Return to play after injury: controlling the load

Returning to play after an injury is not a date on the calendar: it is a process that runs in parallel with recovery and rehabilitation (Ardern et al., 2016). The published literature, including work from elite football, proposes deciding by criteria —risk, risk tolerance and a load that is rebuilt progressively— rather than by days elapsed. This guide summarises that work and says where the evidence is limited.

What does returning to play mean?

The Bern consensus, written by 17 experts, defines return to sport as a continuum with three elements: return to participation, return to sport and return to performance, meaning performing at or above the pre-injury level (Ardern et al., 2016). Defining it requires, at a minimum, the sport and the level being returned to: their example is the English Premier League. The decision is shared among everyone involved, unless there is a risk to health.

The authors state that the research evidence to support these decisions is scarce: it is a consensus of principles, with no numeric criteria.

Deciding with risk and risk tolerance

The StARRT framework (Shrier, 2015) proposes comparing the risk of the activity with the risk tolerance, in three steps: tissue health (what stress the tissue can absorb), the stress the planned activity applies to it, and the risk tolerance modifiers, which are the context. If the risk assessment is lower than the risk tolerance, the decision is to return.

StARRT is a transparent way of reasoning, not a cut-off rule or a validated predictor: Shrier acknowledges that research into predictive algorithms might be beneficial. The Bern consensus adopts it as an aid for synthesising information (Ardern et al., 2016). Training load enters as the stress the activity applies to the tissue.

Why it matters so much in football

The UEFA Elite Club Injury Study followed 3,909 players from 54 teams in 20 European countries over 21 seasons (2001/02 to 2021/22). The 2,636 hamstring injuries were 19% of all recorded injuries, and their share rose from 12% in the first season to 24% in the most recent one. A 25-player squad can expect about eight hamstring injuries per season, and 20% of players miss some training or match play because of them in a season (Ekstrand et al., 2023).

Other numbers from the same study:

• Match incidence was 4.99 per 1,000 hours versus 0.52 in training (RR 9.67).
• 475 injuries (18%) were recurrences, and 69% of those (325) happened within two months of the return.
• Recurrences were nine times more likely in matches than in training (RR 9.25).

These percentages apply to hamstrings, not to every injury. The 24% is the share in the latest season; the average over the 21 seasons is 19%.

Rebuilding chronic load before the match

Buckthorpe (2019), in a narrative review on ACL reconstruction, lists among the problems of the return-to-sport process an insufficient development of chronic load before going back to competition. It is a review, not a study with a quantified effect, and it refers to the ACL.

For hamstrings, Duhig et al. (2016) studied a single Australian football club over two seasons. HSR session means above each player’s own typical level (their two-year average) were associated with more injuries: OR 6.44 in the previous week, 3.06 when summing weeks −1 and −2, and 1.96 when summing weeks −1 to −4. For sRPE and total distance, differences were trivial. It is an observational study from a different sport, and the authors only suggest, from a model, that reducing HSR in the previous week might lower the probability of injury: that was not tested as an intervention.

Hence the idea of raising high-speed load gradually and relative to the individual player, not with one value for everyone.

From control to chaos: phases on the pitch

Taberner, Allen and Cohen (2019), who worked at Everton and Arsenal, propose the control-chaos continuum: five phases (high control, moderate control, control to chaos, moderate chaos and high chaos) that combine the player’s earlier chronic loads measured with GPS and the estimated tissue healing times. They base it on more than a decade of rehabilitation and return to sport in Premier League football.

Those figures (60% and 75%) are the authors’ examples, not validated standards, and the framework has no fixed thresholds. In a published case from the same group, a Premier League player with an open tibia-fibula fracture returned to full team training after 7.5 months and played 90 minutes after 9 months (Taberner et al., 2019). There HSR was defined as above 5.5 m/s and sprinting as above 7 m/s. It is a single case: it illustrates the method, it does not show that it works.

Criteria, not calendar

For Taberner et al. (2019), progression is criteria based, not time dependent. In their example of a central defender with a hamstring injury, return to training was estimated at about six weeks, and the model is shortened or lengthened according to the injury. In the fracture case, the criteria to move up a phase were pain below 2/10, no joint effusion and achievement of the load targets, with daily communication between performance staff, the medical team and the player. Strength and power tests back up clinical reasoning, they do not replace it. The final maximal speed was 9.2 m/s, 103% of the pre-injury value.

The tension is stated in the article itself: early reintegration increases the risk of reinjury, but having key players available benefits the team.

In practice

A staff can organise the return like this: define what returning means (participating, playing, performing), rebuild chronic load with measured external load, raise HSR and sprinting gradually and relative to the individual player, and move up a phase on clinical and performance criteria agreed with the medical team. In Athlo, the detail of an injury can show the load context available around the date it occurred, as a retrospective reading for the staff’s decision ; more on that in health and availability and in GPS load. The software shows the load: the decision stays with the coaching and medical staff.

Limits of the evidence

The Bern consensus admits that the evidence is scarce. StARRT and the control-chaos continuum are reasoning frameworks, not validated algorithms. The fracture case is a single case, Buckthorpe’s review is narrative and concerns the ACL, and Duhig is observational and from a different sport. None of them shows that GPS predicts an injury or that there is an ACWR value or a number of days that works for everyone.

Sources

• Ardern CL, Glasgow P, Schneiders A, et al. (2016) “2016 Consensus statement on return to sport from the First World Congress in Sports Physical Therapy, Bern” British Journal of Sports Medicine 50(14):853–864. https://doi.org/10.1136/bjsports-2016-096278
• Shrier I (2015) “Strategic Assessment of Risk and Risk Tolerance (StARRT) framework for return-to-play decision-making” British Journal of Sports Medicine 49(20):1311–1315. https://doi.org/10.1136/bjsports-2014-094569
• Taberner M, Allen T, Cohen DD (2019) “Progressing rehabilitation after injury: consider the ‘control-chaos continuum’” British Journal of Sports Medicine 53(18):1132–1136. https://doi.org/10.1136/bjsports-2018-100157
• Taberner M, van Dyk N, Allen T, et al. (2019) “Physical preparation and return to sport of the football player with a tibia-fibula fracture: applying the ‘control-chaos continuum’” BMJ Open Sport & Exercise Medicine 5(1):e000639. https://doi.org/10.1136/bmjsem-2019-000639
• Buckthorpe M (2019) “Optimising the late-stage rehabilitation and return-to-sport training and testing process after ACL reconstruction” Sports Medicine 49(7):1043–1058. https://doi.org/10.1007/s40279-019-01102-z
• Ekstrand J, Bengtsson H, Waldén M, Davison M, Khan KM, Hägglund M (2023) “Hamstring injury rates have increased during recent seasons and now constitute 24% of all injuries in men’s professional football: the UEFA Elite Club Injury Study from 2001/02 to 2021/22” British Journal of Sports Medicine 57(5):292–298. https://doi.org/10.1136/bjsports-2021-105407
• Duhig S, Shield AJ, Opar D, Gabbett TJ, Ferguson C, Williams M (2016) “Effect of high-speed running on hamstring strain injury risk” British Journal of Sports Medicine 50(24):1536–1540. https://doi.org/10.1136/bjsports-2015-095679