Nebulized hypertonic-saline vs epinephrine for bronchiolitis; proof of concept study of cumulative sum (CUSUM) analysis
Neeraj Gupta, Ashish Puliyel, Ayush Manchanda, Jacob Puliyel, Neeraj Gupta, Ashish Puliyel, Ayush Manchanda, Jacob Puliyel
Abstract
Objective: To apply cumulative sum (CUSUM) to monitor a drug trial of nebulized hypertonic-saline in bronchiolitis. To test if monitoring with CUSUM control lines is practical and useful as a prompt to stop the drug trial early, if the study drug performs significantly worse than the comparator drug.
Design: Prospective, open label, controlled trial using standard therapy (epinephrine) and study drug (hypertonic-saline) sequentially in two groups of patients.
Setting: Hospital offering tertiary-level pediatric care.
Patients: Children, 2 months to 2 years, with first episode of bronchiolitis, excluding those with cardiac disease, immunodeficiency and critical illness at presentation.
Interventions: Nebulized epinephrine in first half of the bronchiolitis season (n = 35) and hypertonic saline subsequently (n = 29). Continuous monitoring of response to hypertonic-saline using CUSUM control charts developed with epinephrine-response data.
Main outcome measures: Clinical score, tachycardia and total duration of hospital stay.
Results: In the epinephrine group, the maximum CUSUM was +2.25 (SD 1.34) and minimum CUSUM was -2.26 (SD 1.34). CUSUM score with hypertonic saline group stayed above the zero line throughout the study. There was no statistical difference in the post-treatment clinical score at 24 hours between the treatment groups {Mean (SD) 3.516 (2.816): 3.552 (2.686); 95% CI: -1.416 to 1.356}, heart rate {Mean (SD) 136 (44): 137(12); 95% CI: -17.849 to 15.849) or duration of hospital stay (Mean (SD) 96.029 (111.41): 82.914 (65.940); 95% CI: -33.888 to 60.128}.
Conclusions: The software we developed allows for drawing of control lines to monitor study drug performance. Hypertonic saline performed as well or better than nebulized epinephrine in bronchiolitis.
Source: PubMed