Nelson LM, Williams VSL, Carson RT, Baird MJ, Fehnel SE, Johnston JM. Factor analyses of patient-reported outcome measures in Irritable Bowel Syndrome with Constipation (IBSC): results from two randomized, double-blind, placebo-controlled phase 3 trials of linaclotide. Poster presented at the Joint International Neurogastroenterology and Motility Meeting; September 2012. Bologna, Italy. [abstract] Neurogastroenterol Motil. 2012 Sep 1; 24(Suppl. 2):106. doi: 10.1111/j.1365-2982.2012.01997.x.

OBJECTIVE: Factor analyses of patient-reported outcome (PRO) data from two Phase 3 clinical trials of linaclotide for the treatment of irritable bowel syndrome with constipation (IBS-C) were conducted to support previous Phase 2b quantitative work that suggested symptoms elicited from patient interviews belong to one of two underlying domains - abdominal symptoms and bowel symptoms.

METHODS: 1602 patients with IBS-C were analyzed in two Phase 3 multicenter, randomized, double-blind, placebo-controlled, trials of linaclotide (290 lg linaclotide or placebo) administered orally once daily for 12 and 26 weeks, respectively. Ten daily PRO measures addressing abdominal symptoms (pain, discomfort, bloating, fullness, cramping) and bowel symptoms (spontaneous bowel movement [SBM]/complete SBM [CSBM] frequency, unsuccessful bowel movement [UBM] frequency, stool consistency, straining) were assessed using interactive voice response system technology. Principal component analyses (PCA) and exploratory factor analyses (EFA) explored the correlational structure of the PRO symptoms in one random subsample of the dataset, followed by confirmatory factor analysis (CFA) in the second random subsample. Factor loadings and fit statistics were computed to evaluate the models.

RESULTS: PCA of the 10 PRO measures suggested two emergent dimensions accounting for 71.8% of the total variance. The factor loadings and fit statistics from the EFAs indicated that the two-factor model fit the data relatively poorly. Because the standard errors and residual variances associated with UBM Frequency were larger than those associated with the other variables, and because factor loadings for UBM Frequency were relatively small, this variable was removed from the analysis and new EFAs conducted. With this revision, a two-factor EFA model provided the most interpretable solution. When the two-factor CFA model was applied to the correlation matrix without UBM Frequency, and with select correlated error variances, the fit of the model to the data was very good, and the factor load ings were satisfactory (Table). One factor comprised abdominal symptoms: pain, discomfort, bloating, cramping, and fullness; the other factor was bowel symptoms: CSBM frequency, SBM frequency, stool consistency, and straining.

CONCLUSION: The present factor analyses confirm earlier findings that the IBS-C symptoms could be grouped into two factors, or clusters: abdominal symptoms and bowel symptoms.

Share on: