4. Discussion
Tonsillectomy is one of the most common surgeries worldwide. As 26% of tonsillectomies in United States are performed in adults [1], the purpose of this study was to find factors that are associated with PTB with a focus on post-operative pain, a factor which has not been investigated in the adult population.
In this cohort, younger patients had a higher risk of PTB compared to patients over 30 years old. Although there is abundant data regarding the fact that adults have higher tendency for PTB compared to pediatric patients [9], in this study we highlight a specific sub-group at higher risk of PTB, namely those in the 18 to 30 age group.
Similar to previous studies [10-13], males and smoking status were associated with PTB. In this cohort BMI was higher in the PTB group but the difference was not statistically significant. Regarding surgical technique, some studies demonstrated higher bleeding rates with hot dissection technique [14-15], while others found no association, including one Cochrane review [9, 16]. In our study we found a correlation between the use of hot technique and higher bleeding rates. Sutures were unhelpful in reducing the probability of PTB in this cohort.
Patients who underwent tonsillectomy due to recurrent throat infections were more likely to develop PTB, compared to other indications (OSA, halitosis etc.). This is probably due to the fibrosis and scaring in the peritonsillar tissue following recurrent inflammatory states which leads to difficulties with the dissection from the tonsillar bed. This finding was also noted in the pediatric population [18].
Post-operative pain was measured as VAS (1-10) from POD-0 to POD-2 and higher scores on POD-1 and POD-2 were associated with PTB. To the best of our knowledge, this is the first study that implies a relationship between post-operative pain and PTB in the adult population. One hypothesis is that pain reaction involves several mechanisms that lead to increased risk of bleeding. First, surgery induces inflammatory states that results in increase of inflammatory cytokines which lead to vasodilation [19]. Second, uncontrolled pain may lead to anxiety and stress response, both may cause increase in respiratory rate and blood pressure which results in bleeding. Third, increased pain can cause reduced oral intake and dehydration which may impair wound healing and lead to bleeding. On the other hand, this study cannot rule out that pain and PTB are not causally related, but rather have a common cause. One option is that a more difficult operation (e.g., due to recurrent infections) may lead to both higher pain scores and increased risk for PTB. To account for this possibility, a sub-group analysis of correlation between pain and PTB was performed according to indication. There was no difference in our results when analyzing by indication. This analysis cannot completely rule out a common causative agent for increased post-operative pain an PTB risk. A prospective trials that measure the effect of intensive pain control on bleeding risk may clarify the issue. Regardless of causality, increased post-operative pain may serve as a predictor for future bleeding.
This study is mainly limited by its retrospective nature. Tonsillectomies were performed by 3 different surgeons thus small differences in the surgical techniques were encountered. Patients received different types of analgesics (paracetamol, dipyrone, tramadol, oxycodone, non-steroidal anti-inflammatory drugs) and for most of the trial period, there was no uniform protocol for pain control.