JCA-08866; No of Pages 6

Journal of Clinical Anesthesia xxx (2017) xxxxxx

Contents lists available at ScienceDirect

Journal of Clinical Anesthesia

Original Contribution

Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar

spine surgery: A meta-analysis of randomized controlled trials
Zhifeng Zhang a, Haihua Xu b, Yunhui Zhang c, Wei Li c, Yanjie Yang c, Tian Han d, Zhihui Wei d,
Xue Xu c, Jingui Gao c,
a
Department of Pain Treatment, the Second Hospital of Tianjin Medical University, Tianjin 300211, China
b
Department of Urinary Surgery, the Children's Hospital in Tianjin, Tianjin 300074, China
c
Department of Anesthesia, the Second Hospital of Hebei Medical University, Shijiazhuang 050000, China
d
Department of Anesthesia, Baoding NO.2 Central Hospital, Baoding 072750, China

a r t i c l e i n f o a b s t r a c t

Article history: Study objective: Nonsteroidal anti-inammatory drugs (NSAIDs) play a role in pain relief, especially in postoper-
Received 22 May 2017 ative pain caused by inammation. They have demonstrated signicant opioid dose-sparing effects, which help
Received in revised form 24 August 2017 in reducing postoperative effects and opioid side effects. The objective of this meta-analysis was to explore the
Accepted 29 August 2017 role of NSAIDs in reducing postoperative pain at different time intervals and provide reference for medication
Available online xxxx
after lumbar spine surgery by a meta-analysis of randomized controlled trials (RCT).
Design: A meta-analysis study of randomized controlled trials.
Keywords:
Anti-inammatory drugs
Setting: Postoperative recovery area.
Postoperative pain control Patients: Adult patients who have undergone lumbar spine surgery.
Lumbar spine surgery Intervention: Patients received NSAIDs for pain control after lumbar spine surgery.
Measurements: Standardized mean difference (SMD) and 95%CI were used to evaluate the visual analog scale of
postoperative pain.
Main results: Four hundred and eight participants from eight studies were included in this study. The difference
between the NSAIDs group and placebo is signicant in 06, 12, and 24 h groups (overall: SMD = 0.72, 95%CI
0.98 to 0.45; 06 h: SMD = 0.50, 95%CI 0.81 to 0.19; 12 h: SMD = 1.07, 95%CI 1.45 to 0.70; 24 h:
SMD = 1.16, 95%CI 1.87 to 0.45). Heterogeneity and publication bias were observed in the 06 and 24 h
groups.
Conclusion: NSAIDs are effective in postoperative analgesia after lumbar spine surgery. The study type, NSAID
dose, different surgery types, and analgesic type might inuence the efcacy of NSAIDs.
2017 Elsevier Inc. All rights reserved.

1. Introduction physiologic functions [2]. Inducible COX-2 is expressed in cells playing

a role in inammation, such as the macrophages, mononuclear cells,
Nonsteroidal anti-inammatory drugs (NSAIDs) play an important and endothelial cells [3]. COX-3 has been found to have no COX activity
role in pain relief, especially in postoperative pain caused by inamma- [4].
tion. The mechanism underlying NSAID action involves the inhibition of Opioid analgesics are very useful in the management of moderate-
cyclooxygenase (COX) activity to reduce prostaglandin synthesis [1]. In- to-severe pain. However, several side effects of opioid use have been re-
ammatory pain depends on prostaglandin E2 synthesized by COX in ported, including respiratory depression and potential drug addiction
neural cells. COX exists as three distinct isoforms: constitutive COX-1, [57]. Studies and systematic reviews about lumbar spine surgery
inducible COX-2, and COX-3. COX-1 is responsible for the essential have demonstrated that NSAIDs have a signicant opioid dose-sparing
effect, which helps in reducing postoperative effects as well as opioid
Corresponding author: Department of Anesthesia, the Second Hospital of Hebei
side effects [8,9]. New studies and analysis methods have been put for-
Medical University, No.215 West Heping Load, Shijiazhuang 050000, China. ward in recent years. Therefore, conclusions of the related studies
E-mail address: gaojingui@126.com (J. Gao). should be reorganized and re-analyzed. In this study, we conducted a

https://doi.org/10.1016/j.jclinane.2017.08.030
0952-8180/ 2017 Elsevier Inc. All rights reserved.

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030
2 Z. Zhang et al. / Journal of Clinical Anesthesia xxx (2017) xxxxxx

meta-analysis to explore the role of NSAIDs in reducing postoperative Subgroup analysis was conducted according to the type of operation
pain after lumbar spine surgery. We also summarize the side effects of and placebo in this study to assess their inuence. All statistical analyses
NSAIDs in lumbar spine patients and provide reference to clinical were performed using Stata11.0. A P value b 0.05 was considered statis-
medicine. tically signicant.

2. Methods 3. Results

2.1. Search strategy and selection criteria 3.1. Characteristics of the included studies

This study was conducted according to Preferred Reporting Items for Forty-seven studies were obtained through literature screening;
Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Electronic however, 39 studies were not eligible and excluded. Finally, eight
databases (PubMed, EBSCO, Elsevier, Springer, Wiley, Cochrane) were RCTs met the criteria (Fig. 1). Eight studies involving 408 participants
searched systematically up to August 31, 2016, using the search were included in the meta-analysis, and the details are presented in
terms: lumbar, spine, surgery, and NSAIDs. This search was car- Table 1.
ried out without language restriction. Studies were included if they met The outcome measure of pain is depicted as VAS in the included
the following criteria: a. randomized controlled trials; b. trials per- studies. In these studies, COX-2 selective NSAIDs were piroxicam,
formed in humans; patients underwent lumbar spine surgery; c. pain ketorolac, celecoxib, parecoxib, and paracetamol, and the nonselective
score reported with Visual Analog Scale (VAS: 010 cm, 0 = no pain NSAIDs were dexketoprofen and metamizol. Two studies included pa-
and 10 = maximum pain); and d. adult patients. The exclusion criteria tients who underwent spinal fusion surgery and six studies included
were as follows: a. duplication of research literature; b. systematic re- those who underwent lumbar disc surgery. Two studies focused on
views and meta-analysis; and c. results and full study details were un- dose efcacy [13,14]. One study used a VAS of 0100 mm scale [10]. In
available after contacting the authors. six studies, the patients were allowed intravenous morphine as needed.
Two studies used piritramide during the rst 24-h postoperative period
2.2. Data extraction as needed [10,15].
Five studies received Institutional Review Board approval [9,13,15
Two workers were in charge of data extraction. The data included 17]. In ve studies, written informed consent was obtained from pa-
the rst author's name, year of publication, type of study, number of pa- tients who participated in the study [10,1315,18]. In one study, ap-
tients, age(mean), gender, type of operation, VAS score[time, mean, and proval was obtained from ethics committee as well as the
standard deviation(SD)], type of NSAIDs, and placebo. The VAS score of participating patients [9].
0100 mm in one study was transformed into 010 cm VAS [10]. The 0
6, 12, and 24 h groups were divided according to the contents of the in- 3.2. Summary of pain scores
cluded studies.
From the results of meta-analysis, the overall SMD is 0.72 and
2.3. Statistical analysis 95%CI 0.98 to 0.45, and heterogeneity was observed among studies
(I2 = 71.0%, P b 0.0001, Fig. 2). Pain scores were summarized at three
We evaluated the postoperative VAS by using standardized mean time ranges or points: 06, 12, and 24 h (Fig. 2).VAS was signicantly
difference (SMD) and 95%CI. The heterogeneity of the study was esti- lower in the NSAID group (06 h: SMD = 0.50, 95%CI 0.81 to
mated by I2 statistic. I2 statistic shows that the difference between stud- 0.19; 12 h: SMD = 1.07, 95%CI 1.45 to 0.70; 24 h: SMD =
ies was a random error or chance. Random-effect model was utilized for 1.16, 95%CI 1.87 to 0.45). In the time range of 06 h, eight studies
all results where heterogeneity existed across studies [11,12]. We per- were included. Heterogeneity existed between the eight studies accord-
formed Begg's test to assess the possible publication bias among studies. ing to the results of meta-analysis (I2 = 70.2%, P b 0.0001). Five RCTs
If heterogeneity was found, subgroup analyses were conducted. were included in the meta-analysis in the time range of 12 h, and no

Fig. 1. Flow chart of literature screening process.

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030
 Z. Zhang et al. / Journal of Clinical Anesthesia xxx (2017) xxxxxx 3

Table 1
Characteristics of the studies included in the meta-analysis.

Author Year Study type n Study group Outcome measurement

Ryan Michael Garcia16 2013 Prospective, randomized study 22 Celecoxib, pregabalin, and oxycodone VAS
EzequielHCassinelli17 2008 Double-blinded RCT 25 Ketorolac VAS
Ulrich Grundmann10 2006 Double-blinded RCT 80 Parecoxib, paracetamol, metamizol VAS
Matthias Karst15 2003 Double-blinded RCT 34 Celecoxib VAS
Chathchai Pookarnjanamorakot18 2002 Double-blinded RCT 47 Piroxicam VAS
Scott S. Reuben14 1998 Double-blinded RCT 70 Ketorolac VAS
Scott S. Reuben13 1997 Double-blinded RCT 80 Ketorolac VAS
Ebru KELSAKA9 2014 Double-blinded RCT 50 Dexketoprofen VAS

heterogeneity was observed among them (I2 = 0.0%, P = 0.699). In the I2 = 0.0%, P = 0.497). According to the analgesic type, heterogeneity
time range of 24 h, six studies were included in the meta-analysis and existed in the morphine group (morphine: I2 = 57.8%, P = 0.006;
heterogeneity existed between them (I2 = 77.3%, P = 0.001). piritramide: I2 = 0.0%, P = 0.511). Subgroup analysis was not per-
According to the heterogeneity test results, subgroup analysis ac- formed in the 24 h group, as the number of studies was inadequate.
cording to the type of NSAIDs, surgery, study, and analgesics were per- The results of publication bias are presented in Fig. 6. From the g-
formed to assess the reason for heterogeneity in the 06 h group (Figs. ure, it can be seen that publication bias existed in the 06 h group (0
3, 4, and 5). For both COX-1 and COX-2 NSAIDs, the VAS scores in the 6 h: t = 2.15, P = 0.047; 12 h: t = 0.51, P = 0.643; 24 h: t = 1.44, P
study group were lower than those in the control group (COX-1: SMD = 0.224).
= 0.45, 95%CI 0.45 to 0.09; COX-2: SMD = 0.59, 95%CI 1.21
to 0.03). In the 06 h group, nonselective NSAIDs differed from the
COX-2 selective NSAIDs signicantly (P b 0.0001). Heterogeneity 3.3. Side effects
existed in both nonselective NSAIDs and COX-2 selective NSAID groups
(COX-1: I2 = 66.1%, P = 0.001; COX-2: I2 = 78.2%, P b 0.0001). Hetero- The summary of side effects is shown in Table 2. The most common
geneity existed between studies in the lumbar surgery or RCT group, but side effect was nausea or vomiting, followed by shivering. Nausea/
not in the spine surgery or RCT with dose efcacy group (lumbar sur- vomiting was reported in 18 cases from 5 studies and cured in approx-
gery/RCT: I2 = 70.9%, P b 0.0001; spine surgery/RCT with dose efcacy: imately 24 h according to the published studies.

Fig. 2. Forest plot of pain scores in 06, 12, and 24 h groups.

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030
4 Z. Zhang et al. / Journal of Clinical Anesthesia xxx (2017) xxxxxx

Fig. 3. Subgroup analyses according to the type of NSAIDs in 06 h group.

Fig. 4. Subgroup analyses according to the type of surgery and study in 06 h group.

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030
 Z. Zhang et al. / Journal of Clinical Anesthesia xxx (2017) xxxxxx 5

Fig. 5. Subgroup analyses according to the analgesic type in 06 h group.

4. Discussion 4.3. Comparison with published studies and reviews

4.1. Pain score The effect of NSAIDs in postoperative pain management has been
proved in previous studies [1921]. A meta-analysis of RCT by Kitti et
In this meta-analysis, we demonstrated that NSAIDs showed signif- al. showed that NSAIDs provided better analgesia than opioid drugs
icantly better pain management than placebo, especially in the COX-2 alone after lumbar spine surgery, especially nonselective NSAIDs [8].
group. According to these results, the type of surgery, study, and analge- However, different outcomes were obtained in the subgroup analyses.
sic had an effect on postoperative analgesia. The results of Begg's test No signicant difference between nonselective and selective COX-2
suggested that publication bias may exist in the 06 h group. However, NSAIDs was observed in this meta-analysis. Previous studies demon-
owing to the limited number of studies, ubiquitous risks of heterogene- strated that selective COX-2 NSAIDs were more successful in postoper-
ity and publication bias were suspected among studies. This may reduce ative pain control than traditional NSAIDS, with lower opioid
the quality of outcome. In the 12 h group, the combined result was consumption and fewer adverse events, such as nausea and vomiting
steady and can be trusted based on the forest plot and Begg's test. The [22,23].NSAIDs performed better in spine surgery than lumbar disc sur-
VAS of NSAIDs decreased signicantly compared to that of opioid anal- gery based on the results of our meta-analysis. This result was different
gesics in the 12 h pain management group. Although the lack of ade- from that of a study published earlier [8]. Inadequate number of studies
quate studies may lead to heterogeneity and publication bias, we in the spine group may have resulted in this discrepancy. Heterogeneity
suspected that the efcacy of NSAIDs was better than that of opioid existed in the lumbar disc group. RCT with ketorolac dose efcacy has
use alone at 24 h. Using NSAIDs for pain management may be safe and been proved to be benecial in reducing postoperative pain [24]. Mod-
effective after lumbar spine surgery. erate doses of NSAIDs showed benecial effects, decreased incidence of
adverse events, and reduced consumption of opioid drugs [25]. Side ef-
fects were stated in all the included studies. A previous study showed
4.2. Side effects that side effects were affected by different races and age [26]. Therefore,
further study on the inuence of the aforementioned categories on side
Nausea or vomiting was the most common adverse event, and it was effects is needed.
mild in all the reported studies. All the included studies reported ad-
verse events, but only the number of cases of side effects was reported.
The summary of side effects can be a reference for postoperative pain 4.4. Strengths and limitations
control after lumbar spine surgery. Severe side-effects such as platelet
dysfunction and postoperative blood loss were reported in studies, This meta-analysis included eight studies involving 408 cases and
which were not observed in the studies included in this meta-analysis had strong evidences to prove the benecial effects of NSAIDs. Our in-
[16,17]. Choochai et al. demonstrated that COX-2 inhibitors were not re- clusion and exclusion criteria were explicit and strict. Another advan-
lated to the increased risk of postoperative bleeding and platelet dys- tage of this study is that we conducted the subgroup analyses
function [18]. according to the type of study, NSAIDs, surgery, and analgesic.

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030
6 Z. Zhang et al. / Journal of Clinical Anesthesia xxx (2017) xxxxxx

study, dose of NSAIDs, type of surgery, and type of analgesic might inu-
ence the efcacy of NSAIDs. The dose and side effects of NSAIDs summa-
rized in this study will be helpful in future lumbar spine surgery.

References
[1] Bekker A, Cooper PR, Frempong-Boadu A, Babu R, Errico T, Lebovits A. Evaluation of
preoperative administration of the cyclooxygenase-2 inhibitor rofecoxib for the
treatment of postoperative pain after lumbar disc surgery. Neurosurgery 2002;50:
10537 [discussion 1057-8].
[2] Yuan C, Smith WLA. Cyclooxygenase-2-dependent prostaglandin E2 biosynthetic
system in the Golgi apparatus. J Biol Chem 2015;290:560620.
[3] O'Sullivan MG, Huggins Jr EM, Meade EA, DeWitt DL, McCall CE. Lipopolysaccharide
induces prostaglandin H synthase-2 in alveolar macrophages. Biochem Biophys Res
Commun 1992;187:11237.
[4] Brzozowski T, Konturek PC, Konturek SJ, Drozdowicz D, Pajdo R, Pawlik M, et al. Ex-
pression of cyclooxygenase (COX)-1 and COX-2 in adaptive cytoprotection induced
by mild stress. J Physiol Paris 2000;94:8391.
[5] Overdyk FJ. Postoperative opioids remain a serious patient safety threat. Anesthesi-
ology 2010;113:25960 [author reply 260-1].
[6] Chou Roger, Fanciullo GJ, Fine Perry G. Clinical guidelines for the use of chronic opi-
oid therapy in chronic noncancer pain. J Pain 2009;10:11330.
[7] Ripamonti CI, Santini D, Maranzano E, Berti M, Roila F, Group EGW. Management of
cancer pain: ESMO clinical practice guidelines. Ann Oncol 2012;23(Suppl. 7):
vii13954.
[8] Jirarattanaphochai K, Jung S. Nonsteroidal antiinammatory drugs for postoperative
pain management after lumbar spine surgery: a meta-analysis of randomized con-
trolled trials. J Neurosurg Spine 2008;9:2231.
[9] Ebru Kelsea, Guldogus Fuat, Erhan ETNOLU. Effect of intravenous dexketoprofen
use on postoperative analgesic consumption in patients with lumbar disc surgery.
ARI 2014;26:826.
[10] Grundmann U, Wornle C, Biedler A, Kreuer S, Wrobel M, Wilhelm W. The efcacy of
the non-opioid analgesics parecoxib, paracetamol and metamizol for postoperative
pain relief after lumbar microdiscectomy. Anesth Analg 2006;103:21722 [table of
contents].
[11] Yao B, Fang H, Xu W, Yan Y, Xu H, Liu Y, et al. Dietary ber intake and risk of type 2
diabetes: a dose-response analysis of prospective studies. Eur J Epidemiol 2014;29:
7988.
[12] Fleiss JL. The statistical basis of meta-analysis. Stat Methods Med Res 1993;2:
12145.
[13] Reuben SS, Connelly NR, Steinberg R. Ketorolac as an adjunct to patient-controlled
morphine in postoperative spine surgery patients. Reg Anesth 1997;22:3436.
[14] Reuben SS, Connelly NR, Lurie S, Klatt M, Gibson CS. Dose-response of ketorolac as
an adjunct to patient-controlled analgesia morphine in patients after spinal fusion
surgery. Anesth Analg 1998;87:98102.
[15] Karst M, Kegel T, Lukas A, Ludemann W, Hussein S, Piepenbrock S. Effect of celecoxib
and dexamethasone on postoperative pain after lumbar disc surgery. Neurosurgery
2003;53:3316 [discussion 336-7].
[16] Meek IL, Vonkeman HE, Kasemier J, Movig KL, van de Laar MA. Interference of
NSAIDs with the thrombocyte inhibitory effect of aspirin: a placebo-controlled, ex
vivo, serial placebo-controlled serial crossover study. Eur J Clin Pharmacol 2013;
69:36571.
[17] Siddiqui AK, Sadat-Ali M, Al-Ghamdi AA, Mowa HA, Ismail SA, Al-Dakheel DA. The
effect of etoricoxib premedication on postoperative analgesia requirement in ortho-
pedic and trauma patients. Saudi Med J 2008;29:96670.
[18] Teerawattananon C, Tantayakom P, Suwanawiboon B, Katchamart W. Risk of periop-
erative bleeding related to highly selective cyclooxygenase-2 inhibitors: a systemat-
ic review and meta-analysis. Semin Arthritis Rheum 2017;46:5208.
[19] Jirarattanaphochai K, Thienthong S, Sriraj W, Jung S, Pulnitiporn A, Lertsinudom S,
Fig. 6. Begg's funnel plot with pseudo 95% condence limits.
et al. Effect of parecoxib on postoperative pain after lumbar spine surgery: a
bicenter, randomized, double-blinded, placebo-controlled trial. Spine 2008;33:
1329 [Phila Pa 1976].
Owing to the insufcient number of studies, we could not perform [20] Toroudi HP, Borghei Razavi M, Borghei Razavi H, Tabatabayi RM, Tabar YT, Yahyavi
meta-regression analysis to nd out every heterogeneity factor. NSAIDs ST, et al. Comparison of the analgesic effect of ibuprofen with mesalamine after
demonstrated dose efcacy in pain management. Lack of dose-response discectomy surgery in patients with lumbar disc herniation: a double-blind random-
ized controlled trial. Neurol India 2009;57:3059.
analysis is another limitation of our study. [21] Reuben SS, Buvanendran A, Kroin JS, Raghunathan K. The analgesic efcacy of
celecoxib, pregabalin, and their combination for spinal fusion surgery. Anesth
5. Conclusion Analg 2006;103:12717.
[22] Neuss H, Koplin G, Haase O, Reetz C, Mall JW. Preemptive analgesia reduces pain
after radical axillary lymph node dissection. J Surg Res 2010;162:8894.
Our meta-analysis demonstrated that NSAIDs are effective in post- [23] Kashe P, Honarmand A, Safavi M. Effects of preemptive analgesia with celecoxib or
operative pain management after lumbar spine surgery. The type of acetaminophen on postoperative pain relief following lower extremity orthopedic
surgery. Adv Biomed Res 2012;1:66.
[24] Sevarino FBSR, Paige D, Silverman DG. Intravenous ketorolac as an adjunct to pa-
tient-controlled analgesia (PCA) for management of postgynecologic surgical pain.
Table 2 J Clin Anesth 1994;6:237.
Summary of side effects. [25] Cohen BE, Hartman MB, Wade JT, Miller JS, Gilbert R, Chapman TM. Postoperative
pain control after lumbar spine fusion. Patient-controlled analgesia versus continu-
Side effects Incidents ous epidural analgesia. Spine 1997;22:18926 [Phila Pa 1976]. [discussion 1896-7].
Nausea/Vomiting 18 [26] Cepeda MS, Farrar JT, Roa JH, Boston R, Meng QC, Ruiz F, et al. Ethnicity inuences
Shivering 5 morphine pharmacokinetics and pharmacodynamics. Clin Pharmacol Ther 2001;
Urinary retention 3 70:35161.
Postoperative ileus 1

Please cite this article as: Zhang Z, et al, Nonsteroidal anti-inammatory drugs for postoperative pain control after lumbar spine surgery: A meta-
analysis of randomized cont..., Journal of Clinical Anesthesia (2017), https://doi.org/10.1016/j.jclinane.2017.08.030