Ilizarov complications: the real rates, series by series

Dror Paley set the vocabulary in 1990, in a paper on 46 patients and 60 segments lengthened with the Ilizarov technique. A problem is a difficulty resolved without an operation. An obstacle needs a further operation. A complication is an intraoperative injury or any difficulty still unresolved when treatment ends. That series produced 35 problems, 11 obstacles and 27 true complications, 17 minor and 10 major; three of the major ones (one nonunion, two cases of late bowing) got in the way of the goal. The goal was still reached in 57 of 60 segments, and 94% of patients were satisfied.

The taxonomy decides what a clinic's number means, and most Ilizarov frame problems sit in the first two buckets. A centre that counts every pin-site dressing change will report a higher rate than one that counts only events needing surgery, and both can be honest. Marwan's 2020 systematic review applied Paley's grading to 795 cosmetic patients in 11 studies. The 267 lengthened with a classic three- or four-ring Ilizarov frame had 0.7 problems, 1.5 obstacles and 0.2 true complications per patient. Implantable nails scored lower (0.16, 0.23 and 0.02), but the difference was not statistically significant with only 63 nail patients.

When you read any Ilizarov complication rate, including every one below, ask which bucket it counts.

The reference dataset for Ilizarov complications in cosmetic patients is Novikov and colleagues, 2014, from the Ilizarov Centre in Kurgan: 131 healthy adults (mean age 25) lengthened for height, bilaterally, 124 of them (95%) in the tibia alone. Mean gain was 6.9 cm (range 2 to 13). Mean time in the frame was 215 days, 79 of distraction and 116 of consolidation, an external fixation index of 31 days per centimetre.

Forty-eight patients (37%) had 59 complications: 37 soft-tissue and 22 bone-related. Twenty-three patients (18%) needed a reoperation, and bone-related events were far more likely to need one (16 of 22) than soft-tissue events (17 of 37). The outcome was graded excellent in 72 patients (55%), good in 52 (40%), satisfactory in six and poor in one, and 130 of 131 said they were satisfied. Both statements are true at once: most patients got what they came for, and more than a third had something go wrong on the way.

Two details are worth carrying into a consultation. The authors put the practical limit for a tibial lengthening at 5 to 7 cm, beyond which outcomes were less reliable. And older patients tolerated only a slower distraction rate, under 1 mm per day.

Every wire and half-pin is a permanent opening from skin to bone, kept open for months, which is why pin-site infection heads every list of external fixator complications. Ferreira and Marais, reviewing the literature in 2012, put the quoted incidence of pin-track infection at 11.3% to 100% and called it almost inevitable with long-term external fixation. The spread is a definition problem: series that record every red, weeping pin site approach 100%; series that count only infections needing treatment sit near the bottom. One detail favours the Ilizarov design: monolateral and hybrid fixators, built on thick half-pins, showed a much higher incidence than ring fixators built on thin tensioned wires.

The cosmetic series land across that range. Novikov recorded five pin-tract infections needing intervention among 131 patients (3.8%), plus three cases of osteomyelitis (2.3%), all of which cleared. Catagni's 2005 tibial series recorded 26 superficial pin-site infections among 54 patients, and Guerreschi's 2016 ring-fixator series 25 among 63. In Liu's 2020 bone-transport series of 282 trauma and infection patients, who wear frames far longer, pin-site infection reached 65.96%, the single most frequent event.

Most are grade 1 or 2 on the Checketts-Otterburn scale: redness and discharge, treated with better pin care and a short course of oral antibiotics; Ferreira and Marais cite one series in which 94% of infections were mild. Grade 3 means the pin has to be moved; grades 4 to 6 mean the frame comes off or the tract needs surgery. Loosening often comes first and infection second, so a pin that feels loose should be reported the same day. The regimen the review's authors use is a chlorhexidine-alcohol dressing left for seven days, then twice-daily chlorhexidine cleaning, showers once the sites have healed, and the leg elevated whenever the patient is not walking.

The biggest category in every cosmetic series is not infection. It is joints and tendons. Giorgino's 2025 systematic review of 12 studies and 760 aesthetic-lengthening patients logged 425 joint and tendon mentions against 197 infectious ones. In Novikov's 131, the two most frequent single events were a fixed flexion deformity of the knee (14) and ankle equinus (12). Guerreschi's 63-patient tibial series recorded 42 equinus feet.

The mechanism is mechanical. Lengthening the tibia stretches the calf and the Achilles tendon faster than they adapt, so the foot drifts into plantar flexion; lengthening the femur does the same to the hamstrings and the knee. Wires through muscle add friction and pain that discourage movement. Paley's 1997 matched comparison of 32 Ilizarov femoral lengthenings with 32 lengthenings over a nail found knee motion returned to normal 2.2 times faster when the frame came off early: a direct measure of what months in a frame cost the joint.

Nails do not abolish the problem. In Frost's 2023 cohort of 314 FITBONE and PRECICE nail segments, contracture was still the leading joint complication, with 51 recorded, and joint events affected 23% of segments. What the frame adds is duration and pin friction. The countermeasures are unglamorous: daily physiotherapy from the first week, splinting the foot at neutral, sometimes extending the frame to the foot, and slowing distraction when a joint starts to lose range. Ask any Ilizarov clinic how many of its patients finish with a stiff knee or a dropped foot.

Bone-side events are rarer than soft-tissue ones but more likely to need surgery. In Novikov's series 22 of the 59 events were bone-related and 16 of those required a reoperation: six delayed consolidations, five deformities that developed during fixation, four that appeared after the frame came off, two technical problems, one knee subluxation and one delayed fracture.

Axial deviation, the regenerate drifting out of line as it grows, is the classic frame-specific event, because muscle pull is asymmetric and a ring frame resists it imperfectly. The two classic-Ilizarov cosmetic studies pooled by Marwan recorded ten cases of residual axial deviation. In Liu's 282 bone-transport patients, whose frames stayed on for about a year, axial deviation affected 40.78% and joint stiffness 23.76%. The fix is usually a frame adjustment in clinic (a problem, in Paley's terms) and occasionally a repeat osteotomy (an obstacle).

Consolidation misbehaves in both directions. Delayed consolidation showed up in 6 of Novikov's 131 and at 13.48% of Liu's docking sites. Premature consolidation, the bone bridging before distraction is finished, sits in Paley's 1990 list and forces a repeat osteotomy; it does not appear among Novikov's 59 events, and Frost recorded 3 cases in 314 nail segments against 16 of delayed healing. Age and time in the frame were the strongest predictors of complications in Liu's regression, and Frost found the same age effect with nails: relative risk roughly doubled over age 30, and was 1.7 times higher in the tibia than the femur.

The frame carries the load while it is on. The day it comes off, the new bone has to. Simpson and Kenwright reviewed 173 distraction-osteogenesis patients and found 17 fractures in 180 lengthened segments, 9.4%. Six were inside the regenerate, six at the junction between new and old bone, and five elsewhere in the limb; in three patients deformity developed gradually over the six months after removal rather than as a single event.

Paley's 1997 matched comparison is the sharpest illustration of what internal support prevents: six refractures of the distraction bone among the 32 Ilizarov femora, none in the 32 femora lengthened over a nail, which had one broken nail and one broken locking screw instead. Novikov's cosmetic cohort recorded a single delayed fracture in 131 patients, and the protocol is the likely reason: a stress test of the regenerate at the time of removal, then a thigh-to-ankle cast for two to four weeks before full load.

The practical questions are three. Does the clinic test the regenerate before removal, or go by the calendar? What protection is used afterwards, and for how long? And who reviews the X-rays in the six months after removal? Paley lists loss of length, late bowing and refracture as the late complications precisely because they arrive after the patient has stopped expecting anything to go wrong.

Nerve injury is uncommon but real. Novikov recorded common peroneal neuropathy in six of 131 patients (4.6%), treated by slowing distraction; five recovered fully, one did not despite surgical exploration. The nerve wraps around the neck of the fibula, close to the wires of a proximal tibial ring, and the stretch of distraction adds to the risk. Frost's nail cohort logged seven neurological events in 314 segments, a similar order of magnitude.

Then there is the part no rate captures. Novikov's patients wore the frame for a mean of 215 days, range 71 to 390. Marwan's classic-Ilizarov group averaged 42.5 days in the frame per centimetre gained, and the external fixation period across the whole review averaged 201 days, range 24 to 810. Paley's 1990 adults needed 1.7 months per centimetre for a single-level lengthening. Every one of those days includes pin care, hardware in the bed, and the two routine problems Paley listed for extensive cases: pain and difficulty sleeping. Giorgino's review counted 90 pain-and-discomfort mentions and names insomnia as part of the recovery.

Long-term data are thinner but not alarming. Moraal followed 37 patients lengthened as children (mean 4.1 cm) for a mean of seven years: self-esteem was unchanged, perceived appearance improved, and most quality-of-life domains matched the healthy population, with three exceptions: more pain, less vitality and lower gross-motor scores. Twenty-nine of the 37 (78%) would have the treatment again. Those left with a residual discrepancy above 2 cm rated satisfaction 5.1 out of 10 against 7.6. What you live with afterwards is the length and alignment you finish with, not the months in the frame.

Classic ring frames accounted for 267 of the 795 cosmetic patients (33.6%) in Marwan's review, the largest single technique. Three reasons keep the frame in use. Cost: Paley's 1997 comparison found treatment cheaper with the frame than with lengthening over a nail, because rings and wires cost a fraction of an implant; the Kurgan centre's full-course figure in our directory is $8,000 to $20,000, patient-reported, covering surgery, apparatus, hospital stay and removal. Versatility: the frame corrects angulation and rotation while it lengthens, works through infection, and can be adjusted at any point. And it leaves nothing inside the bone.

The trade is not complications versus no complications. Frost's 2023 cohort of 257 patients with 314 motorised nails found that 53% of patients had a complication; device-related events affected 32% of segments, joint events 23%, and infection ran at 0.05 per segment (17 events: six superficial, four deep soft-tissue, three osteomyelitis, four other). A nail moves the risk from the skin and the joints to the device and the removal surgery. A frame keeps the risk visible, at the pin sites, for the whole treatment.

Read this way, Ilizarov complications are predictable rather than mysterious: pin-site infection in a large minority, a stiff knee or a dropped foot in roughly one patient in ten each, a reoperation for any cause in 18%, and a fracture risk after removal that a stress test and a cast reduce. If a clinic offers you a frame for a purely cosmetic lengthening, the Ilizarov method risks above are the checklist: ask why the frame and not a nail or a hybrid, what its own numbers are in each of Paley's three categories, and what its pin-site and splinting protocols say in writing. The method pages on the Ilizarov frame and the Taylor Spatial Frame cover indications, and the surgeon directory lists who performs each.