Tiotropium safety in ‘real‐world’ populations

We read with interest the publication of Schmiedl et al. describing the characteristics of patients treated with tiotropium HandiHaler or Respimat in the Tiotropium Safety and Performance in Respimat (TIOSPIR) study compared with their ‘real-world’ database [1]. The authors noted that patients starting either treatment had similar characteristics but suggest that we may have included a selected population with chronic obstructive pulmonary disease (COPD), who might react differently to treatment (and be at lower risk of cardiac adverse effects) than a ‘real-world’ population. The baseline characteristics of the patients with COPD enrolled in placebo-controlled tiotropium studies using HandiHaler or Respimat were recently compared with those from a range of epidemiological studies. Patients included in tiotropium studies were found to be widely representative of the overall COPD population [2]. While ‘real-world’ observational studies may be useful in generating relevant clinical questions, randomized controlled trials, such as TIOSPIR, use robust methodologies designed and powered to answer these. As Schmiedl et al. note, one of the main reasons for their ‘real-world’ patients notmeeting the inclusion/exclusion criteria of TIOSPIR was a diagnosis of asthma. Indeed, a concomitant diagnosis of asthma was an exclusion criterion in most studies of tiotropium in COPD. On the other hand, tiotropium was investigated in a large clinical programme of patients with asthma (including severe asthma, excluding COPD); a higher risk of adverse cardiac effects in these patients was not identified [3]. Another criterion for exclusion from TIOSPIR was milder COPD [forced expiratory volume in 1 s (FEV1)>70% predicted]. While Schmiedl et al.were right to highlight differences in their ‘real-world’ population from those meeting more stringent criteria in clinical trials, there is no reason to believe that patients with milder disease would be at any additional specific risk for adverse cardiac effects with tiotropium treatment beyond the assessments performed in TIOSPIR. On the contrary, as TIOSPIR was a safety study, and therefore focused on