Pneumologie 2015; 69 - P349
DOI: 10.1055/s-0035-1544791

Effect Size Of Open-Label Versus Double-Blind Administration Of Tiotropium ln Trials lnvestigating Health-Related Quality Of Life ln COPD

H Schmidt 1, H Kögler 1, S Geier 1, T Glaab 1, I Leimer 1
  • 1Boehringer Ingelheim Pharma GmbH & Co. KG

Introduction: Effects of interventions on patient-reported outcomes may be subjective and modulated by patients' expectations regarding treatment efficacy. The gold standard for minimizing such bias is the double-blind [DB] randomized controlled trial. We analyzed the effects of tiotropium (T) on health-related quality of life [HRQOL] in COPD in placebo [PL] -controlled trials and assessed whether the trial design (DB vs open-label [OL]) is a relevant modifier of the effects of T.

Methods: Trials of at least 6 months' duration exploring the effect of T vs PL on HRQOL in COPD (using the St George's Respiratory Questionnaire [SGRQ]) were identified from the Boehringer lngelheim clinical trial database and by a systematic literature search in MEDLlNE (cut-off date: 30 Nov 2011). As clinical endpoint, the mean difference between treatment groups in SGRQ total score [TS] was assessed. Trials were grouped according to whether T had been administered DB or OL. As part of a network meta-analysis we tested for interaction to assess whether the trial design (DB versus OL) is a potential modifier of the effect size or its direction.

Results: In 12 trials T had been administered DB and in three trials T was given OL. The overall effect for mean difference vs PL in SGRQ TS was -2.98 units (95% confidence interval [Cl): -3.49,-2.47). For the subgroup of DB trials, the mean difference vs PL was -3.20 (95% Cl: -3.75, -2.65) vs -1.67 (95% Cl: -3.02,-0.32) for OL trials. The P-value for interaction between subgroup and effect on SGRQ TS was 0.04.

Conclusions: ln patients with COPD, the trial design (DB vs OL) was a statistically significant modifier of the effect of inhaled T on HRQOL, resulting in a substantial underestimation of the effect of T on SGRQ TS when administered OL, compared with DB comparison. Therefore, the validity of studies using non-blinded designs to explore such endpoints must be questioned.

Funded by Boehringer lngelheim.

Presented at the ATS congress meeting 2014