Rehabilitation Regimens Following Surgical Repair of Extensor Tendon Injuries of the Hand—A Systematic Review of Controlled Trials

Abstract

There is no consensus on the most effective rehabilitation regimen following extensor tendon repair of the hand. This systematic review evaluates the outcomes of the various regimens. The Cochrane, MEDLINE, EMBASE, CINAHL, AMED, PEDro, OTseeker databases were searched for any prospective randomised clinical trials comparing rehabilitation regimens for acute extensor tendon injuries in adults. Five papers met the inclusion criteria. The regimens were static immobilisation, dynamic splinting and early active motion (EAM). There was no standard format of reporting. The sample size ranged from 27 to 100 patients. The duration of follow-up ranged from 8 to 24 weeks. Overall, patients’ total active motion improved with time. Early mobilisation regimens (active and passive) achieve quicker recovery of motion than static immobilisation but the long-term outcome appears similar. Given the comparable outcomes between dynamic splinting and EAM, we therefore favour EAM which is simpler and more convenient.

Full text

ORIGINAL ARTICLE
Rehabilitation Regimens Following Surgical Repair
of Extensor Tendon Injuries of the Hand—A Systematic
Review of Controlled Trials
Chye Yew Ng &Joelle Chalmer &
Duncan J. M. Macdonald &Saurabh S. Mehta &
David Nuttall &Adam C. Watts
Received: 5 June 2012 /Accepted: 21 August 2012 / Published online: 15 September 2012
#Society of the Hand & Microsurgeons of India 2012
Abstract There is no consensus on the most effective reha-
bilitation regimen following extensor tendon repair of the
hand. This systematic review evaluates the outcomes of the
various regimens. The Cochrane, MEDLINE, EMBASE,
CINAHL, AMED, PEDro, OTseeker databases were searched
for any prospective randomised clinical trials comparing re-
habilitation regimens for acute extensor tendon injuries in
adults. Five papers met the inclusion criteria. The regimens
were static immobilisation, dynamic splinting and early active
motion (EAM). There was no standard format of reporting.
The sample size ranged from 27 to 100 patients. The duration
of follow-up ranged from 8 to 24 weeks. Overall, patients’
total active motion improved with time. Early mobilisation
regimens (active and passive) achieve quicker recovery of
motion than static immobilisation but the long-term outcome
appears similar. Given the comparable outcomes between
dynamic splinting and EAM, we therefore favour EAM which
is simpler and more convenient.
Keywords Extensor tendon .Immobilisation .Early active
motion .Early passive motion .Dynamic splinting .
Rehabilitation
Introduction
Extensor mechanisms of the hand and wrist have a complex
and intricate anatomy [1,2]. The versatility and delicate
balance of finger motion can be easily impaired by injuries
to the extensor tendons [3]. Following surgical repair of the
extensor tendons, institution of a specific rehabilitation reg-
imen involves consideration of the severity of injury, quality
of the repair, complexity of the regimen and likely compli-
ance of the patient with therapy. The rehabilitation regimens
can be broadly divided into static immobilisation or early
mobilisation [4,5]. Historically immobilisation has been the
mainstay of therapy following extensor tendon injuries as
the repair site is protected from excessive motion which
could threaten the repair. Greater understanding of tendon
biology [6], advancements in surgical technique and evi-
dence for early motion following flexor tendon repair [7,8]
have since prompted interest in employing the principle of
early motion when treating extensor tendon injuries.
Early mobilisation regimens can be further divided into
passive (dynamic) or active mobilisation [9,10]. Passive
mobilisation regimens employ dynamic splinting which
allows glide of the repaired tendon, and in concept all work
of digital extension is done by the elastic mechanism of the
splint. A second static splint is worn outside of exercise
periods. This form of management has been associated with
Electronic supplementary material The online version of this article
(doi:10.1007/s12593-012-0075-x) contains supplementary material,
which is available to authorized users.
C. Y. Ng (*):D. Nuttall :A. C. Watts
Upper Limb Unit, Wrightington Hospital,
Hall Lane, Appley Bridge, Wigan,
Lancashire WN6 9EP, UK
e-mail: chyeng@gmail.com
J. Chalmer
Therapies Department, St George’s Hospital,
London, UK
D. J. M. Macdonald
Department of Trauma & Orthopaedics, Royal Alexandra
Hospital,
Paisley, UK
S. S. Mehta
Department of Trauma & Orthopaedics, Leicester Royal Infirmary,
Leicester, UK
J Hand Microsurg (July–December 2012) 4(2):65–73
DOI 10.1007/s12593-012-0075-x

complications of tendon tethering and discomfort in wearing
the bulky dynamic splint, thus posing questions about pa-
tient compliance [11]. In contrast, early active motion regi-
mens allow concentric contraction of the extensor muscle of
the repaired tendon and a static splint is worn between
exercises to protect the repaired tendon. Both early mobi-
lisation regimens aim to promote earlier restoration of range
of motion and potentially earlier return to work. Neverthe-
less, surgeons and therapists face the difficulty of finding the
balance between protecting the repair site by limiting pre-
mature mobilisation and preventing tendon adhesions due to
prolonged immobilisation [12,13].
The aim of this study is to perform a systematic review of
prospective randomised controlled trials (RCT) available on
the outcomes and complications of various rehabilitation reg-
imens following surgical repair of extensor tendon injuries of
the hand.
Methods
Search Strategy
A literature search was performed on the following electronic
databases: the Cochrane Bone, Joint and Muscle Trauma
Group Specialised Register (accessed 21/5/2011), the
Cochrane Central Register of Controlled Trials (in the
Cochrane Library Issue 2 of 4, Apr 2011), Ovid MEDLINE
(R) (1948 to May Week 2 2011), EMBASE (1902 to May
2011), CINAHL (accessed 22 May 2011), AMED (Allied and
Complementary Medicine) (1985 to May 2011), PEDro -
physiotherapy evidence database (http://www.pedro.org.au)
(accessed 21 May 2011), and OTseeker - The Occupational
Therapy Systematic Evaluation of Evidence Database (http://
www.otseeker.com) (accessed 30 May 2011). The search term
’extensor tendon’was used for both PEDro and OTseeker. In
MEDLINE (OVID WEB) the subject specific search strategy
was combined with all three stages of the optimal trial search
strategy [14]. The full search strategies for MEDLINE are
shown in Appendix 1, the Cochrane Library (Wiley Inter-
Science) in Appendix 2, EMBASE in Appendix 3, CINAHL
in Appendix 4and AMED in Appendix 5.
We searched Current Controlled Trials at www.
controlled-trials.com (accessed 22 May 2011) and the UK
Clinical Research Network Study Portfolio at http://
public.ukcrn.org.uk/search (accessed31May2011)for
ongoing and recently completed trials. We also searched sup-
plements of Journal of Hand Surgery (European) (1990 to
May 2011: http://jhs.sagepub.com), proceedings of Journal
of Bone and Joint Surgery British (2002 to May 2011: http://
www.jbjs.org.uk), abstracts of Orthopaedic Trauma Associa-
tion annual meetings (1996 to 2010: http://www.hwbf.org/ota/
am/), and abstracts of the American Society for Surgery of the
Hand annual meetings (2000 to 2010: http://www.assh.org/
AnnualMeeting/archive/Pages/AbstractBook.aspx). In addi-
tion, the online archives of the following journals were
searched using the term ‘extensor tendon’: Journal of Amer-
ican Academy of Orthopaedic Surgeons (October 1993 to
May 2011), Journal of Plastic, Reconstructive and Aesthetic
Surgery (2005 to May 2011), British Journal of Plastic Sur-
gery (1948 to 2005) and Journal of Hand Therapy (2002 to
April 2011). The bibliographies of all identified studies were
screened for additional relevant studies.
Inclusion and Exclusion Criteria
Twenty-two potentially eligible citations were identified via
the initial search. Full reports of the citations were then
obtained and each was independently assessed (by CN, JC,
DM, SM, AW) for final inclusion. We included any prospec-
tive, randomised or quasi-randomised (for example, allocation
by date of birth or alternation) controlled clinical trials which
compared different rehabilitation regimens for acute extensor
tendon injuries of the hand. Zone I injury (Mallet finger) was
excluded as this had been examined in a Cochrane review
[15]. Thumb was not considered in this review due to its
unique anatomy, compared to the other digits.
The methodological quality of the included trials was
then independently assessed using a rating scheme covering
11 aspects of trial validity (Appendix 6). Each trial was also
assigned a category of effectiveness (Appendix 7). Any
disagreement was resolved by discussion and if no consen-
sus was met, the senior author decided.
Data Extraction and Analysis
Using a data extraction form, the following information was
obtained from the included studies: author, year, Verdan zone
of injuries, outcome measures, method of measurement, du-
ration of follow-up, interventions, number of subjects/hands/
digits and recommendations. Extraction of results from graphs
in trial reports was considered where data were not provided in
the text or tables. Results were collected for the follow-up
times for which these were available.
Meta-analysis was performed using RevMan analysis
software (RevMan 5.1.6) of the Cochrane Collaboration.
Results
Five studies met the inclusion criteria [9,10,16–18]. The
characteristics of the studies have been summarised in Table 1.
The rehabilitation regimens studied were grouped into (1) static
immobilisation; (2) dynamic splinting; and (3) early active
mobilisation (EAM) (Table 2). Bulstrode et al. also studied a
regimen involving immobilisation of the MCPJs but leaving the
66 J Hand Microsurg (July–December 2012) 4(2):65–73

Table 1 Summary of studies included in the systematic review
Papers Zone
of
Injury
Outcome How
measured
(incl timing,
assessor
blinding
Notes (eg.
Level of
significant,
subgroup)
Group A
intervention
N Group B intervention N Group C intervention N Categories
of
effectiveness
Recommendations
Bulstrode
2005
V–VI 1. TAM
[19]
1. Goniometry
at 4, 6,8,12 /
52
Mean and
95 %
confidence
intervals
of TAM
(1) and
grip
strength
(kg)
0 - 4 weeks -
immobilisation with
palmar slab - wrist 30º
ext/MCPjointext/IP
joints ext; week 4 - active
mobilisation all finger
joints hourly, passive
extension of fingers,
splint worn for protection
at night and in danger;
week8-passiveand
resisted flexion begun,
driving allowed; 12
weeks - return to
heavy work.
17 (17 injuries)
(10
completed all
appointments,
4 telephoned,
3lostto
follow-up)
0 - 4 weeks -
immobilisation
with static splint -
wrist 30º ext /
MCP joint ext /
IP joints free. IP
joint ext / flex
hourly; week 4 -
as per Group A
10 pts (13
injuries)
(3 lost
to
follow-
up)
0 - 4 weeks -
static splint -
wrist 45º ext /
MCP joint 50º
flex / IP joints
neutral. Exercises -
4 hourly, active
MCP joint ext /
flex, intrinsic
minus; week
4 - as per
Group A
15 pts
(16
injuries)
2
lost to
follow-
up
2 Greater total active
range at 4 weeks
and greater grip
strength at
12 weeks
in Groups B & C
after simple zone
V–VI extensor
injuries. No
difference in time
spent in therapy
between the
groups.
2. Grip
strength
2. Jamar
dynamometry
12/52
Single assessor,
method of
blinding not
documented
Chester
2002
IV –VIII 1. TAM Goniometry at
4/25, final
follow-up
median 12/
52
p value not
specified
Static splint - wrist 30º ext /
MCPJ 30º flex / IPJ full ext,
between exercises.
Exercises: day 1, hourly
x 5 - active IP joint /
MCP joint ext, active
intrinsic minus; day
5–7 - hourly x 10 - active
MCJP E/F (IPJ extend
ed), active intrinsic
minus; week 2 - active wrist
ext / flex; week 3 -
active gross composite
fist; week 4–6 - discard
splint except nocte,
commence light activities of
daily living; week
6–8-begingripstrength
exercises, return
to work; week 12 - no
further restrictions
30 (19 pts, 29
digits) 11 pts
lost to follow-
up
Day 5–7 - dynamic
splint - wrist 30º ext /
MCP joints neutral
daytime wear; static
splint nocte - wrist 30º
ext / MCP joint 30º
flex / IP joints
neutral; x10 hourly
MCPJ active
flex/passive ext; active
ext/flexof IPJ; week
2 - as for Group A
24 (17 pts,
29
injuries)
7 lost to
follow-
up)
2 Greater range at
4/52 in Group B
however no
significant
difference
between early
active and early
passive motion
regimens at
final follow-up in
simple zone V –
VIII extensor
tendon injuries.
No blinding of
assessors
Hall 2010 V-VI 1.TAM 1. Goniometry
3, 6, 12/52
α00.017
[Estimated
Sd 150 of TAM ]
Bonferroni
adjustment
0 - 3 weeks -
immobilisation with
static splint - wrist 40-
45º ext / MCP joint
0-20º flex / IP joints
neutral; week 3 - graded
mobilisation
9 pts (4 pts
completed, 5
injuries), 5
pts lost to
follow-up at
6 weeks
Day 1–5 - dynamic splint,
wrist40-45ºext/MCP
joint 0º / palmar block
at 30-40º flex; hourly
x 20 - active MCP joint
flex, passive ext with IP
joint extended; passive
supervised wrist
9 pts (5 pts
com
pleted,
8 inju
ries), 4
pts lost
to
follow-
Day 1–5 - static splint,
wrist30ºext/MCP
joint 45º flex / IP
joint free; hourly x
10 active MCP
joint ext / flex with IP
joint neutral, gross
composite flexion in
9 pts ( 11
injuries)
4 Early active motion
patients achieved
greater active
range of motion,
less extension lag
and better self-
report functional
scores than
J Hand Microsurg (July–December 2012) 4(2):65–73 67

Table 1 (continued)
Papers Zone
of
Injury
Outcome How
measured
(incl timing,
assessor
blinding
Notes (eg.
Level of
significant,
subgroup)
Group A
intervention
N Group B intervention N Group C intervention N Categories
of
effectiveness
Recommendations
programme commenced
- not described
tenodesis & IP joint
motion; week 3 - palmar
block removed and full
active flex allowed;
week - 5 splint ceased;
week 6 - as per Group A
up at
6 weeks
limits of splint; week
3 - splint adjusted to
MCP joint 70º flex,
active hook fist
(intrinsic minus);
week 5 - splint
ceased, graded
mobilisation
programme; week
6 - as per Group A
early passive
motion and
immobilisation
protocols.
2. Function 2. Self-report
visual analogue
scale 3, 6, 12/52
3. Extension
lag
3. Goniometry
3,6,12/52
4. Grip
strength
4. Jamar
dymaometry
2/52
No blinding
of assessors
Khandwala
2000
V-VI 1. TAM Goniometry 4,
8/52
p value not
specified
Day 0–2 weeks - Dynamic
extension splint, with
repaired tendon finger
plus one on either side
involved; hourly x 10 -
active MCP joint flex
(IP joint ext), passive
MCP joint ext, free
IP joint movement, no
composite movement
allowed; week 3 -
splint continued, start
composite flex,
continued passive ext;
week 4 - full movement
allowed out of splint
including light
activity; week 6 - passive
flex and ultrasound
commenced as re
quired; week 8 - return to
driving; week 12 - return
to heavy manual work
50 pts, 78 fingers Day 0 - week 2 - all digits
included in static
splint, wrist 30º ext /
MCP joint 45º flex /
IP joints free; hourly
x10-activeextto
neutral, flex to splint;
week 3 - wrist 30º ext /
MCPJ 70º flex - hourly
x10-activeextto
neutral, flex to
splint, intrinsic minus;
week 4 - out of splint,
active ext/flex;
continue as per group
A intervention
50 pts, 84
fingers
2 Both rehabilitation
regimens were
highly effective,
without significant
statistical
difference
between them.
2.Miller's
assessment
of
extensor
tendon
repairs
(1942)
No blinding of
assessors
Mowlavi
2005
V-VI 1. TAM 1. goniometry
4, 6, 8/52
and 6/12
p value not
specified
Day 3–5 - dynamic splint, wrist
30ºext,upto30ºMCPjoint
flex; active flex / passive ext,
17 0 –4 weeks immobilized
with static splint, wrist
30º ext / MCP joint 15-
17 2 Dynamic splinting
of simple,
complete
68 J Hand Microsurg (July–December 2012) 4(2):65–73

IPJs free to mobilise [16]. This regimen was not studied by
other investigators and hence could not be justifiably grouped
with other studies for comparison. The results of the regimen
were summarised (Table 1) but not used in further analysis.
Total Active Motion
The most consistently reported outcome measure is the total
active motion (TAM) which is defined as (MCP+ PIP+DIP)
flexion –(MCP+PIP+DIP) extension lag [19]. Three studies
[16–18] reported TAM as a continuous variable while the
other two [9,10] reported TAM as categories (excellent, good,
fair and poor).
Overall TAM improved with time following extensor
tendon repairs (Fig. 1).
There appeared to be a general trend of better TAM with
regimens involving early mobilisation (dynamic splinting
and EAM) than that of static immobilisation, up to 12 weeks
following surgery. Beyond 12 weeks, the difference
appeared to diminish between the two groups but this ob-
servation was limited by the paucity of data points.
Static Immobilisation Versus Dynamic Splinting
Results (means and standard deviations of TAM reported)
from Hall et al. and Mowlawi et al. studies could be combined
for comparison between static immobilisation and dynamic
splinting (Fig. 2a and b). At 6 weeks, patients who were
rehabilitated using dynamic splinting regained significantly
greater TAM than those with static immobilisation (p00.01
with a mean difference of 30 °). When the results at 8–
12 weeks were combined, dynamic splinting group continued
to achieve significantly greater TAM than static immobilisa-
tion group (p00.004 with a mean difference of 23 °).
Table 2 This table shows the overview of rehabilitation regimens
studied by the various investigators. Blank cells represent regimens
that were not included in that particular trial. The categorisation was
based on the principle of the rehabilitation employed
Regimens Static
immobilisation
Static
immobilisation
(IPJ free)
Early
passive
motion
(dynamic
splinting)
Early
active
motion
(static
splinting)
Mowlawi
et al.
XX
Hall et al. X X X
Bulstrode
et al.
XX X
Khandwala
et al.
XX
Chester et
al.
XX
IPJ Interphalangeal joint
Table 1 (continued)
Papers Zone
of
Injury
Outcome How
measured
(incl timing,
assessor
blinding
Notes (eg.
Level of
significant,
subgroup)
Group A
intervention
N Group B intervention N Group C intervention N Categories
of
effectiveness
Recommendations
free IP joint movement; week 2
- active MCP joint flex to 45º;
week4-startedactiverange,
splint nocte only; week 6 -
splint ceased and passive range
commenced as needed
20º flex / IP joints
neutral; week 4 - splint
reduced to nocte only;
week 6 - splint ceased
and passive range
commenced as needed;
week 8 - strengthen
exercises initiated
extensor tendon
lacerations in
zones V-VI results
in better functional
oucome at 4, 6 and
8/52 compared to
immobilisation.
2. Grip
strength
2. Jamar
dynamometry
8/52 and 6/12
No blinding of
assessors
MCP = metacarpophalangeal
IP = interphalangeal
Ext = extension
Flex = flexion
J Hand Microsurg (July–December 2012) 4(2):65–73 69

Static Immobilisation Versus Early Active Mobilisation
Results from Hall et al. and Bulstrode et al. studies could be
combined for comparison between static immobilisation and
EAM (Fig. 3a and b). At 3-4 weeks, patients who were
rehabilitated using EAM regained significantly greater TAM
than those with static immobilisation (p< 0.00001 with a mean
difference of 80 °). At 12 weeks, EAM group continued to
achieve significantly greater TAM than static immobilisation
group (p00.03 with a mean difference of 19 °).
A Results at 6 weeks.
B Results at 8-12 weeks.
Study or Subgroup
Hall 2010
Mowlawi 2005
Total (95% CI)
Heterogeneity: Chi² = 1.85, df = 1 (P = 0.17); I² = 46%
Test for overall effect: Z = 2.85 (P = 0.004)
Mean
240
216
SD
25
36
Total
5
17
22
Mean
248
247
SD
22
20
Total
8
17
25
Weight
35.0%
65.0%
100.0%
IV, Fixed, 95% CI
-8.00 [-34.69, 18.69]
-31.00 [-50.58, -11.42]
-22.96 [-38.74, -7.17]
Static Dynamic Mean Difference Mean Difference
IV, Fixed, 95% CI
-50 -25 025 50
Study or Subgroup
Hall 2010
Mowlawi 2005
Total (95% CI)
Heterogeneity: Chi² = 0.23, df = 1 (P = 0.63); I² = 0%
Test for overall effect: Z = 2.48 (P = 0.01)
Mean
178
206
SD
42
53
Total
5
17
22
Mean
198
239
SD
39
22
Total
8
17
25
Weight
26.3%
73.7%
100.0%
IV, Fixed, 95% CI
-20.00 [-65.67, 25.67]
-33.00 [-60.28, -5.72]
-29.58 [-53.00, -6.16]
Static Dynamic Mean Difference Mean Difference
IV, Fixed, 95% CI
-100 -50 050 100
Fig. 2 Static immobilisation versus dynamic splinting. aResults at 6 weeks. bResults at 8–12 weeks
Static immobilisation
Dynamic splinting
Early active
TAM
0
30
60
90
120
150
180
210
240
270
0 4 8 12 16 20 24
Weeks
TAM (deg)
Fig. 1 This chart plots the mean TAM values at the varioustime points as reported by the studies using continuousvariables (Bulstrode, Hall, Mowlawi)
70 J Hand Microsurg (July–December 2012) 4(2):65–73

Dynamic Splinting Versus Early Active Mobilisation
Results from Khandwala et al. and Chester et al. studies
were combined. There was no statistically significant differ-
ence in the proportion of fair and poor results, according to
the TAM assessment [19], between patients rehabilitated
using dynamic splinting or EAM (p00.23) (Fig. 4).
Grip Strength
Three studies reported grip strength [16–18]. Hall et al.
found no significant difference in grip strength at 12 weeks
among the three regimens studied [17].Mowlawietal.
reported percentage of grip force measured in injured versus
uninjured hands. Dynamic splinting group was found to
have better grip strength than static immobilisation at
8 weeks but not at 6 months [18]. Bulstrode et al. showed
their results in a bar chart and grip strength was measured in
kilograms [16]. At 12 weeks postoperatively, in the static
immobilisation group, the repaired hand was significantly
weaker than the uninjured hand. However such difference
was not demonstrated in the EAM group.
Demand on Therapist Input
There was no difference in the median number of therapy
visits, comparing dynamic splinting and EAM [10]. Over a
12-week rehabilitation, there was no significant difference
in the overall time spent with the therapists regardless of the
regimens (static immobilisation or EAM) adopted [16].
Complications
Only three cases of tendon re-ruptures were reported in one
study with 100 subjects (3 %) [9]. Two were in the EAM
group and one in dynamic splinting. In addition, there were
two cases of cellulits: one in EAM and one in dynamic
splinting [10].
Study or Subgroup
Chester2002
Khandwala 2000
Total (95% CI)
Total events
Heterogeneity: Not applicable
Test for overall effect: Z = 1.19 (P = 0.23)
Events
0
1
1
Total
29
78
107
Events
0
4
4
Total
29
84
113
Weight
100.0%
100.0%
M-H, Fixed, 95% CI
Not estimable
0.26 [0.03, 2.38]
0.26 [0.03, 2.38]
Dynamic Active Odds Ratio Odds Ratio
M-H, Fixed, 95% CI
0.01 0.1 110 100
Fig. 4 Early active motion versus dynamic splinting
A Results at 3-4 weeks.
B Results at 12 weeks.
Study or Subgroup
Bulstrode2005
Hall 2010
Total (95% CI)
Heterogeneity: Chi² = 0.64, df = 1 (P = 0.43); I² = 0%
Test for overall effect: Z = 2.13 (P = 0.03)
Mean
230
240
SD
27
25
Total
10
5
15
Mean
242
266
SD
31
20
Total
13
11
24
Weight
52.4%
47.6%
100.0%
IV, Fixed, 95% CI
-12.00 [-35.75, 11.75]
-26.00 [-50.90, -1.10]
-18.67 [-35.85, -1.49]
static active Mean Difference Mean Difference
IV, Fixed, 95% CI
-100 -50 050 100
Study or Subgroup
Bulstrode2005
Hall 2010
Total (95% CI)
Heterogeneity: Chi² = 0.01, df = 1 (P = 0.92); I² = 0%
Test for overall effect: Z = 5.46 (P < 0.00001)
Mean
79
110
SD
42
52
Total
10
5
15
Mean
160
188
SD
42
41
Total
13
11
24
Weight
69.0%
31.0%
100.0%
IV, Fixed, 95% CI
-81.00 [-115.62, -46.38]
-78.00 [-129.62, -26.38]
-80.07 [-108.82, -51.31]
Mean Difference Mean Difference
IV, Fixed, 95% CI
-100 -50 050 100
Fig. 3 Static immobilisation versus early active mobilisation. aResults at 3–4 weeks. bResults at 12 weeks
J Hand Microsurg (July–December 2012) 4(2):65–73 71

Discussion
This study reviewed the highest level of evidence available
on the relative merits of the different rehabilitation regimens
following surgical repair of extensor tendon injuries of the
hand. It showed evidence in favour of early mobilisation
regimens in achieving quicker recovery of finger motion
than static immobilisation but the long-term outcome
appeared to be similar. The five prospective randomised
trials that formed the basis of this systematic review includ-
ed only simple tendon lacerations in zones V and VI, except
from Chester et al. which also included zones IV (one digit)
and VII (6 digits) injuries. The numbers were too small to
facilitate meaningful subgroup analysis according to Verdan
zones and the conclusions of this review apply to zones V/
VI injuries only.
There have been two other systematic reviews on this
subject [12,13]. Talsma et al. included four RCT (same as
our inclusion apart from Hall et al.) and one retrospective
comparative study. They concluded that early controlled
mobilisation leads to better functional results than immobi-
lisation early in the rehabilitation. However no differences
were identified between the regimens three months postop-
eratively [12]. In contrast, Sameem et al. applied less strin-
gent criteria and included 17 studies (3 RCT, 4 prospective
series and 10 retrospective series) in total. Notably, studies
by Khandwala et al. and Hall et al. were not included in their
analysis. In spite of that, they arrived at a similar conclusion
in support of dynamic splinting over static immobilisation
[13].
In the present systematic review, the results from the
trials were pooled according to the principle of the regimens
employed. We recognised that there would be institution-
related differences in the design of splintage, frequency of
therapy sessions and specific instructions on exercises
among the trials, even within a similarly-themed regimen.
This is expected to have important implications on the
practical aspects of rehabilitation and potentially on the final
outcome of extensor tendon repairs but a pragmatic ap-
proach has been taken in performing this systematic review.
There is no standardisation in the reporting of TAM,
which may be reported as a continuous or a categorical
variable. Furthermore there is no universally accepted nor-
mal TAM value [19] hence the need of reporting it as a
percentage of the normal contralateral hand. To compound
the issue further, the investigators have not been consistent
when using the denominator of fingers, hands or patients in
their reporting. The outcome of multiple extensor tendon
repairs in one hand cannot be considered as independent
observations [10], as it may lead to an underestimation of
within-group variability and it inflates the sample size [20].
Sauerland et al. outlined options of resolving the problems
by restricting the analysis to only one measurement per
patient or by employing complex statistical modelling (gen-
eralised estimating equations) to analyse all available meas-
urements with adjustment for data dependency [20].
Extensor tendon injuries of the hand affect a predomi-
nantly young male patient population [9]. High rate of loss
of follow-up was a problem faced by many investigators
[10,17]. Therefore the reported results are expected to
reflect only the experience and outcome in patients who
had been compliant with the follow-up. In addition, splint
wear compliance was not addressed formally in any of the
included trials. The recommendations of our review thus
could not be extrapolated to apply to every patient. In
particular, in those patients judged to be unlikely to comply
with therapy involving early mobilisation, static immobili-
sation still has an important role to play [11]. Although static
immobilisation requires minimal therapist input in the first
4 weeks, it is important to appreciate that more input may be
required subsequently [16,18].
The available evidence would suggest that EAM and
dynamic splinting are associated with comparable out-
comes. It has been estimated that 15 min are required to
make a resting splint (for EAM) whilst 45 min are required
to make the two splints required for dynamic splinting (one
static splint for night wear and one outrigger splint) [10].
While there was no difference in the median number of
therapy visits between the two regimens [10], a greater level
of expertise and increased length of appointment time is
anticipated for patients managed with dynamic splinting
regimen.
In the trial by Khandwala et al., two tendon ruptures
occurred in the EAM group and one in dynamic splinting
[9]. Even though the reported incidence of tendon re-rupture
is low, no tendon rupture has been recorded in the static
immobilisation group in a RCT setting.
The published trials have made significant contribution to
our understanding of rehabilitation following extensor ten-
don injuries. However there were limitations in the design,
sample size, standardisation of regimens and reporting of
outcomes. In order to produce more robust data, we recom-
mend that future studies report a minimum dataset to include
the mean and range of patient age; gender and hand domi-
nance; the number of patients, hands and fingers involved;
the zone and extent of injuries; minimum follow-up of
6 months with interval reviews at 4, 8, 12 weeks (as the
usual period for static immobilisation is 4 weeks); actual
values of TAM and grip strength (means and standard devi-
ations) in both injured and uninjured fingers/hands; func-
tion, using a validated patient-rated outcome measure; and
complications particularly rates of tendon re-rupture and
subsequent tenolysis.
Apart from Hall et al. who had employed a self-reported
functional visual analogue scale, all trials had focussed on
measurements of range of motion and grip strength only.
72 J Hand Microsurg (July–December 2012) 4(2):65–73

None had utilised a validated patient reported functional
score such as Disability of Arm, Shoulder and Hand
(DASH) or Patient Evaluation Measure (PEM). In addition,
it would be valuable to prospectively measure the costs of
therapist utilisation and the economic burden produced by
time off from work due to the injuries and subsequent
rehabilitation in the trial recruits.
Only Hall et al. had performed a priori sample-size esti-
mation but their actual loss of follow-up (33 %) was greater
than the 25 % predicted attrition rate. This highlights the
challenge of conducting clinical study in a young, migratory
patient population.
Given the similar results of dynamic splinting and EAM, as
well as the simpler splint design for EAM, we recommend that
future RCT on this subject compare static immobilisation and
EAM in order to keep the number of trial recruits to a mini-
mum. The standard deviations of TAM for static immobilisa-
tion and EAM were roughly 50 and 40 ° respectively [17,18].
We consider 30 ° difference in TAM to be clinically significant
for the two groups. In order to detect this difference between
the two groups with 80 % power and the level of significance
set at 5 %, a minimum of 36 patients are required in each
group. When 33 % of attrition is taken into account, 54
patients need to be recruited into each arm of a RCT.
In conclusion, following zones V/VI extensor tendon
repairs, early mobilisation regimens (active and passive)
achieve quicker recovery of finger motion than static immobi-
lisation but the long-term outcome appears to be similar. Early
mobilisation is suitable for patients who require an early return
to activities and those who would comply with the constraints
of the rehabilitation. Given the comparable outcomes between
dynamic splinting and EAM, the latter may be preferred as it
does not require the same level of technical expertise and may
be less cumbersome for the patient.
References
1. Rockwell WB, Butler PN, Byrne BA (2000) Extensor tendon:
anatomy, injury, and reconstruction. Plast Reconstr Surg 106
(7):1592–1603, quiz 1604, 1673
2. Zilber S, Oberlin C (2004) Anatomical variations of the extensor
tendons to the fingers over the dorsum of the hand: a study of 50
hands and a review of the literature. Plast Reconstr Surg 113
(1):214–221. doi:10.1097/01.PRS.0000091163.86851.9C
3. von Schroeder HP, Botte MJ (1997) Functional anatomy of the
extensor tendons of the digits. Hand Clin 13(1):51–62
4. Evans RB (1986) Therapeutic management of extensor tendon
injuries. Hand Clin 2(1):157–169
5. Eissens MH, Schut SM, van der Sluis CK (2007) Early active wrist
mobilization in extensor tendon injuries in zones 5, 6, or 7. J Hand
Ther 20(1):89–91. doi:10.1197/j.jht.2006.11.003
6. Boyer MI (2005) Flexor tendon biology. Hand Clin 21(2):159–166.
doi:10.1016/j.hcl.2004.11.009
7. Tang JB (2005) Clinical outcomes associated with flexor tendon
repair. Hand Clin 21(2):199–210. doi:10.1016/j.hcl.2004.11.005
8. Vucekovich K, Gallardo G, Fiala K (2005) Rehabilitation after
flexor tendon repair, reconstruction, and tenolysis. Hand Clin 21
(2):257–265. doi:10.1016/j.hcl.2004.11.006
9. Khandwala AR, Webb J, Harris SB, Foster AJ, Elliot D (2000) A
comparison of dynamic extension splinting and controlled active
mobilization of complete divisions of extensor tendons in zones 5
and 6. J Hand Surg Br 25(2):140–146. doi:10.1054/jhsb.1999.0356
10. Chester DL, Beale S, Beveridge L, Nancarrow JD, Titley OG
(2002) A prospective, controlled, randomized trial comparing early
active extension with passive extension using a dynamic splint in
the rehabilitation of repaired extensor tendons. J Hand Surg Br 27
(3):283–288. doi:10.1054/jhsb.2001.0745
11. Russell RC, Jones M, Grobbelaar A (2003) Extensor tendon repair:
mobilise or splint? Chir Main 22(1):19–23
12. Talsma E, de Haart M, Beelen A, Nollet F (2008) The effect of
mobilization on repaired extensor tendon injuries of the hand: a
systematic review. Arch Phys Med Rehabil 89(12):2366–2372.
doi:10.1016/j.apmr.2008.06.019
13. Sameem M, Wood T, Ignacy T, Thoma A, Strumas N (2011) A
systematic review of rehabilitation protocols after surgical repair of
the extensor tendons in zones V-VIII of the hand. J Hand Ther 24
(4):365–373. doi:10.1016/j.jht.2011.06.005
14. Higgins JPT, Green S (2011) Highly sensitive search strategies for
identifying reports of randomized controlled trials in MEDLINE.
In: Cochrane Handbook for Systematic Reviews of Interventions
Version 5.1.Q [updated March 2011]. The Cochrane Library. pp
Available from www.cochrane-handbook.org
15. Handoll HHG, Vaghela MV (2004) Interventions for treating mal-
let finger injuries. Cochrane Database of Systematic Reviews:
Issue 3. Art. No.: CD004574
16. Bulstrode NW, Burr N, Pratt AL, Grobbelaar AO (2005) Extensor
tendon rehabilitation a prospective trial comparing three rehabili-
tation regimes. J Hand Surg Br 30(2):175–179. doi:10.1016/
j.jhsb.2004.10.016
17. Hall B, Lee H, Page R, Rosenwax L, Lee AH (2010) Com-
paring three postoperative treatment protocols for extensor ten-
don repair in zones V and VI of the hand. Am J Occup Ther
64(5):682–688
18. Mowlavi A, Burns M, Brown RE (2005) Dynamic versus static
splinting of simple zone V and zone VI extensor tendon repairs: a
prospective, randomized, controlled study. Plast Reconstr Surg 115
(2):482–487
19. Kleinert HE, Verdan C (1983) Report of the committee on tendon
injuries (International Federation of Societies for Surgery of the
Hand). J Hand Surg Am 8(5 Pt 2):794–798
20. Sauerland S, Lefering R, Bayer-Sandow T, Bruser P, Neugebauer
EA (2003) Fingers, hands or patients? The concept of independent
observations. J Hand Surg Br 28(2):102–105
J Hand Microsurg (July–December 2012) 4(2):65–73 73