Letter in response to Tashkin et al: Limited generalisability of uplift findings to clinical practice.

We appreciate the comments made by Tashkin et al regarding the limited generalisability of the Understanding Potential Long-term Impacts on Function with Tiotropium (UPLIFT) findings. In response, the sample we studied was not designed to be representative of the general chronic obstructive pulmonary disease (COPD) population, but rather the subgroup of patients with severe COPD who are likely to be at greatest risk of serious cardiovascular events. Second, the derived data presented from the Canadian database analysis do not correspond to the UPLIFT inclusion criteria. The UPLIFT study excluded patients with cardiac arrhythmias deemed unstable or life-threatening, or that required either intervention or a change in drug therapy in the last 12 months. In contrast, the Canadian data related to arrhythmias requiring hospital admission in the last 6 months. Third, the cardiovascular and renal comorbidity exclusion criteria in UPLIFT were not used in major long-term COPD trials evaluating LABAS±ICS, such as the landmark TOwards a Revolution in COPD Health study, which had no specific cardiovascular or renal comorbidity exclusion criteria. In our view, the efficacy/safety profile of a COPD medication can only be determined if those patients who are at greatest risk of serious adverse events are studied. If this has not been done, then failing to list the characteristics of patients who were excluded from trial participation in the medication data sheet is unsatisfactory. We propose that the tiotropium data sheet is amended to state that the favourable efficacy/safety profile of tiotropium HandiHaler established in the UPLIFT study applies only to patients without recent cardiovascular or renal comorbidity, as these patients were excluded from the study.