Manley (1981).

Distal extremity fractures in small animals

Aim: To assess the outcomes of distal extremity fractures in dogs treated either conservatively with external coaptation or with surgical stabilisation and to make suggestions based of these for future treatment selection.

Population:	Dogs admitted at a single institute between 1978–1980, with distal extremity fractures, in equal numbers of forelimb and hindlimb involved.
Sample size:	43 dogs.
Intervention details:	12 cases did not meet the inclusion criteria and so only the 31 cases of metabone fractures with follow up data will be reported on this table. n = 31 dogs (6 single bone fractures, 25 multiple bone fractures, 15 fractures involved articular components). Conservative management by external immobilisation (n = 20 dogs) included either Robert Jones bandage (RJB), Mason Meta splint, casting and Thomas splint. Surgical management (n = 11 dogs) included either singularly or with combinations of intramedullary pins, single cerclage wire, tension band plates, interfragmentary compression with screws, internal fixation with a plate. Surgical intervention was recommended when there was marked fragment displacement or involvement of 2 or more bones and if there was articular involvement.
Study design:	Retrospective case series.
Outcome Studied:	Nature of fracture, treatment assessment, and long-term outcomes. Outcomes were determined via a client questionnaire encouraging critical evaluation of their dogs 4–26 months post treatment. Any evidence of residual lameness or swelling associated with the fracture or its treatment was noted as a complication and an unsatisfactory outcome was assigned. No veterinary assessments of these cases to correlate client perception were reported but 2 cases did have revision surgeries. Complications of malunion (n = 3) and osteomyelitis (n = 3) were also reported thus veterinary assessment must have been present for these cases.
Main Findings (relevant to PICO question):	Overall, complications occurred in 10/11 surgically treated cases with 8/11 having persistent lameness. Complications occurred in 11/20 of the conservatively treated cases and 8/20 had persistent lameness according to the questionnaire follow up.  It is not specified how many of the cases had a follow up in person veterinary assessment. 6 cases had one single metabone fracture: 5 were treated conservatively with 3 developing complications (residual lameness, osteomyelitis). 1 was treated surgically and this developed complications (arthrosis and persistent lameness). 25 cases had 2 or more metabone fractures: 15 were treated conservatively. 5/15 conservatively treated cases had persistent lameness. 9/15 developed complications. One refractured after splint removal, one developed malunion requiring corrective osteotomies, and one required amputation of a digit. 10 were treated surgically. 7/10 surgical cases had residual lameness. 9/10 developed complications including malunion, malalignment, persistent drainage, or osteomyelitis. The fracture configurations all varied, and the complications were not statistically affected by the location of the fracture. The degree of fragment displacement and involvement of articular surfaces was a predictor of complications post treatment. 15 cases involved articular surfaces. 13/15 developed complications (residual lameness, osteomyelitis). Out of the 2 that were treated without complication one was conservative and one was surgical. None of the fractures with minimal displacement had unsatisfactory outcomes whereas 25/29 (86%) of all fractures reported in the paper with marked displacement had unsatisfactory outcomes. Unsatisfactory outcomes were also reported in 13/15 (87%) of those with articular involvement.
Limitations:	Retrospective case series are a weaker form of study regarding strength and quality of evidence. The methods of assessing outcome were poor. Client questionnaires – although useful in large numbers (registries) – can be unreliable and biased. There was also no access to the questionnaires for evaluation. Not every case had veterinary assessment and follow up. Patient data that may have had a bearing on the outcomes of the cases was absent e.g. the age and weights of the patients were not recorded. Whether the fracture was open or closed was recorded but the concurrent soft tissue damage or nature of injury was not. Modern surgical repair methods are superior to the techniques implemented during the years of this study, so to draw conclusions about the outcomes of the surgical treated cases is difficult. The study is now over 40 years old. Sample size was relatively small. Surgical intervention versus conservative management was not randomly selected so there may have been a bias to only choose conservative management in more simple fractures.