Kornmayer et al. (2014).

Long-term prognosis of metacarpal and metatarsal fractures in dogs. A retrospective analysis of medical histories in 100 re-evaluated patients

Aim: To retrospectively evaluate cases of metacarpal and metatarsal fracture in dogs with complete clinical and radiographic follow up to determine the long term prognosis.

Population:	Dogs of mixed breed and age with complete clinical and radiographic follow up following metabone fractures within 4months–14 years post injury. Cases from 1990–2007 from a single centre. 37% of animals were under 1 year old and 41% were under 2 years old. Mean age was 2.6 years.
Sample size:	100 dogs.
Intervention details:	55/100 (55%) of the fractures were < 50% displaced. 14/100 (14%) of the fractures were 50–100% displaced displaced grade 2. 31/100 (31%) of the fractures were > 100% displaced. 85/100 (85%) of fractures affected the body of the metabone. 84/100 (84%) of fractures were closed. Cases were divided into 3 groups: Conservative management. Surgical management. Combination of both surgical and conservative management. Treatment selection was made on a case by case basis. Group 1 (conservative management): 67 dogs (68 limbs) with mildly displaced, non-reconstructible multiple bone fractures, physis fractures, or multiple bone fractures which could undergo closed reduction under general anaesthetic and external coaptation. Some with open fractures underwent debridement, ± wound closure. A synthetic splint and bandage was applied for an average of 6 weeks (range 4–12 weeks). The bandages were changed weekly. Group 2 (surgical treatment): 25 dogs (25 limbs) with severely displaced fractures of single bones, re-constructible articular fractures, and multiple bone fractures particularly those involving third and fourth metabone. Dowel pinning and standard dorsal bone plating were most commonly used but there were cases of external fixation, lag screws and medial/lateral bone plates. All cases had a Robert Jones Bandage (RJB) applied for 4–8 weeks, changed weekly. External fixators and most bone plates were removed when healing was radiographically diagnosed.  Group 3 (combined surgical and conservative treatment): 8 dogs with multiple fractures of which not all could be surgically stabilised due to comminution, short bone fragments, or due to specific skin wounds. These cases had a combination of surgery and external coaptation similar to patients in group 1 i.e., synthetic splinted support dressings changed weekly.
Study design:	Retrospective case series.
Outcome Studied:	Outcomes evaluated at average time frame of 4 years post initial presentation. Final radiographic outcomes were interpreted by two investigators using orthogonal views of the affected and contralateral limb. Clinical/functional outcomes were graded as present or absent. 16 dogs were graded by more than one clinician, while 84 dogs were graded by a single clinician. 15 cases were analysed using computed gait analysis using treadmill with force plates owning to patient/client compliance.
Main Findings (relevant to PICO question):	The main findings highlighted below are those relevant to the PICO question. Fracture classification, degree of displacement, metabone number, location on bone and which limb with relation to complications were observed within the study but not reported in detail here. Radiographic outcome reported in respect to complications noted on assessment: Group 1 (conservative treatment): 2/67 (3%) dogs showed early complications of delayed union radiographically. At final radiographic follow up; 9/67 dogs had malunion (13%). Synostosis occurred in 5/67 dogs (7%). Osteoarthritis (OA) was seen in 2/67 dogs (3%). Non union was present in one bone of 1 dog (that had third to fifth metatarsal bone fractures). Group 2 (surgical treatment): 3/25 (12%) dogs developed early complications. 2 developed osteomyelitis and implant loosening. One of which had open shaft fractures ofmetabones (second to fifth) treated with dowel pinning. The other had base fractures of metabones (second to fifth) repaired with cross pins. The third dog had open metabone (second to fifth)fractures and was treated with an external skeletal fixator. Delayed union occurred. At final radiographic follow up; 3/25 (12%) dogs had malunion (12%). 3/25 (12%) dogs had synostosis. Group 3 (combination of conservative and surgical treatment): 3/8 (38%) exhibited complications. 2 cases had implant loosening without bone healing impairment. Both had second metacarpal treated with bone plate and the third metacarpal was treated with external coaptation. The third case had delayed union. This case had open fractures of second to fifth and was repaired with oversized plates. To prevent non union revision surgery to reduce implant size and apply cancellous bone autograft was required. At final radiographic follow up; 2 dogs had malunion (25%). 1 dog showed evidence of OA. Synostosis occurred in 3 dogs. Synostosis in all cases seemed more frequent in proximal metabone fractures and multipart fractures. Functional outcome: Clinical assessment of resolution of lameness: Group 1 (conservative treatment): 65/67 (97%) dogs were free of lameness. Lameness was due to malunion in one case and severe soft tissue injury in another leading to OA. 10 cases had bandage associated problems during recovery – sores, dermatitis, etc. which required treatment but had no significant bearing on final outcomes. Group 2 (surgical treatment): 24/25 (96%) dogs free of lameness. Lameness was due to implant loosening requiring early implant removal and malunion of third metabone. Group 3 (combination of conservative and surgical treatment: 8/8 (100%) dogs free of lameness. Computed gait analysis matched up with clinical assessment for most cases (12/15) but found subtle lameness not detected on visual examination in 3 cases. Statistical results: Final outcomes were generally very good in this paper with only 3/100 (3%) of dogs having lameness reported, 1/100 (1%) having a non union and 3/100 (3%) developing radiographic signs of osteoarthritis. Overall there was a 14/100 (14%) rate of malunion and 19/100 (19%) rate of synostosis but these two complications had no bearing on functional outcome. Synostosis was statistically more common when surgical intervention was involved in terms of fractures of three to four metabones compared to conservative management. Despite a statistically higher incidence of synostosis in surgically treated patients there was no significant difference in outcome between conservative vs. surgically managed cases. The incidence of malunion, non union or OA and functional outcomes were not statistically different in the groups. There was no significant difference in outcomes between metacarpal or metatarsal fractures within the 3 groups; however, the study showed that if complications occurred, this would result in an increased rate of synostosis and malunion. This seemed truer in the metatarsal bones and malunion of the fifth metatarsal bone was more prominent. Complications were also more likely to occur when dealing with open fractures, oblique and comminuted fractures of the base, severely displaced fractures, and when surgery was required. This again was seen more frequently in the metatarsus. No correlation between potential influencing factors and lameness was detected, because lameness was actually rarely found at the final follow up. No statistical conclusion can be made to recommend surgical vs. conservative management but it would appear good outcomes can be achieved from either approach.
Limitations:	Retrospective case series are a weaker form of study regarding strength and quality of evidence. Case treatment was not randomised so it could be argued that groups 2 and 3 were made up of more challenging cases with more severe fracture displacement and therefore comparable outcomes to group one would support the use of surgical intervention in said cases. There are many variations of fracture type in this region with regard to number of metabones, location of the fracture and degree of displacement so it is difficult to interpret the data. There were variable re-evaluation periods in patients so we can draw no conclusions to whether surgery or conservative management in these cases leads to earlier or more prolonged return to function even though long-term outcomes appeared to be favourable. There were discrepancies between the subjective assessment of lameness by clinicians and the vertical ground reactions in 3/15 cases that underwent computed gait analysis so this may indicate a clinician bias or difficulty in seeing subtle persistent lameness’ at long-term follow up of these cases. This could mean that the percentage of successful outcome could be falsely high.