For birds of prey, traumatic injuries are among the most common reasons for admission into rehabilitation programs1 and a leading cause of death.2,3 Birds are particularly susceptible to complex fractures because avian bones have thinner cortices and contain a higher proportion of hydroxyapatite4 than do mammalian bones and their wing bones are subject to high torsional forces.5 Because the bones of the antebrachium (ie, the radius and the ulna) contribute to circumduction of the distal portion of the wing during wing beats, fractures in this region can prevent the pronation and supination movements of the wing required for adequate flight.6
Despite the potentially life-threatening impact of antebrachial fractures on birds of prey, little information is available regarding outcomes for such fractures, particularly the characteristics of these fractures that might influence the ability to return to flight after treatment. Such information would be useful for priority setting in wildlife rehabilitation programs. The objective of the study reported here was to identify characteristics of antebrachial fractures that were associated with a successful outcome (ie, bird released back into the wild after regaining normal flight ability, without being returned to the rehabilitation program within 1 month afterward) for free-ranging birds of prey treated in a rehabilitation program under a triage policy. We hypothesized that a successful outcome would be less likely for birds with open (vs closed) or chronic (vs acute) fractures or concomitant radial and ulnar fractures involving the same third of the antebrachium (vs other types of antebrachial fractures).
Materials and Methods
Source population
In the province of Quebec, Canada, at the time of the study, birds of prey found dead or injured were reported (and, as of 1992, required by law to be reported) by the finder to the governmental wildlife agency. Wildlife officers collected the reported birds, which were subsequently couriered to the Université de Montréal as part of a provincial raptor health program.3 Through this program, live birds were directed to the Clinique des Oiseaux de Proie and dead birds were transported to the Centre Québécois pour la Santé des Animaux Sauvages.
Case selection
Free-ranging birds of prey that were admitted alive to the Clinique des Oiseaux de Proie from August 1986 through December 2015 through the rehabilitation program were eligible for inclusion in the study. Medical records for this period were used to identify birds with an antebrachial fracture (ie, of the radius, ulna, or both) for inclusion in the study. The total number of free-ranging birds of prey that were admitted alive to the clinic for any reason was recorded by species so that the prevalence of antebrachial fractures could be estimated for each bird species that was included in the study.
Data collection
The medical record of each bird with an antebrachial fracture was reviewed to obtain information on species, age, sex, year of admission, fracture characteristics (ie, location, open vs closed, acute vs chronic, and comminuted vs simple), whether comorbidities unrelated to the antebrachial fracture were present, and type of treatment administered. Birds with missing data for evaluated fracture characteristics were excluded from the study.
Identifiable characteristics of bird species were used, whenever possible, to assign birds to age categories. The age of the bird was recorded as juvenile (hatching year or second year), adult (after hatching year or after second year for species with recognizable second-year plumage), or unknown (when age was not recorded). The sex of the bird was determined on the basis of morphological characteristics for dimorphic species; for other species, sex was classified as undetermined if not determined by endoscopic examination, DNA testing, or necropsy. All birds that died or were euthanized during rehabilitation were sexed after death.
Fracture location was classified by the affected bone or bones (radius only, ulna only, or both the radius and ulna) and by third or thirds of the affected bone or bones (proximal, middle, or distal third only or multiple thirds). For birds with concomitant radial and ulnar fractures, location was further classified as involving the same third or different thirds of the antebrachium. A fracture was classified as acute if it had likely occurred < 24 hours prior to admission or if the trauma was witnessed to have occurred during the 24 hours prior to admission. This assessment was alternatively made on the basis of hematoma color, with red to purple color interpreted as an acute fracture and green color as an older (ie, chronic) fracture.7 If there was no indication that the trauma occurred < 24 hours prior to admission and no hematoma was present, the fracture was classified as chronic. Each bone (radius and ulna) was classified as having a comminuted fracture present or not present (ie, simple fracture or no fracture instead). Birds with multiple fractures of the radius or ulna that included comminuted and simple fractures were classified as having a comminuted fracture.
Birds that had 1 or more other problems that were unrelated to the antebrachial fracture (eg, corneal ulceration) were classified as having comorbidities. If the other problems were related to the fracture (eg, an extensive fracture-associated wound on the wing) or if antebrachial fracture was the only problem, birds were classified as having no comorbidities.
Triage policy
Under the triage policy of the rehabilitation program, admitted birds with articular fractures, necrotic bone, or extensive soft tissue damage that impaired function of nerves or ligamentous structures were euthanized. For all other birds, nonsurgical treatments could have included external coaptation (eg, figure-8 bandage) or cage rest, depending on the bird's condition. Surgical treatments could have included stabilization of the fractured bone or bones with nonthreaded intramedullary pins, with or without tie-in to an external fixator, or with polymethylmethacrylate intramedullary implants.8,9 When indicated, partial radiectomy was performed to prevent synostosis. Aseptic technique, prophylactic antimicrobial treatment, and analgesic drugs were used for all surgical procedures, which were performed in a surgical suite. These procedures were performed by interns supervised by faculty clinicians; alternatively, they were performed by residents or faculty clinicians. After surgery or as part of nonsurgical treatment, physical therapy, as described elsewhere,9 was provided on an as-needed basis. In brief, joints were warmed with warm pads and flexion and extension movements were performed for 10 minutes with the bird anesthetized; maximum extension was maintained for 10 seconds.9 Joint range of motion was measured with a goniometer before and after each physical therapy session. Exercise in large aviaries was also considered active physical therapy. Birds were cared for at the clinic until the attending clinician determined that they had sufficiently recovered and were eligible for release.
Release
As part of the rehabilitation program's release policy, each bird deemed eligible for release was transferred to a rehabilitation center (Chouette à voir, Saint-Jude, QC, Canada). At the center, birds were placed in large aviaries where their flight was evaluated; if there was doubt about the bird's endurance or ability to fly at greater heights, the bird was placed in a larger aviary for further evaluation. The largest aviary measured 33 × 10 × 7 m (length × width × height).
To be considered sufficiently rehabilitated for release, a bird was required to be able to fly and perch on high perches (up to 5 m high) and have sufficient endurance (ie, able to continuously fly, without landing or perching for at least 5 minutes, during which respiration and flight quality were evaluated). Some criteria used to evaluate flight differed by species (eg, for Strigiformes, silence of flight was considered critical). Behavior was also assessed, especially in fledglings, and birds that were imprinted on humans were not released. If there was doubt about a bird's ability to kill prey (eg, bird with toe amputation), its hunting ability was tested by placing live prey in a large open container in the aviary. For most birds evaluated during the study period, releasability was assessed by the veterinarian in charge of the rehabilitation program (GF). For birds that were released, the release took place in an open field next to the woods.
Outcome
A goal of the rehabilitation program was to evaluate and monitor the causes of disease and death in birds of prey within the province. Consequently, all birds released were banded on 1 leg with a band provided by the Bird Banding Office (Canadian Wildlife Service). Through an existing international system, lay people or researchers who later found banded birds were able to report the birds, facilitating detection of deaths and reinjuries. Band return information for returned birds was entered into the birds’ records.
On the basis of the available data, the outcome for each treated bird was recorded as unsuccessful if it died or was euthanized at any point during rehabilitation, was kept permanently in captivity (ie, was not released back into the wild), or was returned (as determined through the banding program) within 1 month after release. Outcome was recorded as successful if the bird was released back into the wild after regaining normal flight ability, without being returned to the rehabilitation program within 1 month afterward. Approximately 3% (range, 1.5% to 18% [depending on the species]) of banded birds released by this rehabilitation program were historically returned, with some birds returned as long as 11 years after release.3
Statistical analysis
Descriptive statistics (frequencies and percentages) were computed with the aid of commercially available software.a Birds were excluded from subsequent analyses if they had been euthanized on admission (ie, within 24 hours after admission) to the rehabilitation program, died of reasons unrelated to the antebrachial fracture, or were of a species represented by < 9 birds.
Multivariable logistic regression, performed by use of an open-access statistical program,b was used to identify fracture characteristics (ie, the independent variables of interest) associated with outcome (successful vs unsuccessful; ie, the dependent variable of interest). For independent variables with > 2 categories, the category with the largest number of observations was used as the reference category.
In the first logistic regression model, which included all treated birds except the noted exclusions, fracture characteristics, age category, sex, species, presence of comorbidities, treatment received (surgical vs nonsurgical), and year of admission to the rehabilitation program were all included as independent variables. In a second logistic regression model, which included only a subset of treated birds with fractures involving a single third of the radius, ulna, or both, the potential association between fracture location (ie, proximal, middle, or distal third of the antebrachium) and outcome was explored while simultaneously controlling for whether the fracture was open (vs closed) or acute (vs chronic), age category, sex, species, and presence of comorbidities. This second modeling approach was performed twice with a different reference category used each time for fracture location (fractures of the proximal and distal third of the affected bone). In a third logistic regression model, which included only a subset of treated birds with concomitant radial and ulnar fractures involving the same third or thirds of the antebrachium, fracture location, whether the fracture was open (vs closed) or acute (vs chronic), age category, sex, species, presence of comorbidities, and year of admission were all included as independent variables. The unit of analysis for the study was individual bird, even if a bird had multiple fractures. Values of P < 0.05 were considered significant for all analyses.
Results
Among the 6,528 birds that were admitted alive to the rehabilitation program through the Clinique des Oiseaux de Proie over the study period, 908 (14%) were identified with antebrachial fractures. After exclusion of birds with missing data for evaluated fracture characteristics or that died because of comorbidities, 620 birds admitted alive with antebrachial fractures were included in the study. Twenty-four species were represented, and the prevalence of antebrachial fracture in each species was summarized (Table 1). A successful outcome as defined was achieved for 245 of 620 (40%) admitted birds; no released birds were returned to the rehabilitation program within 1 month after release.
Descriptive statistics for species of free-ranging birds of prey that were admitted alive to a rehabilitation program from August 1986 through December 2015 and ORs for the association between species and a successful outcome* for birds that received treatment for antebrachial fractures.
| Species | Total No. of admitted birds | No. (%) of birds with antebrachial fracture | No (%) of birds with antebrachial fracture that were treated | OR (95% CI) for treated birds (n = 507†) | P value |
|---|---|---|---|---|---|
| American kestrel (Falco sparverius) | 1,039 | 63 (6) | 51 (81) | Reference | — |
| Great horned owl (Bubo virginianus) | 630 | 59 (9) | 50 (85) | 0.48 (0.17-1.35) | 0.17 |
| Sharp-shinned hawk (Accipiter striatus) | 350 | 57 (16) | 49 (86) | 0.74 (0.26-2.10) | 0.57 |
| Merlin (Falco columbarius) | 871 | 57 (7) | 46 (81) | 0.44 (0.16-1.17) | 0.14 |
| Red-tailed hawk (Buteo jamaicensis) | 457 | 56 (12) | 47 (84) | 0.35 (0.12-0.98) | 0.05 |
| Snowy owl (Bubo scandiacus) | 262 | 52 (20) | 46 (88) | 0.82 (0.29-2.35) | 0.72 |
| Rough-legged hawk (Buteo lagopus) | 161 | 44 (27) | 34 (77) | 0.55 (0.17-1.72) | 0.30 |
| Peregrine falcon (Falco peregrinus) | 176 | 23 (13) | 19 (83) | 0.30 (0.07-1.18) | 0.09 |
| Broad-winged hawk (Buteo platypterus) | 273 | 23 (8) | 18 (78) | 0.23 (0.05-1.05) | 0.06 |
| Short-eared owl (Asio flammeus) | 120 | 22 (18) | 19 (86) | 0.88 (0.16-3.26) | 0.65 |
| Barred owl (Strix varia) | 289 | 20 (7) | 17 (85) | 0.72 (0.25-4.60) | 0.67 |
| Cooper's hawk (Accipiter cooperi) | 211 | 20 (9) | 17 (85) | 0.15 (0.003-0.58) | 0.008 |
| Turkey vulture (Cathartes aura) | 93 | 18 (19) | 15 (83) | 0.24 (0.05-1.01) | 0.06 |
| Northern harrier (Circus hudsonius) | 148 | 17 (7) | 16 (94) | 0.28 (0.06-1.23) | 0.10 |
| Eastern screech owl (Megascops asio) | 237 | 16 (7) | 13 (81) | 0.68 (0.13-3.36) | 0.63 |
| Bald eagle (Haliaeetus leucocephalus) | 135 | 15 (11) | 13 (87) | 0.30 (0.06-1.45) | 0.13 |
| Osprey (Pandion haliaetus) | 128 | 13 (10) | 11 (85) | 0.05 (0.003-0.38) | 0.01 |
| Northern saw-whet owl (Aegolius acadicus) | 455 | 13 (3) | 10 (77) | 0.34 (0.06-2.04) | 0.23 |
| Northern goshawk (Accipiter gentilis) | 189 | 11 (6) | 9 (82) | 0.27 (0.04-1.61) | 0.15 |
| Long-eared owl (Asio otus) | 71 | 9 (13) | 7 (78) | 0.20 (0.02-2.35) | 0.20 |
| Red-shouldered hawk (Buteo lineatus) | 70 | 2 (11) | 2 (100) | — | — |
| Great gray owl (Strix nebulosa) | 44 | 5 (11) | 5 (100) | — | — |
| Northern hawk owl (Surnia ulula) | 65 | 3 (5) | 3 (100) | — | — |
| Boreal owl (Aegolius funereus) | 54 | 2 (4) | 2 (100) | — | — |
| Total | 6,528 | 620 (100) | 519 (84) | — | — |
The OR, derived from multivariable analysis, represents the likelihood of a successful outcome for a given species versus American kestrels.
Bird released back into the wild after regaining normal flight ability, without being returned to the rehabilitation program within 1 month afterward.
Birds euthanized on admission (n = 101) and birds of species represented by < 9 birds (12) were excluded.
— = Not applicable.
Note that results for other variables included in the multivariable logistic regression model are shown in Table 2.
Treatment was attempted for 519 of the 620 (84%) birds, and 101 (16%) birds were euthanized on admission to the rehabilitation program. A successful outcome was recorded for 245 of the 519 (47%) treated birds. Characteristics of these birds and their fractures as well as percentages of treated birds with a successful outcome were summarized (Table 2). Among the 519 treated birds, 72 (14%) had comorbidities. Most birds (438/519 [84%]) were treated exclusively with a figure-8 bandage or cage rest (ie, nonsurgical treatments), whereas 81 (16%) birds were treated surgically. Overall, 218 of 438 (50%) birds that received nonsurgical treatment and 27 of 81 (33%) birds that received surgical treatment had a successful outcome.
Selected results of multivariable logistic regression to identify factors associated with a successful outcome* for the birds of Table 1 that that were treated for antebrachial fractures.
| Factor | Total No. of admitted birds | Total No. of treated birds | Total No. of birds with successful outcome | Percentage of admitted birds with successful outcome | Percentage of treated birds with successful outcome | OR (95% CI) for treated birds (n = 507†) | P value |
|---|---|---|---|---|---|---|---|
| Overall | 620 | 519 | 245 | 40 | 47 | — | — |
| Fracture location | |||||||
| Radius only | 71 | 65 | 39 | 55 | 60 | 5.22 (2.43-11.51) | < 0.001 |
| Proximal third only | 34 | 30 | 15 | 44 | 50 | — | — |
| Middle third only | 8 | 8 | 5 | 63 | 63 | — | — |
| Distal third only | 27 | 25 | 18 | 67 | 72 | — | — |
| > I third | 2 | 2 | 1 | 50 | 50 | — | — |
| Ulna only | 225 | 217 | 154 | 68 | 71 | 10.00 (5.69-18.10) | < 0.001 |
| Proximal third only | 47 | 43 | 26 | 55 | 60 | — | — |
| Middle third only | 77 | 76 | 55 | 71 | 72 | — | — |
| Distal third only | 69 | 67 | 51 | 74 | 76 | — | — |
| > I third | 32 | 31 | 22 | 69 | 71 | — | — |
| Radius and ulna involving same third of antebrachium | 265 | 206 | 39 | 15 | 19 | Reference | — |
| Proximal third only | 81 | 62 | 7 | 9 | 11 | — | — |
| Middle third only | 69 | 53 | 12 | 17 | 23 | — | — |
| Distal third only | 77 | 57 | 17 | 22 | 30 | — | — |
| > I third | 38 | 34 | 3 | 8 | 9 | — | — |
| Radius and ulna involving different thirds of antebrachium | 57 | 29 | 12 | 21 | 41 | 2.66 (1.00-2.66) | 0.048 |
| Bilateral antebrachial fracture | 2 | 2 | 1 | 50 | 50 | 3.31 (0.09-119.58) | 0.47 |
| Other fracture characteristics | |||||||
| Open | 259 | 180 | 37 | 14 | 21 | Reference | — |
| Closed | 361 | 339 | 208 | 58 | 61 | 3.78 (2.32-6.27) | < 0.001 |
| Ulna | |||||||
| Comminuted fracture | 173 | 167 | 74 | 43 | 44 | 1.06 (0.63-1.78) | 0.83 |
| No comminuted fracture‡ | — | — | — | — | — | Reference | — |
| Simple fracture | 376 | 313 | 132 | 35 | 42 | — | — |
| No fracture of the ulna | 71 | 65 | 39 | 55 | 60 | — | — |
| Radius | |||||||
| Comminuted fracture | 38 | 20 | 8 | 21 | 40 | 8.59 (0.26-2.68) | 0.80 |
| No comminuted fracture‡ | — | — | — | — | — | Reference | — |
| Simple fracture | 357 | 282 | 83 | 23 | 29 | — | — |
| No fracture of the radius | 225 | 217 | 154 | 68 | 71 | — | — |
| Acute (< 24 h) | 445 | 366 | 166 | 37 | 45 | Reference | — |
| Open | 206 | 144 | 32 | 16 | 22 | — | — |
| Closed | 239 | 222 | 134 | 56 | 60 | — | — |
| Chronic (≥ 24 h) | 175 | 153 | 79 | 45 | 52 | 1.27 (0.76-2.12) | 0.24 |
| Open | 53 | 36 | 5 | 9 | 14 | — | — |
| Closed | 122 | 117 | 74 | 61 | 63 | — | — |
| Bird characteristics | |||||||
| Sex | |||||||
| Males | 120 | 92 | 29 | 24 | 32 | 0.32 (0.15-0.63) | 0.001 |
| Females | 191 | 160 | 72 | 38 | 45 | 0.59 (0.34-1.02) | 0.06 |
| Undetermined | 309 | 267 | 144 | 47 | 54 | Reference | — |
| Age category | |||||||
| Juvenile | 252 | 213 | 101 | 40 | 47 | Reference | — |
| Adult | 178 | 147 | 74 | 42 | 50 | 1.16 (0.68-2.01) | 0.59 |
| Unknown | 190 | 159 | 70 | 37 | 44 | 0.81 (0.46-1.43) | 0.46 |
| Comorbidities | |||||||
| No | 548 | 447 | 190 | 35 | 43 | Reference | — |
| Yes | 72 | 72 | 55 | 76 | 76 | 2.87 (1.49-5.75) | 0.002 |
The OR represents the likelihood of a successful outcome for birds at a given level of a factor versus those at the reference level.
For the radius and ulna, birds with a comminuted fracture were compared with those without a comminuted fracture (ie, those with a simple fracture or no fracture).
Note that the results for species (a variable also included in the multivariable logistic regression model) are shown in Table 1 and that the number of euthanized birds can be derived by subtracting the number of treated birds from the number of admitted birds.
See Table 1 for remainder of key.
Five hundred seven of 519 (98%) treated birds were included in the first logistic regression model to identify fracture characteristics associated with outcome.
The 12 excluded birds were of 4 species represented by < 9 birds. The odds of a successful outcome did not differ significantly (P = 0.44) between birds that received surgical versus nonsurgical treatment (OR, 1.29; 95% CI, 0.67 to 2.50). Although the percentage of treated birds with a successful outcome differed from year to year over the study period (Figure 1), year of admission to the rehabilitation program was not significantly (P = 0.19) associated with outcome (OR, 1.02; 95% CI, 0.99 to 1.05). The odds of a successful outcome did not differ significantly between adult and juvenile birds (P = 0.59) or between birds of unknown age and juvenile birds (P = 0.46; Table 2). Males were significantly (P = 0.001) less likely to have a successful outcome than were birds of undetermined sex. Birds that had comorbidities were significantly (P = 0.002) more likely to have a successful outcome than were birds with only an antebrachial fracture. Among the 20 species included in the logistic regression model, Cooper's hawks (Accipiter cooperi) and ospreys (Pandion haliaetus) were significantly (P = 0.008 and P = 0.01, respectively) less likely to have a successful outcome than were American kestrels (Falco sparverius; Table 1). Birds with closed fractures were significantly (P < 0.001) more likely to have a successful outcome than were those with open fractures. Birds with a radial fracture only, an ulnar fracture only, or concomitant radial and ulnar fractures involving different thirds of the antebrachium were significantly (P < 0.001, P < 0.001, and P = 0.048, respectively) more likely to have a successful outcome than were birds with concomitant radial and ulnar fractures involving the same third or thirds of the antebrachium. The odds of a successful outcome did not differ significantly between birds with chronic versus acute antebrachial fractures (P = 0.24) or between birds with comminuted versus simple fractures of the radius (P = 0.80) or ulna (P = 0.83).
Number of free-ranging birds of prey with antebrachial fractures that were treated in a rehabilitation program per year of admission (gray bars) and the percentage that had a successful outcome (ie, bird released back into the wild after regaining normal flight ability, without being returned to the rehabilitation program within 1 month afterward; black line).
Citation: Journal of the American Veterinary Medical Association 256, 5; 10.2460/javma.256.5.580
Among the 620 birds that were admitted alive, 489 (79%) had fractures that involved a single third of the antebrachium. The percentage of birds with a fracture of the proximal, middle, or distal third of the affected bone that were euthanized on admission was 17% (27/162), 11% (17/154), and 14% (24/173), respectively. Among 489 birds with fractures involving a single third of the radius, ulna, or both, 419 were included in the second logistic regression model; the 70 excluded birds had been euthanized on admission (n = 59) or were of species represented by < 9 birds (11). Birds were significantly more likely to have a successful outcome when the fracture involved the middle third (OR, 2.73; 95% CI, 1.40 to 5.40; P = 0.004) or distal third (OR, 3.90; 95% CI, 2.02 to 7.72; P < 0.001) versus the proximal third of the affected bone. The odds of a successful outcome did not differ significantly (P < 0.28) between birds with fractures involving the middle versus distal third of the affected bone (OR, 0.70; 95% CI, 0.37 to 1.33).
Because the first logistic regression model indicated that birds with concomitant radial and ulnar fractures involving the same third or thirds of the antebrachium were less likely to have a successful outcome than were birds with other types of antebrachial fractures, this subset of birds was examined in greater detail (Table 3). The success rate was 15% (39/265) among all birds admitted alive with concomitant radial and ulnar fractures that involved the same third or thirds of the antebrachium and was 19% (39/206) among the subset that received treatment.
Selected results of multivariable logistic regression to identify factors associated with a successful outcome* for a subset of treated birds in Table 1 that had concomitant radial and ulnar fractures involving the same third or thirds of the antebrachium.
| Factor | Total No. of admitted birds | Total No. of treated birds | Total No. of birds with successful outcome | Percentage of admitted birds with successful outcome | Percentage of treated birds with successful outcome | OR (95% CI) for treated birds (n = 206)† | P value |
|---|---|---|---|---|---|---|---|
| Overall | 265 | 206 | 39 | 15 | 19 | — | — |
| Fracture location | |||||||
| Proximal third only | 81 | 62 | 7 | 9 | 11 | Reference | — |
| Middle third only | 69 | 53 | 12 | 17 | 23 | 8.36 (2.12-38.59) | 0.004 |
| Distal third only | 77 | 57 | 17 | 22 | 30 | 13.93 (3.53-65.23) | < 0.001 |
| > I third | 38 | 28 | 3 | 8 | 11 | 2.06 (0.33-11.55) | 0.41 |
| Fracture type | |||||||
| Open | 165 | 116 | 10 | 6 | 9 | Reference | — |
| Closed | 100 | 90 | 29 | 29 | 32 | 7.61 (2.81-22.96) | < 0.001 |
| Acute (< 24 h) | 184 | 140 | 23 | 13 | 16 | Reference | — |
| Chronic (≥ 24 h) | 81 | 66 | 16 | 20 | 24 | 1.21 (0.45-3.19) | 0.70 |
Birds euthanized on admission were excluded.
See Tables 1 and 2 for remainder of key.
Overall, 206 of 265 (78%) birds with concomitant radial and ulnar fractures that involved the same third or thirds of the antebrachium were included in the third logistic regression model to identify fracture characteristics associated with outcome; the excluded 59 birds had been euthanized on admission. Birds were significantly more likely to have a successful outcome when the fractures were localized to the middle (P = 0.004) or distal (P < 0.001) third versus the proximal third of the affected bones. Birds with closed fractures were significantly (P < 0.001) more likely to have a successful outcome than were those with open fractures. Red-tailed hawks (Buteo jamaicensis; OR, 0.03; 95% CI, 0.001 to 0.35; P = 0.01) and great horned owls (Bubo virginianus; OR, 0.10; 95% CI, 0.01 to 0.74; P = 0.03) were significantly less likely to have a successful outcome than were American kestrels. No significant association with outcome was found for other species. The percentage of birds with a successful outcome was lower for those treated nonsurgically (16% [24/152]) than for those treated surgically (28% [15/54]; not evaluated statistically). Success rates by site of surgical osteosynthesis were as follows: ulna only, 29% (7/24); radius only, 67% (2/3); radius and ulna, 25% (5/20); and ulna with radiectomy, 14% (1/7).
Discussion
In the retrospective study reported here, the prevalence of antebrachial fractures was estimated among free-ranging birds of prey that were admitted alive into a province-wide rehabilitation program over a nearly 30-year period. Several characteristics of those fractures were significantly associated with a successful outcome (ie, bird released back into the wild after regaining normal flight ability, without being returned to the rehabilitation program within 1 month afterward) among birds that received treatment for their fractures. Nearly 20% of the 620 birds included in our study were American kestrels and great horned owls, which likely reflected the high proportion of these species among the bird of prey population in Quebec, rather than a predisposition of these 2 species to antebrachial fractures.
Among species that nest in Quebec, orphan nestlings or juveniles learning to hunt are commonly admitted into our rehabilitation program. In the present study, the highest prevalence of antebrachial fractures was observed in rough-legged hawks (Buteo lagopus) and snowy owls (Bubo scandiacus), which are both migratory species. Because Quebec is south of their breeding grounds, fewer orphaned juveniles of these species are admitted into our rehabilitation program. Therefore, the prevalence of fractures in rough-legged hawks and snowy owls may appear to be higher in Quebec than in the areas where these species breed. In addition, certain species are more likely to be admitted with nontraumatic conditions than with traumatic conditions. For instance, bald eagles (Haliaeetus leucocephalus) in Quebec are particularly prone to being caught by snares used for trapping of fur-bearing animals.10 Thus, in bald eagles, antebrachial fractures would be less likely to be encountered by our rehabilitation program than injuries resulting from snares.
No firm conclusions can be drawn from the present study regarding the outcome of various treatment approaches, given that administered treatments were not assigned randomly. Birds with highly unstable and displaced fractures were more likely to undergo surgery than were other birds. Thus, selection bias likely contributed to the slightly higher success rates for birds that received nonsurgical versus surgical treatment (50% vs 33%). Of note, birds admitted during weekends by interns may have been reevaluated during weekdays by the clinician in charge and subsequently euthanized because of a poor prognosis or deteriorating condition, thus decreasing the success rate for birds that received nonsurgical treatment. However, the results of the various attempted treatment approaches were clinically relevant, given that a success rate of 50% could be considered reasonable for birds that receive nonsurgical treatment. External coaptation with a figure-8 bandage, cage rest, or both appeared to be good treatment options for certain fractures of the radius or ulna alone. Conversely, among birds with concomitant radial and ulnar fractures that involved the same third of the antebrachium, only 16% had a successful outcome after receiving nonsurgical treatment. A prospective study design with randomized treatment assignment would be useful for evaluating whether surgical treatment is more likely to result in a successful outcome than is nonsurgical treatment for birds with concomitant radial and ulnar fractures that involve the same third of the antebrachium. However, surgery may not be an option in some instances, such as when callus has already formed, the fracture is very comminuted, or wide bone fissures are present.
In the present study, the probability of a successful outcome differed among species. This was expected given that the criteria used to assess flight differed among bird species and the decision to release a bird was made mainly on the basis of its quality of flight (eg, maneuverability, endurance, and silentness11) and ability to catch prey. The quality of the flight is considered less critical for scavengers and soaring species than for species that rely on highspeed swooping (eg, peregrine falcon [Falco peregrinus]).6 Strigiformes must fly silently to catch prey, whereas Accipitriformes need to maneuver around obstacles to catch their prey.12 Differences in success rates among species included in the present study could also have resulted from anatomic differences. For instance, results of a previous study13 involving American kestrels suggest that the muscle fiber composition of muscles involved in locomotor function may vary among avian species, which could result in differences in the stresses applied to fracture sites among species. Finally, success rates may have been affected by the ability of various species of free-ranging birds of prey to adapt to captive conditions during rehabilitation. For instance, because in our experience Cooper's hawks and ospreys are often fractious birds, they may sustain traumatic wounds or develop stress-induced conditions secondary to captivity.
Although birds with comorbidities in the present study were significantly more likely to have a successful outcome than were those with no comorbidities, birds that died as the result of comorbidities and not as a result of the antebrachial fracture were excluded from the analysis, which could explain this finding (ie, selection bias). Similarly, the finding that birds of undetermined sex were more likely to have a successful outcome than were males can be explained by the fact birds that died or were euthanized during rehabilitation were sexed after death. Because birds that were sexed on necropsy were not released, dead birds would likely be overrepresented among sexed birds. Hence, presence of comorbidities and undetermined sex should not be considered prognostic factors for a successful outcome in the birds of the present study.
Birds with open fractures were significantly less likely to have a successful outcome following treatment than were birds with closed fractures. This was expected, given that osteomyelitis and avascular bone necrosis are more likely to develop with open fractures than with closed fractures because of extensive soft tissue damage.14
The likelihood of a successful outcome following treatment did not differ significantly between birds with acute versus chronic fractures in the present study. This was likely because birds with severely chronic, contaminated, or devitalized fractures were often euthanized, and all birds euthanized on admission were excluded from the logistic regression portion of the study. In addition, birds with chronic fractures that were selected for rehabilitation likely had callus formation, which could have improved the success rate for birds in this category. Therefore, a more precise definition of chronic fracture would be needed to differentiate malunion from healing fractures and refine prognosis evaluation for future studies.
Birds with antebrachial fractures were significantly more likely to have a successful outcome following treatment when the fracture involved the middle or distal third of the affected bone versus the proximal third, indicating that proximal antebrachial fractures may carry a more guarded prognosis in birds of prey. The cause of this difference could have been related to the complex biomechanical functions of the elbow joint.15,16 The converse is observed in dogs, in which higher rates of delayed union or nonunion for distal versus proximal antebrachial fractures as a result of the limited blood supply in the distal antebrachial area have been reported.17 Interestingly, birds with proximal antebrachial fractures in the present study were more likely to be euthanized on admission than were birds with middle and distal fractures. Thus, the better prognosis associated with middle and distal antebrachial fractures did not appear to be the result of the rehabilitation program's triage protocol.
Birds with concomitant radial and ulnar fractures that involved the same third of the antebrachium were significantly less likely to have a successful outcome following treatment than were birds with other types of antebrachial fractures. This was expected, given that concomitant radial and ulnar fractures may be associated with more extensive soft tissue trauma.18 In addition, synostosis may occur between the radius and ulna, although the frequency of this complication has not been established. In the present study, 19% of treated birds of prey with concomitant radial and ulnar fractures involving the same third of the antebrachium had a successful outcome. In contrast, in a previous report,18 25% of laboratory pigeons that underwent surgical repair of experimentally created (ie, simple, noncontaminated, and acute) concomitant radial and ulnar fractures had a successful outcome (ie, satisfactory bone alignment and fusion on radiographic and histologic examination at predetermined euthanasia time points). Because this previous report18 did not indicate whether the pigeons’ flight was evaluated, it is unknown whether mild flight asymmetry, joint ankylosis, or other complications involving soft tissues were considered, whereas these would be important considerations when evaluating birds of prey for potential release. More stringent criteria for defining a successful outcome would be expected for raptors than for pigeons. Indeed, only birds of prey with a satisfactory flight ability were deemed releasable by our rehabilitation program. Moreover, in a wildlife rehabilitation facility that treats birds with contaminated or chronic fractures and provides ongoing surgical teaching of interns and residents, a lower success rate would be expected than that obtained under experimental conditions. Given limited resources when numerous birds were admitted on the same day, prioritization of treatment in favor of birds with the best anticipated prognosis (ie, birds with a poor prognosis euthanized during rehabilitation) may have also decreased the overall success rate in the present study. Surgical techniques and postoperative management also evolved over the 29-year study period, which could also explain why the success rate was lower in the present study than in the previous experimental study18 of pigeons.
The number of birds treated per year for antebrachial fractures varied over the study period. The observed increase in the number of birds treated between 1986 and 19913 could potentially explain the observed increase in the number of birds treated for antebrachial fracture over this period. Since 1992, when it became mandatory to report injured birds of prey in Quebec,3 the total number of birds admitted each year to the rehabilitation program has remained stable. The 2001–2002 introduction of West Nile virus to Quebec19,20 could explain the high number of birds with antebrachial fractures during these years, given that birds infected with West Nile virus are more prone to traumatic injuries.21 A 2013 recrudescence of human and animal cases of West Nile virus infection in Quebec22 could also explain the increase in the number of birds treated for antebrachial fractures that year. In addition, an increase in the Arctic lemming (Dicrostonyx torquatus) population during the summer of 2013 may have contributed to the high number of snowy owls rehabilitated during the winter of 2013 to 2014,3 when many juveniles were migrating south to Quebec.23 Because snowy owls were among the species with the highest prevalence of antebrachial fractures in the present study, this migration may have contributed to the high number of birds treated for antebrachial fracture in 2013 and 2014. The success rate among birds treated for antebrachial fractures in the present study ranged from 10% to 25% over the study period, except for years when the number of birds treated for antebrachial fractures was particularly low (eg, 1997 and 2000).
The reason red-tailed hawks and great horned owls were less likely to have a successful outcome than were American kestrels following treatment for concomitant radial and ulnar fractures involving the same third or thirds of the antebrachium remains unclear. It is possible that surgery was attempted more frequently for medium-sized birds with a more questionable prognosis as a teaching experience for interns and residents, whereas smaller birds were more likely to be euthanized and therefore excluded from the logistic regression analyses. In addition, speciesspecific anatomic features or differences in body weight may explain why red-tailed hawks and great horned owls had a worse prognosis than American kestrels. Further studies are needed to investigate the potential impact of these factors on the outcome.
A lack of consensus exists regarding the best surgical technique to repair concomitant radial and ulnar fractures in birds. Proposed techniques include surgical repair of the radius only, the ulna only, or both bones or radiectomy. Radiectomy is considered controversial because other less invasive techniques have been described to prevent synostosis.24 In the present study, only 1 of 7 birds with concomitant radial and ulnar fractures that involved the same third of the antebrachium had a successful outcome after radiectomy. Some authors recommend surgical fixation of the radius to achieve proper healing, given that the radius is highly moveable.6,24,25 However, because the ulna is larger than the radius in birds, surgical repair of the ulna allows for use of a larger pin diameter, which results in improved fixator stability. In addition, proper alignment of secondary remiges that insert on the ulnar periosteum is important for flight, with misalignment potentially resulting in increased drag by promotion of turbulence and decreased lift by alteration of the gliding surface.6 Therefore, some authors favor ulnar rather than radial fixation when only a single bone is repaired.4,8,14,18 Ulnar fixation also allows the use of plates in birds weighing ≥ 200 g (0.44 lb), whereas the use of a plate on the radius would be possible only in much larger birds.26 Finally, some clinicians may elect to surgically repair both bones in birds with concomitant radial and ulnar fractures. In a previous experimental study of pigeons,18 fracture union occurred more rapidly when both the radius and ulna were stabilized as opposed to the ulna only, although the pin size used for stabilization of the ulna may have been suboptimal. Additional prospective studies with randomized assignment are needed to evaluate surgical treatments for birds with concomitant radial and ulnar fractures that involve the same third of the antebrachium.
A limitation of the present study was the variation in the types of antebrachial fractures and avian species included, which could have resulted in heterogeneity in healing ability among birds. In addition, because multiple clinicians were involved in the management of these cases, the criteria used to select between surgical and nonsurgical treatments may have varied. Some of the subgroups included small numbers of birds, which may have resulted in insufficient power to detect associations between some fracture characteristics and the outcome. We also acknowledge that success rate estimates for the rehabilitated birds were likely imprecise given that released birds that did not survive may not necessarily have been retrieved and returned to the rehabilitation program after death. Despite these limitations, the characteristics of antebrachial fractures associated with a successful or unsuccessful outcome in the present study could potentially be used by other rehabilitation programs for birds of prey to develop triage criteria that take ethical and financial considerations into account.
Acknowledgments
No external funding was used in this study. The authors declare that there were no conflicts of interest.
The authors thank Guy Beauchamp for statistical advice.
ABBREVIATIONS
| CI | Confidence interval |
Footnotes
Microsoft Excel for Macintosh 2011, version 14.5.7, Microsoft Corp, Redmond, Wash.
R, version 3.1.3, R Foundation for Statistical Computing, Vienna, Austria. Available at: www.r-project.org. Accessed Aug 25, 2018.
References
1. Murphy CJ, Kern TJ, McKeever K, et al. Ocular lesions in free-living raptors. J Am Vet Med Assoc 1982;181:1302–1304.
2. Smith KA, Campbell GD, Pearl DL, et al. A retrospective summary of raptor mortality in Ontario, Canada (1991–2014), including the effects of the West Nile virus. J Wildl Dis 2018;54:261–271.
3. Fitzgerald G. Programme de réhabilitation des oiseaux de proie au Québec: bilan 1986–2013. Nat Can 2015;139:74–81.
4. Hatt JM, Christen C, Sandmeier P. Clinical application of an external fixator in the repair of bone fractures in 28 birds. Vet Rec 2007;160:188–194.
5. de Margerie E. Laminar bone as an adaptation to torsional loads in flapping flight. J Anat 2002;201:521–526.
6. Beaufrère H. A review of biomechanic and aerodynamic considerations of the avian thoracic limb. J Avian Med Surg 2009;23:173–185.
7. Hamdy MK, May KN, Flanagan WP, et al. Determination of the age of bruises in chicken broilers. Poult Sci 1961;40:787–789.
8. Bennett R, Kuzma A. Fracture management in birds. J Zoo Wildl Med 1992;23:5–38.
9. Rubin JA, Runge JJ, Mison M, et al. Surgery. In: Speer BL, ed. Current therapy in avian medicine and surgery. St Louis: Elsevier, 2016;631–667.
10. Fitzgerald G, Tremblay JA, Lemaître J, et al. Captures accidentelles d'aigles royaux et de pygargues à tête blanche par les trappeurs d'animaux à fourrure au Québec. Nat Can 2015;139:82–89.
11. Beaufrère H, Laniesse D. Medicine of strigiformes. In: Speer BL, ed. Current therapy in avian medicine and surgery. St Louis: Elsevier, 2016;566–581.
12. Wagner H, Weger M, Klaas M, et al. Features of owl wings that promote silent flight. Interface Focus 2017;7:20160078.
13. Meyers RA. The morphological basis of folded-wing posture in the American kestrel, Falco sparverius. Anat Rec 1992;232:493–498.
14. Martin H, Ritchie B. Orthopedic surgical techniques. In: Ritchie B, Harrison G, Harrison L, eds. Avian medicine: principles and applications. Lake Worth, Fla: Wingers Publishing, 1994;1137–1169.
15. Werner FW, An KN. Biomechanics of the elbow and forearm. Hand Clin 1994;10:357–373.
16. Youm Y, Dryer R, Thambyrajah K, et al. Biomechanical analyses of forearm pronation-supination and elbow flexion-extension. J Biomech 1979;12:245–255.
17. Nunamaker D. Fractures of radius und ulna. In: Newton C, Nunamaker D, eds. Textbook of small animal orthopedics. Philadelphia: Lippincott Williams & Wilkins, 1985;1140.
18. Newton CD, Zeitlin S. Avian fracture healing. J Am Vet Med Assoc 1977;170:620–625.
19. Drebot MA, Lindsay R, Barker IK, et al. West Nile virus surveillance and diagnostics: a Canadian perspective. Can J Infect Dis 2003;14:105–114.
20. Dubé MC, Bird DM, Dibernardo A, et al. Prevalence of West Nile virus in wild American Kestrels (Falco sparverius) of southern Quebec, Canada. J Wildl Dis 2010;46:603–607.
21. Nemeth NM, Kratz GE, Bates R, et al. Clinical evaluation and outcomes of naturally acquired West Nile virus infection in raptors. J Zoo Wildl Med 2009;40:51–63.
22. Ouhoumanne N, Lowe AM, Fortin A, et al. Morbidity, mortality and long-term sequelae of West Nile virus disease in Quebec. Epidemiol Infect 2018;146:867–874.
23. Robillard A, Therrien JF, Gauthier G, et al. Pulsed resources at tundra breeding sites affect winter irruptions at temperate latitudes of a top predator, the snowy owl. Oecologia 2016;181:423–433.
24. Beaufrère H, Ammersbach M, Nevarez J, et al. Successful treatment of a radioulnar synostosis in a Mississippi kite (Ictinia mississippiensis). J Avian Med Surg 2012;26:94–100.
25. Redig PT. Fractures. In: Samour J, ed. Avian medicine. 2nd ed. Philadelphia: Mosby, 2008;215–248.
26. Gull JM, Saveraid TC, Szabo D, et al. Evaluation of three miniplate systems for fracture stabilization in pigeons (Columba livia). J Avian Med Surg 2012;26:203–212.
