CP_PZ: Calculate promising zone and position of observed effect in...

In IDDI-BE/CPPZ: Calculate conditional power and promising zone at interim analysis

Description

This function calculates the promising zone boundaries at an interim analysis of a
clinical trial

Usage

CP_PZ(
  r,
  n1 = NULL,
  n2 = NULL,
  alpha_1s,
  eff_est,
  eff_planned,
  eff_null,
  SE = NULL,
  p_c = NULL,
  sd_t = NULL,
  sd_c = NULL,
  type,
  pow = NULL,
  f = NULL,
  max,
  plot_effect
)

Arguments

r	Proportion of subjects assigned to active group
n1	Number of patients / events at first interim analysis (NULL if parameter f provided)
n2	Number of patients / events at final analysis
alpha_1s	one-sided alpha
eff_est	effect observed/estimated at interim
eff_planned	planned effect (usually alternative hypothesis, but can be modified to obtain conditional power for any assumprion)
eff_null	effect corresponding with null hypothesis (e.g. 1 for hazard rates, 0 for difference) This accomodates for non-inferiority analyses
SE	standard error: has to be provided if type not equal to "HR". If for instance SE =1 then eff_est and eff_planned correspond with z-scores
p_c	proportion in control group (needs to be provided if type="prop")
sd_t	standard deviation in treatment group (needs to be provided if type="cont")
sd_c	standard deviation in control group (needs to be provided if type="cont")
type	if type="HR", then SE is calculated, if type="general", then SE's have to be provided by user, if type="cont" (continuous) then sd_t and sd_c have to be provided if type="prop" then p_c has to be provided
pow	Power, needed to calculate sample size in second part provided to obtain given power
f	n1/n2 at interim analysis. Only to be provided if n1 and n2 not provided
max	(Maximum sample size)/(original sample size), could be for instance 1.5, 2 or 3
plot_effect	TRUE if plotting boundaries with corresponding effect scale

Value

a list of three vectors

• CP_obs : conditional power, given 1) observed effect at interim, 2) total sample size,
  3) Assumed true effect= "eff_est" parameter

• CP_planned : conditional power, given 1) observed effect at interim, 2) total sample size,
  3) Assumed true effect= "eff_planned" parameter

• n2_inc_new : additional patients needed to obtain given power

• CP_ll lower boundary of promising zone (z-score, b-value, conditional power, effect scale)

• CP_ul upper boundary of promising zone (z-score, b-value, conditional power, effect scale)

References

Lan and Wittes. The B-Value: A Tool for Monitoring Data. Biometrics 1988;44:579-585
Mehta CR, Pocock SJ. Adaptive increase in sample size when interim results are promising: A practical
guide with examples. Statist. Med. 2011;30:3267–3284

Examples

CP_PZ(r=0.5,n1=72,n2=180,alpha_1s=0.025,eff_est=0.7,eff_planned=0.7,eff_null=1,SE=NULL,
      type="HR",max=1.5,pow=0.8       ,plot_effect=1)
CP_PZ(r=0.5      ,n2=180,alpha_1s=0.025,eff_est=0.7,eff_planned=0.7,eff_null=1,SE=NULL,
      type="HR",max=1.5,pow=0.8,f=0.25,plot_effect=1)

IDDI-BE/CPPZ documentation built on Oct. 19, 2020, 3:56 a.m.