Unidentified coagulation disorders in post-tonsillectomy hemorrhage.

Abstract

We conducted a retrospective study of 6, 966 patients who had undergone tonsillectomy or adenotonsillectomy to evaluate the incidence and clinical features of previously unidentified coagulation disorders in patients who experienced postoperative hemorrhage (n = 201). We found that post-tonsillectomy hemorrhage secondary to unidentified coagulation disorders is extremely rare. However, normal coagulation values and an insignificant history do not rule out coagulation disorders. If diffuse, persistent, and bilateral bleeding is not related to arterial hypertension, dissection technique, or local infection, a rapid and detailed analysis of coagulation factors should be considered.

Introduction

Tonsillectomy, with or without adenoidectomy, remains one of the most commonly performed surgical procedures in otolaryngology. For most patients in the pediatric age group, the operation is the first major challenge to their hemostatic system. A decade ago, approximately 285,000 of these surgical procedures were performed in the United States, (1) more than 80,000 in England and Wales, (2) and approximately 200,000 in Germany. (3)

Unlike the case with other surgical procedures, the surface of the large mucosal wound is not closed following tonsillectomy. Instead, it is left to achieve secondary healing. However, during healing, the wound is exposed to bacteria-and enzyme-containing saliva close to the larger vessels of the neck.

Although outpatient tonsillectomy and adenotonsillectomy can be performed safely, a personal or family history of abnormal coagulation, with or without an identified bleeding disorder, certainly warrants overnight observation in an appropriate setting with skilled staff (4) Primary (<24 hr) post-tonsillectomy hemorrhage is acknowledged to be more dangerous than bleeding that occurs later because larger amounts of blood can be swallowed or aspirated during early bleeding. In fact, deaths during the early phase following tonsillectomy have been reported: The American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) suggests that surgeons conduct a coagulation and bleeding work-up on any patient who is suspected of having such an abnormality on the basis of the history and in any patient for whom genetic information is unavailable: However, intra- or postoperative hemorrhage that is the result of a coagulation disorder can occur in patients whose laboratory coagulation values are normal.

Our experience with a dramatic postoperative course following tonsillectomy in an adult with normal coagulation values and an insignificant history encouraged us to conduct a retrospective study to evaluate the incidence and clinical features of previously undetected coagulation disorders that became evident only after the tonsillectomy.

Patients and methods

Between Jan. 1, 1989, and June 30, 2002, 6,966 patients underwent tonsillectomy or adenotonsillectomy at out clinic. This group was made up of 3,574 males (51.3%) and 3,392 females (48.7%), aged 5 months to 93 years (mean: 22.2 yr; median 18 [+ or -] 17.2). Most were between 5 and 29 years of age (figure 1).

[FIGURE 1 OMITTED]

Before surgery, a standardized questionnaire was used to elicit detailed personal and family histories regarding bleeding, intake of drugs that affect coagulation or inhibit platelet function, abnormal epistaxis, bruising or bleeding after injuries or previous surgery (including tooth extraction and circumcision), and medical conditions that affect hemostasis (e.g., liver disease, uremia, and malabsorption). Regular intake of aspirin was permitted until 7 days before surgery.

In addition, all patients were screened by several laboratory tests, including measurement of the complete blood count (CBC), including platelet count; prothrombin time (PT) to evaluate the extrinsic coagulation pathway (factor II, V, VII, and X activity), which were converted to calculate the International Normalized Ratio (INR) (7); and activated partial thromboplastin rime (aPTT) to evaluate the intrinsic coagulation pathway (factor VIII, IX, XI, and XII activity). In children, coagulation values should match age-related values. (8) Measurement of bleeding times was not obtained.

When a detailed analysis of coagulation values was indicated, it included repeated measurements of the CBC, PT, aPTT, bleeding time, von Willebrand's antigen and activity, collagen-binding activity, high-molecular-weight kininogen, and the activity and concentration of factors I, II, V, VII, VIII, IX, X, XI, XII, and XIII.

Tonsillectomy was performed with scissors, a rasp, and a snare. Patients were fitted with a McIvor gag and administered general anesthesia. Adults were prohibited from eating and drinking for at least 6 hours following surgery; the length of the prohibition for children was based on their age. During the prohibition, patients received fluids through an intravenous line. Post-tonsillectomy hemorrhage was treated exclusively with enoral sutures.

Results

Of the 6,966 patients, routine tonsillectomy was performed on 3,076 (44.2%), adenotonsillectomy on 2,160 (31.0%),immediate-abscess tonsillectomy on 1,595 (22.9%), recurrent tonsillectomy to remove remnants of lymphatic tissue on 107 (1.5%), and tonsillectomy to resolve obstruction of the upper airways secondary to infectious mononucleosis in 28 (0.4%).

Post-tonsillectomy hemorrhage occurred in 201 patients (2.9%); 11 of these patients experienced two bleeds and 2 others experienced three bleeds (total bleeds: 216). The incidence of post-tonsillectomy hemorrhage was high among patients between 10 and 29 years of age and low among patients younger than 10 years (figure 1). Of the 216 episodes of bleeding, 168 (77.8%) occurred within 24 hours of surgery--that is, they were primary hemorrhages (figure 2).

[FIGURE 2 OMITTED]

In 199 of the 201 patients, we determined that the bleeding had not been related to a coagulation disorder. We were able to identify a unilateral bleeding vessel in these patients, and we determined that the bleeding was related to either the dissection technique or to arterial hypertension (either temporary or permanent).

A detailed preoperative analysis had been indicated for seven patients--preoperatively in rive children who had repeated abnormal coagulation values prior to surgery and postoperatively in two patients (one child and one adult) who experienced serious post-tonsillectomy bleeding (table 1). Among the rive children with preoperative abnormalities, coagulation values returned to normal in two, two had abnormal values as a result of the presence of lupus anticoagulant, and one had a hereditary deficiency of factors IX and XII. None of these rive experienced any postoperative bleeding. The two patients who underwent a detailed postoperative analysis of coagulation values because of postoperative hemorrhage included a 5-year-old boy and a 40-year-old man. Detailed analysis was indicated because in both, copious bilateral bleeding continued despite all local treatment measures and because the intensity of the bleeding appeared to increase. The 5-year-old boy exhibited blood-tinged sputum immediately after tonsillectomy; he subsequently developed diffuse and intensive bilateral bleeding that required surgical treatment 2 days later (his faucial pillars were sutured together). Although his coagulation values had been normal prior to surgery, detailed laboratory tests revealed that he had a mild factor IX deficiency (activity: 41%). The boy's history was positive; 2 years earlier, he had experienced postadenoidectomy hemorrhage, a fact that his parents had not disclosed until after he experienced his post-tonsillectomy hemorrhage. His post-suture course was uneventful, and purified coagulation factors were not administered. The 40-year-old man had a negative history, but he experienced copious bilateral bleeding that required two surgical procedures under general anesthesia within the first 2 hours following his tonsillectomy. The bleed continued even after the faucial pillars had been sutured together. The sutures were removed, the wounds were covered with gauze packing, and the pillars were sutured together again. A detailed analysis of his coagulation values revealed that he had a factor XIII deficiency (activity: 56%). Despite the fact that he received purified factor XIII, 7 days later he again experienced copious bleeding that required two additional surgical procedures to control. During the second procedure, a packing was sutured over the wounds and removed 2 days later.

Another patient, a 28-year-old man, had been diagnosed with a factor VII deficiency (activity: 20%) during childhood, but he did not undergo detailed analysis or experience any post-tonsillectomy bleeding.

Discussion

Despite our best efforts to avoid it, bleeding remains the most significant complication of tonsillectomy and adenotonsillectomy. The intensity of bleeding sometimes warrants surgical treatment under general anesthesia and/ or the administration of blood transfusions. Among the various risk factors for bleeding that have been suggested (e.g., gender, age, indication for surgery, surgical technique, and hemostasis technique), congenital or acquired coagulation disorders are the most challenging because they are so rare. Controversy surrounds the question as to which is the best method of identifying coagulation abnormalities. Several factors must be taken into account when considering potential hemostatic problems--including the history, coagulation values, screening tests and their reliability, and other aspects.

History. It has been written that a detailed history is the most sensitive tool for predicting post-tonsillectomy hemorrhage. (9) We do not agree, although we certainly believe that taking a history is mandatory. Histories can be inaccurate because of language problems, intentional concealment of information, a desire for surgery, a lack of intelligence, and a failure to recall previous bleeding episodes. Moreover, in the pediatric patient, the reliability of the history has other limitations--particularly that most children have not been exposed to a hemostatic challenge in the past. (10) The 5-year-old boy in whom detailed laboratory testing was indicated had a positive (albeit undisclosed) history of postadenoidectomy bleeding, but his most recent episode was clearly relater to the challenge by surgery itself.

The 40-year-old man had a negative history, but this finding was hot truly relevant because he had not previously undergone any surgical procedure. This is not an unusual circumstance. In a series of 52 adenotonsillectomy patients, Bolger et al found that 40 (76.9%) were undergoing surgery for the first time. (11) Neither does a negative history exclude the possibility that the patient has an acquired or congenital coagulation abnormality. In the series by Bolger et al, 49 of the 52 patients (94.2%) had a negative family history, but 6 patients (11.5%) were later found to have had a hemostatic defect. (11) Moreover, 2 of the patients in that series had denied taking aspirin until after they had experienced postsurgical hemorrhage. Similar findings have been reported by others. (9,12)

Burk et al reported a negative history in two patients, one of whom was known to have von Willebrand's disease and one who had a documenter factor VIII deficiency. (13) De Diego et al described two children with an unremarkable history and normal coagulation values who nevertheless experienced massive post-tonsillectomy hemorrhage. (14) In both cases, a diagnosis of von Willebrand's disease was made after detailed laboratory testing.

Even a positive history appears to have a low predictive value. In a prospective study of 96 tonsillectomy patients, Close et al found that 6 (6.3%) indicated a positive history on a standardized questionnaire. (15) However, none of the 6 experienced post-tonsillectomy hemorrhage, including a 24-year-old woman whose coagulation values were consistent with a variant of von Willebrand's disease. On the other hand, 6 patients who did experience various degrees of bleeding all had a negative history. (15)

In a study of 339 tonsillectomies, Howells et al reported that 10 patients (2.9%) who had a negative personal history experienced secondary (>24 hr) post-tonsillectomy hemorrhage, although one of these patients did have a positive family history. (16) Even so, approximately 25% of all inherited hemostatic defects are the result of a spontaneous mutation or a variable expression of the disease, which clearly limits the value of the family history. (16) In a study by Kang et al, six of seven patients (85.7%) who were newly diagnosed with coagulopathy had no clinical or family history of bleeding. (17)

Coagulation values. During our period of study, tonsillectomy was hOt performed in our clinic on any patient whose hemoglobin concentration was less than 10 g/dl or whose platelet count was less than 150,000/[micro]l. The presence of a functional platelet disorder secondary to diabetes mellitus, hypertension, hyperlipoproteinemia, or smoking (18) would require an evaluation of bleeding rime, which we unfortunately did not perform.

Reliability of screening tests. Several studies have been conducted to evaluate the predictive value of screening methods for post-tonsillectomy hemorrhage. Findings vary (table 2). Deitmer suggested that bleeding time should be used as a screening tool, based on the assumption that most cases of post-tonsillectomy hemorrhage are related to functional platelet disorders. (19) His recommendation differs from that of Schwerdtfeger and Dennebaum, who recommend that bleeding time be measured only in patients who have a positive history. (20) Barber et al showed that bleeding time was not sensitive in predicting posttonsillectomy hemorrhage among patients who had a negative history. (21) It should also be considered that clinical features of functional platelet disorders in children are often missed or misdiagnosed. Finally, regular aspirin intake by children is uncommon.

Lekas et al (12) and Gumprecht and Cichon (22) recommended evaluation of bleeding time in patients with a positive history, whereas Barber et al (21) and Lind (23) argued that it has no predictive value.

The most common hemostatic defects--von Willebrand's disease and platelet dysfunction--can usually be detected by measuring aPTT and bleeding rime, (11,17) although the sensitivity of these tests is limited. (9,15,24) Von Willebrand's disease is acknowledged to be the most common hereditary bleeding disorder. In affected patients, post-tonsillectomy hemorrhage is caused by either a qualitative or quantitative deficiency of von Willebrand' s factor, which is important for platelet aggregation and platelet adhesion to endothelial surfaces. The classic finding in these patients is an increase in aPTT, an increase in bleeding rime, and a decrease in the concentrations of von Willebrand's factor or factor VIII. Another hallmark of this disease is a fluctuation of clinical features and coagulation values. (14)

Ascertainment of PT and INR can detect disturbances of the extrinsic coagulation pathway caused by an intake of oral anticoagulants or secondary to liver disease. These values play a minor role in the detection of coagulation disorders in patients who have a negative history of bleeding. Determination of the aPTT measures the intrinsic pathway of coagulation, and it is useful in diagnosing hemophilias (e.g., deficiencies in factor VIII, IX, or XI). However, aPTT is not sensitive when factor levels exceed 10% of normal. (25)

Lupus anticoagulant is a term used to describe a phenomenon in aPTT testing that occurs when a patient's polyclonal antibodies influence the phospholipid-dependent reaction. In most cases, an abnormally long aPTT is related to the presence of lupus anticoagulant. (26) Polyclonal antibodies were first reported in 1952 by Conley and Hartmann, who detected them in patients who had systemic lupus erythematosus. (26) They can also occur in patients who have minor infections of the upper airway tract, autoimmune deficiencies, Werlhof's disease (idiopathic thrombocytopenic purpura), or malignant lymphomas and in those who take several medications. (13,15,24) The presence of lupus anticoagulant is hot associated with bleeding. We identified lupus anticoagulant in two children who had an abnormal aPTT. Yet despite the benign character of these abnormal coagulation values, the parents of both children chose to refuse surgery. Two other children with lupus anticoagulant did undergo surgery, and neither experienced post-tonsillectomy hemorrhage.

According to the AAO--HNS, screening laboratory tests are indicated only for patients who have a significant history of coagulation disorders. (27) This recommendation is supported by several studies that revealed that abnormal coagulation values are not necessarily associated with bleeding. In a prospective study of 1,603 children, Burk et al observed that 16 of 31 patients (51.6%) who had abnormal coagulation values achieved normalization within 1 week. (13) This coincides with our finding that repeated laboratory tests should be performed before any more detailed and costly tests are indicated. Particularly in children, abnormal aPTT values can occur secondary to repeated infections of the upper airway and then normalize shortly thereafter, as happened in two of our pediatric patients. Because abnormal coagulation values persisted in two other children, a detailed analysis of coagulation values was indicated, and it revealed a mild deficiency of coagulation factors. It is remarkable that post-tonsillectomy hemorrhage did hOt occur in these patients. Of note is the fact that even a diagnosed coagulation-factor deficiency may not necessarily result in postoperative bleeding, as was the case in the previously mentioned 28-year-old man, who had a factor VII deficiency; his postoperative course was uneventful.

Zwack and Derkay reported that laboratory screening tests did not identify two cases of coagulation disorder--von Willebrand's disease and Ehlers-Danlos syndrome. (24) Moreover, they reported that post-tonsillectomy hemorrhage did occur in another patient who had been administered coagulation factors. They concluded that bleeding rates are lower when patients are selectively tested, indicating that routine laboratory tests have a low predictive value for post-tonsillectomy hemorrhage.

In the study by Howells et al, 30 of 39 patients (76.9%) with abnormal coagulation values had only borderline elevations and did not undergo repeat studies. (16) In the 9 patients who did undergo repeat studies, the values of 3 returned to normal and those of 3 others remained prolonged (no post-tonsillectomy hemorrhage). The final 3 patients were found to have lupus anticoagulant, Hageman's factor deficiency, and no disorder, respectively. Post-tonsillectomy hemorrhage occurred in only one of the 39 patients (2.6%).

In their study of 1,603 children, Burk et al found that 15 of 31 patients (48.4%) who had initial laboratory abnormalities continued to experience prolonged abnormalities. (13) Von Willebrand's disease was diagnosed in one of these patients, who was the only patient among 37 who experienced post-tonsillectomy hemorrhage who had an abnormal screening result.

On the other hand, Myssiorek and Alvi recommended that coagulation values be screened routinely. (28) In their report of 14 patients with abnormal PT or aPTT levels who were medically cleared preoperatively, 2 (14.3%) who experienced post-tonsillectomy hemorrhage were later found to have had von Willebrand' s disease.

Kang et al also supported the idea of routine laboratory screening. (17) They studied 1,061 patients who had a normal CBC, PT, and platelet count and found that aPTT values and bleeding times were elevated in 27 patients (2.5%). In this group, 17 (63.0%) experienced a return to normal, 8 (29.6%) had prolonged rimes, and 2 (7.4%) had borderline values; post-tonsillectomy hemorrhage occurred in 4, one, and one, respectively. Von Willebrand's disease was detected in 5 patients, 3 of whom had an additional functional platelet defect or factor V deficiency.

Unidentified coagulation disorder. Compared with screening tests, detailed laboratory tests are costly and should therefore to be ordered with great care. (6,18) It is our strong impression that both of our patients who had a previously unidentified coagulation disorder experienced a characteristic type of post-tonsillectomy hemorrhage: bilateral bleeding of increasing intensity that required repeated local treatments under general anesthesia. Instead of bleeding from a solitary source, these two patients experienced bleeding from the entire surface of the wound.

The estimated incidence of two coagulation disorders--hemophilia B and factor XIII deficiency--is 1 in 50,000 and 1 in 1 million, respectively. (18) According to the literature, we would have been much more likely to have encountered hemophilia A (incidence: 1 in 10,000). The true incidence of von Willebrand's disease is difficult to estimate, but rates of 1 in 500 have been reported. (17,27) However, we did not diagnose any of these abnormalities in any of our patients. The origin of the factor XIII deficiency in our 40-year-old patient remains unclear because his history was insignificant. Congenital factor XIII deficiency is characterized by severe and life-threatening bleeding from the umbilical cord stump and persistent bleeding following injury or surgery during childhood. Unlike the case with hemophilia, joint bleeding is rare in congenital factor XIII deficiency. Acquired factor XIII deficiency is typically associated with liver disease, inflammatory diseases, systemic hematologic diseases, disseminated intravascular coagulation, and septicemia. (29)

Because of the retrospective nature of our study, we were not able to ascertain the predictive value of aPTT and PT because data were not available on patients with borderline or abnormal values who did not undergo surgery. Medicolegal considerations precluded us from operating on these patients.

Conclusions

Based on our findings, we arrived at eight conclusions:

* Patients must complete a standardized questionnaire prior to any surgical procedure, for medicolegal reasons if for no other. However, the surgeon must keep in mind the many factors that can hinder interpretation of results, such as the patient's age, ability to complete questionnaires correctly, intelligence, and desire for surgery, as well as language problems and the intentional concealment of data.

* Normal coagulation values in screening laboratory tests do not exclude the possibility of a coagulation disorder.

* On screening laboratory tests, false-positive results occur frequently. Therefore, a positive test should be repeated 1 week later and the results confirmed before a more detailed and costly laboratory test is ordered.

* Mild hemophilia, von Willebrand's disease, and intake of oral anticoagulants are the most common causes of coagulopathies in clinically normal patients. Such patients can occasionally be identified by the routine measurement of PT, aPTT, and bleeding times, but these coagulopathies can also arise de novo during the postoperative period.

* Abnormal aPTT values are caused in most cases by antiphospholipid antibodies that have no influence on coagulation in vivo.

* Variability in the clinical history and coagulation values is a hallmark of patients with hemophilia.

* A detailed analysis of coagulation values is indicated when copious bilateral bleeding fails to adequately respond to surgical treatment and the intensity of the bleeding is not related to the surgical techniques, arterial hypertension, or the intake of anticoagulants.

* The presence of coagulation disorders in patients who have no significant history and normal coagulation values is extremely rare. However, because post-tonsillectomy hemorrhage is hOt always related to local factors, surgeons should be prepared to treat coagulation disorders when they are present.

Table 1. Characteristics of seven patients who underwent detailed
laboratory analyses

               Preoperative
      Age */   coagulation
Pt.   sex      values                       Results

1      9/M     PT ([dagger]) [down arrow]   Returned to normal

2      4/M     aPTT ([dagger]) [up arrow]   Returned to normal

3      5/M     aPTT [up arrow]              Antiphospholipid antibodies
                                            (lupus anticoagulant)

4      4/M     PT [down arrow]              Factor VII activity: 20%
                                            (lupus anticoagulant)

5      6/M     aPTT [up arrow]              Hereditary factor IX and
                                            XII deficiency

6     40/M     Normal                       Factor XIII deficiency
                                            (activity: 56%)

7      5/M     Normal                       Factor IX deficiency
                                            (activity: 41%)

Pt.   Surgery         PTH ([dagger])

1     AT ([dagger])   No

2     AT              No

3     AT              No

4     AT              No

5     AT              No

6     T ([dagger])    Repeatedly

7     T               Yes

* Age is expressed in years.

([dagger]) PTH = post-tonsillectomy hemorrhage; PT = prothrombin time;
AT = adenotonsillectomy; aPTT = activated partial thromboplastin time;
T = tonsillectomy.

Table 2. Recommendations for preoperative hemostatic assessment
published by others

                                              History      Bleeding
Author                            aPTT/PT *   sufficient   time

Eisenberg et al, (25) 1982        Selective
Kaplan et al, (30) 1985           Selective   Yes
Suchman and Mushlin, (31) 1986    Selective   Yes
Manning et al, (9) 1987           Selective   Yes
Burk et al, (13) 1992             Selective   Yes
Close et al, (15) 1994            Selective   Yes
Howells et al, (16) 1997          Selective   Yes
Zwack and Derkay, (24) 1997       Selective   Yes
Asaf et al, (32) 2001             Selective   Yes
Thomas and Arbon, (33) 1970       Routine
Handler et al, (34) 1986          Routine
Tami et al, (35) 1987             Routine
Bolger et al, (11) 1990           Routine
Smith et al, (36) 1990            Routine
Kang et al, (17) 1994             Routine
Myssiorek and Alvi, (28) 1996     Routine
Gumprecht and Cichon, (22) 1981               Yes          Useful in
                                                           patients
                                                           with a
                                                           positive
                                                           history
Lekas et al, (12) 1982                        Yes          Useful in
                                                           patients
                                                           with a
                                                           positive
                                                           history
Barber et al, (21) 1985                                    Not useful
Lind, (23) 1991                                            Not useful

* a PTT = activated partial thromboplastin time; 17 = prothrombin time

References

(1.) Pokras R, Kozak LJ, McCarthy E, and the National Center for Health Statistics. Ambulatory and inpatient procedures in the United States, 1994. Vital and Health Statistics. Series 13. Data from the National Health survey, No. 132, 1994:11,

(2.) Yardley MP. Tonsillectomy, adenoidectomy and adenotonsillectomy: Are they safe day case procedures? J Laryngol Otol 1992;106:299-300.

(3.) Windfuhr JP. Unanswered questions about tonsillectomy. Presented at the annual meeting of the West German ENT Society; Match 29, 2003; Essen, Germany.

(4.) Clinical indicators compendium 2000. Alexandria, Va.: American Academy of Otolaryngology-Head and Neck Surgery, 2000:19.

(5.) Leuwer R, Petri S, Schulz F, Puschel K. Deaths from tonsillectomy and adenotomy. Laryngol Rhinol Otol 1998;77:669-72.

(6.) Tonsillectomy and Adenoidectomy Inpatient Guidelines. Recommendations of the AAO-HNS Pediatric Otolaryngology Committee. Alexandria, Va.: American Academy of Otolaryngology-Head and Neck Surgery, 1996:1-4.

(7.) Beeser H. Critical evaluation of the so far experience using the WHO model of prothrombin time calibration and outlook for further development. Haemostasis 1991; 18(Suppl 2): 181-2.

(8.) Gallistl S, Muntean W, Leschnik B, Meyers W. Longer aPTT values in healthy children than in adults: No single cause. Thromb Res 1997;88:355-9.

(9.) Manning SC, Beste D, McBride T, Goldberg A. An assessment of preoperative coagulation screening for tonsillectomy and adenoidectomy. Int J Pediatr Otorhinolaryngol 1987; 13:237-44.

(10.) Goddard KE, Broder G, Wenar C. Reliability of pediatric histories. Pediatrics 1961;28:1011-19.

(11.) Bolger WE, Parsons DS, Potempa L. Preoperative hemostatic assessment of the adenotonsillectomy patient. Otolaryngol Head

(12.) Lekas MD, DiBenedetto J, Jr., Smith PS, Kantesaria KP. Surgery on patients with hemostatic disorders. Laryngoscope 1982;92: 873-7.

(13.) Burk CD, Miller L, Handler SD, Cohen AR. Preoperative history and coagulation screening in children undergoing tonsillectomy. Pediatrics 1992;89:691-5.

(14.) de Diego JI, Prim MP, Rodriguez E, et al. Von Willebrand disease as cause of unanticipated bleeding following adeno-tonsillectomy. Int J Pediatr Otorhinolaryngol 1999;49:185-8.

(15.) Close HL, Kryzer TC, Nowlin JH, Alving BM. Hemostatic assessment of patients before tonsillectomy: A prospective study. Otolaryngol Head Neck Surg 1994; 111:733-8.

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

(17.) Kang J, Brodsky L, Danziger 1, et al. Coagulation profile as a predictor for post-tonsillectomy and adenoidectomy (T + A) hemorrhage. Int J Pediatr Otorhinolaryngol 1994;28:157-65.

(18.) Martlew VJ. Peri-operative management of patients with coagulation disorders. Br J Anaesth 2000;85:446-55.

(19.) Deitmer T. [Hemostasis screening before ENT operations]. Laryngorhinootologie 1989;68:188.

(20.) Schwerdtfeger P, Dennebaum R. [Blond coagulation physiologic study in tonsillectomy]. HNO 1984;32:450-3.

(21.) Barber A, Green D, Galluzzo T, Ts'ao CH. The bleeding time as a preoperative screening test. Am J Med 1985;78:761-4.

(22.) Gumprecht TF, Cichon JV. Otolaryngology and von Willebrand's disease. Arch Otolaryngol 1981;107:491-3.

(23.) Lind SE. The bleeding time does hot predict surgical bleeding. Blond 1991;77:2547-52.

(24.) Zwack GC, Derkay CS. The utility of preoperative hemostatic assessment in adenotonsillectomy. Int J Pediatr Otorhinolaryngol 1997;39:67-76.

(25.) Eisenberg JM, Clarke JR, Sussman SA. Prothrombin and partial thromboplastin rimes as preoperative screening tests. Arch Surg 1982;117:48-51.

(26.) Conley CL, Hartmann RC. A hemorrhagic disorder caused by circulating anticoagulants in patients with disseminated lupus erythematosus. J Clin Invest 1952;31:667-84.

(27.) Dana ST, Bogdasarian RS, Burgess LP. Clinical indicators compendium. Alexandria, Va.: American Academy of Otolaryngology-Head and Neck Surgery Bulletin, 1992.

(28.) Myssiorek D, Alvi A. Post-tonsillectomy hemorrhage: An assessment of risk factors. Int J Pediatr Otorhinolaryngol 1996;37: 35-43.

(29.) Egbring R, Kroniger A, Seitz R. Factor XIII deficiency: Pathogenic mechanisms and clinical significance. Semin Thromb Hemost 1996;22:419-25.

(30.) Kaplan EB, Sheiner LB, Boeckmann A J, et al. The usefulness of preoperative laboratory screening. JAMA 1985;253:3576-81.

(31.) Suchman AL, Mushlin AI. How well does the activated partial thromboplastin rime predict postoperative hemorrhage? JAMA 1986;256:750-3.

(32.) Asaf T, Reuveni H, Yermiahu T, et al. The need for routine preoperative coagulation screening tests (prothrombin time PT/ partial thromboplastin rime PTT) for healthy children undergoing elective tonsillectomy and/or adenoidectomy. Int J Pediatr Otorhinolaryngol 2001;61:217-22.

(33.) Thomas GK, Arbon RA. Preoperative screening for potential T&A bleeding. Arch Otolaryngol 1970;91:453-6.

(34.) Handler SD, Millet L, Richmond KH, Baranak CC. Post-tonsillectomy hemorrhage: Incidence, prevention and management. Laryngoscope 1986;96:1243-7.

(35.) Tami TA, Parker GS, Taylor RE. Post-tonsillectomy bleeding: An evaluation of risk factors. Laryngoscope 1987;97:1307-11.

(36.) Smith PS, Orchard PJ, Lekas MD. Predicting bleeding in common car, nose, and throat procedures: A prospective study. R I Med J 1990;73:103-6.

From the Department of Otorhinolaryngology-Plastic Head and Neck Surgery, St. Anna Hospital, Duisburg, Germany (Dr. Windfuhr and Dr. Remmert), and the Department of Otorhinolaryngology--Plastic Head and Neck Surgery, University Hospital, Aachen, Germany (Dr. Chen).

Reprint requests: Jochen P. Windfuhr, MD, Department of Otorhinolaryngology--Plastic Head and Neck Surgery, St. Anna Hospital, Albertus Magnus Str. 33, 47259 Duisburg, Germany. Phone: [phone]; fax: [phone]; e-mail:[e-mail]