Gustilo-Anderson Open Fracture Classification
Gustilo-Anderson Open Fracture Classification
Grade an open fracture I to IIIC from the wound, the energy and the soft-tissue coverage — with the published infection rates and the cohorts they came from, and the reason the grade belongs to the operating theatre rather than the emergency department.
Gustilo-Anderson grade
Stage + wound + coverage + arterial injuryA 6 cm open tibial wound from a motorcycle collision, bone coverable after debridement, no arterial injury, graded in theatre
The grades, and what each one is really asking
- energy beats wound length
- “high-energy trauma (gunshot and farm injuries) regardless of the size of the wound” is type IIIA. A 5 mm entry wound from a high-energy mechanism is a type III and not a type I, and this is the commonest single error in applying the system
- the grade belongs to the theatre
- “Accurate assessment of the fracture can only be performed inside an operating theatre.” The IIIA / IIIB question is whether the bone can be covered, which is answered with the wound open. In 27 children graded in the emergency department from photographs and again after exploration, the theatre grade differed
- what the infection rates rest on
- the type III rates — IIIA 4%, IIIB 52%, IIIC 42% — all come from one series of 87 type III fractures in 75 patients at one hospital between 1976 and 1979. IIIB above IIIC is the giveaway: that ordering is noise, not biology. Modern series give type III as 2.8 to 40.5%
- observers do not agree
- 245 orthopaedic surgeons classifying twelve open tibial fractures from videotape agreed 60% of the time on average, with per-fracture agreement from 42% to 94%. One review: “The GACS only demonstrates moderate interobserver reliability”
- the AO/OTA code is not here, deliberately
- the AO/OTA Fracture and Dislocation Compendium is copyright Lippincott Williams & Wilkins and the OTA’s own page states it “cannot be used commercially or for-profit without permission of the publisher”. No part of that scheme appears on this site
- derivation
- Gustilo RB and Anderson JT, J Bone Joint Surg Am 1976;58:453–458, on 1,025 open fractures of long bones; the type III subdivision added in J Trauma 1984 from 87 type III fractures in 75 patients. Neither paper was reachable in full for this page, and the criteria here are quoted from named reproductions
Worked example
A 6 cm open tibial wound from a motorcycle collision, bone coverable after debridement, no arterial injury, graded in theatre
No arterial injury, so the type IIIC rule does not fire
The bone can be covered, so this is not a type IIIB
The mechanism was high energy, so the wound entry is the high-energy option — and that promotes the injury to type IIIA whatever the wound measures. A 6 cm wound would otherwise read as a type II
Change the mechanism to a low-energy fall with the same 6 cm wound and the answer becomes type II, with a reported infection rate of 2.4% against type IIIA's 4% — both from the original Gustilo series
Now move the grading to the emergency department, before debridement, keeping the low-energy 6 cm wound: the answer becomes provisionally type II, because "without extensive soft-tissue damage" cannot be established through an unopened wound
Add an arterial injury needing repair to any of the above and the answer is type IIIC — the one criterion that overrides every other entry
Infection and amputation by grade, with the cohort
| Grade | Infection rate | Amputation | Cohort |
|---|---|---|---|
| I | 0 to 2% | not reported | Attributed to Gustilo and Anderson’s 1,025 open fractures of long bones, 1976 |
| II | 2.4% | not reported | Same series |
| IIIA | 4% | 0% | 87 type III fractures in 75 patients, Hennepin County Medical Center, 1976–1979 |
| IIIB | 52% | 16% | Same 87 fractures |
| IIIC | 42% | 42% | Same 87 fractures |
| III, any subtype | 2.8 to 40.5% | not reported | “More recent publications” in a 2023 concept review; the individual studies are not named in the text read |
| III, any subtype | 10% to 50% | not reported | Quoted in Wikipedia’s summary with no study named |
How well observers agree, and what that implies
| Study | Observers and material | Agreement |
|---|---|---|
| Brumback and Jones 1994 | 245 orthopaedic surgeons, twelve open tibial fractures on videotaped case presentations | 60% average agreement; 42% to 94% per fracture. Wikipedia renders the same range as 42% to 92% |
| Photographic study, abstract read without a citation | 10 patients, photographic slides before and after operative treatment, number of surgeons not stated | “only moderate agreement”; no numeric kappa reported, so none is quoted |
| Faraj 2002 | 27 children with open tibial fractures, photographed in the emergency department and regraded after exploration | The theatre grading “was different from the initial grading”; the abstract gives no proportion |
| 2023 concept review | narrative | “The GACS only demonstrates moderate interobserver reliability”; the OTA open fracture classification “has not demonstrated any large benefit in interobserver reliability when compared to the GACS” |
A wound you cannot see through, graded in the wrong room
Gustilo and Anderson graded open fractures by the energy of the injury, the extent of the soft-tissue damage and the degree of contamination. Type I is a clean wound under a centimetre. Type II is one to ten centimetres without extensive soft-tissue damage. Type III is everything more severe, subdivided in 1984 into IIIA where the bone can be covered, IIIB where soft tissue has been lost and the periosteum stripped, and IIIC where an artery needs repair. The system caught on because it sorts open fractures roughly by how much trouble they are, and it is still the first thing said about an open fracture almost everywhere.
Two things about it are routinely got wrong. The first is the high-energy clause: type IIIA includes high-energy trauma, gunshot and farmyard injuries among them, regardless of the size of the wound. A five-millimetre puncture from a high-energy mechanism is a type III, and grading it by the skin wound understages it by two steps. The second is the room. The IIIA/IIIB distinction is whether the bone can be covered, which is a surgical judgement made with the wound open and excised; the classification’s own commentators say accurate assessment can only be made in theatre. In 27 children photographed in the emergency department and graded again after exploration, the theatre grade differed from the first one. That is why this page asks when the grade is being assigned and answers “provisionally” where it has to.
The infection rates everyone quotes need their denominator stated. IIIA 4%, IIIB 52% and IIIC 42% all come from one series of 87 type III fractures in 75 patients at one hospital between 1976 and 1979. The internal ordering gives the game away: IIIB had more sepsis than IIIC despite IIIC carrying an arterial injury and a 42% amputation rate. More recent publications put type III infection at 2.8 to 40.5%, a range that comfortably contains all three subtype figures. The grades are useful for ordering severity and for comparing series; they are not a risk calculator.
Observer agreement is the last piece. Two hundred and forty-five orthopaedic surgeons shown videotapes of twelve open tibial fractures agreed on the grade 60% of the time, ranging from 42% to 94% per fracture — and that is raw agreement, not chance-corrected. The British Orthopaedic Association’s own standard for open fractures asks for a photograph before debridement and does not ask for a Gustilo grade at all, which is a reasonable response to all of the above.
A classification is not a score. Garden IV is not one unit worse than Garden III, a Salter-Harris IV is not twice a II, and nothing on this page adds up — these are named categories, not an ordinal measurement, and a mean or a difference taken over them has no meaning. This page states which published category the entered findings fall in and what the literature reports for that category in the cohorts it was measured in. It renders no clinical or surgical decision: not whether to operate, not which implant, not whether to take or withhold an image, and not whether a limb can take weight. Those are the treating clinician’s, and the classification is one input among many. Agreement between observers is part of what a classification means. A category two experienced readers assign differently on the same film is not a measurement of the fracture, and the published kappa values for these systems run from almost nothing to almost perfect. Each page here prints them with the number of observers and the number of radiographs they came from.
Frequently asked questions
What are the Gustilo-Anderson types?
Type I is an open fracture with a clean wound under 1 cm. Type II is a wound of 1 to 10 cm without extensive soft-tissue damage, flaps or avulsions. Type IIIA has adequate soft-tissue coverage of the bone despite extensive laceration, and also includes any high-energy injury regardless of wound size. Type IIIB has extensive soft-tissue loss with periosteal stripping and bone exposure. Type IIIC is an open fracture with an arterial injury requiring repair.
Is a small wound always a type I?
No, and treating it as one is the commonest error with this classification. The published criteria make high-energy trauma — gunshot and farm injuries are the examples given — a type IIIA regardless of the size of the wound. The skin wound is often the least informative feature of a high-energy open fracture.
When should the Gustilo grade be assigned?
After surgical debridement. The IIIA/IIIB distinction is whether the bone can be covered, which is a judgement made with the wound open; summaries of the classification state that accurate assessment can only be performed inside an operating theatre. In practice the grade is usually recorded in the emergency department, and in one series of 27 children the grade assigned after exploration differed from the one assigned before it.
What are the infection rates by Gustilo type?
As originally reported: 0 to 2% for type I, 2.4% for type II, and 4%, 52% and 42% for IIIA, IIIB and IIIC. The three type III figures come from a single series of 87 type III fractures in 75 patients at the Hennepin County Medical Center between 1976 and 1979, which is why IIIB’s rate is higher than IIIC’s. More recent publications give type III infection as 2.8 to 40.5%.
How reliable is the classification between observers?
Poorly. Two hundred and forty-five orthopaedic surgeons asked to classify twelve open tibial fractures from videotaped case presentations agreed 60% of the time on average, with per-fracture agreement ranging from 42% to 94%. A photographic study of 10 patients found only moderate agreement. One review concludes that treatment recommendations and study results based on the classification should be interpreted with caution.
Why is there no AO/OTA code on this page?
Because the AO/OTA Fracture and Dislocation Classification Compendium is copyright Lippincott Williams & Wilkins, and the Orthopaedic Trauma Association’s own page for it states that while the classification may be used for research, educational and medical purposes without asking permission, it cannot be used commercially or for-profit without the publisher’s permission. No part of that scheme is reproduced anywhere on this site.
Related calculators
References
- Gustilo RB, Anderson JT. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg Am. 1976;58:453–458. Citation read in Wikipedia’s reference list for the classification; the paper itself was not reachable. The grade I and II criteria below are therefore quoted from named reproductions.
- Problems in the management of type III (severe) open fractures: a new classification of type III open fractures. J Trauma. August 1984. Abstract read at qxmd.com/r/6471139; the full text was not reachable and the page did not display the author list, which Wikipedia gives as Gustilo RB, Mendoza RM, Williams DN, J Trauma 1984;24:742–746. “Between 1976-1979, 87 Type III open fractures (in 75 patients) were treated at the Hennepin County Medical Center.” “Type IIIA–Adequate soft-tissue coverage of a fractured bone despite extensive soft-tissue laceration or flaps”; “Type IIIB–Extensive soft-tissue injury loss with periosteal stripping and bone exposure”; “Type IIIC–Open fracture associated with arterial injury requiring repair”. “Wound sepsis in the three subtypes were: Type IIIA, 4%, IIIB, 52%; and IIIC, 42%”, with amputation rates “respectively, 0%, 16%, and 42%”. Eighty-seven fractures is the whole denominator for all six of those percentages.
- Wikipedia, Gustilo open fracture classification. The grade table verbatim: grade I “Open fracture, clean wound, wound <1 cm in length”; grade II “Open fracture, wound > 1 cm but < 10 cm in length without extensive soft-tissue damage, flaps, avulsions”; grade IIIA “Open fracture with adequate soft tissue coverage of a fractured bone despite extensive soft tissue laceration or flaps” “or high-energy trauma (gunshot and farm injuries) regardless of the size of the wound”; grade IIIB “Open fracture with extensive soft-tissue loss and periosteal stripping and bone damage”, “Usually associated with massive contamination”; grade IIIC “Open fracture associated with an arterial injury requiring repair, irrespective of degree of soft-tissue injury”. Also: “Infection of the Type III fractures is observed in 10% to 50% of the time” with no study named; “Different studies have shown inter-observer reliability of approximately 60% (ranging from 42% to 92%)”; and “Accurate assessment of the fracture can only be performed inside an operating theatre”.
- Abstract of Brumback and Jones’s 1994 survey, read at orthobullets.com/evidence/8056796; the page did not display the citation, which theplasticsfella.com and Wikipedia both give as Brumback RJ and Jones AL, 1994. “the responses of 245 orthopaedic surgeons to a survey” asked to “classify twelve open fractures of the tibia on the basis of a series of videotaped case presentations”. “The average agreement among the observers for all twelve fractures was 60 percent.” “The over-all agreement for each fracture ranged from 42 to 94 percent.” Note that Wikipedia renders the same range as “42% to 92%”; both are printed here and the primary abstract’s 94 is the one used.
- Abstract read at orthobullets.com/evidence/8377047: “10 patients with open fractures had photographic slides taken of their wound before and after operative treatment” and classified by “a group of orthopaedic surgeons” whose number the abstract does not give. “Kappa analysis demonstrated only moderate agreement among the observers” classifying open fractures with the Gustilo and Anderson system. No numeric kappa is reported, so none is quoted.
- Current Concept Review: Risk Factors for Infection Following Open Fractures. Orthop Res Rev (Dove Medical Press). “Infection rates fell substantially in type I and II fractures (0–2% and 2.4% respectively)”; “type IIIA fractures carry a 4% rate of infection”; “IIIB and IIIC fractures have infection rates as high as 52% and 42%, respectively” — all attributed there to the original Gustilo papers. “More recent publications showed a broader infection rate in type III fractures (2.8–40.5%).” “The GACS only demonstrates moderate interobserver reliability” and “treatment recommendations and study results based upon the GACS should be interpreted with a degree of caution”. Also notes that “in many open fractures, it may not be obvious at the time of initial debridement which tissues are devitalized and which are viable”.
- Faraj AA. The reliability of the pre-operative classification of open tibial fractures in children: a proposal for a new classification. Acta Orthop Belg. 2002;68(1). “Pre- and post-debridement (using the parameters of Gustilo’s classification) wound gradings in 27 children”, with “Pre-operative Polaroid photographs of the wound taken of all these patients in the accident and emergency department”. “Following wound exploration in the operating theater, the wound was classified using Gustilo’s parameters again; this was different from the initial grading.” The abstract gives no proportion regraded and no kappa, so none is quoted.
- AO Surgery Reference. Principles of management of open fractures. “The Gustilo – Anderson classification divides soft-tissue wounding of open fractures into three grades – I, II & III”; “The III grade was later further subdivided into types IIIA, IIIB & IIIC”; of type IIIC, “Gustilo et al. demonstrated a 50% risk of osteomyelitis after such injuries, with amputation (early or late) a frequent outcome”; “Most research shows that the infection rate increases with the severity of the soft-tissue injury”; and, of the AO scheme rather than Gustilo’s, “The detailed categorisation of open fractures by the AO system is most reliably done in the operating room at the completion of primary wound care and surgical excision”. The page’s own criteria tables are images, not text, so nothing further was taken from it.
- British Orthopaedic Association and British Association of Plastic, Reconstructive and Aesthetic Surgeons. Audit Standards for Trauma: Open Fractures, December 2017. “Prior to formal debridement the wound should be handled only to remove gross contamination and to allow photography”; “Photographs of open fracture wounds should be taken when they are first exposed for clinical care, before debridement”; debridement immediately for gross contamination or vascular compromise, within 12 hours for a solitary high-energy fracture and within 24 hours for other low-energy injuries; “Definitive soft tissue closure or coverage should be achieved within 72 hours of injury”. Worth recording that this standard contains NO requirement to assign a Gustilo grade at all — it asks for a photograph instead.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/