Posterior approach — tension band wiring (transverse) or dorsal plate fixation (comminuted/oblique) | intermediate
- Mayo classification (I undisplaced, IIA/B displaced stable, IIIA/B unstable with elbow dislocation) drives fixation choice and prognosis
- Tension band wiring (TBW) converts the triceps extension force into compression at the fracture site — valid only for transverse fractures with less than 80 percent articular surface involvement
- Plate fixation is preferred for comminuted, oblique or osteoporotic fractures, Mayo type III instability, and any fracture involving more than 80 percent of the articular surface
- Hardware prominence causes symptomatic discomfort requiring removal in up to 80 percent of TBW cases — counsel patients pre-operatively
- “TBW is indicated for simple transverse fractures involving less than 80 percent of the articular surface in good bone quality — not for comminuted, oblique or proximal fractures
- “Plate fixation is preferred for comminuted or oblique fractures, fractures involving more than 80 percent articular surface, osteoporotic bone, and Mayo IIIA/B instability patterns
- “Hardware prominence accounts for most of the re-operation rate after TBW — bury K-wire tips and bend them 180 degrees to reduce prominence
- “Avoid leaving K-wires too long or failing to engage the anterior cortex — both lead to K-wire migration, the most feared early complication
When & Why
Indication. Operative fixation of a displaced olecranon fracture (Mayo IIA, IIB, IIIA, IIIB) — pain, loss of active extension, a palpable gap and an inability to extend against gravity — once skin and soft-tissue cover are adequate. The goal is to restore the articular surface and the triceps extensor mechanism so the elbow can move early. Mayo type I fractures stay non-operative (less than 2 mm displacement, intact extensor mechanism): a posterior splint or collar-and-cuff in 45–90 degrees flexion for 2–3 weeks, then early active ROM, with serial radiographs at 1 and 3 weeks. Secondary displacement of greater than 2 mm is a failure of non-operative care and needs surgery. Outcomes are comparable to operative management in genuinely undisplaced fractures (Duckworth et al., JBJS 2014). Three intra-operative decisions frame the case. - Decision 1 — TBW or plate? Reassess the pattern under direct vision. A simple transverse fracture in good bone in a young patient proceeds with TBW; switch to a plate if it is more oblique, more comminuted, or has poorer bone quality than expected on imaging. Changing the planned technique on direct assessment is good surgical judgement — say so in the viva.
- Decision 2 — articular congruity. After every reduction, confirm no articular step-off under direct vision and on fluoroscopy. A step-off of greater than 2 mm is unacceptable; post-traumatic elbow arthritis tracks directly with articular incongruity.
- Decision 3 — elbow stability (Mayo III). After fixing the olecranon in any fracture that had subluxation or dislocation, stress-test under fluoroscopy: valgus that opens more than 3 mm implies MCL injury; varus stress tests the LUCL (posterolateral rotatory instability); posterior stress in extension tests posterior stability. Address each component before closure, or the elbow re-dislocates.
For simple transverse fractures (Mayo IIA) in good bone quality, involving less than 80 percent of the articular surface, in active patients with good bone stock — also oleotomy fixation as part of a distal humerus approach.
For comminuted fractures (IIB/IIIB), oblique patterns, fractures involving more than 80 percent articular surface, osteoporotic or elderly bone, any Mayo III instability, very proximal fragments too small for K-wires, and fractures extending into the coronoid.
For Mayo I (less than 2 mm, intact extensor mechanism) and for low-demand elderly patients with a displaced fracture, where a randomised trial shows non-operative care gives equivalent one-year function with far fewer complications.
Consent specifically for hardware prominence and the high likelihood of a second procedure to remove metal after TBW (up to 80 percent), symptomatic scar over the olecranon tip, a small risk of K-wire migration, loss of fixation or non-union (2–5 percent), infection (1–3 percent) and ulnar nerve symptoms (2–5 percent, usually a resolving neuropraxia).
The Operation
The goal is to expose the fracture through a posterior approach, protect the ulnar nerve, restore an anatomical articular surface, and stabilise it with a construct matched to the pattern — a tension band for a simple transverse fracture, a plate for everything comminuted, oblique, osteoporotic or unstable. The exposure is laid out in full below (and in depth on the posterior approach to the elbow page).
Operative sequence
- Lateral decubitus with the injured elbow uppermost over a padded bolster (allows full elbow extension and easy fluoroscopy); alternatives are prone on chest rolls or supine with the arm across the chest on a hand table.
- Sterile or non-sterile pneumatic tourniquet on the proximal arm, inflated to 250 mmHg; consider release after 90 minutes if prolonged.
- Confirm clear AP (elbow extended) and lateral (90 degrees flexion) fluoroscopy views before draping — the lateral view is the one that confirms K-wire anterior-cortex engagement.
- A 10–12 cm posterior midline incision centred over the olecranon tip, curved slightly to one side of the tip so the scar does not sit directly on the bony prominence.
- Raise full-thickness medial and lateral skin flaps.
- Identify and protect the medial cutaneous nerve of the forearm in the subcutaneous tissue.
- Open the fracture haematoma; identify the fracture plane and confirm the pattern (transverse, simple).
- Irrigate and remove loose clot and small bony debris.
- Preserve all viable periosteum and soft-tissue attachments to the fragment — this maintains the blood supply.
- Locate the ulnar nerve behind the medial epicondyle and protect it with a vessel loop for the whole case.
- Never retract medially without confirming the nerve's position; it runs 1–2 cm medial to the olecranon.
- Reduce the fracture anatomically under direct vision with a pointed (Weber) clamp across the site.
- Confirm anatomical articular congruity — any step-off is unacceptable.
- Hold provisionally with a K-wire or bone clamp while definitive fixation is applied.
- Simple transverse, good bone, young patient: proceed with tension band wiring (Steps 7–10).
- Oblique, comminuted, osteoporotic, very proximal, or any Mayo III instability: switch to plate fixation (Step 11).
- Insert two parallel 1.6 mm (18-gauge) Kirschner wires from the posterior olecranon tip, angled 45 degrees into the intramedullary canal.
- Both wires must penetrate and engage the anterior cortex of the ulna, exiting it by 2–3 mm — confirm on lateral fluoroscopy.
- Divergent or convergent wires do not give adequate rotational stability.
- Drill a transverse hole through the ulna 3–4 cm distal to the fracture, at the ulnar crest.
- Pass a 1.25 mm (18-gauge) stainless-steel wire through this hole, then through the substance of the triceps tendon (not superficially around it) at the olecranon tip, crossing in a figure-of-8 around the K-wire tips.
- Tighten both limbs of the figure-of-8 symmetrically with a wire twister; confirm the fracture compresses on fluoroscopy.
- Avoid asymmetric tightening, which tilts the fragment.
- The completed loop must lie anterior to the K-wire tips — anterior to the axis of rotation — so elbow extension generates compression, not distraction.
- Bend the protruding K-wire tips 180 degrees (shepherd's hook) with a pin bender and bury them into the posterior cortex or triceps tendon.
- Confirm the tips are not prominent on fluoroscopy — this reduces hardware irritation and prevents proximal migration.
For comminuted, oblique, osteoporotic or unstable patterns, after reduction and provisional K-wire fixation:
- Apply a pre-contoured dorsal (or 90-degree lateral) locking plate to the posterior olecranon spanning the fracture; locking screws give angular stability, vital in osteoporotic bone.
- Proximal screws engage the olecranon fragment with 3–4 cortices of purchase; aim for 3–4 bicortical screws distally in the diaphysis.
- A 90-degree lateral plate along the lateral column avoids direct posterior subcutaneous prominence and may lower symptomatic-removal rates.
- For comminuted zones use bridge plating (do not over-compress); small articular fragments may need lag screws perpendicular to the fracture before the plate goes on.
- Confirm articular congruity on fluoroscopy before final tightening.
- Take the elbow through full ROM under fluoroscopy; confirm reduction is maintained through the arc and the construct generates compression in extension.
- Look for impingement or articular step-off; test the extensor mechanism — the patient must extend against gravity.
- In any Mayo III pattern, perform the valgus/varus/posterior stress test and address residual instability before closure.
- Irrigate copiously; close periosteum and deep fascia over the hardware, then subcutaneous tissue and skin.
- Apply a posterior plaster splint at 45–90 degrees flexion and elevate the limb.
Fluoroscopy checkpoints — imaging is mandatory at each of these steps: | Step | View | What to confirm | |------|------|-----------------| | After K-wire insertion | Lateral | Both K-wires engage the anterior cortex; parallel orientation | | After wire tightening | AP and lateral | Fracture compressed, no articular step-off | | After K-wire bending | Lateral | Bent tips not prominent, no loop loosening | | Final check | AP, lateral, oblique | Full reduction, hardware position, no intra-articular penetration | | ROM check | Lateral (dynamic) | No impingement, reduction maintained through the arc | Equipment checklist: | Item | Specification | Purpose | |------|---------------|---------| | Kirschner wires | 1.6 mm (18-gauge) parallel pair | Intramedullary purchase, wire anchor | | Tension band wire | 1.25 mm (18-gauge) stainless steel | Figure-of-8 compression loop | | Wire twister | Kocher clamp or dedicated twister | Symmetric tightening | | Wire bender | Pin bender | 180-degree tip bend | | Weber clamp | Pointed reduction clamp | Provisional reduction hold | | Pre-contoured plate | Dorsal or lateral locking olecranon plate | Plate fixation option | | Fluoroscopy | C-arm | Intra-operative imaging |
Before any reduction, identify and vessel-loop the ulnar nerve behind the medial epicondyle — it is the structure most often harmed by medial retraction. For TBW, both K-wires must engage the anterior ulnar cortex (confirm on lateral fluoroscopy) and be bent 180 degrees and buried — these two measures together prevent the most feared early complication, K-wire migration. If the artery of the fracture pattern does not suit TBW (comminution, obliquity, poor bone, instability), convert to a plate rather than force a tension band that will fail.
The elbow's axis of rotation runs through the centre of the trochlea; triceps contraction applies a posterior distracting force to the olecranon, so a figure-of-8 wire placed anterior to that axis converts that distracting force into compression at the fracture — but only when the fracture is transverse and the posterior cortex is intact. An oblique fracture shears under the wire; a comminuted one has no stable cortex to compress.
Reach for a plate whenever the fracture is comminuted, oblique, osteoporotic, very proximal, involves more than 80 percent of the articular surface, or carries elbow instability (Mayo III). Each of these defeats the tension-band mechanism: comminution splits under compression, osteoporotic bone does not hold K-wires, and a tiny proximal fragment gives no K-wire purchase.
Aftercare & Complications
Rehabilitation | Phase | Timing | Goals | |-------|--------|-------| | Protection | 0–2 weeks | Posterior splint at 45–90 degrees for comfort, elevation, wound care | | Early motion | 2–6 weeks | Splint off at 2 weeks; active-assisted flexion/extension and forearm rotation; no passive stretching (heterotopic ossification risk); target 30–120 degrees by 6 weeks | | Strengthening | 6–12 weeks | Progressive resistance through a functional arc; triceps strengthening from week 6 | | Return to activity | 3–6 months | Full activity, sport and manual work | Extensor mechanism assessment. Test active elbow extension against gravity before discharge and at the first follow-up; any extension lag warrants investigation. Hardware removal. Confirm radiographic union first (typically 12–18 months). Removal is needed in 70–80 percent of TBW cases (symptomatic prominence) versus 15–25 percent of plates — counsel patients that for TBW a second day-case procedure is expected, not a complication. After removal, restrict heavy lifting for 6 weeks because of re-fracture risk through screw holes or K-wire tracts (1–3 percent). Complications
- Incidence
- 70–80 percent TBW; 15–25 percent plate
- Recognition
- Palpable K-wire tips or wire under skin; pain on direct pressure; bursa; skin irritation
- Management
- Removal once union is confirmed (12–18 months). A planned, expected procedure — counsel pre-operatively
- Incidence
- 5–15 percent
- Recognition
- Proximal migration toward the shoulder or distal migration into soft tissue on radiograph; patient feels hardware move
- Management
- Urgent removal if migrating toward neurovascular structures. Prevent at the index operation with a 180-degree bend and anterior-cortex engagement
- Incidence
- 2–5 percent
- Recognition
- Persistent fracture-site pain; palpable motion; progressive radiolucency; hardware loosening
- Management
- Revision fixation with a plate (preferred) and autogenous bone graft; consider total elbow arthroplasty in the elderly with a failed non-union and articular damage
- Incidence
- 1–3 percent
- Recognition
- Wound erythema, discharge, fever, raised CRP/ESR, implant loosening
- Management
- Superficial: oral antibiotics and wound care. Deep: formal debridement, retain a stable implant until union, remove once united; guided by microbiology
- Incidence
- 2–5 percent
- Recognition
- Numbness or paraesthesia in ring and little fingers; intrinsic weakness; Froment sign; usually a resolving neuropraxia over 3–6 months
- Management
- Observe and protect if neuropraxic; EMG/NCS at 3 months if no recovery; neurolysis or decompression by 6 months if none; permanent deficit is rare with careful technique
- Incidence
- 1–3 percent
- Recognition
- Fracture through screw holes or K-wire tracts after removal, typically at 6–12 weeks
- Management
- Avoid heavy lifting for 6 weeks after removal; early re-fracture may need re-fixation. Counsel before the removal procedure
Viva & Exam Focus
MAYOMAYO — olecranon fracture classification
PLATEPLATE — when to choose a plate over tension band wiring
Five danger zones — the structures and mechanisms examiners probe:
Runs medial to the olecranon in the cubital tunnel, 1–2 cm medial, entering between the heads of FCU. At risk during medial retraction and medial plating. Identify and vessel-loop it before any medial dissection; transpose anteriorly if a medial plate is planned.
Parallel wires that do not engage the anterior cortex or are not bent 180 degrees can migrate proximally toward the shoulder or distally into soft tissue. Prevent by engaging the anterior cortex (lateral fluoroscopy) and bending the tip 180 degrees into the olecranon.
Vigorous reduction or over-retraction can disrupt the triceps insertion, causing extensor failure. Handle the fragment and periosteum gently and test active extension against gravity at closure; repair any avulsed fibres with transosseous sutures.
Enters the supinator 10–20 mm distal to the radial head — at risk only if the approach is extended laterally (combined radial head or lateral-column work). Keep the forearm fully supinated for any lateral dissection.
If the figure-of-8 does not pass through the triceps tendon and seat anterior to the axis of rotation, the construct generates distraction, not compression, and will fail. Pass the wire through the tendon substance, seat the loop anterior to the K-wire tips, and confirm compression on fluoroscopy before closure.
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“A 35-year-old cyclist falls directly onto the elbow and has point tenderness over the olecranon. Radiographs show a transverse olecranon fracture displaced 5 mm with a maintained joint line; the elbow is stable. Describe your decision-making and operative technique.”
“A 65-year-old woman with known osteoporosis falls from standing and has a comminuted olecranon fracture (Mayo IIIB) with associated elbow subluxation. How do you manage this?”
“A patient returns at 6 months after TBW of a Mayo IIA fracture with pain directly over the olecranon and a palpable prominence. Radiographs confirm union with prominent K-wire tips subcutaneously. How do you manage this?”
Mayo classification
- Type I: undisplaced (less than 2 mm) — non-operative, splint/cast 2–3 weeks
- Type IIA: displaced, stable, non-comminuted — TBW or plate
- Type IIB: displaced, stable, comminuted — plate preferred
- Type IIIA: unstable (dislocation), non-comminuted — plate plus stability assessment
- Type IIIB: unstable and comminuted — plate plus ligament/coronoid repair as needed
TBW key steps
- Two parallel 1.6 mm K-wires from the tip at 45 degrees — must engage the ANTERIOR cortex
- Transverse drill hole 3–4 cm distal to the fracture for the wire
- Figure-of-8 through the triceps tendon substance, not superficially
- Loop seated anterior to the K-wire tips — anterior to the axis of rotation
- Bend K-wire tips 180 degrees and bury; confirm compression on fluoroscopy
Plate key points
- Pre-contoured dorsal or 90-degree lateral locking plate
- Locking screws mandatory in osteoporotic bone
- Minimum 3–4 bicortical screws distally in the shaft
- Bridge plate across comminution — do not over-compress
- Lateral plate reduces posterior subcutaneous prominence
Hardware removal
- Confirm radiographic union first (minimum 12–18 months for planned removal)
- Symptomatic prominence (pain, bursa) is the main indication
- TBW removal rate 70–80 percent; plate removal rate 15–25 percent
- Restrict heavy lifting for 6 weeks after removal (1–3 percent re-fracture risk)
Complications — numbers to know
- Hardware prominence: 70–80 percent (TBW), 15–25 percent (plate)
- K-wire migration: 5–15 percent
- Non-union: 2–5 percent; infection: 1–3 percent
- Ulnar nerve injury: 2–5 percent (usually resolving neuropraxia)
- Re-fracture after removal: 1–3 percent
Danger zones (5)
- Ulnar nerve — medial to olecranon; identify and vessel-loop FIRST
- K-wire migration — engage anterior cortex and bend tips 180 degrees
- Triceps avulsion — test the extensor mechanism at closure
- Posterior interosseous nerve — at risk only if the approach is extended laterally
- Wire loop slippage — must be anterior to the axis of rotation to compress
Post-op protocol
- Posterior splint at 45–90 degrees immediately post-op
- Splint off at 2 weeks; begin active-assisted ROM
- No passive stretching for 6 weeks (heterotopic ossification risk)
- Target 30–120 degrees by 6 weeks
- Hardware removal planned at 12–18 months for TBW (70–80 percent will need it)
Exam tips
- TBW for simple transverse fractures only — any comminution, obliquity or instability: use a plate
- The tension-band principle works ONLY if the wire loop is anterior to the axis of rotation
- Counsel ALL TBW patients that hardware removal is likely (70–80 percent) — expected, not a complication
- Mayo IIIA/B always needs a plate AND assessment/repair of elbow instability
- Always identify the ulnar nerve BEFORE starting reduction — state this explicitly in the viva
Background & Evidence
Epidemiology. Olecranon fractures are common proximal-forearm injuries, most often from a direct blow or a fall onto the elbow, and span a wide age range — younger active patients (high-energy, sport) and older patients (low-energy falls, often on osteoporotic bone). Many are isolated injuries; complex patterns with elbow dislocation, coronoid or radial-head involvement (Mayo III, terrible-triad) carry the worse prognosis. Surgical anatomy that matters. The olecranon is the proximal posterior ulnar process. The trochlear (semilunar) notch articulates with the trochlea over roughly 180 degrees and is crossed by a transverse bare area — the preferred site for an olecranon osteotomy. The triceps inserts as a broad tendon across the posterior olecranon, extending 2–3 cm distally, so fractures always retain partial triceps attachment and fixing the fragment re-tensions the extensor mechanism; the figure-of-8 wire must pass through that tendon substance. The ulnar nerve runs 1–2 cm medial to the olecranon in the cubital tunnel and enters between the heads of FCU. The MCL anterior bundle (medial epicondyle to the sublime tubercle) is the primary valgus restraint and is at risk in Mayo III dislocations. The posterior interosseous nerve enters supinator 10–20 mm distal to the radial head and is only at risk with lateral extension — keep the forearm supinated. Mayo classification — displacement, comminution and stability drive treatment:
- Pattern
- Undisplaced (less than 2 mm), intact extensor mechanism
- Treatment
- Non-operative — cast or splint at 45–90 degrees for 2–3 weeks, early ROM
- Pattern
- Displaced (greater than 2 mm), stable, non-comminuted
- Treatment
- Operative — TBW (transverse) or plate (oblique or proximal)
- Pattern
- Displaced, stable, comminuted
- Treatment
- Plate fixation preferred over TBW
- Pattern
- Unstable (elbow dislocation or subluxation), non-comminuted
- Treatment
- Plate fixation mandatory; assess and address instability
- Pattern
- Unstable and comminuted
- Treatment
- Plate fixation; may need ligament or coronoid reconstruction
Biomechanics of tension band wiring — and its caveat. The tension-band principle converts the eccentric posterior (triceps) load into anterior compression, but it depends on a transverse fracture with an intact posterior cortex: an oblique fracture shears, a comminuted one splits, and osteoporotic bone will not hold the wires. A biomechanical caveat matters for the exam — cadaveric cyclic-loading work (Hutchinson et al., JBJS Am 2003) showed the classic AO K-wire tension band did not generate measurable compression across the osteotomy gap, and that an intramedullary 7.3 mm cancellous screw with a tension band was about five times stiffer than the K-wire construct. Clinically, treat the tension-band "compression" concept as a useful model rather than a guaranteed mechanical fact: favour passive over aggressive active extension early, and prefer screw-based or plate constructs where stiffness matters (proximal or osteoporotic fractures). Why TBW fails in certain patterns: - Comminuted — no stable cortex to resist compression; fragments split.
- Osteoporotic — K-wires pull out of soft bone; wire tension cannot be maintained.
- Very proximal — too little bone in the proximal fragment for K-wire purchase. Key evidence. Wolfgang (CORR, 1987) established TBW as reliable across 45 displaced fractures (good or excellent in 29 of 30 isolated fractures; 97 percent). Hume and Wiss (CORR, 1992), a prospective randomised trial of 41 fractures, found loss of reduction in 53 percent after TBW versus 5 percent after plating, with better clinical results after plating — arguing for plate fixation whenever the pattern is not a simple transverse fracture. Schliemann (Acta Orthop Belg, 2014) showed no functional or radiographic advantage of a locking plate over TBW for isolated Mayo IIA fractures, at higher cost. Carter (JBJS Essent Surg Tech, 2018) confirmed in a randomised trial that plate and TBW gave equivalent patient- and surgeon-reported outcomes, though implant removal was needed in about half of TBW patients and infection/revision occurred only after plating. In the elderly, Duckworth (Bone Joint J, 2017) showed non-operative care matches surgery for one-year function with far fewer complications. The consistent message: TBW remains reasonable and cost-effective for the simple transverse fracture in good bone (with removal expected), while plating is preferred for anything comminuted, oblique, osteoporotic or unstable.
References
Landmark series — tension band wiring for displaced olecranon fractures
- Series of 45 displaced olecranon fractures treated by TBW over 13 years
- Good or excellent results in 29 of 30 isolated fractures (97 percent)
- True K-wire migration was uncommon and considered avoidable with correct technique
- Established TBW as a reliable standard for displaced olecranon fractures
Randomised trial — TBW vs plate fixation for displaced olecranon fractures
- Prospective randomised study of 41 adult displaced olecranon fractures: TBW versus plate
- Loss of reduction (significant step-off or gap) in 53 percent after TBW versus 5 percent after plate
- Good clinical results in 37 percent (TBW) versus 63 percent (plate); good radiographic results 47 percent versus 86 percent
- Symptomatic metal prominence after TBW in 42 percent; true K-wire migration in only one patient
Randomised trial — non-operative vs operative management in the elderly
- RCT in patients aged 75 and over with isolated displaced olecranon fractures
- Trial stopped early because the operative complication rate reached 81.8 percent (9 of 11)
- No difference in mean one-year DASH: 23 (non-operative) versus 22 (operative)
- No difference in Broberg-Morrey or Mayo Elbow Scores at any time point
TBW vs precontoured locking plate in Mayo type IIA fractures
- Comparative study of 26 isolated Mayo IIA fractures: locking plate (13) versus TBW (13)
- No significant difference in DASH or Mayo Elbow Performance Score at mean 43 months
- Hardware removal needed more often after TBW; locking-plate fixation was significantly more expensive
- Locking plates did not improve functional or radiographic outcome over TBW for simple stable fractures
Biomechanical cyclic-loading of olecranon fixation constructs
- Cadaveric cyclic-loading of a simulated 50 percent transverse fracture across multiple constructs
- A 7.3 mm intramedullary cancellous screw with a tension band was about five times stiffer than K-wire tension-band constructs
- The classic AO K-wire tension band did not generate measurable compression across the osteotomy gap
- Authors recommend passive rather than aggressive active extension early to limit fracture distraction
Randomised trial — plate vs TBW fixation of olecranon fractures
- Randomised trial of plate versus TBW in 67 active adults
- No difference in patient- or surgeon-reported outcomes between groups
- Implant removal required in roughly 1 in 2 patients after TBW
- Infection and revision surgery occurred exclusively after plate fixation
Further references 1. Karlsson MK, Hasserius R, Karlsson C, Besjakov J, Josefsson PO. Fractures of the olecranon: a 15- to 25-year follow-up of 73 patients. Clin Orthop Relat Res. 2002;(403):205–212. PMID: 12360028 2. Duckworth AD, Bugler KE, Clement ND, Court-Brown CM, McQueen MM. Nonoperative management of displaced olecranon fractures in low-demand elderly patients. J Bone Joint Surg Am. 2014;96(1):67–72. PMID: 24382727. DOI: 10.2106/JBJS.L.01137 3. Rommens PM, Küchle R, Schneider RU, Reuter M. Olecranon fractures in adults: factors influencing outcome. Injury. 2004;35(11):1149–1157. PMID: 15488508. DOI: 10.1016/j.injury.2003.12.002