetry would be ineffective because they would be unable to differentiate the results.

Statistical analysis was performed using the R software (R Foundation for Statistical Computing, Vienna, Austria). The I² value was used to estimate statistical heterogeneity. When the I² value was > 50%, heterogeneity was accepted, and a random-effects model was adopted. Otherwise, a fixed-effects model was adopted. A P value of < 0.05 was considered statistically significant.

Results

Study Selection

The study selection process is illustrated in Fig. 1. The electronic search yielded 219 studies in PubMed (Medline), 297 studies in Embase, and 50 studies in the Cochrane Library. No further studies were found through manual search, and 239 duplicate studies were removed. After screening of the titles and abstracts and after a full-text review of the 327 remaining studies, seven studies were finally included in the meta-analysis [18–21, 30–32].

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of the literature selection process

Study Characteristics

All seven articles included in this study were comparative studies published in English between 2001 and 2017 (Table 2). All seven studies used rehabilitation methods consistent with the common AR protocol of early ROM, weight-bearing, and return to sports activity, although the protocol slightly differed from study to study. The detailed rehabilitation protocols for each study are summarized in Table 3. Among them, six studies compared clinical or radiological outcomes between two rehabilitation protocols: AR and NR. In contrast, in the study by Zhu et al., three different rehabilitation protocols were used, and patients were divided into three groups (A, B, and C) [32]. After reviewing the three rehabilitation protocols, we classified group A into the NR group and groups B and C into the AR group.

Table 2.

Summary of the Included Studies

Authors	Journal (year)	Study design	No. of cases	FU period	Age, years	Graft bundle	Technique(femoral tunnel)	Fixation	Treatment for meniscal lesions	Outcome parameters
Murty et al	Knee (2001)	RCS	AR: 14 NR: 14	12 months	AR: 30.1 NR: 33.1	SB	Extra-articular tunnel	F: Staple T: Pull-out suture	AR: MS 7, MR 0 NR: MS 8, MR 1	STSD, Lachman test, effusion, Tegner activity score, tunnel enlargement (T)
Majima et al	CORR (2002)	PCS	AR: 32 NR: 30	AR: mean 43.5 months NR: mean 43.2 months	AR: 25.9 NR: 24.5	SB	Transportal	F: Double staple T: Double staple	AR: MS 7, MR 4 NR: MS 4, MR 4	STSD, IKDC grade, effusion, IKDC activity level, muscle toque (Q/H)
Yu et al	Chin Med J (2005)	RCS	AR: 33 NR: 32	6 months	AR: 31.2 NR: 30.3	SB	Transportal	F: Tendon node fixation to the bottle neck T: Bone bridge fixation by suture	AR: no MRs NR: with MR but no number of cases described	STSD, tunnel enlargement (T)
Vadala et al	KSSTA (2007)	RCT	AR: 20 NR: 25	Mean 10 months	AR: 29 NR: 30	SB	Outside-in	F: CSD T: Interference screw	AR: MS 5, MR 0 NR: MS 8, MR 0	Tunnel enlargement (F/T)
Christensen et al	J Sport Rehabil (2013)	RCT	AR: 19 NR: 17	1, 12, and 24 weeks	AR: 30.1 NR: 33.1	SB	Transportal	F: CSD T: Interference screw	Meniscal tears of > 5 mm were excluded	STSD, IKDC subjective score, ROM, peak isometric force (Q)
Zhu et al	EJOST (2013)	RCT	AR (groups B and C): 30 NR (group A): 15	3, 6, and 12 months	ND	ND	ND	F: CSD T: Interference screw	AR: no MR NR: no MR	ROM, IKDC subjective score, thigh muscle atrophy, tunnel enlargement (F/T)
Gupta et al	JAJS (2017)	RCT	AR: 20 NR: 20	6 weeks 3 and 6 months	AR: 26.5 NR: 28.9	SB	ND	F: CSD T: Interference screw	No difference in the incidence of meniscal tears between groups	Lachman test, pivot shift test, IKDC subjective score, KOOS, Tegner activity score, single-leg hop test

CORR: Clinical Orthopaedics and Related Research; Chin Med J: Chinese Medical Journal; KSSTA: Knee Surgery: Sports Traumatology, Arthroscopy; J Sport Rehabil: Journal of Sport Rehabilitation; EJOST: European Journal of Orthopaedic Surgery and Traumatology; JAJS: Journal of Arthroscopy and Joint Surgery

RCT randomized controlled trial, RCS retrospective comparative study, PCS prospective comparative study, FU follow-up, AR accelerated rehabilitation, NR non-accelerated rehabilitation, SB single-bundle, STSD side-to-side difference, F femoral tunnel, T tibial tunnel, MS meniscectomy, MR meniscus repair, Q quadriceps muscle, H hamstring muscle, CSD cortical suspensory device, ND not described, IKDC International Knee Documentation Committee, KOOS Knee Injury and Osteoarthritis Outcome Score, ROM range of motion

Table 3.

Rehabilitation protocol of the included studies

Authors (publication year)	Group	Range of motion	Weight-bearing	Return to sports
Murty et al. (2001)	AR	ROM start after 2 days (5°–90°)	PWB from 2 days, increasing as tolerable	ND
	NR	Immobilization at 50° flexion for 2 weeks	NWB for 2 weeks, FWB from 2 weeks	ND
Majima et al. (2002)	AR	Immediate (0°–90°)	Immediate full as tolerable	7–9 months
	NR	Start after 1 week (0°–90°)	PWB from 2 weeks, FWB after 4 weeks	12 months
Yu et al. (2005)	AR	Full ROM within 4 weeks	Immediate FWB	6 months
	NR	0°–90° for 6 weeks, full ROM from 6 weeks	PWB for 6 weeks, FWB after 6 weeks	6 months
Vadala et al. (2007)	AR	Immediate	Progressive weight-bearing as tolerated	4–6 months
	NR	Start after 2 weeks	Progressive weight-bearing as tolerated	4–6 months
Christensen et al. (2013)	AR	Immediate	Immediate full as tolerable	6 months
	NR	Start after 1 week (10°–120°)	PWB for 2 weeks, FWB after 2 weeks	6 months
Zhu et al. (2013)	AR (group B)	0°–90° for 1 week, 10° increased per week, nearly normal after 4 weeks	PWB for 2 weeks, FWB for 2 weeks	3 months
	AR (group C)	0°–90° for 2 weeks, 10° increased per week, nearly normal after 8 weeks	PWB after 2 weeks, FWB after 4 weeks	6 months
	NR (group A)	0°–60° for 4 weeks, > 90° after 8 weeks, normal after 12 weeks	NWB for 4 weeks, PWB for 8 weeks, FWB for 12 weeks	12 months
Gupta et al. (2017)	AR	Immediate, full ROM after 3 weeks	NWB for 1 week, FWB after 1 week	ND
	NR	Start after 1 week, full ROM after 7 weeks	NWB for 1 week, PWB after 1 week, FWB after 3 weeks	ND

AR accelerated rehabilitation, NR non-accelerated rehabilitation, ROM range of motion, PWB partial weight-bearing, FWB full weight-bearing, NWB non-weight-bearing, ND not described

Of the included articles, four were randomized controlled studies, and one was a prospective comparative study. The two remaining studies were retrospective comparative studies. The mean follow-up duration was 6 months in three studies, 10–12 months in three studies, and 43 months in the remaining study (Table 2).

The results of the risk of bias assessment are summarized in Tables 4 and 5. Therein, all randomized controlled trials had an overall low risk of bias as assessed using the RoB 2 tool (Table 4). Meanwhile, the other non-randomized studies had an overall moderate risk of bias as assessed using the ROBINS-I tool (Table 5).

Table 4.

Risk of bias assessment for randomized controlled trials using the risk of bias 2 tool

Study (publication year)	Randomization	Intended interventions	Missing data	Measurement of outcome	Reported result	Overall
Vadala et al. (2007)	Low	Low	Low	Low	Low	Low
Christensen et al. (2013)	Low	Low	Low	Low	Low	Low
Zhu et al. (2013)	Low	Low	Low	Low	Low	Low
Gupta et al. (2017)	Low	Low	Low	Low	Low	Low

Table 5.

Risk of bias assessment for non-randomized studies using the risk of bias in non-randomized studies of interventions tool

Study (publication year)	Confounding	Selection bias	Classification of interventions	Intended intervention	Missing data	Measurement of outcomes	Reported result	Overall bias
Murty et al. (2001)	Low	Moderate	Moderate	Low	Low	Low	Low	Moderate
Majima et al. (2002)	Low	Moderate	Low	Low	Low	Low	Low	Moderate
Yu et al. (2005)	Moderate	Moderate	Low	Moderate	Low	Moderate	Low	Moderate

International Knee Documentation Committee Subjective Score

Among the seven included studies, we were able to separately retrieve data on the IKDC subjective scores of the AR and NR groups at the 3- and 6-month follow-ups in three studies [18, 19, 32]. There were 67 cases in the AR group and 51 cases in the NR group. The IKDC subjective score was used in three studies in which pooled data showed a higher score in the AR group than in the NR group by 7.30 points at the 3-month follow-up. The difference was significant (95% confidence interval [CI] 1.55–13.05; P = 0.013; I² = 73%) (Fig. 2).

Fig. 2.

Forest plots of studies reporting the IKDC subjective scores after accelerated and non-accelerated rehabilitation after anterior cruciate ligament reconstruction using hamstring autografts at the 3-month follow-up. Squares represent the mean IKDC subjective score, with the size of the square being proportional to the sample size. Experimental and control indicate the accelerated and non-accelerated rehabilitation groups, respectively. IKDC International Knee Documentation Committee, SD standard deviation, MD mean difference, CI confidence interval

The IKDC subjective score was used in three studies in which pooled data showed a higher score in the AR group than in the NR group by 5.64 points at the 6-month follow-up. The difference was also significant (95% CI 0.11–11.17; P = 0.046; I² = 55%) (Fig. 3).

Fig. 3.

Forest plots of studies reporting the IKDC subjective scores after accelerated and non-accelerated rehabilitation after anterior cruciate ligament reconstruction using hamstring autografts at the 6-month follow-up. Squares represent the mean IKDC subjective score, with the size of the square being proportional to the sample size. Experimental and control indicate the accelerated and non-accelerated rehabilitation groups, respectively. IKDC International Knee Documentation Committee, SD standard deviation, MD mean difference, CI confidence interval

Side-to-side Difference in Anterior Tibial Translation

Four studies were included in the analysis of the STSD in anterior tibial translation at the final follow-up [18, 20, 21, 30]. The follow-up period varied from 6 to 43 months among the studies. The STSD in anterior tibial translation was significantly lower in the NR group than in the AR group (mean difference [MD] 0.59; 95% CI 0.12–1.07; P = 0.015; I² = 0%) (Fig. 4).

Fig. 4.

Forest plots of studies reporting the side-to-side difference after accelerated and non-accelerated rehabilitation after anterior cruciate ligament reconstruction using hamstring autografts at the final follow-up. Squares represent the mean side-to-side difference, with the size of the square being proportional to the sample size. Experimental and control indicate the accelerated and non-accelerated rehabilitation groups, respectively. SD standard deviation, MD mean difference, CI confidence interval

Tunnel Widening (Femoral and Tibial Tunnels)

Four studies reported tunnel widening at the final follow-up [20, 21, 31, 32]. The follow-up periods varied from 6 to 12 months among the studies. An analysis of the overall widening of the femoral or tibial tunnel was performed, followed by a subgroup analysis of the widening of each tunnel.

The overall increase in the femoral or tibial tunnel diameters observed in the AR group was significantly greater than that observed in the NR group (MD 0.48; 95% CI 0.00–0.96; P = 0.0479; I² = 83%) (Fig. 5).

Fig. 5.

Forest plots of studies reporting widening of the femoral and tibial tunnels after accelerated and non-accelerated rehabilitation after anterior cruciate ligament reconstruction using hamstring autografts at the final follow-up. Squares represent the mean increment for the femoral and tibial tunnels, with the size of the square being proportional to the sample size. Experimental and control indicate the accelerated and non-accelerated rehabilitation groups, respectively. SD standard deviation, MD mean difference, CI confidence interval

Two studies were included in the analysis of the widening of the femoral tunnel at the final follow-up [31, 32]. In these studies, pooled data on femoral tunnel widening showed a greater increase in the AR group than in the NR group by 0.69 mm at the final follow-up. The difference was significant (95% CI 0.38–1.00; P < 0.0001; I² = 0%) (Fig. 5).

Four studies were included in the analysis of the widening of the tibial tunnel at the final follow-up [20, 21, 31, 32]. In these studies, pooled data on tibial tunnel widening showed a greater increase in the AR group than in the NR group by 0.37 mm at the final follow-up. However, the difference was not significant (95% CI − 0.38 to 1.13; P = 0.446; I² = 89%) (Fig. 5).

Discussion

The most important finding of this study was that the IKDC subjective score at 3 and 6 months was significantly higher in the AR group than in the NR group. This meta-analysis also showed that the STSD in anterior tibial translation and overall tunnel widening in the AR group were significantly larger than those in the NR group at the final follow-up. However, the difference in the STSD in anterior tibial translation and overall tunnel widening between the two groups was 0.59 mm and 0.48 mm, respectively. These results were not considered to be clinically relevant.

To the best of our knowledge, this is the first meta-analysis to compare the clinical and radiological outcomes between AR and NR after ACLR using hamstring autografts. This study only included studies comparing AR and NR after ACLR using hamstring autografts. Recently, Janssen et al. conducted a systematic review of the clinical outcomes of AR after ACLR using hamstring autografts [22]. They performed a best-evidence synthesis to integrate the results of previous studies and reported that brace-free AR after ACLR using hamstring autografts did not alter the clinical outcomes, including clinical scores and knee laxity (“moderate” level of evidence). Our meta-analysis also showed that there was < 1-mm difference in the STSD in anterior tibial translation between the groups. However, in this study, the IKDC subjective scores at the 3- and 6-month follow-ups were significantly higher in the AR group than in the NR group.

Among the included studies, three reported the IKDC subjective scores, with a slightly different follow-up period among the studies. All three studies reported the IKDC subjective scores at the 3- and 6-month follow-ups after ACLR, which was probably because AR usually aims to allow return to sports by 6 months after surgery. This meta-analysis then revealed that the AR group had higher IKDC subjective scores than the NR group at the 3- and 6-month follow-ups. These findings suggest that AR improved the subjective outcomes earlier in patients who underwent ACLR using hamstring autografts, which can be considered a rationale for an early return to sports.

A recent systematic review demonstrated that femoral and tibial tunnel widening occurred after ACLR using a hamstring tendon autograft and accelerated brace-free rehabilitation [22]. This meta-analysis also showed that the increase in the femoral and tibial tunnel diameters in the AR group was significantly greater than that in the NR group. Previous studies have shown that tunnel widening was significantly greater after ACLR using hamstring autografts than after ACLR using BPTB autografts [33–35]. Cortical suspensory devices used for hamstring graft fixation have been reported to be susceptible to longitudinal (bungee effect) and sagittal motions (windshield wiper effect) because of the distance between the graft and the point of fixation [36, 37]. Moreover, a previous animal study has reported that histological healing of grafts after ACLR was attained by 12 weeks for tendon–bone grafts and by 8 weeks for bone–bone grafts [38]. Therefore, graft motions might be increased by early AR before graft healing is completed, which results in greater tunnel widening after ACLR using hamstring autografts.

However, the MD in the tunnel widening between the AR and NR groups was < 1 mm in this meta-analysis. The subgroup analysis also showed that the MD in the femoral and tibial tunnel widening was 0.69 mm and 0.37 mm, respectively. Although there was a significant difference found in the overall tunnel widening between the two rehabilitation protocols, the difference did not seem to be clinically significant. Therefore, the statistical difference should not be clinically overestimated. Further studies are then required to verify the criteria for clinically relevant tunnel widening.

The same consideration should be applied when interpreting the differences in the STSD in anterior tibial translation between the two rehabilitation protocols. Although the STSD in anterior tibial translation was significantly lower in the NR group than in the AR group, the MD was only 0.59 mm. Previous studies that investigated anterior instability after ACLR have set the clinically significant difference in the STSD in anterior tibial translation to 1 mm [39, 40]. This result is consistent with the small difference found in the STSD in anterior tibial translation. Therefore, the statistical difference in the STSD in anterior tibial translation in this meta-analysis should not be overestimated clinically.

This meta-analysis has some limitations. First, only seven studies with a total of 321 patients were included. Moreover, only three studies were included in the analysis of the IKDC subjective scores as the primary outcome of interest. Therefore, the reliability of the results may be limited by the small sample size. Second, the follow-up periods in the meta-analysis of the IKDC subjective scores were relatively short (3 and 6 months). Nevertheless, these seem to be sufficient to confirm the effectiveness of AR because the aim of the AR protocol is to allow the return to sports within 6 months. Third, each study evaluated the outcome parameters at different follow-up periods (STSD, 6–43 months; tunnel widening, 6–12 months). However, previous studies have reported that tunnel widening mainly occurred within 6 months after ACLR and that no further significant increase occurred thereafter [41, 42]. In addition, Howell et al. reported that the signal of hamstring autografts on serial magnetic resonance imaging was well established by 3 months and remained unchanged at 1 year [43]. Therefore, we decided to perform a meta-analysis of the results at different follow-up periods. Finally, each study used slightly different rehabilitation methods. To solve this problem, we defined common AR and NR protocols before the study and enrolled studies according to these criteria.

Conclusions

This systematic review and meta-analysis indicated that the early subjective clinical outcomes of AR after ACLR using hamstring autografts were superior to those of NR. Although tunnel widening and anterior instability were also significantly larger in the AR group than in the NR group, the difference was not clinically relevant.

Funding

There is no funding source.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

This article does not require ethical approval because it is a meta-analysis of previous studies.

Informed consent

Not applicable for this study.