Table 1—

Rates and rate ratios(RRs) of cause-specific death across randomised controlled trials stratified by any inhaled corticosteroid (ICS) use at baseline

FormoterolNo FormoterolRR (95% CI)#
PatientsnFollow-up TPYDeathsnRate per TTYPatientsnFollow-up TPYDeathsnRate per TTY
Asthma-related death
 All subjects4990623.680.34180989.220.221.57 (0.31–15.1)
 With ICS4600321.770.32139057.210.142.32 (0.30–105)
 No ICS39031.910.5441932.110.481.13 (0.014–88)
Cardiac-related death
 All subjects4990623.680.34180989.290.970.35 (0.12–1.02)
 With ICS4600321.770.32139057.260.840.38 (0.11–1.39)
 No ICS39031.910.5441932.131.440.38 (0.0071–4.7)
Other deaths
 All subjects4990623.6180.76180989.230.322.35 (0.69–12.5)
 With ICS4600321.7170.78139057.220.282.82 (0.67–25)
 No ICS39031.910.5441932.110.481.13 (0.014–88)
Total deaths
 All subjects4990623.6341.44180989.2141.520.95 (0.50–1.92)
 With ICS4600321.7311.43139057.291.261.14 (0.53–2.73)
 No ICS39031.931.6141932.152.410.68 (0.11–3.47)
  • CI: confidence interval; TPY: 1,000 person-yrs; TTY: 1,000 treatment-yrs. #: for asthma, cardiac, “other” or any death for formoterol versus non-long-acting β2-agonist (non-LABA) 28. : two salbutamol as needed randomised patients in the Real life effectiveness of Oxis Turbuhaler as needed in asthmatic patients during 6 months (SD-037-0699) trial who suffered a cardiac-related death actually were exposed to formoterol as needed: one erroneously received formoterol instead of salbutamol at start of treatment, and one used salbutamol for 5 weeks and then exchanged medication with a formoterol-randomised patient and died 2 days later; the physician could not determine which drug was used before death. When applying a worst-case approach and reassigning these two deaths to formoterol, the resulting 10 versus seven cardiac deaths give rates of 0.43 versus 0.76 per TTY and a RR (95% CI) for cardiac death of 0.56 (0.19–1.73) for formoterol versus non-LABA, as calculated using StatXact 8.0.0 (Cytel® Inc., Cambridge, MA, USA) 27.