Primary surgery versus primary rehabilitation for treating anterior cruciate ligament injuries: a living systematic review and meta-analysis

Key points

1. 01No clinically meaningful difference in self-reported knee function at short, medium, or long-term follow-up
2. 02No significant difference in radiological knee osteoarthritis or meniscal injury rates between the two strategies
3. 03Primary rehabilitation showed a trend toward less knee osteoarthritis development, though with very low certainty
4. 04Early surgery showed a trend toward better meniscal status in the long term, but only at low certainty
5. 05Evidence quality was rated low to very low overall; only three small RCTs met inclusion criteria

How it was conducted

Design

Living systematic review and meta-analysis of randomised controlled trials

Databases searched

MEDLINE, EMBASE, CINAHL, Web of Science, CENTRAL, SPORTDiscus plus six trial registries

Studies included

3 RCTs with 9 reports (320 participants total)

Participants

Mean age 29.5 years (SD 7.05); approximately 93% injured during sport

Intervention

Early ACL surgical reconstruction (within 6-10 weeks) plus postoperative rehabilitation

Comparator

Primary supervised rehabilitation with optional delayed ACL reconstruction if instability persisted

What they found

• Self-reported knee function (short-term): SMD -0.25 (95% CrI -0.84 to 0.36), raw mean difference -4.21 IKDC points (95% CrI -14.27 to 6.07), GRADE low - not clinically meaningful (MCID 16.7 points)
• Self-reported knee function (medium-term): SMD -0.10 (95% CrI -0.59 to 0.41), raw mean difference -2.65 IKDC points (95% CrI -15.94 to 10.89), GRADE low - not clinically meaningful
• Self-reported knee function (long-term): SMD -0.21 (95% CrI -1.49 to 0.81), raw mean difference -0.96 IKDC points (95% CrI -5.79 to 3.95), GRADE very low - not clinically meaningful
• Self-reported knee function (all time points combined): SMD -0.27 (95% CrI -0.84 to 0.21), raw mean difference -5.07 IKDC points (95% CrI -15.70 to 3.99), GRADE low
• Radiological knee osteoarthritis (long-term): OR 1.45 (95% CrI 0.30 to 5.17), 95% PI 0.18 to 10.0, risk difference -72 per 1000 patients (95% CrI 144 to -384), GRADE very low
• Meniscal injury (long-term, single study): OR 0.85 (95% CI 0.45 to 1.62), GRADE low - no significant difference
• Health-related quality of life (medium-term): SMD -0.40 (95% CrI -0.88 to 0.09), raw mean difference -5.91 SF-36 points (95% CrI -13.05 to 1.32), GRADE low - not clinically meaningful
• Return to activity (medium-term): SMD -0.10 (95% CrI -0.57 to 0.38), raw mean difference -0.31 Tegner points (95% CrI -1.80 to 1.19), GRADE very low
• Return to activity (long-term): SMD -0.22 (95% CrI -0.92 to 0.39), raw mean difference -0.75 Tegner points (95% CrI -2.92 to 1.23), GRADE very low
• Adverse event graft rupture: OR 2.3 (95% CrI 0.4 to 12.4), risk difference -26 per 1000 (95% CrI 12 to -186), GRADE very low
• Knee stability (medium-term, single study): MD -1.70 (95% CI -2.65 to -0.75), GRADE moderate - favours early surgery
• Patellofemoral cartilage thickness (medium-term, single study): MD 76.00 micrometers (95% CI 10.63 to 141.37), GRADE moderate
• Cost-effectiveness (medium-term, single study): MD 0.04 QALYs (p=0.18), early surgery not cost-effective, GRADE moderate

Limitations

• Only 3 RCTs with 320 total participants; small sample size limits certainty of all estimates
• Only one trial was rated low risk of bias; others had some concerns or high risk of bias
• Surgical techniques and rehabilitation protocols varied across trials, limiting direct comparison
• No data available for professional or elite athletes with the highest functional demands

Why it matters

For patients

Most people with an ACL tear can start with a structured rehabilitation program and achieve similar long-term knee function as those who have immediate surgery, with the option to operate later if instability persists.

For clinicians

Guidelines should shift toward a stepped-care model: offer primary rehabilitation first for patients without serious concomitant injuries, reserving early surgery for cases with functional instability, repairable meniscal tears, or other high-grade ligamentous injuries.

For readers

This living review, updated annually, provides the most current RCT-level synthesis on one of sports medicine's most debated questions, but the low-to-very-low certainty evidence means firm conclusions await larger, longer trials.