Non-PSP group (BPAP study)18 | PSP group (VOICE study)17 | |

Primary analysis | ||

Number of patients | 206 | 253 |

Variables included in IPTW | ||

Age at index, mean (SD) | 50.6 (14.4) | 53.7 (14.3) |

Age at index, median (IQR) | 51.9 (41.4–60.0) | 55.8 (43.7–62.8) |

Bio-naïve status, naïve, n/N (%) | 117/206 (56.8) | 3/253 (1.2) |

Outcome measure | ||

Discontinuation at 48 weeks*, n/N (%) | 33/206 (16.0) | 15/253 (5.9) |

Effect of PSP (logistic regression with IPTW as weight) OR (CI) | Ref | 0.26 (0.11 to 0.57) |

P value | Ref | <0.001 |

Scenario: Include gender along with age and prior biological exposure status | ||

Number of patients | 206 | 103 |

Variables included in IPTW | ||

Age at index, mean (SD) | 50.6 (14.4) | 53.6 (14.2) |

Age at index, median (IQR) | 51.9 (41.4–60.0) | 55.8 (44.9–62.8) |

Female n/N (%) | 131/206 (63.6) | 64/103 (62.1) |

Bio-naïve status, naïve, n/N (%) | 117/206 (56.8) | 3/103 (2.9) |

Outcome measure | ||

Discontinuation at 48 weeks*, n/N (%) | 33/206 (16.0) | 7/103 (6.8) |

Effect of PSP (logistic regression with IPTW as weight), OR (CI) | Ref | 0.25 (0.08 to 0.73) |

P value | Ref | 0.011 |

Scenario: Use a 12-week discontinuation rule instead of the 9-week rule for the PSP group | ||

Number of patients | 206 | 253 |

Variables included in IPTW | ||

Age at index, mean (SD) | 50.6 (14.4) | 53.7 (14.3) |

Age at index, median (IQR) | 51.9 (41.4–60.0) | 55.8 (43.7–62.8) |

Bio-naïve status, naïve, n/N (%) | 117/206 (56.8) | 3/253 (1.2) |

Outcome measure | ||

Discontinuation at 48 weeks*, n/N (%) | 33/206 (16.0) | 12/253 (4.7) |

Effect of PSP (logistic regression with IPTW as weight), OR (CI) | Ref | 0.20 (0.09 to 0.46) |

P value | Ref | <0.001 |

Scenario: Only include patients with prior exposure to biological therapies | ||

Number of patients | 89 | 250 |

Variables included in IPTW | ||

Age at index, mean (SD) | 49.5 (14.1) | 53.8 (14.1) |

Age at index, median (IQR) | 52.3 (38.6–59.9) | 55.8 (44.7–62.8) |

Outcome measure | ||

Discontinuation at 48 weeks*, n/N (%) | 14/89 (15.7) | 15/250 (6.0) |

Effect of PSP (logistic regression with IPTW as weight), OR (CI) | Ref | 0.40 (0.18 to 0.88) |

P value | Ref | 0.022 |

BPAP, Benralizumab Patient Access Programme; IPTW, inverse probability of treatment weighting; ITC, indirect treatment comparison; n/N, number; PSP, patient support programme.