24-Week Predictors of Return to Play After ACL Reconstruction

Key Takeaways

• Among completers, overall 48-week success was 75.4%, and the functional approach was associated with a higher observed success rate than the traditional approach.
• Six predictors remained independently associated with later success across rehabilitation approach, quadriceps strength symmetry, hop performance, psychological readiness, dynamic balance, and proprioception.
• Model discrimination was high, bootstrap correction reduced performance modestly, and the investigators noted that external validation is still needed before implementation.

At 48 weeks, return-to-play success among completers was 89.3% with a functional rehabilitation model and 62.1% with a traditional model among high-sports-demand athletes randomized after primary ACL reconstruction. The Scientific Reports trial focused on nonprofessional athletes who regularly participated in cutting, pivoting, and jumping sports rather than a broader ACL population. The 24-week assessment informed a model for later 48-week success.

This single-blind, assessor-blinded prospective randomized controlled trial was conducted at the First Affiliated Hospital of Xi’an Medical University from April 2024 to May 2025. High-sports-demand status referred to nonprofessional athletes aged 13 to 45 years with Tegner scores above 5 who played level-1 cutting, pivoting, and jumping sports at least three times weekly. Sixty-four participants after primary ACL reconstruction were randomized 1:1 to functional or traditional rehabilitation, and 57 completed follow-up for 89.1% retention. The 24-week assessment included isokinetic strength, hop tests, the modified Star Excursion Balance Test, proprioception, and patient-reported outcomes including IKDC, Lysholm, and ACL-RSI. Successful return to play at 48 weeks required functional testing, psychological readiness, postoperative time of at least eight months, and self-reported return to preinjury sport level.

The final multivariable model retained six independent predictors: rehabilitation approach, 60°/s extensor limb symmetry index, single-leg hop distance, ACL-RSI, mSEBT anterior reach, and 45° proprioceptive error. Functional rehabilitation was associated with an odds ratio of 3.49 versus traditional rehabilitation in the adjusted model. Reported predictive cut points were extensor LSI ≥77.8%, hop distance ≥68.5 cm, ACL-RSI ≥67.5, mSEBT anterior reach ≥63.2 cm, and 45° proprioceptive error ≤6.1°. Effect estimates for the remaining measures were 2.13 per 1% increase in extensor LSI, 1.87 per 1-cm increase in hop distance, 1.69 per 1-point increase in ACL-RSI, 1.39 per 1-cm increase in mSEBT anterior reach, and 0.62 per 1° increase in proprioceptive error. Together, the retained variables spanned rehabilitation exposure, strength, hop performance, balance, psychological readiness, and sensorimotor accuracy.

Discrimination was high, with an AUC of 0.87 and a 95% confidence interval of 0.79 to 0.95; apparent sensitivity was 88.4% and specificity was 78.6%. Bootstrap correction reduced the AUC to 0.81, with a corrected 95% confidence interval of 0.72 to 0.90, and calibration was adequate with Hosmer-Lemeshow P=0.555 and Nagelkerke R²=0.58. Investigators also outlined a preliminary 0-to-8-point scoring tool that grouped participants into high-, moderate-, and low-probability strata. Follow-up losses included one secondary injury after unauthorized return to play in the functional group, one withdrawal for persistent pain at 12 weeks, and relocations in both groups.

The authors cited single-center design, small sample size, complete-case analysis, stepwise modeling, events-per-variable below 10, and possible ACL-RSI criterion contamination, so external validation remains necessary before implementation.