Postoperative Ibuprofen and the Risk of Bleeding After Tonsillectomy With or Without Adenoidectomy

Tonsillectomy with and without adenoidectomy is one of the most commonly performed surgical procedures in the pediatric population. The incidence of adenotonsillectomy has increased over the past four decades (1). This is mainly due to increased awareness of the potential adverse consequences that pediatric sleep disordered breathing (SDB) may have on development and long-term pulmonary and cardiovascular health. SDB has surpassed recurrent tonsillitis as the most common indication for adenotonsillectomy in children (2-5). Over the past decade, according to the CDC's National Health Statistic Report on ambulatory surgery performed in the US, annual rates of tonsillectomy performed with and without adenoidectomy in children aged 15 and younger increased from 287,000 to 530,000 (6, 7). Adenotonsillectomy is the second most common procedure performed on children under the age of 15. Although generally considered a safe procedure, adenotonsillectomy has significant morbidity and potential for complications, particularly in the pediatric population. Complications include postoperative hemorrhage, dehydration, pain, anesthetic complications and airway risks, aspiration, and post-obstructive pulmonary edema (8). In young children, the risk of adenotonsillectomy is more critical due to smaller airways and respiratory reserve, as well as smaller blood volume (5). Postoperative bleeding can be categorized as a primary event, occurring \< 24 hours after surgery, or a secondary event, occurring \>24 hours after tonsillectomy. Additionally, events can be further described by the interventions taken, such as emergency room visits, admission for observation, or return to the operating room to achieve hemostasis. Postoperative bleeding rates, including both primary and secondary events, range from 3.3-20%, with a mean of 4.5% (9). Thus, annually, tens of thousands of children experience exposure to potentially life-threatening postoperative hemorrhage, often requiring readmission, anesthetic exposure, and operative control of hemorrhage. Postoperative pain contributes significantly to post-tonsillectomy morbidity. While narcotics are effective in controlling postoperative pain, they are often contraindicated, particularly in children with sleep disordered breathing, because of their potentially adverse side effects on respiration and the central nervous system (10). Nonsteroidal anti-inflammatories (NSAIDS), which block prostaglandin-induced inflammation and edema, are an attractive therapeutic option because they do not result in respiratory and central nervous system depression, and therefore may reduce the risk of postoperative respiratory depression, nausea and vomiting, excessive sedation and urinary retention. NSAIDS have been shown to be effective analgesics after tonsillectomy (11,12). However, because their mechanism of action may also interfere with platelet aggregation and increase bleeding time, their use is balanced with concern about an increased risk of postoperative hemorrhage. Aspirin, which irreversibly inhibits cyclooxygenase, affects coagulation and bleeding for up to 10 days, has been associated with an increased bleed rate after tonsillectomy (13). However, non-aspirin NSAIDs demonstrate a reversible inhibition of COX-1 and COX-2, and therefore do not have the same severe, prolonged effects on bleeding (14). Ibuprofen, a derivative of propionic acid, is widely used for musculoskeletal pain, but the study of its use for post-tonsillectomy analgesia is limited. In 2011, The American Academy of Otolaryngology published Clinical Practice Guidelines outlining evidence-based selection and perioperative management strategies for tonsillectomy in children. As part of a recommendation that clinicians educate caregivers about the importance of postoperative pain management, Baugh et al advocated the use of ibuprofen postoperatively, stating, "ibuprofen can be used safely for pain control after surgery" (15), citing a 2005 Cochrane Review of NSAIDs and post-tonsillectomy bleeding in support of this guideline (16). The Cochrane Review recently added additional studies to their analysis and results remained similar. The most recent Cochrane Review, published in 2010, evaluated 15 randomized trials comparing NSAIDs with other analgesics or placebo, and determined that NSAID use did not significantly alter the number of perioperative bleeding episodes, both requiring and not requiring surgical intervention; this review did not distinguish between primary and secondary bleeding events (17). Because post-tonsillectomy hemorrhage is an uncommon event, a large number of participants is required to provide an adequate number of events to give a significant result, therefore the large sample size of \>1000 children in the Cochrane Review is admirable. However, sample sizes were not adequate to compare the risk of bleeding with each individual NSAID. Additionally, NSAIDs were given in both oral and parenteral forms, as well as preoperatively, intraoperatively and postoperatively, and the duration of postoperative analgesic use differed between studies. The surgical technique used was not uniform between studies as well. It is our feeling that because of these limitations, the data is not sufficient to broadly implement the Academy's recommendation that ibuprofen can be safely used for post-tonsillectomy analgesia without more carefully controlled, prospective study. Although there are many studies in the literature evaluating NSAID use after tonsillectomy, there are few randomized-prospective trials evaluating the use of ibuprofen, and few trials are powered to adequately assess the risk of postoperative hemorrhage. Designing and executing a study to specifically evaluate ibuprofen after pediatric tonsillectomy and rates of post-operative hemorrhage requiring return to the operating room for control is important to the pediatric otolaryngology community, particularly given the American Academy of Otolaryngology's recent clinical guidelines. The results of such a definitive study would possibly affect the tens of thousands of children at risk for post-tonsillectomy hemorrhage every year. It would affect our own standards of care as well as national and international norms. Preliminary Studies: At the Massachusetts Eye and Ear Infirmary there is no precedent for administering ibuprofen to children after tonsillectomy with or without adenoidectomy. Despite American Academy of Otolaryngology support for the use of ibuprofen in the postoperative setting, pediatric otolaryngologists at our institution are still hesitant to administer ibuprofen postoperatively without additional study of its use. Within the Department of Pediatric Otolaryngology at the Massachusetts Eye and Ear Infirmary there is a precedent for successful completion of randomized, controlled trials involving adenotonsillectomy (19, 20). With principal investigators dedicated to clinical research, as well as research nurses and coordinators at our disposal, a prospective, controlled clinical trial of ibuprofen use after tonsillectomy can be implemented. Additionally, the Massachusetts Eye and Ear Infirmary calculates annual post-tonsillectomy hemorrhage rates, including percentage of patients returning to the emergency room for evaluation for possible post-operative bleeding, as well as the percentage of patients returning to the operating room for control of bleeding. Thus, there is a precedent for collecting data on postoperative bleeding events, as well as a controlled bleed rate to which prospective study hemorrhage rates can be compared. The goal of our study is to determine if postoperative ibuprofen affects the rate of post-tonsillectomy hemorrhage using a non-inferiority trial design, which is intended to show that the effect of ibuprofen is no worse than acetaminophen, which will serve as our control. This differs from an equivalence trial, which aims to demonstrate that the experimental and control group do not differ more than a specified amount. The equivalence margin set for non-inferiority trials is often smaller than the treatment difference for which a placebo-controlled trial is powered, requiring a larger sample size. This will address two issues we have with the current literature evaluating NSAID use after tonsillectomy: inadequate sample size, and unrealistic treatment differences, including up to a 20% difference in bleed rates, required to detect a significant difference between NSAID and control groups. Because our trial will be designed to test non-inferiority, the null hypothesis will assume inferiority: the rate of hemorrhage requiring operative intervention in patients treated with postoperative ibuprofen after tonsillectomy will be increased compared to the rate in patients given acetaminophen.