The Spine Journal 3 (2003) 55–62
Use of ketorolac tromethamine in children undergoing scoliosis surgery: an analysis of complications Michael G. Vitale, MD, MPHa,c*, Julie C. Choe, MPHa, Matthew W. Hwang, MD, MPHb, Rebecca M. Bauer, MPHb, Joshua E. Hyman, MDc, Francis Y. Lee, MD, PhDc, David P. Roye, Jr., MDc a
International Center for Health Outcomes and Innovative Research, College of Physicians and Surgeons, and the Josepth L. Mailman School of Public Health, Columbia University and New York Presbyterian Hospital, 600 West 168th Street, 7th Floor, New York, NY 10032, USA b Columbia University College of Physicians and Surgeons, 630 West 168th Street, New York, NY 10032, USA c Division of Pediatric Orthopedics, Department of Orthopedic Surgery, Columbia University College of Physicians and Surgeons, 3959 Broadway, 8th Floor North, New York, NY 10032, USA Received 1 March 2002; accepted 20 May 2002
Abstract
Background context: Ketorolac Tromethamine (ketorolac) is a nonsteroidal anti-inflammatory drug (NSAID) with proven efficacy in decreasing postoperative pain in various surgical settings, including the treatment of spine deformities. However, some studies have raised questions regarding the potential side effects of this agent, such as increased bleeding and inhibition of bony fusion. Purpose: This study was conducted to determine whether there is any association between the use of ketorolac and postoperative complications in a group of children who underwent scoliosis surgery. Study design/setting: This is a retrospective review of a group of children who underwent spinal fusion between 1989 to 1999 at our institution. Patient sample: Data on a total of 208 children were analyzed in this study. Sixty received ketorolac and 148 did not. Outcome measures: Postoperative transfusion and reoperation rates were the two main outcome measures of interest. Methods: A retrospective review of 208 children who underwent scoliosis surgery was conducted, with a focus on ketorolac use. Univariate analysis and logistic regression were used to quantify the determinants of postoperative complications. Results: Our analyses detected no significant differences in a broad range of socioclinical variables between the two patient groups, including age at surgery, gender, type of scoliosis, surgical approach, use of erythropoietin, levels of curvature and degree of curvature. Analysis of complication rates focusing on postoperative transfusion and revision surgery showed that there were no significant differences between the two groups. Conclusions: In this retrospective study of 208 children undergoing spine surgery, postoperative use of ketorolac did not significantly increase complications, including transfusion and reoperation. © 2003 Elsevier Science Inc. All rights reserved.
Keywords:
Ketorolac tromethamine; Ketorolac; NSAID; Scoliosis; Children; Complications; Side effects; Transfusion; Revision surgery
Introduction
FDA device/drug status: Approved for this indication (ketorolac tromethamine). Nothing of value received from a commercial entity related to this research. * Corresponding author. 600 West 168th Street, 7th Floor, New York, NY 10032, USA. Tel.: (212) 305-5028; fax: (212) 305-4256. E-mail address: mgv16@yahoo.com (M.G. Vitale)
Ketorolac tromethamine (brand name Toradol®, is a registered trademark of Roche Hoffman-LaRoche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA) is a member of the nonsteroidal anti-inflammatory drug (NSAID) class and is a potent analgesic commonly used for short-term management of postoperative pain. Tissue damage, such as from the trauma of a surgical procedure, causes the local release
1529-9430/03/$ – see front matter © 2003 Elsevier Science Inc. All rights reserved. PII: S1529-9430(02)00 4 4 6 - 1
56
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
of a number of chemical mediators of pain and inflammation, including prostaglandins, leukotrienes, bradykinin, histamine, nitric oxide and norepinephrine, that sensitize nearby nociceptors, which results in the subjective sensation of pain [1,2]. Pharmacotherapeutic agents for pain management often target one of these many mechanisms, and rarely is one agent sufficient for effective control of pain in the postoperative setting. Ketorolac, like other NSAIDs, inhibits the enzyme cyclo-oxygenase (COX). COX converts arachidonic acid to cyclic endoperoxides, which are the unstable precursors of prostaglandins, prostacyclin and thromboxane. One mechanism of NSAID efficacy for analgesia is thought to be this inhibition of local production of prostaglandins, but NSAIDs may have central nervous system effects as well. Ketorolac does not share the adverse effects of opioids, which can include respiratory depression, nausea and vomiting, sedation, gastrointestinal stasis, urinary retention and the potential for abuse. This side effect profile makes ketorolac use particularly appealing for postoperative pain management in children. Ketorolac is an effective alternative adjuvant to morphine during general anesthesia in pediatric surgery [3]. Ketorolac also reduces opioid requirements without increased incidence of bleeding and decreased length of hospital stay after tonsillectomies [4,5] and effectively controls pain and reduces opioid requirements, as well as the length of hospital stay, in children who have undergone ureteroneocystostomies [6,7]. In a prospective, randomized, double-blind study, ketorolac was comparable to morphine for postoperative pain relief in children [8]. Ketorolac also significantly decreases opioid requirements for postoperative pain relief in children [9]. The efficacy of ketorolac in postoperative pain management and reducing narcotic requirements in pediatric surgical populations also applies to pediatric orthopedic patients. In the context of a prospective, randomized, double-blind study, Vetter and Heiner [10] demonstrated that a single intraoperative dose of ketorolac to children undergoing various orthopedic surgical procedures decreases the necessary dose of opioids by means of patient-controlled analgesia and provides better pain control than patient-controlled analgesia alone. Children in this study also experienced decreased frequency of urinary retention secondary to opioid use in the first 12 hours after surgery. Sutters et al. [11] compared the use of postoperative morphine patient-controlled analgesia with and without multiple doses of ketorolac by children who had undergone major orthopedic procedures in a prospective, randomized, double-blind, placebo-controlled study. The authors found that the group of children who had received ketorolac required significantly less opioids, used patientcontrolled analgesia less, had lower pain intensity ratings and experienced less narcotic-associated side effects, such as nausea, pruritus and urinary retention. A case-control study conducted by Eberson, et al. [12] compared ketorolac versus morphine in a postoperative pediatric orthopedic setting and found that the ketorolac group required signifi-
cantly fewer doses of morphine, had fewer gastrointestinal side effects and shorter hospital stays. Despite these documented benefits of ketorolac for pain control, the drug has many potential side effects, including platelet function impairment (secondary to the inhibition of the formation of thromboxane A2), gastrointestinal bleeding and perforation and renal impairment [13]. Studies examining children who have undergone various surgical procedures, such as myringotomy, strabismus correction and reconstructive and orthopedic procedures, have reported no increased risk of bleeding complications. However, ketorolac use in tonsillectomies is controversial because of increased bleeding risk, although its use in this case would be a particularly valuable method of postoperative pain management and avoiding narcotic-associated complications, such as airway obstruction and nausea and vomiting [14,15]. The effects of ketorolac, and NSAIDs in general, on bone metabolism and healing are very controversial and are clearly of particular interest when considering their use in orthopedic patients. The evidence has been conflicting. One study examining the effect of the NSAID ibuprofen on fracture healing in adult rats revealed no differences in the rate of bone remodeling and formation in rats [16], whereas another study presented evidence of an adverse effect on bone healing in rabbits [17]. Ketorolac has been shown to have a dose-dependent inhibitory effect on bone repair in adult rabbits [18]. The inhibitory effects of ketorolac on endochondral ossification were confirmed radiologically in the second to fourth week of fracture healing in rabbits [19]. However, ketorolac was found to have no significant adverse effect on healing within the first 3 days after osteotomy in rats when compared with indomethacin and saline [20]. Animal studies specifically studying the effects of NSAIDs on spinal fusion have shown inhibitory effects as well: indomethacin interferes with spinal fusion in adult rats [21], and ketorolac has been shown to have an inhibitory effect on spinal fusion in adult rabbits [22]. Postoperative ketorolac administration in human adult spine surgery patients also has received mixed reviews thus far. Ketorolac is an effective adjunct to patient-controlled morphine, resulting in decreased morphine use and somnolence and better analgesia [23]. In addition, smaller doses of ketorolac are as effective as larger doses in patients who have had spine stabilization surgery [24]. However, a retrospective study of ketorolac in adult spine surgical patients revealed that patients who received ketorolac were five times more likely than patients who did not to result in nonunion (odds ratio [OR] 5.0) [25]. These studies have generated an appropriate caution regarding ketorolac use in the early postoperative period of spine patients. However, the potential benefits of superior analgesia as well as the decreased need for narcotics warrant further investigation of risk versus benefit of ketorolac use in the pediatric spine patient. There are several strategies commonly used to decrease the postoperative pain associated with the surgical treatment
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
of scoliosis in children, including the use of ketorolac. As in adults, ketorolac use in children raises concerns about increased risk of bleeding and inhibition of bony fusion. To date, few studies have examined clinical outcomes of pediatric patients undergoing major orthopedic procedures and receiving ketorolac for postoperative pain relief. Although empirical evidence suggests that side effects of ketorolac would be less pertinent in a pediatric population, these issues have not been specifically examined in children. Because of the great potential benefit of ketorolac use in the pediatric surgical population, especially in orthopedic patients, we examined the incidence of postoperative complications and their association with ketorolac use in children who underwent surgery for scoliosis. Although a retrospective study design poses various sources of bias, we believe it is a logical first step toward a better understanding of issues in this area. Materials and methods Data A retrospective study of hospital medical records was conducted on 208 patients who underwent scoliosis surgery at our institution from 1989 to 1999. Over this time period, some but not all patients were treated with ketorolac as part of their routine postoperative pain management. Of these patients, 60 children had received ketorolac postoperatively, whereas the remaining 148 children did not. Ketorolac was generally used as a salvage medication when inadequate pain control or complications arose while using narcotics. The decision to use ketorolac was not randomized but rather was made on clinical grounds on a case-by-case basis by the senior author. Furthermore, any child with known allergies to opioids or those who experience nausea, itching or poor pain control on maximum dose of opioids were given ketorolac instead. The average dose of ketorolac administered was 0.5 mg/kg, given intravenously every 6 hours, for 2 to 3 days, starting 24 to 28 hours after surgery. Information regarding socioclinical variables, including age, gender, diagnosis, surgical approach, erythropoietin treatment, reoperation, postoperative hematocrit level, length of stay, length of follow-up, postoperative gastrointestinal complications, level of curvature and degree of curvature, was collected on all 208 patients. The endpoints of interest involved two types of complications: 1) increased risk of bleeding, as represented by the need for transfusion, and 2) revision surgery because of pseudarthrosis, infected hardware or hardware failure. It should be noted that although we are interested in the effect of ketorolac on nonunion, other reasons for reoperation, such as infection or decompensation, were also included. Analysis Univariate analyses were performed to quantify the number of patients in each socioclinical category. Table 1 illustrates the demographic and clinical profile of our patient population.
57
Table 1 Demographic and clinical characteristics of ketorolac versus nonketorolac patients
Gender Male Female Diagnosis Idiopathic scoliosis Neurogenic scoliosis Congenital scoliosis Kyphosis Other Surgical approach Posterior Anterior Combined Erythropoietin treatment No Yes Reoperation No Yes Type of reoperation Nonunion Infection Decompensation Gastrointestinal problems* No Yes Transfusion No Yes Level of curvature (mean SD) Degree of curvature (mean SD) Hematocrit level† (mean SD) Length of stay (mean SD) Length of follow-up (mean SD) Age at surgery (mean SD)
Ketorolac (n 60) %
No ketorolac (n 148) %
21.7 78.3
31.8 68.2
55.0 33.3 8.3 1.7 1.7
45.9 27.0 18.9 3.4 4.7
56.7 13.3 30.0
48.6 13.5 37.8
46.7 53.3
52.3 47.7
86.7 13.3
87.1 12.9
37.5 25.0 37.5
52.6 36.8 10.5
86.7 13.3
93.2 6.8
73.3 26.7 9.8 3.5 65.7 22.7 28.8 5.9 10.1 10.8 61.7 23.3 13.5 2.5
78.4 21.6 9.1 4.2 58.4 19.7 28.7 5.2 8.0 5.2 68.7 31.1 13.3 4.3
*Postoperative gastrointestinal complications. † Postoperative hematocrit level.
Bivariate analyses were conducted between the two endpoints of interest (or dependent variables)—the need for transfusion and reoperation—and the various socioclinical variables collected on each study patient. Tables 2 and 3 show the results of the crude OR method of comparison for transfusion and reoperation, respectively. The results of the final logistic regression analysis are shown in Table 4 and 5. For each of these tables, the same seven independent variables were entered into the regression model: ketorolac treatment, erythropoietin treatment, diagnosis (type of scoliosis), surgical approach, length of stay, level of curvature and degree of curvature.
Results Descriptive data Table 1 illustrates the demographic and clinical characteristics of our patient population, stratified by postopera-
58
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
Table 2 The odds of transfusion (crude odds ratio)
Ketorolac No Yes Gender Male Female Diagnosis Idiopathic scoliosis Neurogenic scoliosis Congenital scoliosis Kyphosis Other Surgical approach Anterior Posterior Combined Erythropoietin treatment No Yes Reoperation No Yes Gastrointestinal problems† No Yes Hematocrit level‡ Level of curvature§ Degree of curvature Length of stay¶ Age at surgery# Length of follow-up**
Transfusion (n 48)
No transfusion (n 160)
Crude odds ratio
95% Confidence interval
p value
32 16
116 44
1.00 1.32
— 0.66–2.64
— .44
17 31
43 117
1.00 0.67
— 0.34–1.33
— .25
12 28 7 0 1
89 32 26 6 7
1.00 5.60 0.88 NA* 0.47
— 2.80–11.20 0.36–2.18 NA* 0.06–3.88
— .0001 .78 NA* .48
7 18 23
21 88 51
1.00 0.49 1.97
— 0.25–0.95 1.02–3.79
— .04 .04
26 12
51 63
1.00 0.37
— 0.17–0.81
— .013
41 6
139 21
1.00 0.97
— 0.59–1.58
— .91
43 4 — — — — — —
145 14 — — — — — —
1.00 0.97 0.98 1.12 1.05 1.07 1.003 0.99
— 0.55–1.74 0.92–1.04 1.02–1.22 1.03–1.07 1.02–1.12 0.92–1.09 0.87–1.13
— .93 .45 .02 .0001 .006 .95 .89
*None of the kyphosis cases underwent a transfusion. Postoperative gastrointestinal complications. ‡ Postoperative hematocrit level; estimate for each additional unit increase in hematocrit level. § Estimate for each additional level of increase in curvature involvement. Estimate for each additional degree increase in curvature. ¶ Estimate for each additional day in hospital. # Estimate for each additional year increase in age. **Estimate for each additional month of follow-up. †
tive ketorolac treatment. Of the 208 patients in this study, 60 (29%) were identified as having been treated with ketorolac, whereas 148 (71%) did not receive this drug. The overall patient population was predominantly female, with an average age at surgery of 13.4 years (range, 1.8 to 28.6). Approximately half of the children in each patient group were diagnosed with idiopathic scoliosis and about another third with neurogenic scoliosis. The number of levels of spinal curvature involved was similar in the two groups, but the degree of curvature was higher among ketorolac-treated patients, with an average curvature of 66 degrees (range, 26 to 130 degrees) versus an average of 58 degrees (range, 28 to 110 degrees) among non–ketorolac-treated patients. More than half (53%) of those who received ketorolac were also treated with erythropoietin, whereas less than half (47%) received this drug in the nonketorolac group. The mean postoperative hematocrit level was similar in the two patient groups. The posterior approach was most widely
conducted among these patients. However, the average length of stay was lower in the nonketorolac group with a mean of 8.0 days (range, 1 to 39 days), compared with 10.1 days (range, 4 to 62 days) in the ketorolac group. The length of follow-up was longer in patients who did not receive ketorolac, with a mean of 68.71 months (range, 17.54 to 124.06), compared with 61.67 months (range, 16.69 to 98.10) in children who did not receive this drug. Slightly more patients treated with ketorolac (27%) needed to be transfused, compared with patients in the nonketorolac group (22%), but this was not statistically significant. During this time, approximately 13% in each patient group underwent some type of revision surgery. Bivariate analysis The crude ORs for transfusion and reoperation are shown in Tables 2 and 3, respectively. The reference group for each independent variable was chosen at random and assigned an
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
59
Table 3 The odds of reoperation (crude odds ratio)
Ketorolac No Yes Gender Male Female Diagnosis Idiopathic scoliosis Neurogenic scoliosis Congenital scoliosis Kyphosis Other Surgical approach Anterior Posterior Combined Erythropoietin treatment No Yes Gastrointestinal problems* No Yes Transfusion No Yes Hematocrit level† Level of curvature‡ Degree of curvature§ Length of stay Age at surgery¶ Length of follow-up#
Reoperation (n 27)
No reoperation (n 180)
Crude odds ratio
95% Confidence interval
p value
19 8
128 52
1.00 1.04
— 0.43–2.52
— .94
10 17
49 131
1.00 0.64
— 0.27–1.48
— .29
8 10 7 1 1
93 49 26 5 7
1.00 1.57 2.07 1.35 0.95
— 0.67–3.67 0.80–5.39 0.15–11.98 0.11–8.04
— .29 .13 .79 .96
6 14 7
22 92 66
1.00 1.03 0.61
— 0.46–2.31 0.24–1.51
— .94 .28
7 13
70 62
1.00 2.10
— 0.79–5.59
— .14
23 3
165 15
1.00 1.17
— 0.61–2.26
— .63
21 6 — — — — — —
139 41 — — — — — —
1.00 0.97 1.02 1.04 1.03 1.04 0.98 0.99
— 0.37–2.56 0.94–1.10 0.93–1.16 1.004–1.05 0.99–1.08 0.88–1.09 0.85–1.17
— .95 .71 .50 .02 .10 .70 .94
*Postoperative gastrointestinal complications. † Postoperative hematocrit level; estimate for each additional unit increase in hematocrit level. ‡ Estimate for each additional level of increase in curvature involvement. § Estimate for each additional degree increase in curvature. Estimate for each additional day in hospital. ¶ Estimate for each additional year increase in age. # Estimate for each additional month of follow-up.
OR of 1.00. For instance, in Table 2, the reference group for diagnosis was patients with idiopathic scoliosis. Other types of diagnoses (neurogenic, congenital, kyphosis, other) were compared with the idiopathic cases separately. In Table 2, six independent variables were determined to be significant in predicting transfusion: diagnosis (neurogenic scoliosis cases), surgical approach (both posterior and combined methods), levels of curvature, degree of curvature, length of stay in the hospital after surgery and erythropoietin treatment. More specifically, neurogenic cases were more than five times as likely to undergo a transfusion than idiopathic cases (crude OR 5.6, 95% confidence interval [CI] 2.8 to 11.2). Level of curvature involved (crude OR 1.12/level, 95% CI 1.02 to 1.22) was also a significant predictor of transfusion. Similarly, degree of curvature, estimated for each degree, was a significant independent factor (crude OR 1.05/degree, 95% CI 1.03 to 1.07). In terms of surgical approach, the posterior approach was protective against transfusions in comparison
to the anterior approach (crude OR 0.49, 95% CI 0.25 to 0.950). On the other hand, patients who underwent a combined anterior-posterior approach were almost twice as likely to receive a transfusion compared with those who underwent only an anterior approach (crude OR 1.97, 95% CI 1.02 to 3.79). Erythropoietin was a protective factor against transfusion in our study, with about 70% less likely to require a transfusion (crude OR 0.37, 95% CI 0.17 to 0.81). Finally, transfusion led to a longer stay in the hospital in this setting (crude OR 1.07, 95% CI 1.02 to 1.12). In determining the likelihood for reoperation, as shown in Table 3, only one independent variable was found to be significant: degree of curvature. Degree of curvature was significant in predicting reoperation, without adjusting for any other variables (crude OR 1.03, 95% CI 1.004 to 1.05). Treatment with ketorolac was not a significant independent predictor for either transfusion or reoperation in our patient population.
60
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62 Table 5 Final logistic regression model for reoperation
Table 4 Final logistic regression model for transfusion Adjusted odds ratio Ketorolac No Yes Erythropoietin treatment No Yes Diagnosis Idiopathic scoliosis Neurogenic scoliosis Congenital scoliosis Surgical approach Anterior Posterior Combined Length of stay* Level of curvature† Degree of curvature‡
95% Confidence interval
p value
1.00 1.06
— 0.40–2.78
— .91
1.00 0.52
— 0.21–1.30
— .16
1.00 2.71 3.67
— 0.94–7.83 0.92–14.60
— .07 .06
1.00 1.30 1.27 1.03 1.11 1.03
— 0.29–5.73 0.27–5.99 0.98–1.10 0.97–1.26 1.004–1.06
— .73 .76 .22 .13 .02
*Estimate for each additional day in hospital. Estimate for each additional level of increase in curvature involvement. ‡ Estimate for each additional degree increase in curvature. †
Multivariate analysis Multivariate analyses were conducted in order to adjust for possible confounding and/or interaction effects among the many independent variables in this study. In Table 4, the six factors that were significant for transfusion from the crude analysis in Table 2 were included in the final logistic regression model. Given that ketorolac is the main factor of interest in this study, this seventh variable was also entered into the final model for transfusion. According to this adjusted analysis, degree of curvature (adjusted OR 1.03, 95% CI 1.004 to 1.06) remained as the only significant variable predicting transfusion, even after adjusting for confounding and interaction. Ketorolac treatment did not affect the rate of transfusion. In Table 3, only a single independent variable was significant in predicting reoperation:degree of curvature. However, the same seven factors from Table 4 were entered into the final logistic regression model for reoperation in Table 5. Again, ketorolac treatment did not significantly influence the need for reoperation, even after adjusting for other variables in the model.
Discussion The primary goal of this study was to evaluate the potential inhibitory effects of systemic ketorolac on spine fusions, as well as the potential postoperative bleeding complications in a pediatric population. To our knowledge, this is the largest study to date examining the complications of administering ketorolac postoperatively in children undergoing spine surgery. Although we recognize the limits of a retrospective study, we thought that our study was a rational first step toward a de-
Variable Ketorolac No Yes Erythropoietin treatment No Yes Diagnosis Idiopathic scoliosis Neurogenic scoliosis Congenital scoliosis Surgical approach Anterior Posterior Combined Length of stay* Level of curvature† Degree of curvature‡
Adjusted odds ratio
95% Confidence interval
p value
1.00 0.99
— 0.32–3.10
— .99
1.00 3.79
— 1.17–12.27
— .03
1.00 1.34 1.27
— 0.34–5.30 0.22–7.27
— .68 .79
1.00 0.45 0.62 1.02 0.99 1.02
— 0.09–2.19 0.13–3.09 0.96–1.07 0.84–1.17 0.99–1.06
— .32 .56 .55 .94 .21
*Estimate for each additional day in hospital. Estimate for each additional level of increase in curvature involvement. ‡ Estimate for each additional degree increase in curvature. †
finitive answer to the question of the safety of ketorolac use in pediatric spine surgery patients. As mentioned earlier, the decision to administer ketorolac postoperatively was made by the senior author on a case-by-case basis. The decision was made based on the child’s known allergies to opioids, nausea, itching, poor pain control, somnolence and respiratory depression. Although the children who received ketorolac constituted the more difficult cases in our study, thus limiting the generalizability of the study, we did not find a significantly increased rate of complications in these patients. Ideally, a randomized controlled trial will shed further light on the subject. Based on the known potential complications secondary to ketorolac use, we chose postoperative transfusion and reoperation rates as our two main measures of interest. We chose transfusion, because we thought that it was a reasonable indicator of bleeding that we could measure accurately in a retrospective fashion. Also, the decision to transfuse a patient was based on the senior author’s assessment of bleeding complications. Therefore, we thought that postoperative transfusion would serve well as a proxy for such. Reoperation rates captured those patients who had to undergo a revision surgery for pseudarthrosis, infected hardware or hardware failure. We specifically analyzed the data for pseudarthrosis rates as an indicator for risk of inhibition of bony fusion and found that there was no significant difference between the two patient groups. One potential major limitation of this endpoint as an indicator for nonunion is the length of postoperative follow-up. Our average length of follow-up was almost 62 months for children who received ketorolac and 69 months for those who did not. We thought that this was acceptable, given both the decreased likelihood of children compared with adults to experience nonunion and the retrospective nature of the study.
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
Although the use of ketorolac was not randomized in this study, our analysis detected no significant differences in a broad range of socioclinical variables. Our data revealed a slightly larger proportion of children who received ketorolac did require transfusion; however, the difference was not statistically significant. We also found no difference in the incidence of revision surgery between the two patient groups. However, degree of curvature was a significant predictor for transfusion, whereas erythropoietin treatment was a significant predictor for reoperation in this study according to the final regression models for these two endpoints, respectively. These are concerning endpoints, but given that the decision to treat with ketorolac was not randomized, and that the patient population who received ketorolac was composed primarily of complicated patients, we think that our conclusions about the safety of ketorolac use in this population are valid. A larger, randomized controlled study will provide insight into these endpoints. Overall, these data suggest that postoperative use of ketorolac in children with scoliosis surgery does not significantly increase bleeding and nonunion complications, but a larger population is needed to exclude definitively any association. The results from our retrospective study indicate that further research into the use of ketorolac in the pediatric orthopedic surgical patient population is warranted. Animal studies that have examined the effects of ketorolac on bone healing have largely been conducted using adult animals. One possible solution is to study the effects of systemic ketorolac administration on spine fusion or long bone fracture healing using adolescent animal models. By quantifying the inhibitory effect of ketorolac on bone healing in immature animals, we can clarify further the appropriateness of ketorolac use in pediatric orthopedic patients. Conclusions Our retrospective study indicates that the use of ketorolac for postoperative pain management does not increase risk of bleeding complications or reoperation in a group of children who underwent spine surgery. On the other hand, the degree of preoperative curvature seems to be a significant determinant of transfusion risk. Several other interesting findings were noted in this study. As found in previous studies, children with neurogenic scoliosis tend to have a significantly greater risk of transfusion, as do children with larger curves. Based on the results of our retrospective study, use of ketorolac did not seem to increase this risk, even in these groups of patients. Our next logical step in clinical investigation is the initiation of a prospective, randomized, double-blind, multicenter study, which will allow us to determine definitively whether ketorolac has clinically significant inhibitory effects on spine fusion in children. Included in this follow-up clinical study would be an examination of the commonly accepted effects of ketorolac, potential positive and negative patient outcomes, as well as the potential savings incurred from de-
61
creased incidence of opioid-induced side effects. However, this study provides evidence to support our current use of ketorolac as an adjunct to postoperative pain management in patients after spine surgery, and we continue to use this medication routinely.
References [1] Power I, Barratt S. Analgesic agents for the postoperative period. Nonopiods. Surg Clin North Am 1999;79(2):275–95. [2] Gillis JC, Brogden RN. Ketorolac. A repraisal of its pharmacodynamic and pharmacokinetic properties and therapeutic use in pain management. Drugs 1997;53(1):139–88. [3] Watcha MF, Jones MB, Lagueruela RG, Schweiger C, White PF. Comparison of ketorolac and morphine as adjuvants during pediatric surgery. Anesthesiology 1992;76:368–72. [4] Sutters KA, Levine JD, Dibble S, Savedra M, Miaskowski C. Analgesic efficacy and safety of single-dose intramuscular ketorolac for postoperative pain management in children following tonsillectomy. Pain 1995;61(1):145–53. [5] Agrawal A, Gerson CR, Seligman I, Dsida RM. Postoperative hemorrhage after tonsillectomy: use of ketorolac tromethamine. Otolaryngol Head Neck Surg 1999;120(3):335–9. [6] Gonzalez A, Smith DP. Minimizing hospital length of stay in children undergoing ureteroneocystostomy. Urology 1998;52(3):581–4. [7] Chauhan RD, Idom CB, Noe HN. Safety of ketorolac in the pediatric population after ureteroneocystostomy. J Urol 2001;166:1873–5. [8] Lieh-Lai MW, Kauffman RE, Uy HG, Danjin M, Simpson PM. A randomized comparison of ketorolac tromethamine and morphine for postoperative analgesia in critically ill children. Crit Care Med 1999; 27(12):2786–91. [9] Carney DE, Nicolette LA, Ratner MH, Minerd A, Baesl TJ. Ketorolac reduces postoperative narcotic requirements. J Pediatric Surg 2001;36(1):76–9. [10] Vetter TR, Heiner EJ. Intravenous ketorolac as an adjuvant to pediatric patient-controlled analgesia with morphine. J Clin Anesthesia 1994;6:110–3. [11] Sutters KA, Shaw BA, Gerardi JA, Herbert D. Comparison of morphine patient-controlled analgesia with and without ketorolac for postoperative analgesia in pediatric orthopaedic surgery. Am J Orthopaed 1999;28(6):351–8. [12] Eberson CP, Pacicca DM, Ehrlich MG. The role of ketorolac in decreasing length of stay and narcotic complications in the postoperative pediatric orthopaedic patient. J Pediatric Orthopaed 1999;19:688–92. [13] Reinhart DJ. Minimizing the adverse effects of ketorolac. Drug Safety 2000;22(6):487–97. [14] Rusy LM, Houck CS, Sullivan LJ, et al. A double-blind evaluation of ketorolac tromethamine versus acetaminophen in pediatric tonsillectomy: analgesia and bleeding. Anesthesia Analgesia 1995;80:226–229. [15] Gunter JB, Varughese AM, Harrington JF, et al. Recovery and complications after tonsillectomy in children: a comparison of ketorolac and morphine. Anesthesia Analgesia 1995;81(6):1136–41. [16] Huo MH, Troiano NW, Pelker RR, Gundberg CM, Friedlaender GE. The influence of ibuprofen on fracture repair: biomechanical, biochemical, histologic, and histomorphometric parameters in rats. J Orthopaed Res 1991;9:383–90. [17] Obeid G, Zhang X, Wang X. Effect of ibuprofen on the healing and remodeling of bone and articular cartilage in the rabbit temporomandibular joint. J Oral Maxillofacial Surg 1992;50:843–9. [18] Ho M-L, Chang J-K, Wang G-J. Anti-inflammatory drug effects on bone repair and remodeling in rabbits. Clin Orthopaed Rel Res 1995; 313:270–8. [19] Ho M-L, Chang J-K, Wang G-J. Effects of ketorolac on bone repair: a radiographic study in modeled demineralized bone matrix grafted rabbits. Pharmacology 1998;57:148–59.
62
M.G. Vitale et al. / The Spine Journal 3 (2003) 55–62
[20] Reikeraas O, Engebretsen L. Effects of ketorolac tromethamine and indomethacin on primary and secondary bone healing. An experimental study in rats. Arch Orthopaed Trauma Surg 1998;118:50–2. [21] Dimar JR, Ante WA, Zhang YP, Glassman SD. The effects of nonsteroidal anti-inflammatory drugs on posterior spinal fusions in the rat. Spine 1996;21:1870–6. [22] Martin GJ, Boden SD, Titus L. Recombinant human bone morphogenetic protein-2 overcomes the inhibitory effect of ketorolac, a nonsteroidal anti-inflammatory drug (NSAID), on posterolateral lumbar intertransverse process spinal fusion. Spine 1999;24:2188–94.
One Hundred Years Ago in Spine . . . The 1903 Nobel Prize for Physics was awarded to Antoine Becquerel (1852–1908), Marie Curie (1867–1934) and Pierre Curie (1859–1906). Marie Curie won a second Nobel Prize, for Chemistry, in 1911. They
[23] Reuben SS, Connelly NR, Steinberg R. Ketorolac as an adjunct to patient-controlled morphine in postoperative spine surgery patients. Reg Anesthesia 1997;22(4):343–6. [24] 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. Anesthesia Analgesia 1998;87: 98–102. [25] Glassman SD, Rose SM, Dimar JR, Puno RM, Campbell MJ, Johnson JR. The effect of postoperative nonsteroidal anti-inflammatory drug administration on spinal fusion. Spine 1998;23(7):834–838.
isolated radium from the study of pitchblend, a material rich in uranium and thorium, and showed it to be extremely radioactive. References [1] Becquerel AH. Sur les radiations émisesparphosphorescence. CR Acad Sci (Paris) 1896;10:420–1. [2] Curie P, Curie MS. Sur une substance nouvelle radio-active, continue dans la pitchblend. CR Acad Sci (Paris) 1898;127:175–8,1215–7.
David Fardon Knoxville, TN