Dr Reid notes that 15.2% of children in our rehydration trial were unable to perform oral rehydration therapy (ORT) and surmises that the conclusion of our study, that ORT is as effective as intravenous fluids (IVF),1 was unsupported. We disagree. These patients were analyzed as ORT failures, and still our results demonstrated effectiveness for the treatment of moderately dehydrated children. The American Academy of Pediatrics practice parameter for gastroenteritis2 recommends ORT as the initial therapy for mild and moderately dehydrated children. However, a survey by Ozuah et al3 demonstrated that emergency medicine providers who are very familiar with this practice parameter use IVF 75% of the time, and those who are less familiar use IVF 90% of the time for moderately dehydrated children. The fact that only 15% of these children were unable to tolerate ORT demonstrates that we can potentially reduce the amount of IVF administered to children by >50%.
We limited the study enrollment period until 8 PM for practical purposes so that the 4-hour study would be completed before midnight (when our research staff and attending coverage was more limited). It is true, however, that ORT requires a cooperative patient and parent to succeed. Additional study of the ability of patients and families to perform ORT in the middle of the night is warranted.
Furthermore, Dr Reid raises the issue of intention-to-treat versus treatment-received analysis. Intention-to-treat analysis is the favored approach and is the main result presented in our trial because it avoids bias associated with nonrandom loss of participants. However, a simultaneous presentation of the treatment-received analysis is appropriate, because it describes what happened to the patients who were compliant with the therapy.4 In our trial, 3 patients in the ORT group dropped out because the parents truly didn't want ORT, and 3 patients in the IVF group did not receive the IVF. It is these patients who were excluded in the treatment-received approach. We were very careful to clearly state when a result was based on the treatment-received analysis. Dr Reid states that what matters to physicians is the intention-to-treat approach. We believe that a thorough evaluation of all data is critical to practitioners making treatment decisions. In the intention-to-treat approach, there was an 18% greater hospitalization rate for patients who received IVF, although this finding was just short of our a priori definition of statistical significance. In the treatment-received analysis, this finding was significant.
REFERENCES
Related articles in Pediatrics:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||