International Journal of Rheumatic Diseases 2017

ORIGINAL ARTICLE

Local steroid injection versus wrist splinting for carpal

tunnel syndrome: A randomized clinical trial
Ho SO,1 Vincent C. H. CHUNG,2 Joseph C. K. CHENG1 and Ronald M. L. YIP3
1
Kwong Wah Hospital, 2The Chinese University of Hong Kong, and 3Integrated Diagnostic and Medical Centre, Tung Wah Group of
Hospitals, Hong Kong, China

Abstract
Aim: We conducted a prospective randomized parallel clinical trial comparing the efficacy of local steroid injec-
tion and nocturnal wrist splinting in patients with carpal tunnel syndrome (CTS).
Methods: The well-validated and disease-specific Boston Carpal Tunnel Questionnaire (BCTQ) was employed
and its score at 4 weeks after treatment was used as the primary outcome measure. Important secondary out-
comes included patient satisfaction, the change of an objective finger dexterity test and the side effects.
Results: Twenty-five patients in the local steroid group and 25 patients in the wrist splinting group completed
the study procedures. At 4 weeks after treatment, there was significant improvement of the BTCQ scores in both
the steroid group and splinting group. There was improvement of the finger dexterity test only in the steroid
group but not in the splinting group. However, there was no statistically significant difference in the changes of
BTCQ scores between the two groups after treatment. Patient satisfaction score was higher in the steroid group.
Patients in the steroid group took fewer painkillers after treatment. Four patients developed side effects after
splinting and three after local steroid injection, which was not statistically significant.
Conclusion: Although local steroid injection and nocturnal wrist splinting were equally effective in the treat-
ment of patients with CTS, only the former improved objective hand function. Local steroid injection also
resulted in better patient satisfaction and less painkiller use without causing more side effects.
Key words: local steroid injection, carpal tunnel syndrome, wrist splinting.

BACKGROUND the workplace. In the USA, between 400 000 and

500 000 patients underwent surgical decompression
Carpal tunnel syndrome (CTS) is caused by the pressure
annually which equates to an economic cost of in
and consequent compression on the median nerve
excess of $2 billion.1 CTS was found to account for
within a confined anatomical area at the wrist, referred
more morbidity, measured by cases and working days
to as the carpal tunnel. It is very common and has an
lost, than any other illness in the US working popula-
important socio-economic impact. The prevalence of
tion.3 CTS may also occur as a work-related disorder
CTS in the UK is 7–16%.1 Its incidence has been esti-
leading to compensation claims.
mated at 88–125 per 100 000 in the USA.2 Economic
Despite the vast burden of the disease, there is no
consequences include the direct financial implications
consensus regarding its best management. Local steroid
of treatment and the indirect cost of absenteeism from
injection and wrist splinting are among the most popu-
lar non-surgical interventions. They are commonly
employed in mild to moderate cases of CTS. For severe
Correspondence: Dr Ho So, Department of Medicine and Geria-
cases, they can also offer relief of complaints during the
trics, Kwong Wah Hospital, 25 Waterloo Road, Kowloon,
Hong Kong, China. waiting period for surgery or when there are contra-
Email: h99097668@hotmail.com indications for surgery. However, the number of studies

© 2017 Asia Pacific League of Associations for Rheumatology and John Wiley & Sons Australia, Ltd
H. So et al.
that compare these two methods is limited, and they
are mostly either retrospective in design or prospective
but non-randomized.4,5 Therefore, we conducted a
prospective randomized parallel clinical trial comparing
the efficacy of local steroid injection and wrist splinting
in patients with CTS using a well-validated outcome
measure with a 4-week follow-up.
PATIENTS AND METHODS
Patient recruitment
Patients attending the medical clinic of a local hospital
(Kwong Wah Hopital) complaining of finger or hand
numbness were referred to the electro-neuro-diagnostic
unit for nerve conduction test. Consecutive patients
with clinical and electrophysiological features of CTS
were invited to participate in the study. Clinical features
were pain, paresthesia or weakness in the median nerve
distribution for at least 3 months. The neurodiagnostic
criteria were based on the American Academy of Neu-
rology summary statement, which further classified the
abnormalities as follows: (i) mild abnormality, that is,
abnormal comparative tests or prolonged median distal
sensory latency (> 3.5 ms) but normal median distal
motor latency (DML); (ii) moderate abnormality, that
is, prolonged median DSL and DML (> 4.2 ms); and
(iii) severe abnormality, that is, absence of median sen-
sory nerve action potential or absent compound muscle
action potentials.6
Patients were excluded if they had any recognized
causes of CTS including inflammatory arthritis, diabetes
mellitus, hypothyroidism, renal failure, polyneuropathy
and history of significant local trauma. Other exclusion
criteria included age younger than 18 years, pregnancy
and previous treatments of CTS, namely injection,
splinting and surgery. Patients with motor impairment
or thenar muscle atrophy were also excluded and
referred to the Orthopaedic Unit for assessment of
surgery.
The study was approved by the Kowloon West Cluster
Research Ethics Committee. All potential participants
were informed of the objectives and procedures of this
study, as well as the possible complications. The
patients who have given informed written consent were
interviewed by a single investigator within 12 weeks
after the nerve conduction velocity (NCV) test. Their
demographical data including age, gender and body
mass index (BMI) were recorded. The duration of the
symptoms, medical comorbidities and employment sta-
tus were documented. Examination of the hand was
performed by the same investigator focusing on the
2 International Journal of Rheumatic Diseases 2017
sensory loss at the tips of digits one, two, three, or the
medial side of digit four, and weakness or atrophy of
the abductor pollicis brevis or opponens pollicis. Sensa-
tion was assessed using pinpricks. Sensory and motor
function was recorded to be either normal or impaired.
A well-validated test for finger dexterity, namely the
nine-hole peg test was performed.7 In patients with
bilateral CTS, the most symptomatic hand was
included. In case both hands were equally symp-
tomatic, the dominant hand was included. Recruited
patients were asked to complete the Boston Carpal Tun-
nel Questionnaire (BCTQ). They were then allocated to
one of the two treatment arms according to the ran-
domization procedure using sequentially numbered
opaque sealed envelopes (SNOSE).8 Two sets of equal
number of sealed opaque envelopes containing a sheet
of paper marked Steroid Injection or Splinting were shuf-
fled very thoroughly. The envelopes were then marked
on the front with a unique number sequentially starting
from one. Patients were thus randomly assigned to one
of the two treatment arms according to what was
marked in these envelopes.
Intervention
The local injection of steroid was performed by the
same investigator after the randomization. Using a
sterile technique, 20 mg methylprednisolone acetate
premixed with lidnocaine was injected using a
25-guage 9 5/8” needle. The needle was inserted medi-
ally to the palmaris longus tendon at the distal palmar
crease in the wrist at an angle of 45° to the forearm. The
steroid was injected at approximately 1 cm below the
skin. The needle was repositioned if there was any resis-
tance to injection, or any pain or paresthesia in the
median nerve territory.9
Comparator
After randomization, the hands of the patients in the
splinting group were splinted in a neutral position with
a standard cotton–polyester splint. Patients were
instructed to use the splints during nighttime for
1 month. They were also encouraged to mark each
night they had worn the splints on a calendar to ensure
compliance.
Outcome
The primary outcome measure was clinical improve-
ment in terms of symptoms and function at 4 weeks.
The BCTQ was used as the measure assessing the clini-
cal response. It is a self-administered disease-specific
questionnaire for assessing severity of symptoms and

functional status of patients with CTS. The evidence
base of the psychometric properties indicates that the
BCTQ is a valid, reliable, responsive and acceptable
instrument.10 It has been recommended by the Ameri-
can Academy of Orthopedic Surgeons as the disease-
specific instrument used to assess patients’ responses to
CTS treatment for research.11 It essentially assesses
severity of symptoms and functional status based on
two scales. The symptom severity scale (SSS) is com-
prised of 11 questions, and the functional status scale
(FSS) is comprised of eight questions. The assessment
of each question is on a scale of 1–5 points, in which 1
indicates no symptom and 5 indicates severe symp-
toms. Each scale generates a final score (sum of individ-
ual scores divided by number of items) which ranges
1–5. The questionnaire has Chinese validity.12 It has
been shown that the ordinality of response of the
Chinese version agreed with the original instrument.
Test-retest reproducibility showed no significant differ-
ence between tests. The Pearson correlation coefficient
ranged from 0.83 to 0.93 and the internal consistency
was good.
Secondary outcome measures included patient satis-
faction, nine-hole peg test, duration of sick leave of
employed patients, use of analgesics, side effects and
prognostic factors. Patients were interviewed again by a
single investigator blinded to the treatment assignment
at 1 month after treatment. They were asked to complete
the BCTQ and rate their satisfaction regarding the treat-
ment received. Assessment of patient satisfaction was
classified on a five-point scale as completely satisfied,
almost satisfied, moderately satisfied, somewhat satis-
fied, and dissatisfied. The nine-hole peg test which objec-
tively reflects finger dexterity was repeated. The patient
was instructed to pick up the pegs one at a time using the
diseased hand, put them into the holes and then remove
the pegs one at a time as fast as possible. The time used
for the whole process was recorded. Duration of sick
leave for employed patients and concomitant use of
analgesics were counted in days and recorded. Finally,
an adverse effect report form was filled in. Patients were
asked if there was any adversity related to the received
treatment and any side effects thought to be related were
documented. The association between the potential
prognostic factors and the outcomes was assessed. These
factors included age, duration of symptoms, presence of
sensory deficit and NCV severity.13–16
Sample size estimation and statistical analysis
Based on previous studies, the minimum clinically
important difference is 0.74 for the total BCTQ score
International Journal of Rheumatic Diseases 2017
Steroid injection versus splinting for CTS
after surgery.17 With 90% statistical power, 5% signifi-
cance level, and two-sided statistical tests, and assuming
a standard deviation of 0.8 for the BCTQ score, the
study could detect a true difference of at least 0.74 point
on the BCTQ between the two intervention groups
when 50 patients were randomized.10,18
Version 20.0 of the SPSS statistical package was used
(SPSS Inc., Chicago, IL, USA). Descriptive statistics were
presented as frequencies, means with standard devia-
tion or medians with ranges as appropriate. Baseline
characteristics of patients between two treatment groups
were compared by Chi-square test for categorical vari-
ables, independent samples t-test for continuous vari-
ables with normal distribution or Mann–Whitney
U-test for nonparametric continuous variables. BCTQ
scores (SSS score, FSS score and total score) before treat-
ment and at 4 weeks after treatment were compared
with Wilcoxon signed ranks test. Differences of the
BCTQ scores between the two treatment groups were
investigated with independent samples t-test. Regarding
the secondary outcomes, Mann–Whitney U-test was
used to compare the differences of the patient satisfac-
tion scores, nine-hole peg test time and the durations of
sick leave and analgesics used, while Fisher exact test
was used to analyze the side effects of the two treat-
ments. For multivariate analysis, multiple linear regres-
sion was used to compute the differences of outcome
measures between the two treatment groups before and
after the respective treatments with adjustment for base-
line variables. Potential predictive factors on treatment
response were analyzed. Results were considered statis-
tically significant if the P-value was less than 0.05.
RESULTS
Characteristics of the study population
Patients were recruited over a period of 8 months from
October 2013 to May 2014. Twenty-five patients in the
local steroid injection group and 25 in the splinting
group completed the study procedures. The mean age
of the patients was 57.3 years (SD = 9.34). There was a
clear female predominance (43/50, 86%). Many
patients were working (29/50, 58%) and many were
overweight (29/50, 58%).19 The median duration of
symptoms before treatment was 104 weeks (range:
12–1040). Out of the 50 patients, 16 (32%) had mild,
33 (66%) had moderate and one (2%) had severe NCV
abnormality. The baseline median SSS score was 2
(range: 1.27–3.64), while the baseline median FSS score
was 1.44 (range: 1–3.5). The baseline median nine-hole
peg test time was 24.75 sec (range: 17.2–109.81).
3
H. So et al.

Table 1 shows the baseline characteristics of the
patients after randomization by treatment groups. The
two groups were essentially comparable in terms of
demographic and clinical characteristics. There was also
no statistically significant difference in the baseline
BCTS scores and the nine-hole peg test time between
the two groups.
Efficacy
There were statistically significant improvements after
local steroid injection in the SSS score (P < 0.001), FSS
score (P = 0.022) and the overall score (P < 0.001).
The improvements in the SSS (P = 0.001), FSS
(P = 0.05) and overall score (P < 0.001) after wrist
splinting were also statistically significant. There was
improvement in the nine-hole peg test time only in the
steroid group (P = 0.038) but not in the splinting
group (P = 0.065).
Table 2 shows the treatment outcomes of the patients
by study group. There was no statistically significant dif-
ference between the two groups in terms of the SSS
score change, FSS score change, overall score change
and nine-hole peg test time change. However, the
patient satisfaction score was significantly higher in the
steroid group. After treatment, there was also signifi-
cantly less painkiller use in the steroid injection group
(P = 0.042) but not in the splinting group. There was
no difference in the duration of sick leave taken related
to CTS.
Multivariate analysis showed that the changes in the
BCTQ scores were not significantly affected by age,
duration of symptoms, presence or absence of sensory
deficit and NCV severity. However, the presence of
sensory loss and young age were found to be the inde-
pendent predictors of improvement in the nine-hole
peg test time.
Side effects
Four patients in the splinting group reported discomfort
when wearing the splint, while 3 patients in the steroid
injection group reported short-lasting pain after the
injection. The occurrence rates of the side effects were
not statistically significantly different.
DISCUSSION
Despite the importance of CTS, there is no universally
accepted treatment strategy. Surgical treatment is gener-
ally preferred in severe cases of CTS. A Cochrane review
investigating surgical treatment of CTS showed surgical
treatment relieves symptoms significantly better than
splinting, but no conclusion could be drawn in people
with mild symptoms and if surgical treatment is better
than steroid injection.20 On the other hand, surgical
treatment is relatively costly and carries risks of signifi-
cant complications that may last several months and
give rise to further work absence.21,22 For non-surgical
interventions many modalities have been trialed but
only a few have shown discernible benefits.1 Local ster-
oid injection and wrist splinting are among the most
popular options. Systematic reviews of randomized
controlled trials have concluded that local steroid injec-
tion provides greater clinical improvement at 1 month
compared with placebo and that there is weak evidence
that a splint worn at night is more effective than no
treatment in up to 3 months.4,11,23

Table 1 Baseline characteristics of the study subjects after randomization by study group
Characteristic Steroid injection Splinting P-value
Mean age  SD years 57.32  9.12 57.28  9.75 0.988
Gender: female/male 21/4 22/3 1
Employment: yes/no 17/8 12/13 0.152
Overweight: yes/no 12/13 17/8 0.152
Median duration of symptoms (range) weeks 78 (12–1040) 104 (39–1040) 0.078
Right hand/left hand 13/12 16/9 0.39
Sensory loss: yes/no 16/9 12/13 0.254
NCV severity: 1/2/3 7/17/1 9/16/0 0.527
Median baseline SSS (range) 2.27 (1.27–3.64) 2 (1.36–3.45) 0.69
Median baseline FSS (range) 1.5 (1–2.75) 1.38 (1–3.5) 0.853
Median baseline overall score (range) 1.88 (1.14–2.76) 1.89 (1.18–3.48) 0.892
Median baseline nine-hole peg test (range) sec 25.16 (17.43–109.81) 24.47 (17.2–34.68) 0.299
NCV, nerve conduction velocity; SSS, symptom severity scale; FSS, functional status scale.

4 International Journal of Rheumatic Diseases 2017


Table 2 Treatment outcomes of the study subjects after randomization by study group
Mean SSS change  SD
Mean FSS change  SD
Mean overall score change  SD
Median patient satisfaction score (range)
Median nine-hole peg test change (range)
Side effects:† yes/no
Change of treatment:‡ yes/no
†All 3 patients receiving steroid injection reported pain while 4 who received splinting reported discomfort. ‡Six patients in the steroid injection
group opted to receive splinting while 3 in the splinting wanted to have steroid injection after the study period. SSS, symptom severity scale; FSS,
functional status scale.
We conducted the first prospective randomized paral-
lel clinical trial comparing head-to-head the effective-
ness of local steroid injection and wrist splinting with a
sizeable sample using a validated outcome measure.
Our patients were mostly middle-aged working women,
who are the characteristic population at risk of CTS.4,24
The duration of symptoms before treatment was long.
This might reflect the typical low level of awareness of
the disease in the community. There were significant
clinical improvements after either local steroid injection
or wrist splinting at 1 month. Although there were no
statistically significant differences in the improvements
of symptom or functional scores between the two treat-
ment groups, the objective assessment of hand dexterity
and patient satisfaction were better after steroid
injections.
Prior to this study, there were two randomized con-
trolled trials comparing local steroid injection and
splinting. Celiker et al. conducted an unblinded ran-
domized clinical trial with an 8-week follow-up of 33
hands in 23 patients, comparing the efficacy of steroid
injection against splinting combined with the use of a
non-steroidal anti-inflammatory drug.25 It showed that
both treatment regimes resulted in significant improve-
ment but there was no difference between the two inter-
ventions. These findings resonate with those from ours.
On the other hand, in the other trial of 120 patients by
Sevim et al. comparing steroid injection and splinting,
only splinting provided symptomatic relief while injec-
tions of steroids were ineffective at 1-year follow-up.26
However, the conclusions of these two studies are lim-
ited either by the small sample size or the lack of a vali-
dated outcome measure.
Ultrasound-guided local steroid injection has gained
increasing popularity in the past few years. It potentially
minimizes the risk of damage to the median nerve or
the surrounding vulnerable structures, and increases the
International Journal of Rheumatic Diseases 2017
Steroid injection versus splinting for CTS
Steroid injection Splinting P-value
0.670  0.614 0.38  0.475 0.07
0.190  0.364 0.21  0.342 0.873
0.443  0.426 0.29  0.351 0.22
5 (1–5) 3 (1–5) 0.04
2.56 ( 9.47 to 7.73) 1.19 ( 7.46 to 7.91) 0.16
3/22 4/21 1
6/19 3/22 0.463
efficacy by ensuring placement of the injected steroid in
the carpal tunnel. However, in a randomized prospec-
tive study comparing ultrasound-guided and blind ster-
oid injection in CTS, SSS and FSS of the BTCQ
improved significantly in both groups at 6 weeks after
treatment.27 There was no significant difference
between the two groups in symptoms or function as
well as side effects.
This study was conducted in the general medical
unit of a local hospital. It recruited a representative
sample from the general patient population present-
ing with symptoms of CTS. The external validity of
the results in patient populations using non-tertiary
healthcare services could be optimized. However, the
open label design of the study means the potential
ascertainment bias introduced by unblinding is not
excluded. The overall BCTQ score improvement was
numerically better in the steroid injection group. It is
possible that the study is underpowered to detect the
true difference between the two treatments as the
sample size was calculated using previous data from
patients receiving surgical treatment. Further, we have
not looked into the relative long-term efficacy and
cost effectiveness of the two treatments. The actual
choice of treatment in daily practice will depend also
on these unresolved issues.
To conclude, our study findings suggest that both
local steroid injection and nocturnal wrist splinting are
effective in alleviating the symptoms of CTS at 1 month
of follow-up. However, objective improvement in hand
function could be shown only after steroid injection,
which also offered better patient satisfaction. Thus, in
patients with mild to moderate CTS, we advocate the
use of local steroid injection for quick relief of symp-
toms and restoration of function. Nocturnal wrist
splinting is a valid alternative, especially for patients
who do not want an injection.
5
H. So et al.

ACKNOWLEDGEMENTS 13 Gelberman RH, Aronson D, Weisman MH (1980) Carpal-car-

pal tunnel syndrome. Results of a prospective trial of steroid
We have nothing to declare or acknowledge. injection and splinting. J Bone Joint Surg 62A, 1181–4.
14 Weiss AP, Sachar K, Gendreau M (1994) Conservative
management of carpal tunnel syndrome: a reexamination
CONFLICT OF INTEREST of steroid injection and splinting. J Hand Surg 19, 410–15.
There is no conflict of interest. 15 Graham RG, Hudson DA, Solomons M, Singer M (2004)
A prospective study to assess the outcome of steroid injec-
tions and wrist splinting for the treatment of carpal tunnel
REFERENCES syndrome. Plast Reconstr Surg 113 (2), 550–6.
16 Visser LH, Ngo Q, Groeneweg SJ, Brekelmans G (2012)
1 Prime MS, Palmer J, Khan WS, Goddard NJ (2010) Is there
Long term effect of local corticosteroid injection for carpal
light at the end of the tunnel? controversies in the diagno-
tunnel syndrome: a relation with electrodiagnostic sever-
sis and management of carpal tunnel syndrome. Hand 5
ity. Clin Neurophysiol 123 (4), 838–41.
(4), 354–60.
17 Ozyurekoglu T, McCabe SJ, Goldsmith LJ, LaJoie AS
2 Stevens JC, Sun S, Beard CM, O’Fallon WM, Kurland LT
(2006) The minimal clinically important difference of the
(1988) Carpal tunnel syndrome in Rochester Minnesota,
carpal tunnel syndrome symptom severity scale. J Hand
1961 to 1980. Neurology 38, 134–8.
Surg 31 (5), 733–8.
3 Leigh JP, Miller TR (1998) Occupational illnesses within two
18 Hagebeuk EE, de Weerd AW (2004) Clinical and electro-
national data sets. Int J Occup Environ Health 4, 99–113.
physiological follow-up after local steroid injection in the
4 Marshall S, Tardif G, Ashworth N (2007) Local corticos-
carpal tunnel syndrome. Clin Neurophysiol 115 (6), 1464–8.
teroid injection for carpal tunnel syndrome. Cochrane
19 World Health Organization (1995) Physical status: the use
Database Syst Rev 18, 001554.
and interpretation of anthropometry. Report of a WHO
5 Ucan H, Yagci I, Yilmaz L, Yagmurlu F, Keskin D, Bodur H
Expert Committee. WHO Technical Report Series 854.
(2006) Comparison of splinting, splinting plus local steroid
WHO, Geneva.
injection and open carpal tunnel release outcomes in idio-
20 Verdugo RJ, Salinas RA, Castillo JL, Cea JG (2008) Surgical
pathic carpal tunnel syndrome. Rheumatol Int 27 (1), 45–51.
versus non-surgical treatment for carpal tunnel syndrome.
6 Visser LH, Smidt MH, Lee ML (2008) High-resolution
Cochrane Database Syst Rev 4, 01552.
sonography versus EMG in the diagnosis of carpal tunnel
21 Benson LS, Bare AA, Nagle DJ, Harder VS, Williams CS,
syndrome. J Neurol Neurosurg Psychiatry 79, 63–7.
Visotsky JL (2006) Complications of endoscopic and open
7 Mathiowetz V, Weber K, Kashman N, Volland G (1985)
carpal tunnel release. Arthroscopy 22, 919–24.
Adult norms for the nine hole peg test of finger dexterity.
22 Atroshi I, Larsson GU, Ornstein E, Hofer M, Johnsson R,
Occup Ther J Res 5 (1), 24–37.
Ranstam J (2006) Outcomes of endoscopic surgery com-
8 Doig GD, Simpson F (2005) Randomization and alloca-
pared with open surgery for carpal tunnel syndrome
tion concealment: a practical guide for researchers. J Crit
among employed patients: randomized controlled trial.
Care 20, 187–93.
BMJ 332, 1473–6.
9 Weiss AP, Sachar K, Gendreau M (1994) Conservative
23 O’Connor D, Marshall S, Massy-Westropp N (2003) Non-
management of carpal tunnel syndrome: a reexamina-
surgical treatment (other than steroid injection) for carpal
tion of steroid injection and splinting. J Hand Surg 19
tunnel syndrome. Cochrane Database Syst Rev 1, 03219.
(3), 410–15.
24 Ly-Pen D, Andreu JL, Blas G, Olaso AS, Millan I (2005)
10 Leite J, Jerosch-Herold C, Song F (2006) A systematic
Surgical decompression versus local steroid injection in
review of the psychometric properties of the Boston
carpal tunnel syndrome. Arthritis Rheum 529 (2), 612–19.
Carpal Tunnel Questionnaire. BMC Musculoskelet Disord
25 Celiker R, Arslan S, Inanici F (2002) Corticosteroid injection
7, 78.
vs. nonsteroidal antiinflammatory drug and splinting in car-
11 Keith MW, Masear V, Chung KC et al. (2010) American
pal tunnel syndrome. Am J Phys Med Rehabil 81 (3), 182–6.
academy of orthopaedic surgeons clinical practice guide-
26 Sevim S, Dogu O, C ß amdeviren H et al. (2004) Long-term
line on the treatment of carpal tunnel syndrome. J Bone
effectiveness of steroid injections and splinting in mild and
Joint Surg 92, 218–19.
moderate carpal tunnel syndrome. Neurol Sci 25, 48–52.
12 Fok MW, Leung HB, Lee WM (2007) Evaluation of a Hong
27 Ustun N, Tok F, Yagız AE et al. (2013) Ultrasound-guided
Kong Chinese version of a self-administered questionnaire
vs. blind steroid injections in carpal tunnel syndrome: a
for assessing symptom severity and functional status of
single-blind randomized prospective study. Am J Phys Med
carpal tunnel syndrome: cross-cultural adaptation and
Rehabil 92, 999–1004.
reliability. Hong Kong Med J 13 (5), 342–7.

6 International Journal of Rheumatic Diseases 2017