Dr. Windfuhr responds.

Our study was undertaken to evaluate the incidence of delayed hemorrhage in patients who underwent tonsillectomy in our department. We chose a follow-up period of 3 months so that we might include all patients with post-tonsillectomy hemorrhage, including those whose hemorrhages occurred more than 10 days postoperatively. In recent years, several patients who had undergone tonsillectomy elsewhere required surgical treatment for postoperative hemorrhage under general anesthesia in our emergency department.

It might be that the small number of patients (602) in our study led to a statistical bias. Our latest contribution to EAR, NOSE & THROAT JOURNAL, which appears in this issue (page 626), concerns the incidence of post-tonsillectomy hemorrhage in 4,848 adults and children. This study confirmed what we previously reported in our earlier article: primary hemorrhage was clearly more common than secondary hemorrhage in both age groups.

Based on his clearly broad clinical experience, Dr. Geraghty is skeptical of the high rate of primary hemorrhage that we reported in November. However, there is no consensus in the literature regarding the relative incidence of primary and secondary hemorrhage. Some authors have published reports in which primary hemorrhage was more common, some have reported that it was less common, and some have reported that rates were similar. * (Comparisons are also hindered by the fact that there is no common definition of post-tonsillectomy hemorrhage.) Lee reported a study of more than 3,000 patients in which 23 experienced primary hemorrhage and 48 experienced secondary hemorrhage; none of the latter group required surgical treatment, but 19 of the former group did (J Otolaryngol 1985; 14:176-8). Since we included in our results only those patients whose posttonsillectomy hemorrhage required surgical treatment under general anesthesia, Lee's findings were consistent with ours. But if one looks at only the total numbe r of postoperative bleeds regardless of the need for intervention, Lee's findings more closely parallel the experience of Dr. Geraghty.

The incidence of primary hemorrhage is generally considered to be associated with the technical aspects of the surgery-specifically with the type of hemostatic technique. Most tonsillectomies in our clinic are performed by supervised ENT surgeons in training. Dr. Geraghty suggests that primary hemorrhage might be the result of a technical failure on the part of the surgeon. However, it has not been shown that the level of the surgeon's training is an independent risk factor for primary hemorrhage. Although one might expect that tonsillectomies performed by surgeons in training would result in a higher rate of complications, this has not been shown to be the case.

The overall rate of complications in our prospective study reported in the November issue (16/602; 2.66%) was identical to the rate reported in our retrospective study published in the current issue (129/4,848; 2.66%). Rates reported by others range from 0.3 to 6.1%.

Dr. Ulbrich and I strongly agree with Dr. Geraghty that the rate of ligation of the external carotid artery in our prospective study was unusually high. However, one of the two patients in our prospective study who experienced post-tonsillectomy hemorrhage several weeks after surgery was one of ten patients in whom the indication for tonsillectomy was the presence of cancer. In our latest contribution, we excluded patients with this indication. The other patient in our earlier study who ultimately required ligation of the external carotid artery had a history of post-trauma splenectomy and had experienced an unusual degree of intraoperative bleeding that required an unusual number of sutures. In the judgment of the senior surgeon, the intensity of the bleeding 5 days after surgery was such that electrosurgical means were inadequate, and the only appropriate treatment left to us was ligation of the external carotid artery. Another surgeon might have tried to embolize the bleeding vessel via interventional radi ology, but a facility for doing so is not available at our hospital. Ligation of the external carotid artery is the last resort in our management protocol. We believe that the unusually high incidence of such ligation in our prospective study represented, as Dr. Geraghty suggested, simple bad luck.

Our method of achieving hemostasis without electrosurgical means might influence our rate of both primary and secondary hemorrhage. Some authors have found a relationship between primary hemorrhage and suture ligation as well as between secondary hemorrhage and electrosurgery, and they concluded that the use of electrocautery decreases the incidence of primary hemorrhage. Conversely, Mutz and Simon (Wien Klin Wochenschr 1993;105:520-2) and Myssiorek and Alvi (Int J Pediatr Otorhinolaryngol 1996;37:35-43) primarily used electrosurgical means, and they reported a higher rate of secondary hemorrhage. Others found no statistically significant difference between the two methods.

We greatly appreciate Dr. Geraghty's constructive remarks. We hope that our reply will to some extent lessen the degree of his disbelief in our results. Obviously, the fact that our results differ from those of other studies does not mean that our data are unreliable; rather it suggests that more studies are indicated, in fact, we are planning another prospective study of bipolar electrocautery for hemostasis to explore its influence on the rate of primary hemorrhage.

Jochen P. Windfuhr, MD

Department of Otorhinolaryngology-Head and Neck Surgery

St. Anna Hospital

Duisburg, Germany

* Journal citations for studies not referenced in this response are available on request at [e-mail].