Research · Injury & Recovery

The Cast Comes Off at Six Weeks. The Refracture Clock Runs to Ten.

By Muhammad Dzulhisham · MSc Sports Coaching, NSCA-CSCS

A ten-year-old breaks a forearm in September. Cast for six weeks, comes off in late October, the follow-up film looks fine, and the message to the family is that the bone has healed. Football resumes in early November.

A Finnish registry study suggests the first week of November is roughly when the risk peaks.

How often it actually happens

Researchers reviewed 20,749 fractures in children under 16 from a hospital fracture register, identified every case of a second fracture in the same bone within two years, and excluded pathological fractures and systemic conditions. One hundred genuine refractures remained.¹

The headline reading is reassuring: refracture is uncommon, and an anxious parent can be told so. The second reading is the one that should change practice, because the rate is not evenly spread and neither is the timing.

The timing is the finding

Median time to refracture, by location:¹

Seventy-three days is about ten and a half weeks. A distal forearm fracture is commonly immobilised for four to six. The median refracture therefore lands roughly four to six weeks after the cast came off — which is to say, after the child returned to sport, once they had stopped being careful with it, and at the point where everyone involved had mentally closed the file.

Compare that with the distal humerus at a median of 426 days and the practical lesson is that "how long until this is safe" has no single answer. The bone matters more than the calendar.

The same study found that displaced fractures requiring closed reduction carried a relative risk of refracture of 8.0 (95% CI 4.5–14) against stable fractures.¹ A reduction in the emergency department is, among other things, information about the following six months.

What treatment choice does to the risk

A retrospective cohort of 837 children treated for diaphyseal forearm fractures across two university hospitals compared treatment modalities, adjusting for age.² Against flexible intramedullary nailing, the relative risk of refracture within a year was 9.8 (95% CI 5.9–16.3) for nonoperative treatment and 2.5 (1.2–5.3) for closed reduction; malunion risks were similarly elevated.

Those are dramatic numbers and they need their caveat immediately: 90% of that cohort received nailing, 6% closed reduction and 4% nonoperative treatment. When a comparison group is 4% of the sample and treatment was not randomised, the children treated nonoperatively differed systematically from those who went to theatre. This is a signal about who gets which treatment as much as about what each treatment does.

The contrast with a different bone is instructive. A prospective multicentre cohort followed 136 adolescent athletes aged 14 to 18 with comminuted or severely shortened midshaft clavicle fractures, 69 treated nonoperatively and 67 operatively.³ At two years there were no differences in nonunion, delayed union, symptomatic malunion, refracture, clinically significant complications, rates of return to sport, or patient-reported outcomes. The only significant difference was timing: 10.3 weeks to return with surgery against 13.5 weeks without.³

Three weeks earlier, at the cost of an operation, with identical outcomes at two years. That is a trade-off a family can actually weigh, and it is a very different conversation from the forearm one.

Stress fractures play by different rules

A traumatic fracture has a clean start date. A bone stress injury does not, and the return timeline reflects it. A multicentre case series of 110 adolescents aged 10 to 19 with tarsal navicular bone stress injuries — mean age 14.7, 65% female — found the injuries concentrated in cross-country and track (32%) and gymnastics and dance (27%).⁴

Nonoperative treatment with protected weightbearing was trialled in every patient and succeeded in 85%. The 15% who went to surgery were older, had higher BMI, and were far more likely to have a visible fracture line on imaging — 88% against 38% in the nonoperative group. Median return to full sport after fixation was around five months.⁴

The relevant coaching point is that this injury arrives through training load rather than contact, which means the athlete returns to the same environment that produced it. Nothing about the healing timeline addresses that.

What this changes in practice

Do not treat cast removal as the end of the injury. For the distal forearm, the median refracture sits about a month after immobilisation ends. The weeks of highest risk are the weeks everyone has relaxed.

Ask which bone, and whether it was reduced. Both-bone diaphyseal forearm at 3.76% and a closed reduction carrying eight times the refracture risk of a stable fracture are the two facts that should change how cautiously a return is staged.

Stage the return by contact risk, not just by symptoms. A child can be pain free and full range long before the bone tolerates a fall onto an outstretched hand — the same gap between pain-free and actually ready that shows up in soft-tissue injury. Non-contact conditioning, then contact, is a sequence the timing data supports.

For a stress fracture, fix the load. Eighty-five per cent of navicular injuries resolved without surgery, but the training pattern that caused them does not resolve on its own.

Let the surgical conversation be about timing, honestly. In adolescent clavicle fractures, surgery bought about three weeks and changed nothing at two years. Families deserve that framed as the trade-off it is.

Where this evidence stops

The registry study is retrospective and counts only radiographically confirmed refractures at the same hospital system; a child who refractured elsewhere, or whose second injury was never imaged, is invisible to it. It also reports timing, not causation — nothing in it establishes that returning to sport caused those refractures rather than merely coinciding with them.

The forearm treatment comparison has a 4% nonoperative arm and no randomisation. The clavicle study is the strongest design here, and it addresses one bone in one age band with one fracture pattern. The navicular series is retrospective with no control group.

Most importantly, none of these studies tested a return-to-sport protocol. They describe when refractures happen, not what would have prevented them. Clearance after a fracture is a clinical decision, and the honest contribution from this literature is a question worth putting to the clinician: given this bone and this fracture, when does the risk window actually close?