Defining Minimal Clinically Important Differences (MCID) in Chronic Cough: Analyses of Objective Cough Counts from a Phase 2 Randomized Controlled Trial

Abstract

Chronic cough (CC) is a condition resulting in negative physical, emotional and social outcomes. Objective measures to quantify frequency of cough are necessary in clinical trials to examine treatment efficacy; however, the degree of cough reduction required to be considered a minimal clinically important difference (MCID) has not been defined. The objective was to define the MCID for objective cough counts (OCCs). Pooled data from a Phase 2 RCT of an investigational treatment for CC were analyzed. Enrollees were non-smokers, had refractory/unexplained CC for ≥1yr, and baseline cough severity VAS ≥40mm. The 24-hour OCC (VitaloJAK™, Vitalograph, Buckingham, UK; Baseline, Week4) and Patient Global Impression of Change (PGIC; Week4) data were analyzed. MCIDs were defined using distribution-based (MCID-D; ½ standard deviation and standard error of measurement) and anchor-based (MCID-A; receiver operating characteristic curves [ROC] and PGIC) analyses. Analyses from n=253 participants (mean age 60.2; 76% female) resulted in MCID-D estimates of 11.3% to 19.7% for OCC. MCID-A/ROC analyses indicated that mean OCC reduction at Week4 of 38% had the best sensitivity/specificity for predicting PGIC responses of “somewhat improved,” “improved,” or “very much improved”. Among those reporting themselves as “somewhat improved” and “improved” vs. “very much improved,” the corresponding %reductions in OCC were 29.7% and 57.5%, respectively. Taken together, these results provide guidance on the degree of change in OCC that can be considered clinically meaningful to help guide treatment decisions and drug development. Specifically, clinicians may consider OCC reductions in the range of 20-30% as the MCID.