- Email: [email protected]
Management of secondary hemorrhage following pediatric adenotonsillectomy
International Journal of Pediatric Otorhinolaryngology 40 (1997) 115-124
ELSEVIER
Management of secondary hemorrhage following pediatric adenotonsillectomy Danesh B. Irani, Robert G. Berkowitz * Department
of Otolaryngology,
Royal
Children’s Hospital, Flemington Victoria 3052, Australia
Road,
Parkville,
Melbourne,
Received 10 September 1996; received in revised form 13 February 1997; accepted 18 February 1997
Abstract A retrospective study was performed of all patients requiring admission to the Royal Children’s Hospital, Melbourne over a 12 year period with secondary haemorrhage following adenotonsillectomy, to determine what percentage of these children received blood transfusions or were returned to the operating room to secure hemostasis, and to identify factors predictive of the need for major intervention. There were 163 children who presented from 2 to 15 days following surgery. Initial management in all cases was establishment of intravenous access,and 151 received intravenous or oral antibiotics. One hundred and forty one were managed without the need for major intervention (87%), including five who had silver nitrate cautery to the tonsillar fossae. Major intervention was required in 22 cases (13%): 5 patients were returned to the operating room for hemostasis; 15 received blood transfusions and 2 underwent both. All surgery was required within 12 h of admission and all blood transfusions within 24 h. The highest rates of major intervention were in those with fresh bleeding at the time of presentation (38%) and hemoglobin levels less than 100 g/l (36%). For those requiring admission with secondary haemorrhage, a period of observation of 24 h would probably be adequate in the majority of casesto identify those children who will require major intervention by surgery or transfusion. 0 1997 Elsevier Science Ireland Ltd. Keywords:
Adenotonsillectomy; Tonsillectomy; Secondary haemorrhage
* Corresponding
author. Tel.: + 61 3 93456476; fax: + 61 3 93455595.
0165-5876/97/$17,00 0 1997 Elsevier Science Ireland Ltd. All rights reserved, PII SO165-5876(97)00025-6
116
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
1. Introduction Adenotonsillectomy is one of the most common surgical procedures performed in the pediatric age group and haemorrhage, which may be classified as primary (intra-operative), reactionary (appearing within the first 24 h post-operatively), or secondary (occurring after the first 24 h) [16], remains the most frequent complication [2,1 l] and its management is often clinically challenging. Secondary haemorrhage rates following pediatric adenotonsillectomy reported in the literature range from 0.5% to 9.3% [1,3,7,8,10,12,13,15,17,18,20], with a mean rate of 1.5%, as outlined in Table 1. Rates as high as 10% have been quoted by Fox in 1952 [5] and Neivert in 1945 [14] but significantly the patients in these older studies were given salicylates as post-operative analgesia. Prior to consenting to adenotonsillectomy for their children, parents are informed of the risk of secondary haemorrhage, and the usually quoted figure of 2% is often of concern to many parents. It is therefore important to know both from a clinical and medico-legal point of view what proportion of patients with secondary haemorrhage actually develop serious sequelae or require major intervention as a consequence. We therefore undertook a review of our experience with secondary haemorrhage following adenotonsillectomy in pediatric patients to determine the likelihood of significant intervention being required. We also attempted to identify factors that were predictive of the need for major intervention and would thereby allow admission criteria for children presenting with bleeding to be developed. As our institution is the major pediatric tertiary referral centre for a population of about 5 million, this series included both patients who had their initial surgery performed at our institution and also those that underwent surgery at other hospitals, while a small number of cases that had their initial operation at our hospital may have presented elsewhere with bleeding. Hence no attempt has been made to calculate the secondary haemorrhage rate for this group of children.
2. Methods A retrospective chart review was carried out on all patients requiring inpatient treatment at the Royal Children’s Hospital, Melbourne, Australia, for secondary haemorrhage from the tonsillar fossa following adenotonsillectomy or tonsillectomy alone, over a 12 year period from January 1st 1984 to December 31st 1995. All patients presenting to the emergency department were admitted except for those in whom there was a history of only very minor bleeding and normal findings on oropharyngeal examination. There were 163 patients (76 males, 87 females) with ages ranging from 1 to 16 years (mean 7.1) and weights ranging from 12.2 to 74.5 kg (mean 26.1). Details of the original surgery, clinical and laboratory findings at presentation, and subsequent management including the need for intervention were recorded.
1990 1991 1988 1993 1983 1985 1986 1995 1985 1993 1979
PI
Totals
PI
[71 1141 I171 191
WI 1161 1111
[I91 131
Year
hemorrhage
Author
Table 1 Secondary
<3 <3 Children Children Children <14 Children 2-17 <16
Patient
rates
pediatric
age (years)
following of patients
190 200 6842 7743 430 3240 1445 255 247 413 108 21 113
No.
adenotonsillectomy
I 2 80 105 6 48 36 7 7 16 10 318
No.
as reported of secondary
in the recent hemorrhages
literature
0.5 1.0 1.2 1.4 1.4 1.5 2.5 2.7 2.8 3.9 9.3 Mean
Rate
1.5
of secondary
hemorrhage
(%)
No. of patients
52 83 20 8 163 7.0 days
Time of presentation (days)
2-5 6-10 11-15 Not stated Total
Mean time of presentation
6.9 days
45 73 15 8 141
No major intervention
Table 2 Time of presentation with secondary hemorrhage following adenotonsillectomy required
5.7 days
5
22 (13%) 7.6 days
2 2 1
8.3 days
15
4 7 4
Transfusion
required Surgery
7 (14%) 10 (12%) 5 (25%)
Total
Major intervention
3.5 days
2
1 1
Both
D.B. Irani, R.G. Berkowitz /Int. J. Pediatr. Otorhinolaryngol.
119
40 (1997) 115-124
All patients underwent observation, with insertion of an intravenous line and fluid replacement as required. Antibiotics were administered in most cases. Where clinically indicated and where there was sufficient patient co-operation, silver nitrate cautery was attempted using topical local anesthetic. Major intervention was defined as the need for blood transfusion or return to the operating room to secure hemostasis under general anesthetic. 3. Results One hundred and sixty-three patients met the criteria for inclusion in this study of which 109 had undergone adenotonsillectomy and 54 tonsillectomy alone. Initial surgery was carried out at the Royal Children’s Hospital in 85 cases, while 78 children were referred following surgery at other centers. The surgery was performed by an otolaryngology resident in 61 cases, and by an attending in 92. The identity of the surgeon was not recorded for 10 patients. The technique of tonsillectomy was diathermy in 59 cases, dissection in 53, and not stated in 51. The indication for surgery was recurrent tonsillitis in 86 cases, obstructive sleep apnoea in 30, and both problems in 14. No indication was documented for 35 children. The time of presentation following surgery was recorded in 155 cases, and ranged from 2 to 15 days (mean 7.0), with only 20 patients (13%) presenting with bleeding after the 10th post-operative day. Details are outlined in Table 2. Of those patients presenting with secondary haemorrhage at day 2-5, 14% required major intervention; at day 6- 10, 12% underwent major intervention; and at day 1 1- 15, the major intervention rate was 25%. The oropharyngeal findings at initial presentation are listed in Table 3. Only 5% of children with clot, and 12% of those with exudate only seen on oropharyngeal examination needed major intervention, whereas 38% of patients with fresh blood visible in the oropharynx ultimately required surgery or blood transfusion. Table 4 outlines the first hemoglobin estimation made at our institution and includes the major intervention rate for each hemoglobin range. Hemoglobin levels were between 54 and 164 g/l (mean 1 II). The time blood was taken with respect to bleeding and transfusion however was not accurately recorded in the charts. Table 3 Oropharyngeal Oropharyngeal findings
Fresh blood or ooze Exudate only Clot only Total
findings at initial presentation No. of patients
No major intervention required
Major intervention
Total
Surgery
required Transfusion
Both
26
16
10 (38%)
4
4
2
75 62 163
66 59 141
9 (12%) 3 (5%) 22 (13%)
~ I 5
9 2 15
2
No. of patients
12 35 41 41 22 163 111 g/l
Hemoglobin level g/l
<80 81-100 101-120 >I20 Not done Total
Mean hemoglobin level
Table 4 Initial hemoglobin levels
114 g/l
6 24 45 44 22 141
No major intervention
required
5
22 (13%)
111 g/l
1 1 1 2
92 g/l
required
Surgery
(50%) (31%) (4.3%) (6.4)
6 11 2 3
Total
Major intervention
2 105 g/l
86 g/l
1
1
Both
15
5 9 1
Transfusion
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
121
115-124
Table 5 Time of major intervention following admission with secondary hemorrhage Time of intervention (h)
<12 12-24 >24 Total
Major intervention required Total
Surgery
Transfusion
Both
17 (77%) 5 (23%)
5
2
22
5
10 5 15
2
No child who underwent a coagulation profile was found to have a significant underlying coagulopathy as a cause for their secondary haemorrhage. Initial management in all 163 cases was establishment of intravenous access, and 151 patients received either intravenous or oral antibiotics. One hundred and forty one patients (87%) needed no major intervention including 5 who received silver nitrate cautery to the tonsillar fossae under local anesthetic, and had no significant bleeding thereafter. Major intervention was required in 22 cases (13%). Five patients were returned to the operating room for hemostasis, 15 received blood transfusion, and two children underwent both surgery and transfusion. Their details with respect to time of presentation, oropharyngeal findings and hemoglobin levels (including mean) have been outlined in Tables 2-4. The time of major intervention following admission is listed in Table 5. All surgery was carried out within 12 h of admission, while transfusion was required within 12 h in 12 cases, and between 12 and 24 h in the remaining five. No major intervention was carried out more than 24 h after admission. No child needed to return to the operating room on more than one occasion. Of the 17 blood transfusions administered, there were 12 between January 1st 1984 and December 31st 1989, but only five from January 1st 1990 to December 31st 1995. The mean stay in hospital for those patients not requiring intervention was 3.2 days (range 2-4), and 4.7 days (range 3-7) for the group needing major intervention. No other significant morbidity was recorded and there were no deaths.
4. Discussion The major intervention rate for patients admitted with secondary haemorrhage following adenotonsillectomy in this series was 13%, and compares favourably to the recent literature which reports figures from 17% to 56% [3,4,6,7,1 l] as outlined in Table 6, although only two of those cited were exclusively pediatric series. The low rate in our series may be due to a liberal policy of admitting patients presenting with secondary haemorrhage who could otherwise have been managed as outpatients. Other authors, such as Handler et al. [7], while reporting much higher rates of intervention for children admitted with secondary haemorrhage
1990 1988 1962 1984 1986
[41 t31
171 Total
[lOI
161
Year
6 intervention
Author
Table Major patients
age (years)
for
All ages Children 2 years-adult 2-71 years Children
Patient
rates
with
secondary
of patients
3340 6842 loo 1150 1445 12877
No.
admitted
29 80 7 64 36 216
No.
of secondary
hemorrhage hemorrhages 5 25 3 32 20 85
No.
of interventions
17 31 43 50 56 Mean
Rate
39
of intervention
(%)
D.B.
Irani,
R.G.
Berkowitz
/ Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
123
(56%), excluded those with relatively minor bleeding who were observed for several h and then discharged if no clot was noted in the tonsillar fossa, and no further bleeding occurred. A return to the operating room for hemostasis is often advised for children with secondary haemorrhage following adenotonsillectomy who have active bleeding or clot in the tonsillar at the time of admission. In this retrospective study we found that the approach to this group of children was more conservative with only 10 of the 26 (38%) patients with fresh bleeding and three of the 62 (5%) with clot in the tonsillar fossa requiring major intervention. Had all the patients who were found to have oozing or clot in the tonsillar fossa been routinely taken back to the operating room for hemostasis, the major intervention rate for all patients in this series would have been 59%. Not surprisingly, the highest rates of major intervention were in those children with fresh bleeding at the time of presentation (38O), and where hemoglobin levels were less than 100 g/l (36%). Of the 17 blood transfusions administered, there were 12 between January 1st 1984 and December 31st 1989, but only 5 in the following 6 year period. The indication for transfusion was acute blood loss associated with cardiovascular compromise. The mean hemoglobin of the children receiving blood was 86 g/l, however hemoglobin levels quoted in this series were the initial assays performed at our institution, and a number of children had commenced blood transfusion prior to their arrival. Our current approach of administering plasma expanders, close monitoring of vital signs, and judicious surgical intervention aims to reduce the number of unnecessary transfusions and the associated risk of viral infection and other diseases transmitted by blood. There were no documented cases of coagulopathy amongst the patients admitted with secondary haemorrhage. While we do not routinely screen our patients for a coagulopathy prior to adenotonsillectomy, Kang et al. [93 did find that an initially abnormal coagulation profile may identify those more likely to bleed after surgery, and suggested that a coagulation profile may be a valuable screening tool for children undergoing adenotonsillectomy. All surgery was required within 12 h of admission and all blood transfusion within 24 h. No major intervention was required more than 24 h after admission in this series and it would appear that after a period of observation of 24 h it is probably safe to discharge the majority of cases, if no further bleeding is noted and there are no other ongoing problems. Eighty-seven per cent of the children presented at day 10 or earlier, with the latest presentation being 15 days following adenotonsillectomy. However, while Chowdhury et al. [3] in a study of 80 children with secondary bleeds, noted a wide range of timing in readmission for delayed haemorrhage from 28 h to 21 days post-operatively, most authors agree that secondary haemorrhage is uncommon after the 14th post-operative day [3,17,19], and usually occurs from 5 to 10 days after surgery. In conclusion, while the literature indicates that it is appropriate to inform parents of a 2% risk of secondary haemorrhage following adenotonsillectomy, our
124
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
major intervention rate of 13% in this series suggests that the overall risk of the need for blood transfusion or return to the operating room for hemostasis as a consequence of secondary bleeding is in the order of 0.3% for each child undergoing adenotonsillectomy in our institution. This is a useful figure to quote to parents, helping them make an informed decision to accept adenotonsillectomy for their child.
References [l] Berkowitz, R.G. and Zalzal, G.H. (1990) Tonsillectomy in children under 3 years of age. Arch. Otolaryngol. Head Neck Surg. 116, 6855686. [2] Carmody, D., Vamadevan, T. and Cooper, S.M. (1982) Post-tonsillectomy haemorrhage. J. Laryngol. Otol. 96, 635-638. [3] Chowdhury, K., Tewfik, T.L. and Schloss, M.D. (1988) Post-tonsillectomy and adenoidectomy haemorrhage. J. Otolaryngol. 17, 46649. [4] Colclasure, J.B. and Graham, S.S. (1990) Complications of outpatient tonsillectomy and adenoidectomy: a review of 3340 cases. Ear Nose and Throat J. 69, 155-160. [5] Fox, S.L. (1952) Bleeding Following tonsillectomy. Laryngoscope 62, 414-425. [6] Haase, F.R. and Noguera, S.T. (1962) Hemostasis in tonsillectomy by electrocautery. Arch. Otolaryngol. 75, 125-126. [7] Handler, S.D., Miller, L., Richmond, K.H. and Baranak, CC. (1986) Post tonsillectomy haemorrhage: incidence, prevention and management. Laryngoscope 96, 1243-1247. [8] Helmus, C. (1979) Tonsillectomy and adenoidectomy in the one and two-year-old child. Laryngoscope 89, 17641771. [9] Kang, J., Brodsky, L., Danziger, I., Volk, M. and Stanievich, J. (1995) Coagulation profile as a predictor for post-tonsillectomy and adenoidectomy haemorrhage. Int. J. Pediatr. Otorhinolaryngol. 32, 261-263. [lo] Kendrick, D. and Gibbin, K. (1993) An audit of the complications of paediatric tonsillectomy, adenoidectomy and adenotonsillectomy. Clin. Otolaryngol. 18, 115- 117. [l l] Kristensen, S. and Tveteras, K. (1984) Post-tonsillectomy haemorrhage. A retrospective study of 1150 operations. Clin. Otolaryngol. 9, 347-350. [12] Lee, I.N. (1985) Outpatient management of T and A procedure in children. J. Otolaryngol. 14, 176-177. [13] Mum, I. and Simon, H. (1993) Hemorrhagic complications after tonsillectomy and adenoidectomy. Experiences with 7743 operation in 14 years. Weiner Klinische Wochenschrift 105, 520-522. [14] Neivert, H. (1945) Late secondary tonsillar haemorrhage. Arch. Otolaryngol. 42, 14- 18. [15] Nicklaus, P.J., Herzon, F.S. and Steinle, E.W. 4th. (1995) Short-stay outpatient tonsillectomy. Arch. Otolaryngol. Head Neck Surg. 121, 521-524. [16] Roberts, C., Jayaramachandran, S. and Raine, C.H. (1992) A prospective study of factors which may predispose to post-operative tonsillar fossa haemorrhage. Clin. Otolaryngol. 17, 13- 17. [17] Segal, C., Berger, G., Basker, M. and Marshak, M. (1983) Adenotonsillectomies on a Surgical Day-Clinic Basis. Laryngoscope 93, 1205-1207. [18] Siodlak, M.Z., Gleeson, M.J. and Wengraf, C.L. (1985) Post-tonsillectomy secondary haemorrhage. Ann. R. Coll. Surg. Eng. 67, 167- 168. [19] Tami, T.A., Parker, G.S. and Taylor, R.E. (1987) Post-tonsillectomy bleeding: An evaluation of risk factors. Laryngoscope 97, 1307- 1311. [20] Wiatrak, B.J., Myer, C.M. 3rd and Andrews, T.M. (1991) Complications of adenotonsillectomy in children under 3 years of age. Am. J. Otolaryngol. 12, 170- 172.
ELSEVIER
Management of secondary hemorrhage following pediatric adenotonsillectomy Danesh B. Irani, Robert G. Berkowitz * Department
of Otolaryngology,
Royal
Children’s Hospital, Flemington Victoria 3052, Australia
Road,
Parkville,
Melbourne,
Received 10 September 1996; received in revised form 13 February 1997; accepted 18 February 1997
Abstract A retrospective study was performed of all patients requiring admission to the Royal Children’s Hospital, Melbourne over a 12 year period with secondary haemorrhage following adenotonsillectomy, to determine what percentage of these children received blood transfusions or were returned to the operating room to secure hemostasis, and to identify factors predictive of the need for major intervention. There were 163 children who presented from 2 to 15 days following surgery. Initial management in all cases was establishment of intravenous access,and 151 received intravenous or oral antibiotics. One hundred and forty one were managed without the need for major intervention (87%), including five who had silver nitrate cautery to the tonsillar fossae. Major intervention was required in 22 cases (13%): 5 patients were returned to the operating room for hemostasis; 15 received blood transfusions and 2 underwent both. All surgery was required within 12 h of admission and all blood transfusions within 24 h. The highest rates of major intervention were in those with fresh bleeding at the time of presentation (38%) and hemoglobin levels less than 100 g/l (36%). For those requiring admission with secondary haemorrhage, a period of observation of 24 h would probably be adequate in the majority of casesto identify those children who will require major intervention by surgery or transfusion. 0 1997 Elsevier Science Ireland Ltd. Keywords:
Adenotonsillectomy; Tonsillectomy; Secondary haemorrhage
* Corresponding
author. Tel.: + 61 3 93456476; fax: + 61 3 93455595.
0165-5876/97/$17,00 0 1997 Elsevier Science Ireland Ltd. All rights reserved, PII SO165-5876(97)00025-6
116
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
1. Introduction Adenotonsillectomy is one of the most common surgical procedures performed in the pediatric age group and haemorrhage, which may be classified as primary (intra-operative), reactionary (appearing within the first 24 h post-operatively), or secondary (occurring after the first 24 h) [16], remains the most frequent complication [2,1 l] and its management is often clinically challenging. Secondary haemorrhage rates following pediatric adenotonsillectomy reported in the literature range from 0.5% to 9.3% [1,3,7,8,10,12,13,15,17,18,20], with a mean rate of 1.5%, as outlined in Table 1. Rates as high as 10% have been quoted by Fox in 1952 [5] and Neivert in 1945 [14] but significantly the patients in these older studies were given salicylates as post-operative analgesia. Prior to consenting to adenotonsillectomy for their children, parents are informed of the risk of secondary haemorrhage, and the usually quoted figure of 2% is often of concern to many parents. It is therefore important to know both from a clinical and medico-legal point of view what proportion of patients with secondary haemorrhage actually develop serious sequelae or require major intervention as a consequence. We therefore undertook a review of our experience with secondary haemorrhage following adenotonsillectomy in pediatric patients to determine the likelihood of significant intervention being required. We also attempted to identify factors that were predictive of the need for major intervention and would thereby allow admission criteria for children presenting with bleeding to be developed. As our institution is the major pediatric tertiary referral centre for a population of about 5 million, this series included both patients who had their initial surgery performed at our institution and also those that underwent surgery at other hospitals, while a small number of cases that had their initial operation at our hospital may have presented elsewhere with bleeding. Hence no attempt has been made to calculate the secondary haemorrhage rate for this group of children.
2. Methods A retrospective chart review was carried out on all patients requiring inpatient treatment at the Royal Children’s Hospital, Melbourne, Australia, for secondary haemorrhage from the tonsillar fossa following adenotonsillectomy or tonsillectomy alone, over a 12 year period from January 1st 1984 to December 31st 1995. All patients presenting to the emergency department were admitted except for those in whom there was a history of only very minor bleeding and normal findings on oropharyngeal examination. There were 163 patients (76 males, 87 females) with ages ranging from 1 to 16 years (mean 7.1) and weights ranging from 12.2 to 74.5 kg (mean 26.1). Details of the original surgery, clinical and laboratory findings at presentation, and subsequent management including the need for intervention were recorded.
1990 1991 1988 1993 1983 1985 1986 1995 1985 1993 1979
PI
Totals
PI
[71 1141 I171 191
WI 1161 1111
[I91 131
Year
hemorrhage
Author
Table 1 Secondary
<3 <3 Children Children Children <14 Children 2-17 <16
Patient
rates
pediatric
age (years)
following of patients
190 200 6842 7743 430 3240 1445 255 247 413 108 21 113
No.
adenotonsillectomy
I 2 80 105 6 48 36 7 7 16 10 318
No.
as reported of secondary
in the recent hemorrhages
literature
0.5 1.0 1.2 1.4 1.4 1.5 2.5 2.7 2.8 3.9 9.3 Mean
Rate
1.5
of secondary
hemorrhage
(%)
No. of patients
52 83 20 8 163 7.0 days
Time of presentation (days)
2-5 6-10 11-15 Not stated Total
Mean time of presentation
6.9 days
45 73 15 8 141
No major intervention
Table 2 Time of presentation with secondary hemorrhage following adenotonsillectomy required
5.7 days
5
22 (13%) 7.6 days
2 2 1
8.3 days
15
4 7 4
Transfusion
required Surgery
7 (14%) 10 (12%) 5 (25%)
Total
Major intervention
3.5 days
2
1 1
Both
D.B. Irani, R.G. Berkowitz /Int. J. Pediatr. Otorhinolaryngol.
119
40 (1997) 115-124
All patients underwent observation, with insertion of an intravenous line and fluid replacement as required. Antibiotics were administered in most cases. Where clinically indicated and where there was sufficient patient co-operation, silver nitrate cautery was attempted using topical local anesthetic. Major intervention was defined as the need for blood transfusion or return to the operating room to secure hemostasis under general anesthetic. 3. Results One hundred and sixty-three patients met the criteria for inclusion in this study of which 109 had undergone adenotonsillectomy and 54 tonsillectomy alone. Initial surgery was carried out at the Royal Children’s Hospital in 85 cases, while 78 children were referred following surgery at other centers. The surgery was performed by an otolaryngology resident in 61 cases, and by an attending in 92. The identity of the surgeon was not recorded for 10 patients. The technique of tonsillectomy was diathermy in 59 cases, dissection in 53, and not stated in 51. The indication for surgery was recurrent tonsillitis in 86 cases, obstructive sleep apnoea in 30, and both problems in 14. No indication was documented for 35 children. The time of presentation following surgery was recorded in 155 cases, and ranged from 2 to 15 days (mean 7.0), with only 20 patients (13%) presenting with bleeding after the 10th post-operative day. Details are outlined in Table 2. Of those patients presenting with secondary haemorrhage at day 2-5, 14% required major intervention; at day 6- 10, 12% underwent major intervention; and at day 1 1- 15, the major intervention rate was 25%. The oropharyngeal findings at initial presentation are listed in Table 3. Only 5% of children with clot, and 12% of those with exudate only seen on oropharyngeal examination needed major intervention, whereas 38% of patients with fresh blood visible in the oropharynx ultimately required surgery or blood transfusion. Table 4 outlines the first hemoglobin estimation made at our institution and includes the major intervention rate for each hemoglobin range. Hemoglobin levels were between 54 and 164 g/l (mean 1 II). The time blood was taken with respect to bleeding and transfusion however was not accurately recorded in the charts. Table 3 Oropharyngeal Oropharyngeal findings
Fresh blood or ooze Exudate only Clot only Total
findings at initial presentation No. of patients
No major intervention required
Major intervention
Total
Surgery
required Transfusion
Both
26
16
10 (38%)
4
4
2
75 62 163
66 59 141
9 (12%) 3 (5%) 22 (13%)
~ I 5
9 2 15
2
No. of patients
12 35 41 41 22 163 111 g/l
Hemoglobin level g/l
<80 81-100 101-120 >I20 Not done Total
Mean hemoglobin level
Table 4 Initial hemoglobin levels
114 g/l
6 24 45 44 22 141
No major intervention
required
5
22 (13%)
111 g/l
1 1 1 2
92 g/l
required
Surgery
(50%) (31%) (4.3%) (6.4)
6 11 2 3
Total
Major intervention
2 105 g/l
86 g/l
1
1
Both
15
5 9 1
Transfusion
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
121
115-124
Table 5 Time of major intervention following admission with secondary hemorrhage Time of intervention (h)
<12 12-24 >24 Total
Major intervention required Total
Surgery
Transfusion
Both
17 (77%) 5 (23%)
5
2
22
5
10 5 15
2
No child who underwent a coagulation profile was found to have a significant underlying coagulopathy as a cause for their secondary haemorrhage. Initial management in all 163 cases was establishment of intravenous access, and 151 patients received either intravenous or oral antibiotics. One hundred and forty one patients (87%) needed no major intervention including 5 who received silver nitrate cautery to the tonsillar fossae under local anesthetic, and had no significant bleeding thereafter. Major intervention was required in 22 cases (13%). Five patients were returned to the operating room for hemostasis, 15 received blood transfusion, and two children underwent both surgery and transfusion. Their details with respect to time of presentation, oropharyngeal findings and hemoglobin levels (including mean) have been outlined in Tables 2-4. The time of major intervention following admission is listed in Table 5. All surgery was carried out within 12 h of admission, while transfusion was required within 12 h in 12 cases, and between 12 and 24 h in the remaining five. No major intervention was carried out more than 24 h after admission. No child needed to return to the operating room on more than one occasion. Of the 17 blood transfusions administered, there were 12 between January 1st 1984 and December 31st 1989, but only five from January 1st 1990 to December 31st 1995. The mean stay in hospital for those patients not requiring intervention was 3.2 days (range 2-4), and 4.7 days (range 3-7) for the group needing major intervention. No other significant morbidity was recorded and there were no deaths.
4. Discussion The major intervention rate for patients admitted with secondary haemorrhage following adenotonsillectomy in this series was 13%, and compares favourably to the recent literature which reports figures from 17% to 56% [3,4,6,7,1 l] as outlined in Table 6, although only two of those cited were exclusively pediatric series. The low rate in our series may be due to a liberal policy of admitting patients presenting with secondary haemorrhage who could otherwise have been managed as outpatients. Other authors, such as Handler et al. [7], while reporting much higher rates of intervention for children admitted with secondary haemorrhage
1990 1988 1962 1984 1986
[41 t31
171 Total
[lOI
161
Year
6 intervention
Author
Table Major patients
age (years)
for
All ages Children 2 years-adult 2-71 years Children
Patient
rates
with
secondary
of patients
3340 6842 loo 1150 1445 12877
No.
admitted
29 80 7 64 36 216
No.
of secondary
hemorrhage hemorrhages 5 25 3 32 20 85
No.
of interventions
17 31 43 50 56 Mean
Rate
39
of intervention
(%)
D.B.
Irani,
R.G.
Berkowitz
/ Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
123
(56%), excluded those with relatively minor bleeding who were observed for several h and then discharged if no clot was noted in the tonsillar fossa, and no further bleeding occurred. A return to the operating room for hemostasis is often advised for children with secondary haemorrhage following adenotonsillectomy who have active bleeding or clot in the tonsillar at the time of admission. In this retrospective study we found that the approach to this group of children was more conservative with only 10 of the 26 (38%) patients with fresh bleeding and three of the 62 (5%) with clot in the tonsillar fossa requiring major intervention. Had all the patients who were found to have oozing or clot in the tonsillar fossa been routinely taken back to the operating room for hemostasis, the major intervention rate for all patients in this series would have been 59%. Not surprisingly, the highest rates of major intervention were in those children with fresh bleeding at the time of presentation (38O), and where hemoglobin levels were less than 100 g/l (36%). Of the 17 blood transfusions administered, there were 12 between January 1st 1984 and December 31st 1989, but only 5 in the following 6 year period. The indication for transfusion was acute blood loss associated with cardiovascular compromise. The mean hemoglobin of the children receiving blood was 86 g/l, however hemoglobin levels quoted in this series were the initial assays performed at our institution, and a number of children had commenced blood transfusion prior to their arrival. Our current approach of administering plasma expanders, close monitoring of vital signs, and judicious surgical intervention aims to reduce the number of unnecessary transfusions and the associated risk of viral infection and other diseases transmitted by blood. There were no documented cases of coagulopathy amongst the patients admitted with secondary haemorrhage. While we do not routinely screen our patients for a coagulopathy prior to adenotonsillectomy, Kang et al. [93 did find that an initially abnormal coagulation profile may identify those more likely to bleed after surgery, and suggested that a coagulation profile may be a valuable screening tool for children undergoing adenotonsillectomy. All surgery was required within 12 h of admission and all blood transfusion within 24 h. No major intervention was required more than 24 h after admission in this series and it would appear that after a period of observation of 24 h it is probably safe to discharge the majority of cases, if no further bleeding is noted and there are no other ongoing problems. Eighty-seven per cent of the children presented at day 10 or earlier, with the latest presentation being 15 days following adenotonsillectomy. However, while Chowdhury et al. [3] in a study of 80 children with secondary bleeds, noted a wide range of timing in readmission for delayed haemorrhage from 28 h to 21 days post-operatively, most authors agree that secondary haemorrhage is uncommon after the 14th post-operative day [3,17,19], and usually occurs from 5 to 10 days after surgery. In conclusion, while the literature indicates that it is appropriate to inform parents of a 2% risk of secondary haemorrhage following adenotonsillectomy, our
124
D.B.
Irani,
R.G.
Berkowitz
/Int.
J. Pediatr.
Otorhinolaryngol.
40 (1997)
115-124
major intervention rate of 13% in this series suggests that the overall risk of the need for blood transfusion or return to the operating room for hemostasis as a consequence of secondary bleeding is in the order of 0.3% for each child undergoing adenotonsillectomy in our institution. This is a useful figure to quote to parents, helping them make an informed decision to accept adenotonsillectomy for their child.
References [l] Berkowitz, R.G. and Zalzal, G.H. (1990) Tonsillectomy in children under 3 years of age. Arch. Otolaryngol. Head Neck Surg. 116, 6855686. [2] Carmody, D., Vamadevan, T. and Cooper, S.M. (1982) Post-tonsillectomy haemorrhage. J. Laryngol. Otol. 96, 635-638. [3] Chowdhury, K., Tewfik, T.L. and Schloss, M.D. (1988) Post-tonsillectomy and adenoidectomy haemorrhage. J. Otolaryngol. 17, 46649. [4] Colclasure, J.B. and Graham, S.S. (1990) Complications of outpatient tonsillectomy and adenoidectomy: a review of 3340 cases. Ear Nose and Throat J. 69, 155-160. [5] Fox, S.L. (1952) Bleeding Following tonsillectomy. Laryngoscope 62, 414-425. [6] Haase, F.R. and Noguera, S.T. (1962) Hemostasis in tonsillectomy by electrocautery. Arch. Otolaryngol. 75, 125-126. [7] Handler, S.D., Miller, L., Richmond, K.H. and Baranak, CC. (1986) Post tonsillectomy haemorrhage: incidence, prevention and management. Laryngoscope 96, 1243-1247. [8] Helmus, C. (1979) Tonsillectomy and adenoidectomy in the one and two-year-old child. Laryngoscope 89, 17641771. [9] Kang, J., Brodsky, L., Danziger, I., Volk, M. and Stanievich, J. (1995) Coagulation profile as a predictor for post-tonsillectomy and adenoidectomy haemorrhage. Int. J. Pediatr. Otorhinolaryngol. 32, 261-263. [lo] Kendrick, D. and Gibbin, K. (1993) An audit of the complications of paediatric tonsillectomy, adenoidectomy and adenotonsillectomy. Clin. Otolaryngol. 18, 115- 117. [l l] Kristensen, S. and Tveteras, K. (1984) Post-tonsillectomy haemorrhage. A retrospective study of 1150 operations. Clin. Otolaryngol. 9, 347-350. [12] Lee, I.N. (1985) Outpatient management of T and A procedure in children. J. Otolaryngol. 14, 176-177. [13] Mum, I. and Simon, H. (1993) Hemorrhagic complications after tonsillectomy and adenoidectomy. Experiences with 7743 operation in 14 years. Weiner Klinische Wochenschrift 105, 520-522. [14] Neivert, H. (1945) Late secondary tonsillar haemorrhage. Arch. Otolaryngol. 42, 14- 18. [15] Nicklaus, P.J., Herzon, F.S. and Steinle, E.W. 4th. (1995) Short-stay outpatient tonsillectomy. Arch. Otolaryngol. Head Neck Surg. 121, 521-524. [16] Roberts, C., Jayaramachandran, S. and Raine, C.H. (1992) A prospective study of factors which may predispose to post-operative tonsillar fossa haemorrhage. Clin. Otolaryngol. 17, 13- 17. [17] Segal, C., Berger, G., Basker, M. and Marshak, M. (1983) Adenotonsillectomies on a Surgical Day-Clinic Basis. Laryngoscope 93, 1205-1207. [18] Siodlak, M.Z., Gleeson, M.J. and Wengraf, C.L. (1985) Post-tonsillectomy secondary haemorrhage. Ann. R. Coll. Surg. Eng. 67, 167- 168. [19] Tami, T.A., Parker, G.S. and Taylor, R.E. (1987) Post-tonsillectomy bleeding: An evaluation of risk factors. Laryngoscope 97, 1307- 1311. [20] Wiatrak, B.J., Myer, C.M. 3rd and Andrews, T.M. (1991) Complications of adenotonsillectomy in children under 3 years of age. Am. J. Otolaryngol. 12, 170- 172.