Her er hva som står om statistikk i NIPU, som for en statistiker leses som HR 0.734. Klarer ikke å finne noe tilsvarende for INITIUM, hvis det finnes er det vel ikke så mye å diskutere.
Statistical analyses
Under the null hypothesis, the PFS hazard ratio (HR), for ipilimumab and nivolumab in combination with UV1 (ipi/nivo/UV1) vs ipilimumab and nivolumab (ipi/nivo) is assumed to be 1.00. Under the alternative hypothesis, the HR is assumed to be 0.60. To test the null hypothesis with 80% power and a 1-sided alpha level of 0.10, a total of 69 PFS events are required. Based on the INITIATE trial [28], with 12 months median follow-up it is expected that 69% of patients treated with ipi/nivo will have progressed and, with a HR of 0.60, it is further expected that 51% of patients treated with ipi/nivo/UV1 will have progressed. With an expected accrual rate of 5 patients per month, a total N=118 patients randomized into the trial over a 24-month period and followed for a minimum of 2–3 months after the last patient is randomized will yield the 69 PFS events required.Statistical analyses
Under the null hypothesis, the PFS hazard ratio (HR), for ipilimumab and nivolumab in combination with UV1 (ipi/nivo/UV1) vs ipilimumab and nivolumab (ipi/nivo) is assumed to be 1.00. Under the alternative hypothesis, the HR is assumed to be 0.60. To test the null hypothesis with 80% power and a 1-sided alpha level of 0.10, a total of 69 PFS events are required. Based on the INITIATE trial [28], with 12 months median follow-up it is expected that 69% of patients treated with ipi/nivo will have progressed and, with a HR of 0.60, it is further expected that 51% of patients treated with ipi/nivo/UV1 will have progressed. With an expected accrual rate of 5 patients per month, a total N=118 patients randomized into the trial over a 24-month period and followed for a minimum of 2–3 months after the last patient is randomized will yield the 69 PFS events required.