To the Editor.
I read with interest the study by Ramnath et al1that uses ultrasonography to detect children with parapneumonic effusion who would benefit from surgical intervention. When researching this topic for a review article2 more than 15 years ago, I was frustrated by the paucity of pediatric data. Much current clinical practice is often based on adult data and/or dogma, with pediatric experience added. Discussions about the “proper” approach often generate more heat than light, so the present paper is of great assistance. I have questions in two areas.
First, the endpoint is hospital discharge. However, there are no descriptions of discharge criteria, and whether they were consistent between the two institutions.
Second, it is intriguing that in the nonoperative group, children treated with antibiotics alone had a shorter length of stay than those treated with antibiotics and tube drainage. The authors show a 47.6% drop in hospital stay for children with high-grade ultrasounds with an operative approach. However, within the nonoperative group, a 42.7% drop in hospital stay is attained by eliminating the chest tube, and just treating with antibiotics! Are the authors able to analyze the length of hospitalization for patients in the antibiotic-alone group, compared with the operative group, in both low-grade and high-grade categories? I realize that this was not the intent of the paper, but the data raise questions about the need for an operative approach at all. Given that McLaughlin et al3 showed no long-term differences in lung function between children treated with or without surgery, such an analysis would further justify a surgical approach in children with high-grade ultrasound findings.
- Ramnath RR,
- Heller RM,
- Ben-Ami T,
- et al.
- Bye MR
- McLaughlin FJ,
- Goldmann DA,
- Rosenbaum DM,
- Harris GBC,
- Schuster SR,
- Strieder DJ
To the Editor.
The report by Ramnath et al1 regarding parapneumonic effusions in children was very interesting and highlights the empiric nature of treating such effusions, particularly in children. The lack of prospective, controlled trials is a major contributor to the lack of consensus on treatment.2
It appears that the extended hospitalization for patients with high-grade effusions in the nonoperative group as a whole (16.4 days) was attributable to the longer stay of the nonoperative subgroup with chest tube drainage (19.9 days) as the subgroup treated with antibiotics alone did not appear to have prolonged hospitalization compared with the operative group (11.4 vs 8.6 days, respectively). Did the authors analyze this relationship?
As the authors state, because of the retrospective nature of this report, the very interesting finding of a 75% increase in hospital stay with the use of tube thoracostomy drainage (as compared with antibiotics alone) is difficult to explain, although it does strongly suggest that tube thoracostomy drainage alone is, at best, not likely to improve outcome. However, presumably those patients with operative drainage also had chest tubes. Were these used for shorter periods than those with tubes only?
With the length of hospitalization being the outcome variable for this report, it is unfortunate that there is no mention of criteria used for hospital discharge, especially as it may vary between the two institutions contributing to the study. For example, were patients changed to oral antibiotics or discharged with intravenous antibiotics? Were patients discharged more aggressively with intravenous antibiotics at one institution (or by one attending service, etc) than the other? It would also be helpful to the reader to have some quantitative information on what type of operative drainage (video thoracostomy vs formal thoracotomy) was used and whether this varied between the institutions. The authors also state that they did not analyze the data with respect to pleural fluid chemistry, oxygen requirements, or febrile days. Surely this information was available and would be extremely interesting to report. Especially the predictive nature of pleural fluid chemistry, such as pH, and the objective measure of number of febrile days as an outcome variable could be quite valuable to the reader.
I wholeheartedly agree with the authors in their call for a prospective randomized trial, not only of sonography use, but of treatment in general for pediatric parapneumonic effusions.
We would like to thank Drs Bye and Colombo for their careful reading of our manuscript. The points that they address are well-taken and are an unavoidable result of the retrospective nature of this investigation.
Discharge criteria were based on the ability of the patients to be cared for at home. Although hospital discharge is a valid endpoint that carries implications regarding rapidity of recovery, with clinical and financial ramifications, it does not reflect return to normal activity. Obviously discharge criteria will vary not only between institutions but also individually with each patient, the age of the patient, and the family environment. It is hoped that by incorporating a larger number of patients in the study group the influence of some of these unavoidable variables might be lessened, but clearly this is a shortcoming of this retrospective study, which we acknowledge.
Regarding comparison of the length of hospital stay between the group treated with antibiotics-alone, and the group treated with nonoperative drainage, we were also surprised at the direction of our data points. Again, because this was a retrospective study, many clinical parameters among these patients were not controlled. It is possible that the presence of a tube in a high-grade empyema not only was unsuccessful in effecting adequate drainage but also introduced a foreign body into the undrained cavity and, coupled with the invasiveness of the procedure, may have worsened the situation, resulting in a longer hospital stay than in the group treated with antibiotics alone.
However, comparison of the antibiotics alone group within the high-grade category and the operative group in the same high-grade category yields a reduction of 2.8 days achieved by operative treatment. In this retrospective study many factors, both objective and subjective, might have entered into the decision to undertake an antibiotic-alone versus an operative treatment in some of these patients. Without a randomized, prospective trial, it is not possible to determine whether antibiotics alone would have sufficed in the group that underwent operative treatment, nor to determine whether operative treatment would have further shortened the hospital stay of those patients treated with antibiotics alone. We agree with Dr Bye and the data that he cites1 to support his opinion that, in the long-term, the children eventually recover completely. The goal of our inquiry rather was to determine whether time to recovery may be shortened, and the endpoint that we chose was hospital discharge. Patients who had operative drainage with postoperative chest tube placement are a different subgroup, in whom adequate drainage had already been achieved, and the chest tube was not placed into the untreated, highly loculated empyema cavity.
In our patient population we encountered only one child who had a tube thoracostomy. Data on this particular treatment method was insufficient to draw valid conclusions, and this patient was excluded from the study group.
We share Dr Colombo's interest in the results of the empyema fluid chemistries. However, we found on chart review that the chemistries were obtained inconsistently, and for this reason we did not include these data points in our results. Again, this is a shortcoming of this retrospective review.
We thank both Drs Bye and Colombo for their incisive comments and concur that our retrospective study raises intriguing questions and points to the need for a prospective, randomized clinical trial. We are in the process of organizing such a prospective study, hoping to answer some of the questions raised by the retrospective analysis of patient outcomes.
- Copyright © 1998 American Academy of Pediatrics