Evaluation of post-tonsillectomy bleeding in the adult population.

Abstract

A retrospective review of a consecutive series of 685 adult patients undergoing tonsillectomy was conducted. Determinations were made of the post-tonsillectomy bleeding rate, the need for intervention to control bleeding, and the blood transfusion rate. Statistical analysis was used to determine whether bleeding rates differed according to three criteria: gender, indication for tonsillectomy, and age. Post-tonsillectomy bleeding occurred in 35 patients (5.1%); five of these patients experienced bleeding during the first 24 hours postoperatively, and the remaining 30 experienced delayed bleeding. When it occurred, the mean time lapse between tonsillectomy and bleeding was 6.9 days ([+ or -]4.1). Twenty of the 35 patients (57.1%) required a procedure to control their bleeding, but no patient required a transfusion. There was no statistically significant difference in bleeding rates based on gender, the indication for surgery (chronic tonsillitis, obstructive sleep apnea syndrome, or to rule out neoplasia), and age. These results indicate that (1) post-tonsillectomy bleeding occurs in approximately 1 of 20 adults independent of individual patient characteristics, (2) more than half of patients who bleed are likely to require a procedure to control their hemorrhage, and (3) the need for transfusion is distinctly unlikely.

Introduction

Tonsillectomy with or without adenoidectomy is one of the most frequently performed surgical procedures in the United States. [1] Although these procedures are usually performed on pediatric patients, a significant proportion of adults also undergoes the procedure, most often as a treatment for chronic recurrent tonsillitis or as part of a surgical airway augmentation in the treatment of obstructive sleep apnea syndrome. A less common indication is the need to rule out malignancy.

When patients satisfy clinical criteria and consider tonsillectomy as an option for treatment, much of the subsequent discussion centers on the procedure's morbidity. Other important considerations are postoperative pain, the anticipated length of absence from work, and the procedure's risks. One of the more important risks of tonsillectomy is postoperative bleeding. Traditionally, postoperative bleeding is classified as one of two types: immediate ([less than]24 hr) and delayed ([greater than or equal to]24 hr). Reported rates for delayed bleeding range from 0.1 to 8.1%, but most of these rates are derived from studies of the pediatric population. [2-4]

Otolaryngologists at the author's institution have suspected for some time that adults who undergo tonsillectomy, especially for chronic tonsillitis, have a higher postoperative bleeding rate than do children. Therefore, this study was undertaken to examine the incidence of post-tonsillectomy bleeding specifically in the adult patient population.

Methods

The author examined the procedural database of a large academic otolaryngology practice. All patients who had undergone tonsillectomy, with or without adenoidectomy, between 1982 and 2000 were identified. Only patients who were aged 16 years or older at the time of surgery were included in this study; for purposes of this study, patients 16 and 17 years old were classified as adults. A retrospective review of each patient's medical record provided demographic information. The indication for tonsillectomy was identified by the ICD-9 diagnosis code. Patients were grouped into one of three categories, depending on the primary indication for tonsillectomy: (1) chronic tonsillitis or pharyngitis, (2) obstructive sleep apnea syndrome, or (3) to rule out malignancy.

A second database that contained information on all otolaryngology admissions during the same time period was examined, and all patients who had been admitted to the hospital with a diagnosis of post-tonsillectomy bleeding were identified. For each such patient, the medical record was further examined to ascertain four factors: (1) the length of time that had passed between the tonsillectomy and the onset of the bleeding, (2) the need for a procedure to control the hemorrhage, (3) the need for a blood transfusion, and (4) elements of the complete blood count.

Statistical testing was performed with the SPSS statistical package, version 10.0 (SPSS, Inc.; Chicago). Descriptive statistics were obtained for categorical variables. A [[chi].sup.2] analysis was used to compare two characteristics (the patient's gender and indication for surgery) of those patients who had and had not bled. The Student's t test was used to examine the influence of age on the incidence of post-tonsillectomy bleeding.

Results

A total of 685 adults had undergone tonsillectomy during the 18-year study period: 281 men and 404 women (table). Their mean age at the time of tonsillectomy was 29.9 years ([+ or -] 10.9). Tonsillectomy alone was performed on 624 patients (91.1%), and tonsillectomy with adenoidectomy was performed on the remaining 61 (8.9%). The most common indication for surgery was chronic tonsillitis.

Bleeding rates. A total of 35 patients (5.1%) experienced post-tonsillectomy bleeding. Five patients (0.7% of all patients; 14.3% of those who bled) experienced immediate postoperative bleeding, and 30 (4.4 and 85.7%, respectively) experienced delayed bleeding. The mean interval between surgery and the onset of bleeding was 6.9 days ([+ or -]4.1). Delayed bleeding occurred within a range of 1 to 16 days postoperatively, and more cases of bleeding occurred on postoperative day 7 or 8 than any other 2-day day period (figure).

Bleeding control. Of the 35 patients who experienced post-tonsillectomy bleeding, 20(57.1%) required a procedure to control the hemorrhage; the remaining 15 patients were managed with observation alone. No patient required a transfusion. The mean hematocrit nadir during the admission for bleeding was 38.3%, and only one patient had a hematocrit below 30.0%. The mean length of hospital stay for postoperative bleeding was 1.6 days ([+ or -]0.8).

Variables. There was no significant difference in bleeding rates between the genders (p = 0.199; [[chi].sup.2] exact significance). Similarly, no difference in bleeding rate was identified according to the primary indication for tonsillectomy (p = 0.130; [[chi].sup.2] exact significance). Finally, there was no statistically significant difference in the mean age of patients who bled and those who did not (p = 0.616; Student's t test).

Discussion

Traditionally, adult tonsillectomy has been recommended for patients with chronic recurrent tonsillitis and for those who are persistent carriers of streptococcal spp. [5] To some degree, these indications have been extrapolated from the pediatric patient population. The diagnosis of chronic tonsillitis usually hinges on the frequency and severity of episodes, but the criteria to satisfy the diagnosis have varied from institution to institution. Less common indications for adult tonsillectomy include halitosis, chronic cryptic debris, and as an adjunct to uvulopalatopharyngoplasty for patients with obstructive sleep apnea syndrome. Tonsillectomy has been shown to be quite effective in reducing the clinical burden of chronic tonsillitis and pharyngitis in adults. [6] Appropriately selected adults who undergo tonsillectomy for chronic recurrent tonsillitis require significantly fewer courses of antibiotics and significantly fewer physician visits for tonsillitis/pharyngitis than those who do not undergo tonsillectomy.

Post-tonsillectomy bleeding is a long-standing and as-yet-unsolved problem. This bleeding can be especially distressing because patients have no direct way of visualizing or controlling their bleeding. They often swallow or aspirate their blood, and they are usually still in a fair amount of expected postoperative pain. Many patients start and stop bleeding spontaneously. Some have recurrent episodic bleeding, persistent venous bleeding, or frank arterial hemorrhage. Because it is difficult to predict which patients will experience bleeding that is significant enough to require operative control, the author's institution has consistently followed a policy of admitting and observing all patients who experience post-tonsillectomy hemorrhage. Given that more than 50% of the post-tonsillectomy bleeds in this study required some sort of intervention to control their bleeding, this period of observation appears to be warranted. Furthermore, patients are generally more comfortable and feel more secure in the hospit al setting following a post-tonsillectomy hemorrhage.

Our post-tonsillectomy bleeding rate is somewhat higher than that reported by Wei et al, who found that approximately 3.6% of adult patients (age range: 21 to 30 years) experienced post-tonsillectomy bleeding. [7] In addition, Wei et al found that approximately 50% of patients who bled (both children and adults) returned to the operating room for control of their hemorrhage--a finding that is in line with the finding of this study (57.1%).

Many investigators have attempted to reduce post-tonsillectomy bleeding by varying their surgical techniques, hemostatic methods, and postoperative care practices. Yet none of these proposed variations has made an appreciable impact on the overall rate. For example, intraoperative steroid treatment has not been found to improve post-tonsillectomy bleeding rates. [8] Similarly, although oral antibiotics administered during the postoperative period tend to reduce postoperative pain and halitosis, they have not been found to have a significant impact on preventing post-tonsillectomy bleeding. [9-11] Some investigators believe that the anti-inflammatory ketorolac might predispose patients to post-tonsillectomy bleeding, but other studies have not shown this. [4,12-14] Because of this uncertainty, the author's institution has been reluctant to use ketorolac for postoperative analgesia in adult tonsillectomy patients. Unfortunately, almost all of the ketorolac studies have been performed on children, and it is not clear whether these data can be easily extrapolated to adults.

Additional debate has centered on surgical technique. Some investigators advocate cold steel rather than cautery, citing decreased postoperative pain as a major advantage. [15] Because 11 of the 12 otolaryngologists at the author's institution perform tonsillectomy with an electrocautery technique, the author was unable to examine for differences in bleeding rates according to technique. Electrocautery is often preferred for adult patients with chronic tonsillitis because these patients often have additional scarring and hypervascularity as a result of their tonsillitis. Electrocautery tonsillectomy causes less intraoperative bleeding, but it can result in higher postoperative pain scores. [16] Recent studies have suggested that microbipolar cautery dissection tonsillectomy might have advantages in terms of postoperative pain and bleeding rates, at least in the pediatric population. [17]

One potential limitation of this study lies in the method of identifying patients who bled after tonsillectomy. It is possible that they underwent observation in the outpatient setting. Given that it is difficult to predict which patients will experience a serious post-tonsillectomy bleed following a "sentinel" bleed, the author's institution has adhered to a policy of hospital admission for a period of observation after any incident of post-tonsillectomy bleeding. Therefore, one would expect that this potential underestimation of the post-tonsillectomy bleeding rate is limited. It is also possible that patients who bled sought care for their bleeding at another facility or that they did not bring their bleeding to the attention of their treating physician. These selection biases would tend to underestimate the post-tonsillectomy bleeding rate determined by these data, and, therefore, the data obtained in this study might reflect a low-end estimate of post-tonsillectomy bleeding in the adult population.

References

(1.) Derkay CS. Pediatric otolaryngology procedures in the United States: 1977-1987. Int J Pediatr Otorhinolaryngol 1993;25:1-12.

(2.) Randall DA, Hoffer ME. Complications of tonsillectomy and adenoidectomy. Otolaryngol Head Neck Surg 1998:118:61-8.

(3.) Howells RC II, Wax MK, Ramadan HH. Value of preoperative prothrombin time/partial thromboplastin time as a predictor of postoperative hemorrhage in pediatric patients undergoing tonsillectomy. Otolaryngol Head Neck Surg 1997;117:628-32.

(4.) Agrawal A, Gerson CR, Seligman I, Dsida RM. Postoperative hemorrhage after tonsillectomy: Use of ketorolac tromethamine. Otolaryngol Head Neck Surg 1999;120:335-9.

(5.) Fry TL, Pillsbury HC. The implications of "controlled" studies of tonsillectomy and adenoidectomy. Otolaryngol Clin North Am 1987;20:409-13.

(6.) Mui S, Rasgon BM, Hilsinger RL. Jr. Efficacy of tonsillectomy for recurrent throat infection in adults. Laryngoscope 1998;108:1325-8.

(7.) Wei JL, Beatty CW, Gustafson RO. Evaluation of posttonsillectomy hemorrhage and risk factors. Otolaryngol Head Neck Surg 2000:123:229-35.

(8.) Ohlms LA, Wilder RT, Weston B. Use of intraoperative corticosteroids in pediatric tonsillectomy. Arch Otolaryngol Head Neck Surg 1995;121:737-42.

(9.) Grandis JR, Johnson JT, Vickers RM, et al. The efficacy of perioperative antibiotic therapy on recovery following tonsillectomy in adults: Randomized double-blind placebo-controlled trial. Otolaryngol Head Neck Surg 1992:106:137-42.

(10.) Telian SA, Handler SD, Fleisher GR, et al. The effect of antibiotic therapy on recovery after tonsillectomy in children. A controlled study. Arch Otolaryngol Head Neck Surg 1986;112:610-5.

(11.) Lee WC, Duignan MC, Walsh RM, McRae-Moore JR. An audit of prophylactic antibiotic treatment following tonsillectomy in children. J Laryngol Otol 1996;110:357-9.

(12.) Bailey R, Sinha C, Burgess LP. Ketorolac tromethamine and hemorrhage in tonsillectomy: A prospective, randomized, double-blind study. Laryngoscope 1997;107:166-9.

(13.) Judkins JH, Dray TG, Hubbell RN. Intraoperative ketorolac and posttonsillectomy bleeding. Arch Otolaryngol Head Neck Surg 1996;122:937-40.

(14.) Rusy LM, Houck CS, Sullivan LJ, et al. A double-blind evaluation of ketorolac tromethamine versus acetaminophen in pediatric tonsillectomy: Analgesia and bleeding. Anesth Analg 1995:80:226-9.

(15.) Nunez DA, Provan J, Crawford M. Postoperative tonsillectomy pain in pediatric patients: Electrocautery (hot) vs cold dissection and snare tonsillectomy--a randomized trial. Arch Otolaryngol Head Neck Surg 2000;126:837-41.

(16.) Leach J, Manning S, Schaefer S. Comparison of two methods of tonsillectomy. Laryngoscope 1993;103:619-22.

(17.) Pizzuto MP, Brodsky L, Duffy L, et al. A comparison of microbipolar cautery dissection to hot knife and cold knife cautery tonsillectomy. Int I Pediatr Otorhinolaryngol 2000:52:239-46.

Table. Characteristics of patients who
did not experience post-tonsillectomy
bleeding
Variable                       Overall  No bleeding  Bleeding
Total (n [%])                      685  650 (94.9)   35 (5.1)
Gender
  Male (n [%])                     281  263 (93.6)   18 (6.4)
  Female (n [%])                   404  387 (95.8)   17 (4.2)
Indication for surgery
  Chronic tonsillitis (n [%])      506  481 (95.1)   25 (4.9)
  OSAS * (n [%])                   147  141 (95.9)    6 (4.1)
  Rule out neoplasia (n [%])        32   28 (87.5)    4 (12.5)
Mean Age (yr)                     29.9     29.9       30.8
(*)Obstructive sleep apnea syndrome.

[Graph omitted]