Response to Letter: Is Antibiotic Alternative to Appendectomy?
We thank Dr Tez for his interest in our paper1 and are grateful for the opportunity to clarify the remarks. As he correctly states, nonoperative treatment of appendicitis is not a new concept; the Ochsner–Sherren method2 or its equivalents have been used for a hundred years. But, no high-level evidence has previously been presented in this field in children.
In regards to Dr Tez's specific questions:
In our pilot randomized controlled trial (RCT), a clinical suspicion of perforated appendicitis on the basis of generalized peritonitis was an exclusion criterion and patients with such a clinical diagnosis were not randomized. Perforated appendicitis is a diagnosis that can only be made reliably at surgery or even histology and therefore could not be accurately determined at the time of randomization. Hence we used a pragmatic approach in our pilot RCT that reflects surgical decision making as performed day-to-day by surgeons around the world. It would be methodologically incorrect to exclude patients after the intervention as this may introduce bias into the study. We only know for certain what the disease severity was in children who had an operation and it is entirely possible that some of the children who did not receive an operation also had perforated appendicitis. Because of the randomization process we assume that the groups were similar and to remove patients posthoc would render the groups unbalanced. We have therefore included them in the study and analyzed the results on an intention-to-treat basis.
We do not state that there was no patient without appendicitis in the nonoperative (antibiotic) group. We have no way of knowing this, as we did not operate on them. We only know that the two groups of patients were very similar at the time of randomization and that they, as a group, probably had a similar disease profile. The negative appendectomy rate at our institution has been between 2.7% and 4.3% for the last 5 years. Statistically, it is therefore likely that one or two patients of the 50 randomized would not have acute appendicitis. In fact we do know that one case in the nonoperative group definitely did not have acute appendicitis—the patient who failed antibiotic treatment and had a negative appendectomy on day 2 after randomization.
We acknowledge that the effect of the antibiotic treatment and the spontaneous resolution are two entities that need to be addressed. It is not possible to separate these two effects in our trial. It is probably correct to assume that a proportion of the patients in the nonoperative group would have resolved without antibiotics, as it is likely that a proportion of the patients randomized to surgery would have done the same. The way to differentiate between these two effects would be to randomize patients to nonoperative treatment with either antibiotics or placebo. This trial may take place in the future but we believe it would currently pose ethical issues to propose such a study before the effect of nonoperative treatment with antibiotics has been tested against surgery.
We share, strongly, the concern for increasing antibiotics resistance because of the introduction of nonoperative treatment with antibiotics for acute appendicitis in children (and in adults). We do not support the routine use of this treatment regimen outside the scope of a RCT. Only when this treatment modality has been tested in a large RCT with adequate sample size, will it be of benefit to discuss if this treatment modality is of any use in a larger scale.
In regards to the role of bacteria in the pathogenesis of acute appendicitis, it was interesting to learn about the role of Fusobacterium.