- Email: [email protected]
Safety of perioperative ketorolac administration in pediatric appendectomy
j o u r n a l o f s u r g i c a l r e s e a r c h o c t o b e r 2 0 1 7 ( 2 1 8 ) 2 3 2 e2 3 6
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Association for Academic Surgery
Safety of perioperative ketorolac administration in pediatric appendectomy Hibbut-ur-Rauf Naseem, MD,a,* Robert Michael Dorman, MD,a,b George Ventro, MD,a,b David H. Rothstein, MD, MS,a,b and Kaveh Vali, MDa,b a b
Department of Pediatric Surgery, Women and Children’s Hospital of Buffalo, Buffalo, New York Department of Surgery, University at Buffalo, State University of New York, Buffalo, New York
article info
abstract
Article history:
Background: Recent studies in adults undergoing gastrointestinal surgeries show an
Received 12 February 2017
increased rate of complications with the use of ketorolac. This calls into question the safety
Received in revised form
of ketorolac in certain procedures. We sought to evaluate the impact of perioperative
20 April 2017
ketorolac administration on outcomes in pediatric appendectomy.
Accepted 24 May 2017
Methods: The Pediatric Health Information System database was queried for patients aged
Available online xxx
5-17 y with a primary diagnosis of appendicitis and a primary procedure of appendectomy during the period 2010-2014. Patients with procedures suggesting incidental appendec-
Keywords:
tomy, those records with data quality issues, deaths, and extracorporeal membrane
Primary appendicitis
oxygenation were excluded. Variables recorded included age, sex, race, ethnicity, discharge
Nonsteroidal anti-inflammatory
year, complex chronic conditions, geographic region, intensive care unit admission, me-
drugs Pediatric surgery
chanical ventilation, and whether appendicitis was coded as complicated. The exposure variable was ketorolac administration on the day of or day after operation. The primary outcomes of interest were any surgical complications during the initial encounter, postoperative length of stay (LOS), total cost for the initial visit, any readmission to ambulatory, observation, or inpatient status within 30 d, and readmission with a diagnosis of peritoneal abscess or other postoperative infection or with transabdominal drainage performed. Results: A total of 78,926 patients were included in the analysis cohort. Mean age was 11.4 y (standard deviation 3.3 y), the majority were males (61%), White (70%), and non-Hispanic (65%). Few had a complex chronic condition (3%) or required mechanical ventilation (2%) or an intensive care unit admission (1%). Patients with complicated appendicitis comprised 28% of the cohort. Most (73%) received ketorolac on postoperative day 0-1; those with complicated appendicitis were more likely to receive ketorolac. In all, 2.6% of the cohort had a surgical complication during the index visit, 4.3% were readmitted within 30 d, and 2% had a postoperative infection or transabdominal drainage (1% in the uncomplicated group and 5% in the complicated group). Median postoperative LOS was 1 d and mean cost was $9811 $9509. On bivariate analysis, ketorolac administration was associated with a decrease in same-visit surgical complications (P ¼ 0.004) and cost ($459 decrease, P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection. On
* Corresponding author. Department of Pediatric Surgery, Women and Children’s Hospital of Buffalo, 219 Bryant Street, Buffalo, NY 14222. Tel.: þ1 (479) 595 2817; fax: þ1 (716) 888 3850. E-mail address: [email protected] (H.-u. Naseem). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.05.087
naseem et al safety of perioperative ketorolac administration in pediatric appendectomy
233
multivariate analysis, ketorolac administration was associated with a significant decrease in any complication (adjusted odds ratio 0.89, 95% confidence interval 0.80-0.99) and cost (analysis of variance P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection. Conclusions: Based on a large, contemporary data set from children’s hospitals, ketorolac administration in the immediate postoperative period after appendectomy for appendicitis is common and was not associated with an increase in postoperative LOS, postoperative infection, or any-cause 30-d readmission. Ketorolac was, however, independently associated with a lower overall rate of postoperative complications and cost in this population. ª 2017 Elsevier Inc. All rights reserved.
Introduction Acute appendicitis is the most common gastrointestinal surgical urgency in the pediatric population.1 Several recent publications have supported same day discharge as safe and appropriate after laparoscopic appendectomy for uncomplicated appendicitis.2,3 One of the barriers to same day discharge is postoperative pain control.3,4 With increasing scrutiny of the use of narcotic pain control, there is a greater need for alternatives.5 Ketorolac is a well-known nonsteroidal antiinflammatory drug and has been FDA approved for the management of acute pain since 1989.6 The bleeding risks of ketorolac and NSAIDS have been well studied.7-9 Ketorolac has been theorized to impair wound healing and scar tissue formation through its anti-inflammatory mechanisms, which has been demonstrated in several animal studies.10,11 In addition, a recent study in adults undergoing gastrointestinal surgeries by M. Kotagal et al. showed an increased rate of complications ranging from increased emergency department visits, increased readmission, and increased risk of reintervention, calling into question the safety of ketorolac in certain procedures.12 We aimed to reproduce these findings among pediatric patients undergoing appendectomy. We hypothesized that ketorolac would be an independent predictor of increased complications such as abscess formation, readmission, and reoperation.
Methods We conducted a retrospective cohort study of Pediatric Health Information System (PHIS) database, a comparative pediatric database administered by the Children’s Hospital Association (Overland Park, KS) that collects clinical and resource utilization data in inpatient, ambulatory, and emergency department settings at over 45 US children’s hospitals. We included encounters for subjects aged 5-17 y who were discharged between 2010 and 2014 with a primary diagnosis of appendicitis (International Classification of Diseases, Ninth Revision [ICD-9] codes 540, 540.0, 540.1, 540.9) who had an appendectomy coded (ICD-9 procedure codes 47.0, 47.01, 47.09). We excluded incidental appendectomy (47.11 and 47.19) as well as deaths, extracorporeal membrane oxygenation (ECMO), and records with data quality issue because of hospital day and operative day being mismatched and therefore creating inaccurate perioperative timeline for ketorolac exposure. Deaths and ECMO were excluded to minimize confounding factors in the clinical
decision to prescribe ketorolac. We assumed that both these outcomes were because of severity of disease and comorbidities rather than ketorolac administration. Variables recorded included age, sex, race, ethnicity, discharge year, complex chronic conditions (CCC), geographic region, intensive care unit (ICU) admission, mechanical ventilation, and whether appendicitis was coded as complicated. The primary predictor was ketorolac administration on postoperative day 0 or day 1. The primary outcome of interest was readmission with postoperative infection within 30 d. Secondary outcomes were surgical complications during the initial encounter, postoperative length of stay (LOS), total cost for the initial visit, and any readmission to ambulatory, observation, or inpatient status within 30 d. “Any surgical complication” was determined by the presence of a PHIS-generated surgical complication flag for that visit, which was triggered by the presence of one of a number of diagnosis codes chosen by PHIS. These included ICD-9 codes 99x (complications of surgical and medical care, necrotizing enterocolitis, including device, transplant, and transfusion-related complications), as well as complications specific to certain procedures, respiratory failure after trauma and surgery (518.5x), obstetric complications, and postprocedural fever (780.62). The total adjusted costs were based on the reported charges, adjusted for the Ratio of Cost to Charges submitted by each hospital on their respective Medicare cost reports, then adjusted by the Centers for Medicare and Medicaid Services wage/price index for the hospital’s location (13). The postoperative infection on readmission was determined when a peritoneal abscess (ICD-9 code of 567.22), generic postoperative infection (ICD-9 diagnosis code 998.5 or 998.59), or percutaneous abdominal drainage (ICD-9 procedure code 54.91) were coded. Descriptive statistics and bivariate comparisons used ttest, Fisher exact, and Pearson Chi-squared tests where appropriate. Multivariate analysis of binary outcomes used a binary logistic regression model and multivariate analysis of continuous outcomes used a general linear model. Statistics performed using Minitab version 17.3.1. Adjusted odds ratios (aOR) were calculated with corresponding 95% confidence intervals (CI).
Results A total of 79,509 encounters met inclusion criteria. A total of 583 (0.7%) were excluded including 455 for incidental
234
j o u r n a l o f s u r g i c a l r e s e a r c h o c t o b e r 2 0 1 7 ( 2 1 8 ) 2 3 2 e2 3 6
Table 1 e Demographics and outcomes for cohort with bivariate analysis of complicated appendicitis and ketorolac administration. Variable Age, y (mean)
Cohort
Complicated
P
Ketorolac
11.4 3.3
10.8 3.3
<0.001
11.4 3.3
P 0.087
5-11 y
57%
64%
<0.001
57%
0.554
Male
61%
60%
0.003
60%
<0.001
White
70%
65%
<0.001
70%
<0.001
Black
8%
9%
<0.001
7%
<0.001
Other
22%
26%
<0.001
22%
0.213
Hispanic ethnicity
36%
41%
<0.001
38%
<0.001
2010-2011
37%
36%
0.017
36%
<0.001
3%
4%
<0.001
3%
<0.001
South
43%
43%
0.028
45%
<0.001
West
25%
29%
<0.001
28%
<0.001
Northeast
11%
10%
<0.001
9%
<0.001
Midwest
20%
18%
<0.001
18%
<0.001
Mechanical ventilation
2%
2%
0.001
1%
<0.001
Any ICU admission
1%
2%
<0.001
1%
<0.001 <0.001
CCC
Complicated appendicitis
28%
e
e
29%
Ketorolac POD 0-1
73%
75%
<0.001
e
3%
7%
<0.001
3%
Adjusted cost, $
9811 9509
15,694 15,013
<0.001
9689 7224
Postoperative LOS, d
Same-visit surgical complication
e 0.004 <0.001
2.29 13.39
4.90 4.28
<0.001
2.23 2.56
0.192
Any readmission within 30 d
4%
8%
<0.001
4%
0.364
Readmission with intra-abdominal infection within 30 d
2%
5%
<0.001
2%
0.144
appendectomy, 126 for data quality issues, two for death, and three for ECMO. A total of 78,926 were included in the analysis cohort. The mean age of patients undergoing appendectomy was 11.4 y (SD 3.3 y) with a slight skew to the younger age group (57% ages 5-11 y). The majority was male (61%), White (70%), and non-Hispanic (64%). A larger percentage of were in the South region of the United States (43%). Few had a CCC (3%) or required mechanical ventilation (2%) or an ICU admission (1%). In all, 2.6% of the cohort had a surgical complication during the index visit, 4.3% were readmitted within 30 d, and 2% had a postoperative infection or transabdominal drainage (1% in the uncomplicated group and 5% in the complicated group). Median postoperative LOS was 1 d (interquartile range [IQR] 1-3) and mean cost was $9811 $9509 (Table 1). Patients with complicated appendicitis comprised 28% of the cohort and were slightly younger (10.8 versus 11.4 y, P < 0.001), more often listed as Hispanic ethnicity (41% versus 36%, P < 0.001), with a slight preponderance to having a CCC (4% versus 3%, P < 0.001). On bivariate analysis, patients with complicated appendicitis had a higher rate of ketorolac administration on POD 0-1 (75% versus 73%, P < 0.001), with a slightly higher rate of ICU admission (2% versus 1%, P < 0.001) and more frequent same-visit surgical complications (7% versus 3%, P < 0.001). Patients with complicated appendicitis also had significantly higher adjusted cost ($15,694 versus $9,811, P < 0.001), median postoperative LOS (4 [IQR 3-6] versus
1 [IQR 1-1] d, P < 0.001), readmission within 30 d (8% versus 4%, P < 0.001), and readmission with intra-abdominal infection within 30 d (5% versus 2%, P < 0.001; Table 1). Most (73%) of the cohort received ketorolac on postoperative day 0-1; on bivariate analysis, ketorolac administration was associated with Hispanic ethnicity, variability by region of United States, decreased association with mechanical ventilation (1% versus 2%, P < 0.001), and increased use in later years (74.1% in 2012-2014 versus 71.2% in 2010-2011, P < 0.001). Ketorolac was associated with a decrease in same-visit surgical complications (2.9% versus 2.5%, P ¼ 0.004) and cost ($459 decrease, P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection (Table 1). On multivariate analysis, males (aOR 1.2, CI 95% [1.1, 1.3]), Hispanic ethnicity (aOR 1.5, CI [1.0, 1.3]), and complicated appendicitis (aOR 4.0, 95% CI [3.6, 4.5]) had greater odds of readmission with intra-abdominal infection within 30 d. Ketorolac administration on POD 0-1 did not have greater odds of readmission with intra-abdominal infection within 30 d. On separate multivariate analyses, ketorolac was associated with a significant decrease in any complication (aOR 0.89, 95% CI [0.80-0.99]) and cost (analysis of variance P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection (Table 2). In a sensitivity analysis of ketorolac given only on postoperative day 0 (62% of the cohort), qualitatively identical results were obtained.
naseem et al safety of perioperative ketorolac administration in pediatric appendectomy
Table 2 e Multivariate analysisepredictors of readmission with postoperative infection. Variable
Variable subgroup
OR (95% CI)
Age 5-11 y
1.04 (0.93, 1.15)
Male
1.23 (1.10, 1.37)
Race
Black
Reference
White
1.07 (0.87, 1.31)
Other
1.18 (0.93, 1.49)
Hispanic ethnicity
1.15 (1.02, 1.30)
2010-2011
1.01 (0.91, 1.12)
CCC
1.02 (0.77, 1.34)
US region
Midwest
Reference
Northeast
0.70 (0.57, 0.87)
South
0.93 (0.80, 1.08)
West
0.75 (0.63, 0.88)
Mechanical ventilation
1.14 (0.79, 1.66)
Any ICU admission
0.78 (0.50, 1.20)
Complicated appendicitis
4.06 (3.65, 4.52)
Ketorolac POD 0-1
1.02 (0.91, 1.15)
235
effects of ketorolac. In a meta-analysis of six randomized controlled trials with 480 patients, ketorolac did not show a significant difference in anastomotic dehiscence.13 The authors concluded no difference but because of insufficient power. In multiple retrospective observational studies, non-steroidal anti-inflammatory drugs such as ketorolac have shown to have a detrimental effect on anastomotic dehiscence.17-19 Ketorolac use in our study showed a statistically significant decrease in cost of $459 but this amount would only make a small dent in the greater than $9000 cost of hospitalization after an appendectomy. There was no significant difference in LOS with ketorolac use. This correlates well with a study by Manworren et al. that showed decreased morphine equivalents and pain scores with ketorolac use that did not produce the expected decrease in LOS.14 Similar to Kotagal et al., we used a national database (pediatric instead of adult). We noted patterns of regional variation in the use of ketorolac, which may be due to random variation or may correlate with aversion due to cost or perceived risk of the drug. The Hispanic population was more likely to receive the drug as well, which may correlate with the regional variation of more frequent ketorolac use in the South region of the United States. These conjectures may explain the trend but further research needs to assess the use of analgesics across different demographics.
Discussion Limitations Findings In this large pediatric database study, we found that ketorolac administration on POD 0 or 1 after appendectomy did not show increased complications including readmission, postoperative infection, and reinterventions within 30 d. This also applied when Ketorolac was administered on POD 0 only. Ketorolac administration was also associated with fewer same-visit complications and cost. These results are contrary to the study by Kotagal et al., which showed increased rates of emergency department visits, readmissions, and reinterventions associated with GI operations involving anastomosis. While adult studies show a range of 15%-30% ketorolac use after gastrointestinal surgery,12,13 recent pediatric studies show rates closer to 50% after appendectomy.14 Our study had an overall 73% use of ketorolac in the perioperative period, with an increase over the time period. This shows that despite the concern for ketorolac-related complications in gastrointestinal surgery, most surgeons do not view this as translated to pediatric appendectomy patients. About one-quarter of the patients did not receive ketorolac, which may have been for a variety of reasons including concern for intraoperative bleeding, surgeon hesitation because of perceived risk of ketorolac, and other medical contraindications to ketorolac. The rate of readmission with intra-abdominal infection in our study was 2% for uncomplicated and 5% for complicated. This is consistent with previous studies showing postappendectomy intra-abdominal infection ranging from 1% to 7%.15,16 The effects of ketorolac on complications, readmissions, and reinterventions have not been studied in the pediatric population in relation to gastrointestinal surgeries, including appendicitis. Adult studies have shown conflicting data on the
The limitations of this study include its retrospective design and the nature of administrative databases. The accuracy of coded diagnoses and procedures depends on the quality of documentation and interpretation of medical coders. This is partially mitigated by ongoing quality-control review within the PHIS database to minimize systematic errors. As part of this review, there are checks on billing data in PHIS to match billing data in a particular hospital, but specificity of the review is not clarified. This would be limiting to the primary analysis of the study if there were mistakes in documenting the time of ketorolac administration. These errors are likely minimized with the large cohort of patients. It was noted that there was a possible discrepancy in the dataset, with 2% of patients coded for mechanical ventilation and only 1% of patients admitted to the ICU. It is possible that mechanical ventilation in the postoperative period includes those patients that were extubated in Post Anesthesia Care Unit. Another possibility is that hospitals may have different criteria to code for ICU admission, but we are unable to precisely determine that through the database. While it is impossible to know if we captured exactly the populations we sought, we found that approximately 614/ 100,000 children aged 5-17 y in the inpatient database discharged within the study period met our inclusion criteria, which is a reasonable prevalence of appendicitis in hospitalized children. PHIS also offers a simplified all surgery-related complication flag which only spans the index admission. This may preclude complications that occur postdischarge. To cover these potential complications, we also included any readmission and readmission with postoperative infection in the analysis.
236
j o u r n a l o f s u r g i c a l r e s e a r c h o c t o b e r 2 0 1 7 ( 2 1 8 ) 2 3 2 e2 3 6
Several potentially important variables were either unable to be parsed from this database or were not included in our study design. These include events that occurred at hospitals different from that in the primary encounter or details of intraoperative events (e.g., complications). We limited our focus to ages 5-17 y. While appendicitis does occur outside of this range, its presentation is typically more advanced in younger children and less likely to be captured by this database for older patients. Our study included 30 d of postoperative follow-up and therefore may miss a late-developing complication.
Conclusion Based on a large, contemporary data set of children’s hospitals, ketorolac administration in the immediate postoperative period after appendectomy for appendicitis is common and was not associated with an increase in postoperative LOS, postoperative infection, or any-cause 30-d readmission. Ketorolac was, however, independently associated with a lower overall rate of postoperative complications and cost in this population. Ketorolac is safe to use in the perioperative period for pediatric appendectomy and should be encouraged with the goal of reducing narcotic pain medication in this population.
Acknowledgment Authors’ contributions: H.-u.-R.N, R.M.D., and G.V. contributed to study design, analyzing data, writing the article, and approving the final version of the article. D.H.R. and K.V. contributed to conceiving and designing the study, interpreting the data, providing critical revisions, and approving the final version of the article.
Disclosure The authors have no personal, commercial, political, academic or financial disclosures.
references
1. Aguayo P, Alemayehu H, Desai AA, Fraser JD, St Peter SD. Initial experience with same day discharge after laparoscopic appendectomy for nonperforated appendicitis. J Surg Res. 2014;190:93e97. 2. Alkhoury F, Burnweit C, Malvezzi L, et al. A prospective study of safety and satisfaction with same-day discharge after laparoscopic appendectomy for acute appendicitis. J Pediatr Surg. 2012;47:313e316.
3. Alkhoury F, Malvezzi L, Knight CG, et al. Routine same-day discharge after acute or interval appendectomy in children: a prospective study. Arch Surg. 2012;147:443e446. 4. Kehlet H, Wilmore DW. Evidence-based surgical care and the evolution of fast-track surgery. Ann Surg. 2008;248:189e198. 5. Nelson LS, Juurlink DN, Perrone J. Addressing the opioid epidemic. Jama. 2015;314:1453e1454. 6. Levin RA. Food and Drug Administration; 2009. http://www. accessdata.fda.gov/drugsatfda_docs/nda/2010/ 022382Orig1s000MedR.pdf. 7. Singer AJ, Mynster CJ, McMahon BJ. The effect of IM ketorolac tromethamine on bleeding time: a prospective, interventional, controlled study. Am J Emerg Med. 2003;21:441e443. 8. Chin KR, Sundram H, Marcotte P. Bleeding risk with ketorolac after lumbar microdiscectomy. J Spinal Disord Tech. 2007;20:123e126. 9. Møiniche S, Rømsing J, Dahl JB, Trame`r MR. Nonsteroidal antiinflammatory drugs and the risk of operative site bleeding after tonsillectomy: a quantitative systematic review. Anesth Analg. 2003;96:68e77. 10. Reuter BK, Asfaha S, Buret A, Sharkey KA, Wallace JL. Exacerbation of inflammation-associated colonic injury in rat through inhibition of cyclooxygenase-2. J Clin Invest. 1996;98:2076e2085. 11. Krischak GD, Augat P, Claes L, Kinzl L, Beck A. The effects of non-steroidal anti-inflammatory drug application on incisional wound healing in rats. J Wound Care. 2007;16:76e78. 12. Kotagal M, Hakkarainen TW, Simianu VV, Beck SJ, AlfonsoCristancho R, Flum DR. Ketorolac use and postoperative complications in gastrointestinal surgery. Ann Surg. 2016;263:71e75. 13. Burton TP, Mittal A, Soop M. Nonsteroidal antiinflammatory drugs and anastomotic dehiscence in bowel surgery: systematic review and meta-analysis of randomized, controlled trials. Dis Colon Rectum. 2013;56:126e134. 14. Manworren RCB, McElligott CD, Deraska PV, et al. Efficacy of analgesic treatments to manage Children’s postoperative pain after laparoscopic appendectomy: retrospective medical record review. AORN J. 2016;103:317.e1e317.e11. 15. Garcell HG, Arias AV, Sandoval CAP, et al. Incidence and etiology of surgical site infections in appendectomies: a 3year prospective study. Oman Med J. 2017;32:31e35. 16. Arnbjo¨rnsson E. Laparoscopic or open appendectomy for pediatric appendicitis? MOJ Surg. 2016;3:00032. 17. Gorissen KJ, Benning D, Berghmans T, et al. Risk of anastomotic leakage with non-steroidal anti-inflammatory drugs in colorectal surgery. Br J Surg. 2012;99:721e727. 18. Holte K, Andersen J, Jakobsen DH, Kehlet H. Cyclooxygenase 2 inhibitors and the risk of anastomotic leakage after fast-track colonic surgery. Br J Surg. 2009;96:650e654. 19. Hakkarainen TW, Steele SR, Bastaworous A, et al. Nonsteroidal anti-inflammatory drugs and the risk for anastomotic failure: a report from Washington State’s Surgical Care and Outcomes Assessment Program (SCOAP). JAMA Surg. 2015;150:223e228.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Association for Academic Surgery
Safety of perioperative ketorolac administration in pediatric appendectomy Hibbut-ur-Rauf Naseem, MD,a,* Robert Michael Dorman, MD,a,b George Ventro, MD,a,b David H. Rothstein, MD, MS,a,b and Kaveh Vali, MDa,b a b
Department of Pediatric Surgery, Women and Children’s Hospital of Buffalo, Buffalo, New York Department of Surgery, University at Buffalo, State University of New York, Buffalo, New York
article info
abstract
Article history:
Background: Recent studies in adults undergoing gastrointestinal surgeries show an
Received 12 February 2017
increased rate of complications with the use of ketorolac. This calls into question the safety
Received in revised form
of ketorolac in certain procedures. We sought to evaluate the impact of perioperative
20 April 2017
ketorolac administration on outcomes in pediatric appendectomy.
Accepted 24 May 2017
Methods: The Pediatric Health Information System database was queried for patients aged
Available online xxx
5-17 y with a primary diagnosis of appendicitis and a primary procedure of appendectomy during the period 2010-2014. Patients with procedures suggesting incidental appendec-
Keywords:
tomy, those records with data quality issues, deaths, and extracorporeal membrane
Primary appendicitis
oxygenation were excluded. Variables recorded included age, sex, race, ethnicity, discharge
Nonsteroidal anti-inflammatory
year, complex chronic conditions, geographic region, intensive care unit admission, me-
drugs Pediatric surgery
chanical ventilation, and whether appendicitis was coded as complicated. The exposure variable was ketorolac administration on the day of or day after operation. The primary outcomes of interest were any surgical complications during the initial encounter, postoperative length of stay (LOS), total cost for the initial visit, any readmission to ambulatory, observation, or inpatient status within 30 d, and readmission with a diagnosis of peritoneal abscess or other postoperative infection or with transabdominal drainage performed. Results: A total of 78,926 patients were included in the analysis cohort. Mean age was 11.4 y (standard deviation 3.3 y), the majority were males (61%), White (70%), and non-Hispanic (65%). Few had a complex chronic condition (3%) or required mechanical ventilation (2%) or an intensive care unit admission (1%). Patients with complicated appendicitis comprised 28% of the cohort. Most (73%) received ketorolac on postoperative day 0-1; those with complicated appendicitis were more likely to receive ketorolac. In all, 2.6% of the cohort had a surgical complication during the index visit, 4.3% were readmitted within 30 d, and 2% had a postoperative infection or transabdominal drainage (1% in the uncomplicated group and 5% in the complicated group). Median postoperative LOS was 1 d and mean cost was $9811 $9509. On bivariate analysis, ketorolac administration was associated with a decrease in same-visit surgical complications (P ¼ 0.004) and cost ($459 decrease, P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection. On
* Corresponding author. Department of Pediatric Surgery, Women and Children’s Hospital of Buffalo, 219 Bryant Street, Buffalo, NY 14222. Tel.: þ1 (479) 595 2817; fax: þ1 (716) 888 3850. E-mail address: [email protected] (H.-u. Naseem). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.05.087
naseem et al safety of perioperative ketorolac administration in pediatric appendectomy
233
multivariate analysis, ketorolac administration was associated with a significant decrease in any complication (adjusted odds ratio 0.89, 95% confidence interval 0.80-0.99) and cost (analysis of variance P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection. Conclusions: Based on a large, contemporary data set from children’s hospitals, ketorolac administration in the immediate postoperative period after appendectomy for appendicitis is common and was not associated with an increase in postoperative LOS, postoperative infection, or any-cause 30-d readmission. Ketorolac was, however, independently associated with a lower overall rate of postoperative complications and cost in this population. ª 2017 Elsevier Inc. All rights reserved.
Introduction Acute appendicitis is the most common gastrointestinal surgical urgency in the pediatric population.1 Several recent publications have supported same day discharge as safe and appropriate after laparoscopic appendectomy for uncomplicated appendicitis.2,3 One of the barriers to same day discharge is postoperative pain control.3,4 With increasing scrutiny of the use of narcotic pain control, there is a greater need for alternatives.5 Ketorolac is a well-known nonsteroidal antiinflammatory drug and has been FDA approved for the management of acute pain since 1989.6 The bleeding risks of ketorolac and NSAIDS have been well studied.7-9 Ketorolac has been theorized to impair wound healing and scar tissue formation through its anti-inflammatory mechanisms, which has been demonstrated in several animal studies.10,11 In addition, a recent study in adults undergoing gastrointestinal surgeries by M. Kotagal et al. showed an increased rate of complications ranging from increased emergency department visits, increased readmission, and increased risk of reintervention, calling into question the safety of ketorolac in certain procedures.12 We aimed to reproduce these findings among pediatric patients undergoing appendectomy. We hypothesized that ketorolac would be an independent predictor of increased complications such as abscess formation, readmission, and reoperation.
Methods We conducted a retrospective cohort study of Pediatric Health Information System (PHIS) database, a comparative pediatric database administered by the Children’s Hospital Association (Overland Park, KS) that collects clinical and resource utilization data in inpatient, ambulatory, and emergency department settings at over 45 US children’s hospitals. We included encounters for subjects aged 5-17 y who were discharged between 2010 and 2014 with a primary diagnosis of appendicitis (International Classification of Diseases, Ninth Revision [ICD-9] codes 540, 540.0, 540.1, 540.9) who had an appendectomy coded (ICD-9 procedure codes 47.0, 47.01, 47.09). We excluded incidental appendectomy (47.11 and 47.19) as well as deaths, extracorporeal membrane oxygenation (ECMO), and records with data quality issue because of hospital day and operative day being mismatched and therefore creating inaccurate perioperative timeline for ketorolac exposure. Deaths and ECMO were excluded to minimize confounding factors in the clinical
decision to prescribe ketorolac. We assumed that both these outcomes were because of severity of disease and comorbidities rather than ketorolac administration. Variables recorded included age, sex, race, ethnicity, discharge year, complex chronic conditions (CCC), geographic region, intensive care unit (ICU) admission, mechanical ventilation, and whether appendicitis was coded as complicated. The primary predictor was ketorolac administration on postoperative day 0 or day 1. The primary outcome of interest was readmission with postoperative infection within 30 d. Secondary outcomes were surgical complications during the initial encounter, postoperative length of stay (LOS), total cost for the initial visit, and any readmission to ambulatory, observation, or inpatient status within 30 d. “Any surgical complication” was determined by the presence of a PHIS-generated surgical complication flag for that visit, which was triggered by the presence of one of a number of diagnosis codes chosen by PHIS. These included ICD-9 codes 99x (complications of surgical and medical care, necrotizing enterocolitis, including device, transplant, and transfusion-related complications), as well as complications specific to certain procedures, respiratory failure after trauma and surgery (518.5x), obstetric complications, and postprocedural fever (780.62). The total adjusted costs were based on the reported charges, adjusted for the Ratio of Cost to Charges submitted by each hospital on their respective Medicare cost reports, then adjusted by the Centers for Medicare and Medicaid Services wage/price index for the hospital’s location (13). The postoperative infection on readmission was determined when a peritoneal abscess (ICD-9 code of 567.22), generic postoperative infection (ICD-9 diagnosis code 998.5 or 998.59), or percutaneous abdominal drainage (ICD-9 procedure code 54.91) were coded. Descriptive statistics and bivariate comparisons used ttest, Fisher exact, and Pearson Chi-squared tests where appropriate. Multivariate analysis of binary outcomes used a binary logistic regression model and multivariate analysis of continuous outcomes used a general linear model. Statistics performed using Minitab version 17.3.1. Adjusted odds ratios (aOR) were calculated with corresponding 95% confidence intervals (CI).
Results A total of 79,509 encounters met inclusion criteria. A total of 583 (0.7%) were excluded including 455 for incidental
234
j o u r n a l o f s u r g i c a l r e s e a r c h o c t o b e r 2 0 1 7 ( 2 1 8 ) 2 3 2 e2 3 6
Table 1 e Demographics and outcomes for cohort with bivariate analysis of complicated appendicitis and ketorolac administration. Variable Age, y (mean)
Cohort
Complicated
P
Ketorolac
11.4 3.3
10.8 3.3
<0.001
11.4 3.3
P 0.087
5-11 y
57%
64%
<0.001
57%
0.554
Male
61%
60%
0.003
60%
<0.001
White
70%
65%
<0.001
70%
<0.001
Black
8%
9%
<0.001
7%
<0.001
Other
22%
26%
<0.001
22%
0.213
Hispanic ethnicity
36%
41%
<0.001
38%
<0.001
2010-2011
37%
36%
0.017
36%
<0.001
3%
4%
<0.001
3%
<0.001
South
43%
43%
0.028
45%
<0.001
West
25%
29%
<0.001
28%
<0.001
Northeast
11%
10%
<0.001
9%
<0.001
Midwest
20%
18%
<0.001
18%
<0.001
Mechanical ventilation
2%
2%
0.001
1%
<0.001
Any ICU admission
1%
2%
<0.001
1%
<0.001 <0.001
CCC
Complicated appendicitis
28%
e
e
29%
Ketorolac POD 0-1
73%
75%
<0.001
e
3%
7%
<0.001
3%
Adjusted cost, $
9811 9509
15,694 15,013
<0.001
9689 7224
Postoperative LOS, d
Same-visit surgical complication
e 0.004 <0.001
2.29 13.39
4.90 4.28
<0.001
2.23 2.56
0.192
Any readmission within 30 d
4%
8%
<0.001
4%
0.364
Readmission with intra-abdominal infection within 30 d
2%
5%
<0.001
2%
0.144
appendectomy, 126 for data quality issues, two for death, and three for ECMO. A total of 78,926 were included in the analysis cohort. The mean age of patients undergoing appendectomy was 11.4 y (SD 3.3 y) with a slight skew to the younger age group (57% ages 5-11 y). The majority was male (61%), White (70%), and non-Hispanic (64%). A larger percentage of were in the South region of the United States (43%). Few had a CCC (3%) or required mechanical ventilation (2%) or an ICU admission (1%). In all, 2.6% of the cohort had a surgical complication during the index visit, 4.3% were readmitted within 30 d, and 2% had a postoperative infection or transabdominal drainage (1% in the uncomplicated group and 5% in the complicated group). Median postoperative LOS was 1 d (interquartile range [IQR] 1-3) and mean cost was $9811 $9509 (Table 1). Patients with complicated appendicitis comprised 28% of the cohort and were slightly younger (10.8 versus 11.4 y, P < 0.001), more often listed as Hispanic ethnicity (41% versus 36%, P < 0.001), with a slight preponderance to having a CCC (4% versus 3%, P < 0.001). On bivariate analysis, patients with complicated appendicitis had a higher rate of ketorolac administration on POD 0-1 (75% versus 73%, P < 0.001), with a slightly higher rate of ICU admission (2% versus 1%, P < 0.001) and more frequent same-visit surgical complications (7% versus 3%, P < 0.001). Patients with complicated appendicitis also had significantly higher adjusted cost ($15,694 versus $9,811, P < 0.001), median postoperative LOS (4 [IQR 3-6] versus
1 [IQR 1-1] d, P < 0.001), readmission within 30 d (8% versus 4%, P < 0.001), and readmission with intra-abdominal infection within 30 d (5% versus 2%, P < 0.001; Table 1). Most (73%) of the cohort received ketorolac on postoperative day 0-1; on bivariate analysis, ketorolac administration was associated with Hispanic ethnicity, variability by region of United States, decreased association with mechanical ventilation (1% versus 2%, P < 0.001), and increased use in later years (74.1% in 2012-2014 versus 71.2% in 2010-2011, P < 0.001). Ketorolac was associated with a decrease in same-visit surgical complications (2.9% versus 2.5%, P ¼ 0.004) and cost ($459 decrease, P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection (Table 1). On multivariate analysis, males (aOR 1.2, CI 95% [1.1, 1.3]), Hispanic ethnicity (aOR 1.5, CI [1.0, 1.3]), and complicated appendicitis (aOR 4.0, 95% CI [3.6, 4.5]) had greater odds of readmission with intra-abdominal infection within 30 d. Ketorolac administration on POD 0-1 did not have greater odds of readmission with intra-abdominal infection within 30 d. On separate multivariate analyses, ketorolac was associated with a significant decrease in any complication (aOR 0.89, 95% CI [0.80-0.99]) and cost (analysis of variance P < 0.001) but was not associated with readmission, postoperative LOS, or postoperative infection (Table 2). In a sensitivity analysis of ketorolac given only on postoperative day 0 (62% of the cohort), qualitatively identical results were obtained.
naseem et al safety of perioperative ketorolac administration in pediatric appendectomy
Table 2 e Multivariate analysisepredictors of readmission with postoperative infection. Variable
Variable subgroup
OR (95% CI)
Age 5-11 y
1.04 (0.93, 1.15)
Male
1.23 (1.10, 1.37)
Race
Black
Reference
White
1.07 (0.87, 1.31)
Other
1.18 (0.93, 1.49)
Hispanic ethnicity
1.15 (1.02, 1.30)
2010-2011
1.01 (0.91, 1.12)
CCC
1.02 (0.77, 1.34)
US region
Midwest
Reference
Northeast
0.70 (0.57, 0.87)
South
0.93 (0.80, 1.08)
West
0.75 (0.63, 0.88)
Mechanical ventilation
1.14 (0.79, 1.66)
Any ICU admission
0.78 (0.50, 1.20)
Complicated appendicitis
4.06 (3.65, 4.52)
Ketorolac POD 0-1
1.02 (0.91, 1.15)
235
effects of ketorolac. In a meta-analysis of six randomized controlled trials with 480 patients, ketorolac did not show a significant difference in anastomotic dehiscence.13 The authors concluded no difference but because of insufficient power. In multiple retrospective observational studies, non-steroidal anti-inflammatory drugs such as ketorolac have shown to have a detrimental effect on anastomotic dehiscence.17-19 Ketorolac use in our study showed a statistically significant decrease in cost of $459 but this amount would only make a small dent in the greater than $9000 cost of hospitalization after an appendectomy. There was no significant difference in LOS with ketorolac use. This correlates well with a study by Manworren et al. that showed decreased morphine equivalents and pain scores with ketorolac use that did not produce the expected decrease in LOS.14 Similar to Kotagal et al., we used a national database (pediatric instead of adult). We noted patterns of regional variation in the use of ketorolac, which may be due to random variation or may correlate with aversion due to cost or perceived risk of the drug. The Hispanic population was more likely to receive the drug as well, which may correlate with the regional variation of more frequent ketorolac use in the South region of the United States. These conjectures may explain the trend but further research needs to assess the use of analgesics across different demographics.
Discussion Limitations Findings In this large pediatric database study, we found that ketorolac administration on POD 0 or 1 after appendectomy did not show increased complications including readmission, postoperative infection, and reinterventions within 30 d. This also applied when Ketorolac was administered on POD 0 only. Ketorolac administration was also associated with fewer same-visit complications and cost. These results are contrary to the study by Kotagal et al., which showed increased rates of emergency department visits, readmissions, and reinterventions associated with GI operations involving anastomosis. While adult studies show a range of 15%-30% ketorolac use after gastrointestinal surgery,12,13 recent pediatric studies show rates closer to 50% after appendectomy.14 Our study had an overall 73% use of ketorolac in the perioperative period, with an increase over the time period. This shows that despite the concern for ketorolac-related complications in gastrointestinal surgery, most surgeons do not view this as translated to pediatric appendectomy patients. About one-quarter of the patients did not receive ketorolac, which may have been for a variety of reasons including concern for intraoperative bleeding, surgeon hesitation because of perceived risk of ketorolac, and other medical contraindications to ketorolac. The rate of readmission with intra-abdominal infection in our study was 2% for uncomplicated and 5% for complicated. This is consistent with previous studies showing postappendectomy intra-abdominal infection ranging from 1% to 7%.15,16 The effects of ketorolac on complications, readmissions, and reinterventions have not been studied in the pediatric population in relation to gastrointestinal surgeries, including appendicitis. Adult studies have shown conflicting data on the
The limitations of this study include its retrospective design and the nature of administrative databases. The accuracy of coded diagnoses and procedures depends on the quality of documentation and interpretation of medical coders. This is partially mitigated by ongoing quality-control review within the PHIS database to minimize systematic errors. As part of this review, there are checks on billing data in PHIS to match billing data in a particular hospital, but specificity of the review is not clarified. This would be limiting to the primary analysis of the study if there were mistakes in documenting the time of ketorolac administration. These errors are likely minimized with the large cohort of patients. It was noted that there was a possible discrepancy in the dataset, with 2% of patients coded for mechanical ventilation and only 1% of patients admitted to the ICU. It is possible that mechanical ventilation in the postoperative period includes those patients that were extubated in Post Anesthesia Care Unit. Another possibility is that hospitals may have different criteria to code for ICU admission, but we are unable to precisely determine that through the database. While it is impossible to know if we captured exactly the populations we sought, we found that approximately 614/ 100,000 children aged 5-17 y in the inpatient database discharged within the study period met our inclusion criteria, which is a reasonable prevalence of appendicitis in hospitalized children. PHIS also offers a simplified all surgery-related complication flag which only spans the index admission. This may preclude complications that occur postdischarge. To cover these potential complications, we also included any readmission and readmission with postoperative infection in the analysis.
236
j o u r n a l o f s u r g i c a l r e s e a r c h o c t o b e r 2 0 1 7 ( 2 1 8 ) 2 3 2 e2 3 6
Several potentially important variables were either unable to be parsed from this database or were not included in our study design. These include events that occurred at hospitals different from that in the primary encounter or details of intraoperative events (e.g., complications). We limited our focus to ages 5-17 y. While appendicitis does occur outside of this range, its presentation is typically more advanced in younger children and less likely to be captured by this database for older patients. Our study included 30 d of postoperative follow-up and therefore may miss a late-developing complication.
Conclusion Based on a large, contemporary data set of children’s hospitals, ketorolac administration in the immediate postoperative period after appendectomy for appendicitis is common and was not associated with an increase in postoperative LOS, postoperative infection, or any-cause 30-d readmission. Ketorolac was, however, independently associated with a lower overall rate of postoperative complications and cost in this population. Ketorolac is safe to use in the perioperative period for pediatric appendectomy and should be encouraged with the goal of reducing narcotic pain medication in this population.
Acknowledgment Authors’ contributions: H.-u.-R.N, R.M.D., and G.V. contributed to study design, analyzing data, writing the article, and approving the final version of the article. D.H.R. and K.V. contributed to conceiving and designing the study, interpreting the data, providing critical revisions, and approving the final version of the article.
Disclosure The authors have no personal, commercial, political, academic or financial disclosures.
references
1. Aguayo P, Alemayehu H, Desai AA, Fraser JD, St Peter SD. Initial experience with same day discharge after laparoscopic appendectomy for nonperforated appendicitis. J Surg Res. 2014;190:93e97. 2. Alkhoury F, Burnweit C, Malvezzi L, et al. A prospective study of safety and satisfaction with same-day discharge after laparoscopic appendectomy for acute appendicitis. J Pediatr Surg. 2012;47:313e316.
3. Alkhoury F, Malvezzi L, Knight CG, et al. Routine same-day discharge after acute or interval appendectomy in children: a prospective study. Arch Surg. 2012;147:443e446. 4. Kehlet H, Wilmore DW. Evidence-based surgical care and the evolution of fast-track surgery. Ann Surg. 2008;248:189e198. 5. Nelson LS, Juurlink DN, Perrone J. Addressing the opioid epidemic. Jama. 2015;314:1453e1454. 6. Levin RA. Food and Drug Administration; 2009. http://www. accessdata.fda.gov/drugsatfda_docs/nda/2010/ 022382Orig1s000MedR.pdf. 7. Singer AJ, Mynster CJ, McMahon BJ. The effect of IM ketorolac tromethamine on bleeding time: a prospective, interventional, controlled study. Am J Emerg Med. 2003;21:441e443. 8. Chin KR, Sundram H, Marcotte P. Bleeding risk with ketorolac after lumbar microdiscectomy. J Spinal Disord Tech. 2007;20:123e126. 9. Møiniche S, Rømsing J, Dahl JB, Trame`r MR. Nonsteroidal antiinflammatory drugs and the risk of operative site bleeding after tonsillectomy: a quantitative systematic review. Anesth Analg. 2003;96:68e77. 10. Reuter BK, Asfaha S, Buret A, Sharkey KA, Wallace JL. Exacerbation of inflammation-associated colonic injury in rat through inhibition of cyclooxygenase-2. J Clin Invest. 1996;98:2076e2085. 11. Krischak GD, Augat P, Claes L, Kinzl L, Beck A. The effects of non-steroidal anti-inflammatory drug application on incisional wound healing in rats. J Wound Care. 2007;16:76e78. 12. Kotagal M, Hakkarainen TW, Simianu VV, Beck SJ, AlfonsoCristancho R, Flum DR. Ketorolac use and postoperative complications in gastrointestinal surgery. Ann Surg. 2016;263:71e75. 13. Burton TP, Mittal A, Soop M. Nonsteroidal antiinflammatory drugs and anastomotic dehiscence in bowel surgery: systematic review and meta-analysis of randomized, controlled trials. Dis Colon Rectum. 2013;56:126e134. 14. Manworren RCB, McElligott CD, Deraska PV, et al. Efficacy of analgesic treatments to manage Children’s postoperative pain after laparoscopic appendectomy: retrospective medical record review. AORN J. 2016;103:317.e1e317.e11. 15. Garcell HG, Arias AV, Sandoval CAP, et al. Incidence and etiology of surgical site infections in appendectomies: a 3year prospective study. Oman Med J. 2017;32:31e35. 16. Arnbjo¨rnsson E. Laparoscopic or open appendectomy for pediatric appendicitis? MOJ Surg. 2016;3:00032. 17. Gorissen KJ, Benning D, Berghmans T, et al. Risk of anastomotic leakage with non-steroidal anti-inflammatory drugs in colorectal surgery. Br J Surg. 2012;99:721e727. 18. Holte K, Andersen J, Jakobsen DH, Kehlet H. Cyclooxygenase 2 inhibitors and the risk of anastomotic leakage after fast-track colonic surgery. Br J Surg. 2009;96:650e654. 19. Hakkarainen TW, Steele SR, Bastaworous A, et al. Nonsteroidal anti-inflammatory drugs and the risk for anastomotic failure: a report from Washington State’s Surgical Care and Outcomes Assessment Program (SCOAP). JAMA Surg. 2015;150:223e228.