You are on page 1of 12

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Available online at www.sciencedirect.com

The Surgeon, Journal of the Royal Colleges


of Surgeons of Edinburgh and Ireland
www.thesurgeon.net

Review

Acute fractures of the scaphoid bone: Systematic review and


meta-analysis
S. Alshryda a,*, A. Shah b, S. Odak b, J. Al-Shryda c, B. Ilango d, S.R. Murali b
a

Department of Orthopaedics, James Cook University Hospital, 7 Finchlay Court, Middlesbrough TS5 8EL, UK
Departments of Trauma and Orthopaedics, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan WN6 9EP, UK
c
Department of Surgery, Collm Klinik Oschatz GmbH, Parkstr. 1, Oschatz, Germany
d
Department of Trauma and Orthopaedics, Fairfield General Hospital, Old Rochdale Road, Bury, Greater Manchester BL9 7TD, UK
b

article info

abstract

Article history:

Background: The scaphoid fractures account for 50%e80% of all carpal bone fractures in

Received 11 January 2012

young individuals. Non-union of the fracture occurs in approximately 5%e10% of undis-

Received in revised form

placed scaphoid fractures. Current management varies significantly among different pla-

14 March 2012

ces and surgeons.

Accepted 21 March 2012

Objectives: The purpose of this review is to investigate the evidence of the effectiveness and

Available online 15 May 2012

safety of various treatments of acute scaphoid fractures.


Methodology: Systematic review and metanalysis of all the randomised and quasi-

Keywords:

randomised trials comparing different treatments of acute scaphoid fractures.

Scaphoid fracture

Results: Thirteen RCTs (Published 18 times) have met our inclusion criteria. The followings

Scaphoid cast

have been investigated:

Colles cast

1. Colles cast versus scaphoid cast.

Open reduction

2. Above elbow versus below elbow scaphoid cast.

Internal fixation

3. Colles cast with the wrist in flexion versus Colles cast with the wrist in extension.

Non-union

5. Operative versus non-operative treatment.


6. Union rate versus time to union.
Conclusion: Scaphoid fracture can be treated by Colles cast for up to 12 weeks. The wrist
should not be in flexion. There is no advantage of an above elbow cast over a below elbow
cast. Operative treatment for scaphoid does not provide a higher union rate in undisplaced
fractures, but may do in displaced fracture. Open approach seems to be superior to
percutaneous fixation.
2012 Royal College of Surgeons of Edinburgh (Scottish charity number SC005317) and
Royal College of Surgeons in Ireland. Published by Elsevier Ltd. All rights reserved.

* Corresponding author.
E-mail addresses: sattar26@doctors.org.uk (S. Alshryda), orthopaedics@gmail.com (A. Shah), saurabhsodak@gmail.com (S. Odak),
Jalal78@alice.de (J. Al-Shryda), bilango@aol.com (B. Ilango), Raj.Murali@wwl.nhs.uk (S.R. Murali).
1479-666X/$ e see front matter 2012 Royal College of Surgeons of Edinburgh (Scottish charity number SC005317) and Royal College of
Surgeons in Ireland. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.surge.2012.03.004

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Introduction
The fracture of the scaphoid bone account for 50%e80% of all
carpal bone fractures in young and active individuals. A
common cause is a fall onto an outstretched hand with the
wrist bent back. The fracture can sever the blood flow to the
bone, causing the fracture to heal slowly (delayed union) or
not at all (non-union). Early evaluation and appropriate
treatment are therefore important in avoiding complications
such as avascular necrosis (AVN) and carpal collapse. Nonunion of the fracture occurs in approximately 5%e10% of
fractures of undisplaced scaphoid. The incidence increases to
up to 90% for displaced proximal pole fractures.1 Recent
studies have shown that non-union can lead to carpal collapse
and post-traumatic arthritis. For this reason, treatment is
recommended for all scaphoid non-union, even if the patient
is asymptomatic.2,3
Current management varies significantly among different
places and surgeons and it is quite usual to find different
treatment approaches even within the same centre.
Short scaphoid cast, long scaphoid cast, Colles cast and
scaphoid splint have all been advocated to treat undisplaced
scaphoid fractures. A long (above elbow) cast can lead to
stiffness of the elbow and the DVLA (Driver and Vehicle
Licensing Agency) does not allow patients to drive while
having such a cast. Short casts may be suboptimum treatment
according to some authors.4e7 Moreover, there is still
disagreement on how long the fracture should be protected in
plaster.
If the scaphoid fracture is displaced, the risk of non-union
is higher and surgery may be recommended to reposition and
fix the bones into place.8 However, there is more than one way
to do the surgery (percutaneous approach, open volar
approach or open dorsal approach). Even with surgery, fractures of the scaphoid bone can take long time to heal, and it is
possible that a non-union or avascular necrosis will occur.8
For this reason some have advocated additional interventions such as electrical stimulation or therapeutic ultrasound
to promote healing.9,10
The diversity of scaphoid fracture clinical presentations,
the complex anatomical and functional properties and the
wide range of possible interventions make acute management of scaphoid fractures a complex issue. Although it is
common to use closed methods to treat a stable and undisplaced fracture, proponents of operative treatment claim
surgery allows early return of function and should be regarded as an alternative to conservative treatment in patients in
whom immobilisation in a cast for three months is unacceptable for reasons related to sports, social life or work.11
This review aims to identify interventions for which there
is evidence of effectiveness and safety from randomised
controlled trials.

219

Study selection criteria


Types of studies
Randomised controlled trials (RCTs) and quasi-randomised
(for example, allocation by hospital number or date of birth)
trials comparing different treatments of acute scaphoid
fractures were considered within this review. We excluded
trials that investigated interventions for non united scaphoid
fractures.

Types of interventions included


Any intervention used to treat acute scaphoid fractures. For
practical reasons they were grouped under the following three
headings.
1. Non-operative interventions:
! Types of cast: (Cast versus splints, scaphoid cast versus
Colles cast, below elbow (short) cast versus above elbow
(long) cast).
! Adjuvant interventions such as ultrasound and electromagnetic pulses.
2. Non-operative versus operative interventions.
3. Operative interventions.

Types of outcome measures


The primary outcome measure within the review was the
union rate of fractures. The secondary outcome measures
were: time to union, pain scale, range of movement, grip
strength, return to work/pre-injury sport activities, general
quality of life outcome, functional wrist outcome measure,
patient satisfaction and complications e.g. avascular necrosis,
osteoarthritis, need for further operations.

Search methods
We searched the Bone, Joint and Muscle Trauma Group Specialised Register, the Cochrane Central Register of Controlled
Trials (The Cochrane Library, current issue), MEDLINE
(1966eJune 2010), EMBASE (1980eJune 2010), CINAHL
(1982eJune 2010), Library and Information Science Abstracts
(from 1969eJune 2010), Science Citation Index (ISI Web of
Science 1987eJune 2010). Search strategies can be found in
Appendix 1. No language restrictions were applied. The
bibliographies of retrieved trials and other relevant publications, including reviews and meta-analyses, were examined
for additional articles. The following websites were searched
to identify unpublished and ongoing studies: Current
Controlled Trials (www.controlled-trials.com); Centre Watch
(www.centerwatch.com); Trials Central (www.trialscentral.
org/ClinicalTrials.aspx); The UK National Research Register
(www.nrr.nhs.uk). Journal of Bone and Joint Surgery e British
Volume and American Volume (www.ejbjs.org), and the
American Academy of Orthopaedic Surgeons (www.aaos.org)
were searched manually.

Data collection and analysis

Materials and methods


Study selection
The review was conducted in accordance with guidelines
described in the Cochrane handbook for systematic review
and meta-analysis of interventions.12

Two authors (SO and AS) independently applied the search


strategy to select references from the above mentioned databases. The article titles and abstracts were then reviewed

220

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

independently by SO and AS. When there was an issue


requiring clarity, the full article was retrieved for further
scrutiny. The two authors independently assessed each full
study report to see whether it met the reviews inclusion
criteria. Where necessary, authors were contacted for more
information and clarification of data. If there was disagreement regarding inclusion, the senior authors (BI & RM) would
compare findings and if no consensus could be reached, the
study was excluded.

Assessment of methodological quality of included studies


The review authors used a modification of the generic evaluation tool used by the Cochrane Bone, Joint and Muscle

Trauma Group13 (Tables 1 and 2). Two authors (SO and AS)
assessed the methodological quality of each study. Disagreement was resolved by the senior authors (BI & RM) consensus.
The total quality assessment score (QAS) was reported for
each study to give an overall impression of the study but was
not used to weight studies in meta-analyses.

Data extraction and management


A data extraction form was designed and agreed by the review
team. A pilot test using five articles was performed to ensure
the forms consistency. The form was then refined accordingly. Initially, three authors (SA, AS, and SO) extracted data
independently and reviewed extracted data jointly.

Table 1 e Quality assessment items and possible scores.


A. Was the assigned treatment adequately concealed prior to allocation?
2 method did not allow disclosure of assignment.
1 small but possible chance of disclosure of assignment or unclear.
0 quasi-randomized or open list/tables.
B. Were the outcomes of participants who withdrew described and included in the analysis (intention to treat)?
2 withdrawals well described and accounted for in analysis.
1 withdrawals described and analysis not possible.
0 no mention, inadequate mention, or obvious differences and no adjustment.
C. Were the outcome assessors blinded to treatment status?
2 effective action taken to blind assessors.
1 small or moderate chance of unblinding of assessors.
0 not mentioned or not possible.
D. Were the treatment and control group comparable at entry? (Likely confounders may be age, weight or co morbidity).
2 good comparability of groups, or confounding adjusted for in analysis.
1 confounding small; mentioned but not adjusted for.
0 large potential for confounding, or not discussed.
E. Were the participants blind to assignment status after allocation?
2 effective action taken to blind participants.
1 small or moderate chance of unblinding of participants.
0 not possible, or not mentioned (unless double-blind), or possible but not done.
F. Were the treatment providers blind to assignment status?
2 effective action taken to blind treatment providers.
1 small or moderate chance of unblinding of treatment providers.
0 not possible, or not mentioned (unless double-blind), or possible but not done.
G. Were care programmes, other than the trial options, identical?
2 care programmes clearly identical.
1 clear but trivial differences.
0 not mentioned or clear and important differences in care programmes.
H. Were the inclusion and exclusion criteria clearly defined?
2 clearly defined.
1 inadequately defined.
0 not defined.
I. Were the interventions clearly defined?
2 clearly defined interventions are applied with a standardised protocol.
1 clearly defined interventions are applied but the application protocol is not standardised.
0 intervention and/or application protocol are poorly or not defined.
J. Were the outcome measures used clearly defined?
2 clearly defined.
1 inadequately defined.
0 not defined.
K. Were diagnostic tests used in outcome assessment clinically useful?
2 optimal.
1 adequate.
0 not defined, not adequate.
L. Was the surveillance active, and of clinically appropriate duration?
2 active surveillance and appropriate duration.
1 active surveillance, but inadequate duration.
0 surveillance not active or not defined.

221

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Table 2 e Characteristics of included studies.


Study

Intervention

Adolfsson 2001

53

Alho 1975

99

Arora 2007

44

Bond 2001

30

Clay 1991

291

Dias 2005

88

Gellman 1989

51

(1) Non-operative BE cast.


(2) Operative percutaneous Acutrack screw.
(1) Non-operative BE scaphoid cast
(2) Non-operative AE scaphoid cast
(1) Non-operative BE scaphoid cast
(2) Operative percutaneous Acutrack screw.
(1) Non-operative AE scaphoid cast for 6 weeks
then BE scaphoid cast until union
(2) Operative percutaneous Acutrack screw.
(1) Non-operative scaphoid cast.
(2) Non-operative Colles cast
(1) Non-operative BE Colles cast
(2) Operative treatment using Herbert screw
(1) Non-operative BE scaphoid cast
(2) Non-operative AE scaphoid cast
(1) Non-operative Colles cast in 20# of wrist
flexion (2) Non-operative Colles cast in 20# of
wrist extension
(1) Non-operative standard cast
(2) Non-operative standard cast low intensity
ultrasound adjunct.
(1) Non-operative Colles cast
(2) Operative percutaneous Acutrack screw.
(1) Non-operative scaphoid cast.
(2) Non-operative FRPSC
(1) Non-operative BE scaphoid cast
(2) Operative treatment using Herbert screw
(1) Non-operative BE scaphoid cast
(2) Operative treatment using Herbert screw

Hambidge 1999

121

Mayr 1999

30

McQueen 2008

60

Petty 1999

34

Saeden 2001

85

Vinnars 2007

29

Disagreements were resolved by consensus or consultation


with the senior reviewers (BI & RM). If necessary, authors of
individual trials were contacted directly to provide
clarification.

Measures of treatment effect


Continuous data was recorded as a mean, standard deviation
and group size for each trial arm, with treatment effect being
reported as the mean difference. We used the weighted mean
difference to summarise across trial findings since outcomes
were measured consistently across trials (without recourse to
using standardised mean differences). Dichotomous data
were expressed as proportions or risks, with the treatment
effect reported as a relative risk or risk difference.
Missing data was sought from the authors. Where this was
not possible or data was missing through loss to follow up,
intention-to-treat principles were used. No attempts of
imputation were made. Trials with multiple arms (using
different doses of TXA) were conflated to a single comparator
of TXA or placebo. Review Manager (RevMan 5, The Nordic
Cochrane Centre, Copenhagen), was used to present study
findings and combine the estimates of the treatment effects.
Summary estimates of the overall treatment effects are
provided in the form of a forest plot. The Mantel-Haenszel (HM) method was used to combine studies using fixed effects
model. The presence of statistical heterogeneity was assessed
through Q and I2 statistics, a value of >50% being considered
as substantial heterogeneity.

Non-union

Displaced fractures

CT Scan

No

Plain x ray or
tomogram
CT Scan

Yes

Plain x ray

No

Plain x ray

Yes

CT Scan

Yes

Plain x ray

No

Plain x ray

Yes

CT Scan

No

Plain x ray

Yes

Plain x ray

No

CT Scan

Yes

Plain x ray

No

No

Subgroup analysis
The following sub-group analyses were prospectively planned.1 The site of fractures (proximal, middle or distal)2; displaced fracture versus undisplaced fractures3; dorsal versus
volar approach4; type of metal ware fixation used (K-wires,
Herbert screws, Aquatrack screws etc). Data on the first three
subgroups were lacking and it was not possible to conduct
these analyses.

Results
Description of studies
One hundred and thirty five citations were identified as
potentially relevant studies. Subsequent scrutiny led to the
exclusion of 99 of these citations. Full publications were
obtained for 36 citations. These were assessed and 19 further
citations were excluded.
The remaining 17 citations referred to 13 trials4,9,11,14e23
included in the study (Fig. 1). Some have been published
more than once with or without extra information, sometimes
under different corresponding authors. The data from the
original publication was used; however, extra useful information from subsequent publication was also used. A note has
been added to verify this where necessary. The multiple
references of the same study were grouped together under the
original (primary) publication which was denoted by an asterix.

222

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

and subsequently, a different technique was used to fix the


fracture. Although, he excluded this patient from his analysis, he should have analysed this patient in the operative
group. On the other hand Dias et al.19 did not exclude
participants who did not complete their follow up from the
analysis.
Terminology can be a source of bias. The definition of
a displaced fracture was not clear in most trials. Bond
et al.17 considered a fracture displaced when there is a gap
of more than 1 mm. Dias et al.19 gave a thorough description of shape, configurations and stability of his series,
however, he stated that displacement is difficult to quantify
radiologically and, at best, can be used as a relative indication for surgery.
Some authors used the name above elbow casts, others
called it long cast. Some referred to scaphoid cast as thumb
spica, although some authorities consider these are not the
same. Moreover, some authors calculated the ROM and grip
strength as a percentage of the contralateral hand, others gave
an actual value. All these led to significant heterogeneity for
the compared outcomes.
Figure 1 e Flow chart of study selection.

Effects of non-operative interventions


Colles cast versus scaphoid cast

Five trials looked at different types of non-operative


treatments while the rest compared non-operative to operative treatment. There was not a randomised or quasirandomised controlled trial comparing different types of
operative treatment. However, the review identified a single
cadaveric biomechanical study comparing four different
screw fixations for scaphoid fractures.24 There was varying
degrees of methodological quality of the included papers.
Three trials were considered to have adequate concealment
prior to allocation while this was not clear in the other
trials. For obvious reasons, most trials have been open label
trials.
Most of the included trials did not have comparable groups
at entry points. Adolfsson et al.14 included undisplaced distal
oblique and waist fractures and excluded displaced and
longitudinal fractures. Bond and Vinnars excluded undisplaced fractures.17,23 Although, McQueen et al.21 did not
exclude the displaced fractures, there were only seven
undisplaced scaphoid fractures in her series. Dias and Saeden
included displaced fractures.11,19 Adolfsson, Arora, Bond and
McQueen performed their fixation percutaneously while Dias,
Saeden and Vinnars used an open technique.11,14,16,17,19,21,23
Some authors believe that open technique is superior as it
ensures adequate reduction of displaced fractures. However,
opponents claim that open techniques can risk vascular
supply to the scaphoid and cause AVN and delayed or nonunion. The length of follow up, the mode of diagnosing nonunion and types of secondary outcomes varied among
different trials.
Another source of bias was the way that original authors
dealt with protocol non-adherents such as missing data and
loss follow up. For example Adolfsson et al.14 encountered
a participant, who was randomized for operative treatment,
but his operation got complicated by the guide wire loosing

One trial only compared Colles cast to the scaphoid cast.18


The trial recruited 291 patients, 148 in the Colles cast group
and 143 in the scaphoid group. The main outcome was the
union rate. The union was diagnosed on clinical and radiological bases (plain X-ray only). There was no significant
difference between the two treatment groups in term of union
rate between the two groups (RD: 0%, 95% CI $0.07e0.07,
P 0.92).

Above elbow versus below elbow scaphoid cast


Two studies compared an above elbow to a below elbow
scaphoid cast.4,15 The studies provided sufficient data for
meta-analysis. There was no statistically significant difference in the union rate (P 0.48) or time to union (P 0.17)
(Fig. 2, Fig. 3). There was no significant difference in complications rate (P 0.53).

Colles cast with the wrist in flexion versus Colles cast with the
wrist in extension
One trial investigated the effect of the wrist position on
scaphoid fracture healing.20 Fifty eight participants with acute
fractures of the scaphoid were randomly allocated for
conservative treatment in Colles cast with the wrist immobilised in either 20# flexion or 20# extension. There was not
a significant difference in the union rate (RD: 0.04%, 95% CI
$0.07e0.15, P 0.47). However, there was a significant
reduction in the wrist extension in the flexed wrist group by
about 12# (95% CI 7.77# e16.23# , P 0.00001). Although there
were more complications and less grip strength in the flexed
wrist group, these did not reach statistical significance level
(P 0.30 and 0.07 respectively).

Scaphoid cast versus focused rigidity primary scaphoid cast


One trial compared the focused rigidity cast to other standard
casts in five common injuries.22 A sub-group of 35 patients

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

223

Figure 2 e Trials of above elbow vs. below elbow cast: Forest plot of union rate.

had scaphoid fractures. Sixteen had standard scaphoid cast


and 19 used the focused rigidity primary scaphoid cast
(FRPSC). There was 100% union rate in both groups at 5.2
weeks and 4.4 weeks respectively. Although the study
showed no significant difference in the main outcomes,
there is a weakness in this study, making its value very low.
It is a small study in which various type of injuries, including
soft tissue injuries, have been included. It was not clear
whether the participants did have confirmed scaphoid
fractures. This impression was strengthened by the short
time for healing. There was insufficient data to analyse
other outcomes.

The effect of adjunct ultrasound treatment with standard


scaphoid cast
A single study investigated the effects of low intensity, pulsed
ultrasound on acute undisplaced scaphoid fracture healing.9
Thirty participants (15 in each arm) with scaphoid fracture
treated with standard scaphoid cast but one group received
adjunct pulsed ultrasound treatment for 20 min a day. All
fractures united but the ultrasounded group had significantly
shorter time to heal by about 19 days, (95% CI: 7.75e29.85 days,
P 0.0009).

Effects of operative versus non-operative interventions


Seven trials compared operative versus non-operative treatments.11,14,16,17,19,21,23 Outcomes included union rate, time to
union, range of movement, grip strength, complications and
cost.
Pooled data showed a slightly higher overall union rate in
the operative group and this was statistically significant (RD
6%, 95% CI: 1%e11%, P 0.01) (Fig. 4). However, there was
significantly higher rate of complications in the operative
group (RD 10%, 95% CI: 5%e16%, P 0.0004) (Fig. 5). There is
a significant heterogeneity of I2 of 56% and 81% respectively.
Subgroup analysis showed that there was no significant
difference in union rate between operative and non-operative

treatment among the trials that included undisplaced


scaphoid fractures only, while there was a higher union rate in
the operative treatment in the trials that included displaced
scaphoid fractures (Figs. 6 and 7).
Open technique was superior to percutaneous technique in
achieving union in scaphoid fractures. Trials that compared
open technique to non-operative treatment had a higher
union rate (RD 9%, 95% CI: 3%e15%, P 0.005, I2 65%)
whereas trials compared percutaneous fixation to non-operative showed no significant difference between the two
intervention (RD 1%, 95% CI: $6%e8%, P 0.77, I2 14%)
(Fig. 8).
Two trials provided data about the cost effectiveness of
operative treatment versus non-operative treatment.16,23 As
expected the data was non-parametric and analysed in
a separate table (Table 3).
The ROM, grip strength and return to work data could
not be pooled safely because they have been reported in
various ways. We summarised the results in (Table 4).
Cautious analysis of the result shows that there is no
substantive difference in the ROM, but there is a consistent
trend that operation may improve grip strength and early
return to work. However, there are many confounding
factors for the latter and it should be interpreted with
caution.

Union rate versus time to union


An important clinical and research question is the average
length of time for a fracture scaphoid to unite and when
a delayed union becomes a non-union. All trials provided data
about time to union; we tabulated cumulative union rates
against time to union whenever a study provided useful data
to do so. For non-operative interventions 59.6% of fractures
united within 8 weeks, 90.1% united within 12 weeks and
92.4% of fractures united within 24 weeks. Immobilisation of
a fractured scaphoid beyond 12 weeks may help only 2.3% to
achieve union (Table 5).

Figure 3 e Trials of above elbow vs. below elbow cast: Forest plot of time to union.

224

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Figure 4 e Trials of operative vs. Non-operative intervention: Forest plot of union rate.

Discussion
This review specifically seeks to investigate the evidence for
the effectiveness and safety of various treatment options of
acute scaphoid fractures. The main questions we wanted to
address are:
1. Are there significant differences between different types of
non-operative treatments for acute fractures of the
scaphoid in terms of outcome and safety?
2. Are there significant differences between different types of
operative treatments for acute fractures of the scaphoid in
terms of outcome and safety?
3. Are there significant differences between operative
compared to non-operative treatments for the acute fractures of the carpal scaphoid in terms of outcome and
safety?
There were five non-operative treatment modalities that
have been investigated by RCTs: a below elbow scaphoid cast,
an above elbow scaphoid cast, Colles cast, a focused rigidity
primary scaphoid cast and ultrasound adjunct to a scaphoid
cast. There was no difference in the union rate.
Clay et al.18 investigated 291 patients with confirmed
scaphoid fractures, 148 in the Colles cast group and 143 in the
scaphoid cast group. The main outcome was the union rate.
The union was diagnosed on clinical and radiological bases
(plain x ray only). He showed that there was no significant
difference between the scaphoid cast and Colles cast in
achieving scaphoid union. However, the diagnosis of nonunion was not robust, partly because authors relied on plain
x-ray and clinical examination only. Eighteen patients in the
scaphoid group and 19 patients in the Colles cast group were
labelled as having probable union. Although the authors
stated that these patients remained asymptomatic at
one year, we do not know what happened to them after this

period e they may well have asymptomatic non-union. It


would have been extremely valuable if these patients had had
a CT scan to confirm union. The study showed there was
a high non-union rate in the proximal pole fractures (31%).
There were more proximal pole fractures in the scaphoid cast
group (9 versus 2). The authors rightly concluded that there
were too few proximal pole fractures to investigate them
further. Pooled data from other trials supported Clays
conclusion that the Colles cast achieved a comparable union
rate to that of the scaphoid cast. Our personal view is that
Colles cast is adequate when it is not made of plaster of Paris.
The latter is weak and gets broken when patient do certain
tasks such as driving.
Pooled individual participants data from other studies
confirmed Clays conclusion that Colles cast is adequate for
treating scaphoid fracture.
A Colles cast with wrist in flexion had no advantage over
a Colles cast with wrist in extension in achieving union and it
was associated with significantly more problems. There was
decreased in wrist ROM and grip strength in the participants
that were treated with Colles cast with wrist in flexion. Hence
it is not recommended.
Several authors advocated the use of the above elbow cast
to avoid shearing effects of pronation and supination on the
scaphoid fractures. Two trials investigated this issue (Alho
1975; Gellman 1989) with 151 participants.4,15 There was
significant heterogeneity between the two trials as evident by
high I2 (70%) and therefore the result should be interpreted
with caution. The two studies failed to show that above elbow
cast produces a higher union rate. Although, there was some
residual elbow stiffness this was resolved at subsequent
follow up (Figs. 2 and 3). Gellman et al.4 also concluded that an
initial period of immobilisation of six weeks in an above elbow
cast improved the time to union (9.5 week versus 12.7). We
express our reservation on this conclusion for the following
reasons: His trial was underpowered to support this conclusion, they excluded the 6 delayed unions from the short

Figure 5 e Trials of operative vs. non-operative intervention: Forest plot of complications rate.

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

225

Figure 6 e Subgroup analysis of union rate by fracture displacement: Forest plot.

thumb spica group but they did not do the same with the other
arm of the trial. Moreover, time to union as an outcome is
closely related to the frequency of assessment for the union.
They evaluated the union every 3e4 weeks which casts doubt
about the accuracy of the time to union they showed.
Low intensity, pulsed ultrasound may speed scaphoid
fracture healing (43.2 % 10.9 versus 62 % 19.2 days). This has
been promoted by a single small trial published in a nonorthopaedic journal in German.9 It may have a role in
promoting healing but this needs further investigation by
a larger RCT.
We were able to find seven completed trials (another three
ongoing/unclassified trials) that compared operative versus
non-operative treatment. The pooled data from 403 favours
operative treatment over the non-operative treatment in
terms of union rate. The overall risk difference is 6% only (95%
CI 1%e11%, P 0.01). There is a high heterogeneity as shown
by a high I2 value of 56%. In addition to the play of chance,
heterogeneity may arise between trials because of variations
in clinical practice or trial methodology. Heterogeneity was

explored further by sub-group analysis which uncovered two


interesting findings. Operative intervention did not increase
union rate in undisplaced fractures, however, in the trials that
recruited patients with displaced and undisplaced fractures,
operative intervention provided a higher union rate. Unfortunately authors of these trials did not provide sub-group
analysis of their series to confirm that this extra union rate
was due to a higher union rate among the displaced fractures.
The funnel plot showed a similar finding. Although, the
studies are distributed symmetrically around the mean risk
difference, careful scrutiny shows the trials that did not
exclude the displaced fractures are located on the right side
with (McQueen 2008, Dias 2005) dominated the effect.19,21
Contrary to the expectation, open technique was superior
to percutaneous technique in achieving union in scaphoid
fractures. Percutaneous fixation is thought to be associated
with less soft tissue damage, hence no interference with blood
supply and bone healing. But our study did not support this
assumption. There is a higher union rate in the trials that used
open technique than the trials that used percutaneous

Figure 7 e Subgroup analysis of union rate by fracture displacement: Funnel plot.

226

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Figure 8 e Subgroup analysis of union rate by surgical technique: Forest plot.

technique (Fig. 7). Our finding might be explained by the fact


that open fixation may achieve a better reduction and stimulate increase blood supply to the area. Moreover, the percutaneous fixation is performed blindly and may be associated
with more damage to scaphoid bone blood supply.
Operative intervention was associated with a significantly
higher complications rate in the operative group (Fig. 6). There
is a significant heterogeneity (I2 81%) in reporting these
complications. McQueen et al.,21 contrary to all other trials,
showed there were more complications in the non-operative
groups. She reported three malunions and DISI (Dorsal Intercalated Segment Instability). This has not been reported by
others. This may be explained by the fact that other authors
may not have looked specifically for these complications or
they were not clinically important. The nature of these
complications favours the first assumption. The clinical
effects of such complications tend to be delayed for several
years which is far longer than most trials follow up duration.
Saeden and Vinnars reported a large number of radiologically
diagnosed osteoarthritis in the operative group.11,23 It is well
known that radiological features of osteoarthritis do not
correlate well with symptoms and these high rates may be
irrelevant, but this is remains to be proven. Dias 2005 reported
many soft tissue problems (sensitive and hypertrophic scars)
that several surgeons may regard them as within normal
healing process.19
Dias recommended an aggressive conservative treatment
which involves treating these types of fractures nonoperatively, and early identification of fractures that show
signs of non-union and treating them operatively at around 6e8
weeks.19 Pooled data from our review (Table 5) showed that

around 60% of the non-operatively treated scaphoid fractures


had healed by 8 weeks, but this increased to 90% at 12 weeks.
Decision making at 12 weeks rather than 8 weeks may save
about 30% of unnecessary operations. Appropriate radiological
tests such as CT-scan or MRI scan may be necessary.
Other outcomes (ROM, grip strength and time to return to
work) are summarised in (Table 4). Cautious analysis of the
table showed there is no substantive difference in the ROM
between the operative and the non-operative groups. In fact,
some of the trials showed a better ROM in the non-operative
group.11,16,23 There is a consistent trend that surgery
improved grip strength and time to return to work. However,
we think it is unsafe to pool the data to produce an overall
effect and provide a clinical recommendation. Although,
return to work is a very important outcome, it is impossible to
standardise it. Patients may have different types of jobs,
usually need different grip strength to do their jobs, different
places provide different support and flexibility for their
workers, health and safety issues, availability of transport to
work and hand dominance plays a role in the timing of return
to work. All these confounding factors that none of the trials
we studied adjusted for, made it impossible to prove that
operative intervention speeds return to work. In a recent
metanalysis,25 authors tried to pool these outcomes to
produce a useful recommendation; however, we have a strong
reservation on doing so for the reasons mentioned above and
the fact that visual presentation of these outcomes in a single
table give a better and safer account of these outcomes. Our
views have been shared by Yins systemic review.26
There is insufficient data to produce any useful
recommendation on cost effectiveness of surgery. Two trials

Table 3 e Summary findings of cost analysis in trials reported on cost.


Study ID

Intervention

Cost median (range)

Notes

Arora 2007

Non-operative
Operative
Non-operative
Operative

2363.1 (%1394.7) Euros


2097.1(/325) Euros
2507 Euros (810e3739)
3155 Euros (2018e4663)

Non-operative treatment cost is relatively


similar in both trials. But the operative
cost is surprisingly low in Aroras study.

Vinnars 2007

227

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

Table 4 e Summary findings of range of movement, grip strength and return to work.
Study ID

Intervention

Adolfsson 2001

Non-operative
Operative
Non-operative

87% (NS) *
94% (NS)
140# *

83% (NS)
88%(NS)
42 kg % 10.2 kg

Unclear
6 weeks
8 % 7 weeks

Operative
Non-operative
Operative
Non-operative
Operative
Non-operative
Operative

132# *
124# % 7#
139# % 6#
93.4% (NS)*
94.1% (NS)*
94% (NS) *
98% (NS) *

46 kg % 10.7 kg
36 kg % 3 kg
40 kg % 2.5 kg
91.6% (NS)*
99.1% (NS)*
95% (NS) *
102% (NS) **

1 % 1 week
15 % 0.7 weeks
8 % 0.7 weeks
6 weeks
5 weeks
11.4 weeks
3.8 weeks

Non-operative
Operative
Non-operative

98% (NS)
96% (NS)
29/35 had
normal ROM
29/40 had
normal ROM

88% (NS)
95% (NS)
27/35 had
normal strength
30/40 had
normal strength

15 % 10 weeks
6 % 3 weeks
10.5 weeks

Arora 2007

Bond 2001
Dias 2005
McQueen 2008

Saeden 2001
Vinnars 2007

Operative

Range of movement

provided data about the cost effectiveness of operative


treatment versus non-operative treatment (Arora 2007 from
Austria and Vinnars 2007 from Sweden) (Table 3).16,23
Non-operative treatment cost is relatively similar in both
trials (2363 Euros and 2507 Euros respectively), but the
operative cost is surprisingly low in Arora et al.16 (2097 Euros
and 3155 Euros). Careful analysis of the cost breakdown

Grip strength

Return to work

Notes
* NS normal side
* Flexion Extension
arc at 3 months.

* At one year
* At one year
** May reflect hand
dominance and
researchers adjustment.

5.5 weeks

showed the main difference was due to time off work


compensation. Participants in the operative group were off
work for 7.5 % 8.1 days versus 54.6 % 49.6 days in the
non-operative group. The estimated compensation was
200 % 217 Euros versus 1453 % 1321 Euros respectively. We
think it unrealistic to expect patients to go back to work
within a week from their surgery. Also, Arora et al.16 excluded

Table 5 e Cumulative union rate per unit time (weeks) of treatment.


Study ID
Adolfsson 2001
Alho 1975
Arora 2007
Bond 2001
Clay 1991
Dias 2005
Gellman 1989
Hambidge 1999
Mayr 1999
McQueen 2008
Petty 1999
Saeden 2001
Vinnars 2007
Cumulative union rate
Cumulative union

Intervention
Scaphoid cast
Operative
BE Scaphoid cast
AE Scaphoid cast
Scaphoid cast
Operative
AE Scaphoid cast
Operative
Scaphoid cast
Colles cast
Colles cast
Operative
BE Scaphoid cast
AE Scaphoid cast
Flexed Wrist in Colles cast
Extended wrist in Colles cast
Scaphoid cast no US
Scaphoid cast with US
Colles cast
Operative
Scaphoid cast
FRPSC
Scaphoid cast
Operative
Scaphoid cast
Operative
Non-operative
All types

<6 W

8W

10 W

12 W

22/26
20/23

0/23
2/21

16 W

18 W

22 W

24/26
22/23

26/26
22/23

51/53
41/47
19/23

0/23
20/21

>24 W

23/23
20/21

11/11
14/14
129/143
133/148
35/44
43/44
21/23

34/44
43/44
17/23
26/28

28/28
53/58
55/63
15/15

15/15
26/30
29/30
15/15
19/19

59.6%
77.5%

34/35
40/40
90.1%
91.9%

28/30
31/32
35/35
35/35
92.4%
93.4%

228

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

self-employed and unemployed patients which may have an


impact on cost effectiveness analysis. We believe that
operations to fix scaphoid fracture cost more, but
unfortunately the data was not strong enough to support our
belief.
There was no RCT that compared different operative
modalities head to head. The data was not sufficient to
perform among studies sub-group analysis. One cadaveric
study Toby et al.24 compared a Herbert screw, an AO 3.5-mm
cannulated screw, a Herbert-Whipple screw, an Acutrak
cannulated screw, and a Universal Compression screw using
cyclical bending loads. The study demonstrated that
commonly used screws for fixation of the scaphoid vary
significantly in their ability to resist cyclical bending loads.
This study was excluded as it did not meet our inclusion
criteria.

Authors conclusions
Scaphoid fracture can be treated by a Colles cast for up to 12
weeks. The wrist should not be in flexion. A scaphoid cast is
as good as a Colles cast but is less convenient for the patient
and the staff. This should achieve a union rate of 90%. There
is no advantage of an above elbow cast over a below elbow
cast.
There is no evidence that operative treatment for undisplaced scaphoid fracture provide a higher union rate and it is
associated with a higher complication rate. However, there is
indirect evidence that it may provide a higher union rate in
displaced fractures.
We recommend a non aggressive approach when all
undisplaced scaphoid fractures are treated non operatively for
12 weeks and to consider surgery at 12 weeks if there is no
signs of union.
The exception for this recommendation is proximal
scaphoid fracture which is rare and associated with a high
non-union rate. The evidence is not sufficient to advocate for
or against surgery. The high non-union rate of proximal pole
fractures may justify the risk of surgical intervention.
However, the last statement needs further exploration by
a well designed research studies.
Operative treatment of scaphoid fracture is growing in
orthopaedic practice and there may be a role for such growth.
However, the correct indications should be identified further
by targeted research. A possible example is to conduct a large
retrospective analysis of non united scaphoid fractures to
identify subgroups with poor prognostic factors for non-union
such as displacement or site of the fracture, then to set
a properly designed RCT to investigate whether early surgery
offer a higher union rate in these sub-group of patients.
Moreover, the best operative approach needs to be explored
further comparing percutaneous approach versus open
fixation.

Appendix A. Supplementary data


Supplementary data related to this article can be found online
at doi:10.1016/j.surge.2012.03.004.

references

1. Brinker M. Review of orthopaedic trauma. London: WB Saunders;


2001.
2. Miric D, Vucetic C, Senohradski K, Micunovic L. Incidence and
severity of degenerative changes in the wrist in
pseudoarthrosis of the scaphoid bone. Srp Arh Celok Lek 2001
MareApr;129(3e4):61e5.
3. Ruby LK, Stinson J, Belsky MR. The natural history of scaphoid
non-union. A review of fifty-five cases. J Bone Jt Surg Am 1985
Mar;67(3):428e32.
4. Gellman H, Caputo RJ, Carter V, Aboulafia A, McKay M.
Comparison of short and long thumb-spica casts for nondisplaced fractures of the carpal scaphoid. J Bone Jt Surg Am
1989 Mar;71(3):354e7.
5. Broome A, Cedell CA, Colleen S. High plaster immobilisation
for fracture of the carpal scaphoid bone. Acta Chir Scand 1964
Aug;128:42e4.
6. Verdan C. Fractures of the scaphoid. Surg Clin North Am 1960
Apr;40:461e4.
7. Petracic B. Immobilization of the scaphoid bone (alterative
between upper and forearm plaster). Handchir Mikrochir Plast
Chir 1985 Jan;17(1):53e4.
8. Herbert TJ, Fisher WE. Management of the fractured scaphoid
using a new bone screw. J Bone Jt Surg Br 1984
Jan;66(1):114e23.
9. Mayr E, Rudzki MM, Rudzki M, Borchardt B, Hausser H,
Ruter A. Does low intensity, pulsed ultrasound speed healing
of scaphoid fractures? Handchir Mikrochir Plast Chir 2000
Mar;32(2):115e22.
10. Divelbiss BJ, Adams BD. Electrical and ultrasound stimulation
for scaphoid fractures. Hand Clin 2001 Nov;17(4):697e701
[xexi].
11. Saeden B, Tornkvist H, Ponzer S, Hoglund M. Fracture of the
carpal scaphoid. A prospective, randomised 12-year followup comparing operative and conservative treatment. J Bone Jt
Surg Br 2001 Mar;83(2):230e4.
12. Higgins JPT, Green S. Cochrane handbook for systematic reviews
of interventions 4.2.5 [updated May 2005] 2005.
13. Madhok R, Shaw LJ, Elliott J, Gillespie LD. Bone, Joint and
Muscle Trauma Group (formerly the Musculoskeletal Injuries
Group). About the Cochrane Collaboration (Collaborative
Review Groups (CRGs)) 2007, Issue 1. Cochrane Database Syst
Rev 2007.
14. Adolfsson L, Lindau T, Arner M. Acutrak screw fixation versus
cast immobilisation for undisplaced scaphoid waist fractures.
J Hand Surg [Br] 2001 Jun;26(3):192e5.
15. Alho A, Kankaanpaa. Management of fractured scaphoid
bone. A prospective study of 100 fractures. Acta Orthop Scand
1975 Nov;46(5):737e43.
16. Arora R, Gschwentner M, Krappinger D, Lutz M, Blauth M,
Gabl M. Fixation of nondisplaced scaphoid fractures: making
treatment cost effective. Prospective controlled trial. Arch
Orthop Trauma Surg 2007 Jan;127(1):39e46.
17. Bond CD, Shin AY, McBride MT, Dao KD. Percutaneous screw
fixation or cast immobilization for nondisplaced scaphoid
fractures. J Bone Jt Surg Am 2001 Apr;83-A(4):483e8.
18. Clay NR, Dias JJ, Costigan PS, Gregg PJ, Barton NJ. Need the
thumb be immobilised in scaphoid fractures? A randomised
prospective trial. J Bone Jt Surg Br 1991 Sep;73(5):828e32.
19. Dias JJ, Wildin CJ, Bhowal B, Thompson JR. Should acute
scaphoid fractures be fixed? A randomized controlled trial. J
Bone Jt Surg Am 2005 Oct;87(10):2160e8.
20. Hambidge JE, Desai VV, Schranz PJ, Compson JP, Davis TR,
Barton NJ. Acute fractures of the scaphoid. Treatment by cast
immobilisation with the wrist in flexion or extension? J Bone Jt
Surg Br 1999 Jan;81(1):91e2.

t h e s u r g e o n 1 0 ( 2 0 1 2 ) 2 1 8 e2 2 9

21. McQueen MM, Gelbke MK, Wakefield A, Will EM, Gaebler C.


Percutaneous screw fixation versus conservative treatment
for fractures of the waist of the scaphoid: a prospective
randomised study. J Bone Jt Surg Br 2008 Jan;90(1):66e71.
22. Petty ACWC. A randomized, controlled comparison of
adjustable focused rigidity primary casting technique with
standard plaster of Paris/synthetic casting technique in the
management of fractures and other injuries. J Orthopaedic
Nurs 1999;2(2):95e102.
23. Vinnars B, Ekenstam FA, Gerdin B. Comparison of direct and
indirect costs of internal fixation and cast treatment in acute
scaphoid fractures: a randomized trial involving 52 patients.
Acta Orthop 2007 Oct;78(5):672e9.

229

24. Toby EB, Butler TE, McCormack TJ, Jayaraman G. A


comparison of fixation screws for the scaphoid during
application of cyclical bending loads. J Bone Jt Surg Am 1997
Aug;79(8):1190e7.
25. Buijze GA, Doornberg JN, Ham JS, Ring D, Bhandari M,
Poolman RW. Surgical compared with conservative treatment
for acute nondisplaced or minimally displaced scaphoid
fractures: a systematic review and meta-analysis of
randomized controlled trials. J Bone Jt Surg Am 2010
Jun;92(6):1534e44.
26. Yin ZG, Zhang JB, Kan SL, Wang P. Treatment of acute
scaphoid fractures: systematic review and meta-analysis. Clin
Orthop Relat Res 2007 Jul;460:142e51.