Calculates power, sample size, or minimum detectable hazard ratio for survival studies. This helps researchers design adequately powered studies to detect meaningful differences in survival between groups.
Usage
powersurvival(
data,
calc_type = "sample_size",
study_design = "simple",
hazard_ratio = 0.7,
power = 0.8,
alpha = 0.05,
sample_size = 200,
allocation_ratio = 1,
prob_event = 0.5,
accrual_time = 1,
follow_up_time = 3,
median_survival = 5,
loss_followup = 0.05
)Arguments
- data
The data as a data frame. Note that power analysis typically doesn't require actual data - it's used for study planning.
- calc_type
Select what to calculate. 'Power' calculates statistical power given sample size and hazard ratio. 'Sample Size' determines required sample size for desired power and hazard ratio. 'Hazard Ratio' calculates the minimum detectable effect size given sample size and power.
- study_design
Select the study design complexity. 'Simple' assumes a fixed follow-up period for all subjects. 'Complex' allows for accrual period and variable follow-up times.
- hazard_ratio
The hazard ratio to detect. Values < 1 indicate protective effects (treatment better than control); values > 1 indicate harmful effects (control better than treatment).
- power
The probability of detecting an effect if one exists (1 minus the Type II error rate). Conventional values are 0.8 or 0.9.
- alpha
The Type I error rate (probability of falsely rejecting the null hypothesis). Conventional value is 0.05.
- sample_size
The total number of subjects across all groups. For sample size calculation, this is a starting value for the search algorithm.
- allocation_ratio
The ratio of control group size to treatment group size. 1 indicates equal allocation. Values > 1 mean more subjects in the control group; values < 1 mean more in the treatment group.
- prob_event
The overall probability of observing the event (e.g., death) during the study period. This affects the number of events observed, which is crucial for power.
- accrual_time
The period over which participants are recruited, in years. Only used for complex designs.
- follow_up_time
The additional follow-up period after accrual ends, in years. Only used for complex designs.
- median_survival
The median survival time in the control group, in years. Used to estimate the baseline hazard rate. Only used for complex designs.
- loss_followup
The annual rate of loss to follow-up (attrition). Only used for complex designs.