Full Length Article

Journal of Hand Surgery

(European Volume)
A randomized controlled trial comparing 0(0) 1–6
! The Author(s) 2018
ketorolac and triamcinolone injections in Reprints and permissions:
sagepub.com/journalsPermissions.nav
adults with trigger digits DOI: 10.1177/1753193418756808
journals.sagepub.com/home/jhs

Mabel Qi He Leow1, Aik Siew Robyn Hay2, Shu Li Ng2,

Muntasir Mannan Choudhury2, Huihua Li3,
Duncan Angus McGrouther2 and Shian Chao Tay1,2,4

Abstract
We assessed the efficacy of ketorolac trometamol injections compared with triamcinolone acetonide injec-
tions in trigger digits. Patients with trigger digits were randomized to receive either ketorolac or triamcino-
lone. They were followed up at 3, 6, 12 and 24 weeks, and monitored for resolution of triggering, pain and total
active motion. One hundred and twenty-one patients with single trigger digits were recruited (59 ketorolac,
62 triamcinolone). At 6 weeks, 54% of patients in the triamcinolone group had complete resolution of trigger,
whereas no patients in the ketorolac group had resolution. At 12 weeks, 58% of patients in the triamcinolone
group had complete resolution of trigger compared with 6.7% in the ketorolac group. At 24 weeks, both
groups had comparable rates of resolution at 26% and 25%, respectively. Patients in the triamcinolone
group had significantly better resolution of pain at 3, 6 and 12 weeks. But at 24 weeks, there was no significant
difference in pain between both groups. Significantly less flexion deformity was reported at 3 weeks and
6 weeks in the triamcinolone group. In the short term, ketorolac was less effective in relieving symptoms of
trigger digit than triamcinolone.
Level of evidence: I

Keywords
Triamcinolone, ketorolac, steroid, trigger finger

Date received: 26th July 2017; revised: 11th January 2018; accepted: 12th January 2018

phospholipid, ketorolac acts via the cyclooxygenase

Introduction
Steroid injection is a common and generally safe
treatment for triggering of digits, with response
rates ranging from 45% to 80% (Choudhury and
Tay, 2014; Lambert et al., 1992; Murphy et al.,
1995). Some of the reported side effects include tran-
sient hyperglycaemia (Wang and Hutchinson, 2006),
immune suppression resulting in increased predis-
position to infection (Yam et al., 2009), skin atrophy
(Friedman et al., 1988; Schoepe et al., 2006) and
hypopigmentation (Brinks et al., 2010; Friedman
et al., 1988).
Use of non-steroidal anti-inflammatory drugs
(NSAIDs) has been proposed as a substitute for
triamcinolone (Shakeel and Ahman, 2012). Unlike
triamcinolone, which reduces inflammation by
inhibiting the release of arachidonic acid from
pathway and inhibits COX-2 dependent prostaglan-
dins in the cells. The aims of this prospective rando-
mized trial were to assess the efficacy of ketorolac
injections compared with triamcinolone injections to
treat trigger digit in adults and to determine the
safety profile of ketorolac injections over 24 weeks.
1
Biomechanics Laboratory, Singapore General Hospital, Singapore
2
Department of Hand Surgery, Singapore General Hospital,
Singapore
3
Division of Medicine, Singapore General Hospital, Singapore
4
Duke-NUS Medical School, Singapore
Corresponding Author:
Shian Chao Tay, Department of Hand Surgery, Singapore General
Hospital, The Academia, 20 College Road, 169856 Singapore,
Singapore.
Email: tay_sc77@yahoo.com
2 Journal of Hand Surgery (Eur) 0(0)
Methods
Institutional ethics review board approval was
obtained before conducting the study. Approval was
obtained from the Health Sciences Authority for off-
label use of ketorolac trometamol with lidocaine in
the setting of this trial.
All patients who presented to the clinic between
January 2014 and June 2016 were screened for
recruitment into the trial. Adults (21 years and
above) presenting with one trigger digit requiring
injection treatment were included. The exclusion
criteria were patients with ketorolac or triamcino-
lone allergy, pregnant or breastfeeding women,
and previous triamcinolone injection to the same
digit.
Patients who were eligible for recruitment were
given information on the study and consented.
Recruited patients were randomized into one of two
treatment groups. Computerized randomization was
used and sealed envelopes containing the treatment
groups were opened after the patient had consented
to participate in the study. Patients in the study group
(ketorolac group) were treated with 0.5 ml parenteral
ketorolac trometamol (30 mg/ml) þ 0.5 ml parenteral
lidocaine 1% (10 mg/ml). Patients in the control group
(triamcinolone group) were treated with 0.5 ml paren-
teral triamcinolone acetonide (10 mg/ml) þ 0.5 ml
parenteral lidocaine 1% (10 mg/ml). Using an asep-
tic technique, a combined intrasynovial and extrasy-
novial injection was given at the level of the A1 pulley
using a 25-gauge needle and a 3 ml syringe. Only a
single injection was administered to each affected
digit.
This was a single blinded study and patients were
not informed of the drugs they were given. A trained
research coordinator carried out the assessments
and data collection.
Patients were examined at 3, 6, 12 and 24 weeks
from time of the first injection. Demographic data,
past medical history, presenting symptoms, pain
score (rated on a scale of 0 for no pain and 10 for
the worse possible pain), site of pain, tenderness and
swelling at the A1 pulley, severity of trigger digit and
total active motion (TAM) were recorded at baseline.
At subsequent data collection time-points, severity of
trigger digit, pain, tenderness and swelling at the A1
pulley and TAM were obtained. Swelling at the A1
pulley was assessed by palpating the A1 pulley with
the distal pulp of the finger. At 24 weeks, patients
were assessed for their satisfaction, whether they
would be willing to undergo the same injection
again and perceived improvement. Perceived
improvement was patients’ self-reported perception
of improvement in symptoms.
The severity of the trigger digit was quantified
using a four-grade system. Grade 1 (pre-triggering)
patients had pain, a history of catching, but catching
was not demonstrable on physical examination and
tenderness over the A1 pulley. Grade 2 patients had
demonstrable catching but could actively extend the
digit. Grade 3 patients had demonstrable catching
requiring passive extension (Grade 3A) or inability
to actively flex (Grade 3B). Recurrence of trigger
digit was defined as presence of the signs and symp-
toms of trigger digit (e.g. triggering or pain over the
A1 region), after the patient had achieved complete
resolution following the injection. There were no
changes in outcome measures throughout the dur-
ation of the study.
Any side-effects or adverse events related to the
drug were monitored and recorded. The minimum
sample size to detect a difference in one severity
grade at significance level of 0.05, power of 0.8 and
standard deviation of 0.6 was 90 patients. Taking into
consideration an anticipated dropout rate of 25%, 121
patients were recruited into the study.
Continuous variables are reported as median and
interquartile range. Categorical data are reported
with frequency, together with proportions. The
Mann–Whitney U test was used to compare unpaired
continuous variables between two groups. The
Wilcoxon signed rank test was carried out to com-
pare paired continuous variables between the two
groups. Fisher’s exact test was used to compare
the distribution of categorical data between different
groups of patients. An intention to treat analysis was
conducted, and all participants who were recruited
were included in the data analysis.
Results
One hundred and fifty patients were assessed for eli-
gibility, and 121 were recruited. Fifty-nine were ran-
domized to the ketorolac group and 62 were
randomized to the triamcinolone group. Three
patients in the ketorolac group dropped out at 6
weeks as they opted for surgery. Forty-seven
patients in the ketorolac group and 52 patients in
the triamcinolone group completed the final evalu-
ation. The CONSORT diagram is shown in Figure 1.
The mean age of the patients was 60 years (range
44–87). There was no statistical difference between
the ketorolac and triamcinolone groups in terms of
age, gender, race, hand dominance, diabetes or other
medical history. There was a significant difference
(p = 0.002) in the hand affected by trigger digit
between the two groups. There was no difference in
mechanical symptoms, such as triggering, stiffness,
clicking and locking between both groups (Table 1).
Leow et al. 3

Figure 1. Flow chart of patient disposition.

FU: follow-up.

Table 1. Continued
Table 1. Patient demographics and clinical data by
groups. Ketorolac Triamcinolone
Ketorolac Triamcinolone Demographics (n = 59) (n = 62)
Demographics (n = 59) (n = 62) Hand dominance
Gender Right 55 (93%) 59 (95%)
Female 36 (61%) 40 (65%) Left 4 (7%) 3 (4.8%)
Male 23 (39%) 22 (36%) Hand involved
Race Right 25 (42%) 44 (71%)
Chinese 50 (85%) 53 (86%) Left 34 (58%) 18 (29%)
Malay 2 (3%) 3 (5%) Medical history
Indian 6 (10%) 5 (8%) Diabetes mellitus 15 (25%) 9 (15%)
Others 1 (2%) 1 (2%) Hypertension 25 (42%) 29 (47%)
Smoking 6 (10%) 6 (10%) Hyperlipidaemia 27 (46%) 29 (47%)
(continued)
4 Journal of Hand Surgery (Eur) 0(0)

Outcomes Table 2. Comparison of trigger digit grade in ketorolac

and triamcinolone groups.
Severity of trigger digit
Ketorolac Triamcinolone
A comparison between patients with no demon- Grade (n = 59) (n = 62) p-value
strable triggering (Grade 0) with other patients with
symptoms (Grades 1–3) was conducted. At 3 weeks, Baseline 0.4198
6 weeks and 12 weeks, there was significantly 1 4 (7%) 5 (8%)
higher resolution of trigger digit in the triamcinolone 2 35 (59%) 39 (63%)
group (p < 0.0001). However, at 24 weeks, both 3A 13 (22%) 7 (11%)
groups had comparable rates of resolution (p = 0.15) 3B 7 (12%) 11 (18%)
(Table 2). 3 weeks <0.0001
0 0 (0%) 19 (37%)
Pain 1 14 (28%) 20 (39%)
2 30 (59%) 10 (19%)
At baseline, the percentage of patients without pain
in the ketorolac and triamcinolone group was com- 3A 4 (8%) 1 (2%)
parable at 12% and 4.8%, respectively. At 3 weeks, 3B 3 (6%) 2 (4%)
6 weeks and 12 weeks, the percentage of patients 6 weeks <0.0001
without pain was significantly higher in the triamcino- 0 0 (0%) 27 (54%)
lone group compared with the ketorolac group 1 16 (33%) 15 (30%)
(p < 0.001). At 24 weeks, the percentage of patients 2 22 (46%) 5 (10%)
without pain in the ketorolac and triamcinolone 3A 6 (13%) 2 (4%)
group was comparable (55% and 48%, respectively, 3B 4 (8%) 1 (2%)
p = 0.548). 12 weeks <0.0001
The median pain scores between the ketorolac and 0 3 (7%) 28 (58%)
triamcinolone group was comparable at baseline.
1 19 (42%) 13 (27%)
For the ketorolac group, the median pain score
2 17 (38%) 5 (10%)
decreased steadily from 5 to 0 over 24 weeks,
whereas for the triamcinolone group, the median 3A 4 (9%) 2 (4%)
pain score decreased rapidly from 5 to 0 at 3 weeks 3B 2 (4%) 0 (0%)
and increased up to a median of 1 at 24 weeks. 24 weeks 0.1490
Patients in the triamcinolone group had better reso- 0 12 (26%) 13 (25%)
lution of pain at 3, 6 and 12 weeks (p < 0.001), but at 1 14 (30%) 27 (52%)
24 weeks, there was no significant difference in pain 2 16 (34%) 9 (17%)
between both groups (p = 0.37). 3A 3 (6%) 2 (4%)
3B 2 (4%) 1 (2%)
A1 pulley tenderness and swelling p-values calculated by Fisher’s exact test. Significant differences
shown in bold type.
Tenderness at the A1 pulley on palpation was noted
in 88% and 97% in the patients’ hands in the ketor-
olac and triamcinolone groups, respectively, at base- at 64% and 65% for ketorolac and triamcinolone
line. There was significantly less tenderness at the groups, respectively (p = 0.66).
A1 pulley in the triamcinolone group compared with
the ketorolac group at 3, 6 and 12 weeks (p < 0.001).
Total active motion and flexion deformity
However, at 24 weeks, the percentage of patients
with tenderness at the A1 pulley was comparable There was no significant difference in TAM between
(47% for ketorolac, and 46% for triamcinolone, the two groups at all time points. At baseline, the
p = 1.0) percentage of patients with flexion deformity at the
At baseline, the percentages of patients with swel- proximal interphalangeal joint was 14% and 26% for
ling and/or induration of the A1 pulley were 97% for ketorolac and triamcinolone groups, respectively
both the ketorolac and triamcinolone groups. At 3, (p = 0.11). The percentage of patients with flexion
6 and 12 weeks, patients in the triamcinolone group deformity at the proximal interphalangeal joint was
had significantly less swelling at the A1 pulley only significantly lower in the triamcinolone group at
(p < 0.001 at weeks 3 and 12, p < 0.05 at week 6). 3 weeks and 6 weeks (p < 0.05). By week 24, there
At 24 weeks, the percentage was comparable was no difference between the groups (p = 1.0)
Leow et al.
Patient subjective responses
At week 24, 90% of patients in the triamcinolone
group and 79% of patients in the ketorolac group
reported that they were willing to undergo same
treatment (p = 0.16). Patient satisfaction was rec-
orded at 24 weeks, and no significant difference
between the groups was found (81% for both
groups). However, there was a significant difference
in patients’ perceived improvement in their condition.
From 3 to 12 weeks, the triamcinolone group
reported significantly higher perceived improve-
ments (p < 0.001). At 24 weeks, both groups of
patients had similar perceived improvement of 80%
from baseline (p = 0.33).
Conversion to surgery
Three patients in the ketorolac group elected for sur-
gery and exited the trial within 24 weeks giving the
ketorolac group a surgical rate of 5.1%. Two of the
patients had persistent or worsening pain, and the
third had persistent triggering. None of the patients
in the triamcinolone group opted for surgery.
Complications
There were no complications associated with the use
of triamcinolone or ketorolac in this study.
Discussion
Our study showed that triamcinolone injection pro-
vided better short-term outcomes in terms of sever-
ity, resolution of triggering and reduction in pain,
swelling and flexion deformity at 12 weeks.
However, the longer term outcomes of ketorolac
and triamcinolone at 24 weeks were similar. This
was attributed to both improvements in outcomes
in the ketorolac group and recurrence of symptoms
in the triaminolone group. Triamcinolone was most
effective at 12 weeks with nearly 60% of patients
having complete resolution of trigger. However,
between 12 to 24 weeks, the recurrence rate was
56% for triamcinolone. This high rate of recurrence
with triamcinolone injection was also reported by
Rozental et al. (2008), with 56% of patients reporting
recurring symptoms at a median of 5.6 months.
Patients who were injected with ketorolac did not
experience any improvement of trigger until the
twelfth week and achieved a resolution rate of 25%
at 24 weeks with no recurrences noted within the
study period. As the half-life of ketorolac is 4–6
hours (Beacon Pharmaceuticals, 2017), it is possible
that the resolutions in the ketorolac group at 12 and
5
24 weeks could be attributed to the natural reso-
lution of symptoms. However, there has been no
study on the natural history of trigger digits without
treatment. Further, given the high recurrence in the
triamcinolone group, it is unlikely that the reduction
of symptoms in the ketorolac group at 24 weeks was
due to the natural history. It is also possible that what
was observed may also be due to an unexplained
effect of ketorolac.
Our response rate with ketorolac at 12 weeks was
6.7% and compares unfavourably with a previous
study that reported a response rate of 56% using
diclofenac injection (Shakeel and Ahmad, 2012). The
lower response rate could be attributable to the drug
dosage. The dose of ketorolac used in this study
(15 mg) was a quarter of the dose typically used for
injection (Chung, 2002). Shakeel and Ahmad (2012)
used 12.5 mg of diclofenac, which was half of the
dose usually used for injection. Other possible rea-
sons could be the combination of lidocaine in the
injections, which reduced the drug concentration,
whereas in the study of Shakeel and Ahmad (2012),
the drugs were given neat. Also, intrasynovial and
extrasynovial injections were given in our study,
whereas Shakeel and Ahmad (2012) gave all injec-
tions intrasynovially.
We chose ketorolac for our trial instead of diclo-
fenac because there were reports of diclofenac caus-
ing local skin and soft tissue irritation (Chung, 2002).
Ketorolac has traditionally been indicated for the
short-term management of moderate to severe
acute post-operative pain, and given intravenously
or intramuscularly (Chung, 2002).
Despite the lower resolution rate and prevalence
of pain up to 12 weeks in the ketorolac group, a sur-
prising 79% of patients in the ketorolac group
reported that they were willing to undergo the
same treatment; this was lower than the 90% in the
triamcinolone group, but the differences were not
significant. As expected, patients in the triamcinolone
group had significantly higher rates of perceived
improvements up to 12 weeks, but at 24 weeks, the
rate of perceived improvements was similar. Despite
these differences, the overall satisfaction rate
between the two groups was comparable at 81%.
Although response to ketorolac is slower than that
to triamcinolone, it may have a role in the treatment of
trigger digits where steroid injections are contraindi-
cated, such as in patients with uncontrolled diabetes.
However, patients should be aware that ketorolac may
take several months to show its effect. Triamcinolone
also has its limitations, mainly the high recurrence of
triggering. It is clear that ketorolac cannot be used on
its own to replace steroid injections. However, it is
possible that it might reduce recurrence if given in
6 Journal of Hand Surgery (Eur) 0(0)
combination with steroid injections, which can be
explored in future studies.
Acknowledgements The authors would like to thank Ms
Kai Li Chan, Dr Geraldine Leow, Nursing Officer Yeam Shin
Yen, Staff Nurse Li Chan Tou, Staff Nurse Datchayini D/O
Thiyanamurthi and Ms Yen Lee Yoong for their invaluable
contribution to the study. We are grateful to the partici-
pants in the study. We also acknowledge Zeus Lim for his
assistance in manuscript submission.
Declaration of conflicting interests The authors
declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding The author(s) disclosed receipt of the following
financial support for the research, authorship, and/or pub-
lication of this article: This study was funded by the follow-
ing grants: (1) The Biomechanics Laboratory Programme
Grant and the Surgery Academic Clinical Program seed
grant [Reference no: GRSG13NSL]. Both grants are
funded by the SingHealth Duke-NUS Academic Medical
Centre, facilitated by the Joint Office of Academic
Medicine (JOAM). It is an initiative of Surgery Academic
Clinical Programme (SURG ACP), hosted at the Singapore
General Hospital (SGH). (2) Singapore Ministry of Health’s
National Medical Research Council under its Centre Grant
[NMRC/CG/016/2013].
Ethical approval The ethics for this study were cleared
by the SingHealth Research Centralised Institutional
Review Board (CIRB Ref: 2013/258/D) and Health
Sciences Authority (HSE Ref: HPRG/CTB 78:10/13-073).
References
Beacon Pharmaceuticals. Ketorolac 30 mg/ml solution for
injection. UK: Beacon Pharmaceuticals: 2017. https://
www.medicines.org.uk/emc/medicine/20922 (accessed 25
March 2017).
Brinks A, Koes BW, Volkers AC, Verhaar JA, Bierma-Zeinstra
SM. Adverse effects of extra-articular corticosteroid injec-
tions: a systematic review. BMC Musculoskelet Disord.
2010, 11: 1.
Choudhury MM, Tay SC. Prospective study on the management of
trigger finger. Hand Surg. 2014, 19: 393–7.
Chung CL. The use of injectable nonsteroidal anti-inflammatory
drugs in local accident & emergency practice. Hong Kong J
Emerg Med. 2002, 9: 65–71.
Friedman SJ, Butler DF, Pittelkow MR. Perilesional linear atrophy
and hypopigmentation after intralesional corticosteroid ther-
apy. Report of two cases and a review of the literature. J Am
Acad Dermatol. 1988, 19: 537–41.
Lambert MA, Morton RJ, Sloan JP. Controlled study of the use of
local steroid injection in the treatment of trigger finger and
thumb. J Hand Surg Br. 1992, 17: 69–70.
Murphy D, Failla JM, Koniuch MP. Steroid versus placebo injection
for trigger finger. J Hand Surg Am. 1995, 20: 628–31.
Rozental TD, Zurakowski D, Blazar PE. Trigger finger: prognostic
indicators of recurrence following corticosteroid injection.
J Bone J Surg. 2008, 90: 1665–72.
Schoepe S, Schäcke H, May E, Asadullah K. Glucocorticoid
therapy-induced skin atrophy. Experiment Derm. 2006, 15:
406–20.
Shakeel H, Ahmad TS. Steroid injection versus NSAID injection for
trigger finger: a comparative study of early outcomes. J Hand
Surg Am. 2012, 37: 1319–23.
Wang A, Hutchinson D. The effect of corticosteroid injection for
trigger finger on blood glucose level in diabetic patients.
J Hand Surg Am. 2006, 31: 979–81.
Yam A, Teoh LC, Yong FC. Necrotising fasciitis after corticosteroid
injection for trigger finger: a severe complication from a ‘safe’
procedure. J Hand Surg Eur. 2009, 34: 689–90.