• 
• 

Three Part Question

[Are children under 4 years old presenting to emergency departments with spiral fracture of the tibia (with or without radiological evidence)] [best managed in cast, boot or with no immobilization] [in respect to fracture healing, pain control and complication rates?]

Clinical Scenario

You are seeing a 2 year old boy who has been non-weightbearing since sustaining a twisting injury whilst running. X-ray has revealed a toddler fracture of his left tibia (non-displaced spiral fracture). You are planning to place him in an above knee backslab but when you explain this to his mother she becomes upset because she feels this will be very distressing to him. You wonder if there are any other treatment options.

Search Strategy

Medline using Pubmed Interface

Search terms: "toddler's fracture" OR "toddler fracture" OR (spiral AND fracture AND (tibia OR tibial)) – limited to English

Inclusion:
• 4 years and under
• Presenting with spiral fracture of tibia with or without radiological evidence
• Management - cast, splint or no immobilisation
• Observational studies, control trials, comparative studies, cohort studies, case-control studies
• In English
• Human trials

Exclusion:
• Systematic review

Search Outcome

161 studies identified
1 duplicates removed
160 studies screened against title and abstract
134 studies excluded
24 studies assessed for full-text eligibility
20 studies excluded
10 Wrong study design
6 Wrong outcomes
2 Wrong patient population
1 Wrong comparator
1 Wrong intervention

4 studies included

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
John S Adamich and Mark W Camp 2017 Canada	Children presenting to Emergency Department and orthopaedic outpateints of Hospital for Sick Children in Toronto (large paediatric teaching hospital) between Jan 2009 and December 2014, aged 0-4 years with confirmed fracture of tibia on initial x-ray with less than 2mm of diplacement. Exclusion criteria were perinatal injury, multiple fractures, non-accidental injury, underlying bone disease, complicated fractures and fractures followed up at outside institutions. Patients were identified by a search of the institution’s radiology database and following this review of the electronic records. 159 children were initially treated with a cast (above or below knee backslab or circular cast) and 12 children with no cast. This was at the choice of the treating physician.	Single centre, retrospective cohort study with parallel groups.	Achieving radiological or clinical union	All 159 children treated with a cast and 12 children treated without a cast achieved clinical union	• The study is only a single centre observational study and also did not set out to directly answer the question asked in this systematic review but instead the results have been extrapolated to do so. It is not a randomised controlled trial and there are a number of biases present. • There are a number of confounders that have not been adjusted for including age, severity of fracture and mechanism of injury which may all be confounding variables that could affect choice of mobilisation method, fracture healing and risk of complications. • The decision about whether to place in a cast or not was down to the clinician treating and there are likely to represent different patient groups (for instance age, weight-baring status and degree of distress on presentation amongst other factors would likely have influenced which treatment type clinicians chose). • Children followed up in other places were excluded from the study and it is not clear if these children shared the same characteristics as children followed up locally or not. • There is no blinding of patients or parents as to intervention (although this would be impossible to achieve. There was also no blinding of the assessors of the outcomes as to which invervention group the patient was in. • Only a very small number of children were treated without a cast. • This study did not look at children diagnosed with a toddler fracture without radiological findings.
			Refracture	None of the children in either group sustained a refracture
			Cast-related complications	No child who was placed in any form of cast had any cast related complication
Jennifer M. Bauer, Steven A. Lovejoy 2017 USA	Children aged 9 years to 4 months who had at least 1 tib-fib x-ray showing isolated, non-displaced spiral tibial fracture with intact fibula or those without the appearance of any injury being followed up by orthopaedics at the study centre – an orthopaedic department at a large dedicated children’s hospital in USA (having either presented to outside institutions or local ED). Children were excluded if they had a fibula fracture or buckle, transvers, physeal or displaced tibial fracutres, there was suspicion of non-accidental injury or the injury was the result of polytrauma or there was loss of follow up before clinic discharge. Patients were identified by a search of the institution’s radiology database and following this review of the patient’s notes to obtain the data as well as reviews of repeat x-rays. 139 children were initially treated with a cast or splint (either above or below knee), 46 with a controlled ankle movement boot and and 7 with no immobilisation.	Single centre, retrospective cohort study with parallel groups.	Fracture displacement	No children treated with or without a cast or a controlled ankle movement boot had fracture displacement	• The study is only a single centre observational study. It is not a randomised controlled trial and there are a number of biases present. • There are a number of confounders that have not been adjusted for including age, severity of fracture and mechanism of injury which may all be confounding variables that could affect choice of initial mobilisation method, fracture healing and risk of complications. • The decision about whether to place in a cast, controlled ankle movement boot or no immobilisation was down to the clinician treating and there are likely to represent different patient groups (for instance age, weight-baring status and degree of distress on presentation amongst other factors would likely have influenced which treatment type clinicians chose). All children not immobilised were missed fractures and in particular this group are likely to differ. • It is unclear what percentage of eligible population included as children lost to follow up before clinic discharge were excluded and whether this group differed to other children. • There is no blinding of patients or parents as to intervention (although this would be impossible to achieve). There was also no blinding of the assessors of the outcomes as to which intervention group the patient was in. • Only a very small number of children were treated without a cast. • Immobilisation type could be changed at follow up appointment and in fact 55 of the children placed in a cast or splint were changed to a boot at follow up and this may make comparison of the groups more difficult. However none of the children who were not immobilised were subsequently immobilised. • Time to weight-baring was influenced by timing of clinic appointment as this was when this information was recorded.
			Skin breakdown	3/130 children treated with a cast had skin breakdown. No children in the other two groups had skin breakdown
			Time to weight baring	98% of all children were weight-baring by 4 weeks. There was a significantly earlier return to weight-bearing if initial immobilization was a boot rather than a short leg cast (2.5 vs. 2.8 wk, P= 0.04). 100% of children not initially placed in any form of immobilisation were weight-baring by 2.5 weeks.
Karuna Sapru and Jamie G. Cooper 2014 Scotland	All children between 9 months and 36 months presenting to the Emergency Department of the Royal Aberdeen Children’s Hospital between October 2008 and July 2010 who had x-ray of tibia and fibula and x-ray confirmed toddler fracture or clinically felt to have toddler fracture. Exclusion criteria were patients with features inconsistent with a diagnosis of toddler fractre in their history (e.g. significant mechanism of injury), on examination (e.g. clinical deformity, signs of infection), or on radiographic examination (e.g.displaced fracture of the tibia and fibula). 19 children were treated with a cast (either backslab or circular cast) and 10 children were treated without a cast.	Single centre, retrospective cohort study with parallel groups	Suffered any complication with treatment	No children in either group suffered any complications.	• The study is only a single centre observational study and also did not set out to directly answer the question asked in this systematic review but instead the results have been extrapolated to do so. It is not a randomised controlled trial and there are a number of biases present. • There are a number of confounders that have not been adjusted for including age, severity of fracture and mechanism of injury which may all be confounding variables that could affect choice of mobilisation method, fracture healing and risk of complications. • The decision about whether to place in a cast or not was down to the clinician treating and there are likely to represent different patient groups (for instance age, weight-baring status and degree of distress on presentation amongst other factors would likely have influenced which treatment type clinicians chose). • Children with an initial non-radiologically confirmed toddler fracture were much more likely to have no immobilisation then children with a radiologically confirmed fracture. • It is not recorded if any eligible children were not included because they were lost to follow up (although this is the orthopaedic department following up children incorporating a large geographical area and so most children were likely to have been followed up in Aberdeen. • There is no blinding of patients or parents as to intervention (although this would be impossible to achieve. There was also no blinding of the assessors of the outcomes as to which intervention group the patient was in. • This was a very small study with only 43 children included. • It is not recorded if any children had their treatment type changed during follow up.
Abigail M. Schuh, Kathryn B. Whitlock and Eileen J. Klein 2016 USA	All patients age 9 months to 3 years presenting to the paediatric emergency department (PED) at Seattle Children’s Hospital (SCH) between January 1, 2008, and December 31, 2012, with radiographic evidence of toddler’s fracture. Patients were excluded if they had a history of metabolic bone disease (including rickets and osteogenesis imperfecta), if the fracture was thought to be caused by non-accidental trauma, if there was an associated fibular fracture or displacement requiring reduction by orthopaedics, or if fracture was not diagnosed at the time of the patient’s visit where the radiograph visualizing the fracture was initially obtained. Patients were identified by searching computerised radiological databases for children with a confirmed toddler fracture and data obtained by reviewing the medical records for eligible patients. Children were initially immobilised with either a splint or cast (50 children), controlled ankle movement boot (18 children) or no initial immobilisation (7 children).	Single centre, retrospective cohort study with parallel groups	Skin breakdown	13/50 children (26%) initially placed in a cast developed skin breakdown. No children initially placed in controlled ankle movement boot or having no initial form of immobilisation had skin breakdown	• The study is only a single centre observational study and also did not set out to directly answer the question asked in this systematic review but instead the results have been extrapolated to do so. It is not a randomised controlled trial and there are a number of biases present. • Immobilisation method was frequently changed at follow up. 50% of children with no initial form of immobilisation, 27.3% placed in a controlled ankle movement boot and 10.4% of children placed in a cast had a change in therapy type at follow up but the study does not describe what they were changed to which makes interpreting the results challenging. • The decision about whether to place in a cast or not was down to the clinician treating and there are likely to represent different patient groups (for instance age, weight-baring status and degree of distress on presentation amongst other factors would likely have influenced which treatment type clinicians chose). • There was also no blinding of the assessors of the outcomes as to which intervention group the patient was in. • This study did not look at children diagnosed with a toddler fracture without radiological findings.
			Return to ED due to pain	6/50 912%) children initially placed in a cast returned due to pain (2 due to skin breakdown, 4 due to ill-fitting casts. No children initially placed in a controlled ankle movement boot or having no initial form of immobilisation returned due to pain.
			Number of repeat x-rays	Children initially placed in a cast had a mean of 1.3 repeat x-rays (95% CIs 1-1.6). Children initially placed in a controlled movement boot had a mean of 0.5 repeat x-rays (95% CIs 0.2-0.9). Children who had no initial immobilisation had a mean of 0.4 repeat x-rays (95% CIs 0.1-1.3). P value was <0.001.
			Length of immobilisation for children who had local orthopaedic follow up (43 initially placed in cast, 8 in controlled ankle movement boot and 7 with no initial immobilisation)	Children initially placed in a cast had a mean length of immobilisation of 27.5 days (95% CIs 26.0-29.1). Children initially placed in a controlled ankle movement boot had a mean length of immobilisation of 27.0 (95% Cis 23.5-30.9). Children who had no form of initial immobilisation were immobilised for a mean of 4.1 days (95% Cis 2.8-5.9). P value was <0.001.

Comment(s)

The studies identified are all observational studies and for the most part the evidence is indirect as the studies were not designed to answer the question posited here. The studies all strongly suggest that there good healing of toddler fractures irrespective of the treatment used but it is unclear whether more severe toddler fractures were more likely to be immobilised or not. Only 1 study looked at whether there was any difference in relation to pain control and this did not do so in a robust and good quality way and so evidence in relation to this is severely lacking.

Clinical Bottom Line

There appears to be no difference in healing between immobilisation with cast or controlled-ankle movement boot or no immobilisation but there is not sufficient evidence in relation to pain control.

References

1. John S Adamich and Mark W Camp Do toddler's fractures of the tibia require evaluation and management by an orthopaedic surgeon routinely? European Journal of Emergency Medicine June 2017 Epublished ahead of print
2. Jennifer M. Bauer, Steven A. Lovejoy Toddler’s Fractures: Time to Weight-bear With Regard to Immobilization Type and Radiographic Monitoring Journal of Paediatric Orthopaedics January 2017 Epublished ahead of print
3. Karuna Sapru and Jamie G. Cooper Management of the Toddler’s fracture with and without initial radiological evidence European Journal of Emergency Medicine 2014 21:451–454
4. Abigail M. Schuh, Kathryn B. Whitlock and Eileen J. Klein Management of Toddler’s Fractures in the Pediatric Emergency Department Pediatric Emergency Care July 2016. Volume 32, Number 7. 452-454