SAMPLE SIZE AND POWER ANALYSIS
CALCULATOR AND GRAPHER

Health Decision Strategies, LLC.

Data Entry Page
--  Beta version, for demonstration only -
   
             

Enter data in yellow boxes, hit "submit button" and answers will appear
on the next page in blue boxes -- graphs of data below.

 
   
  WTP_1 WTP_2 WTP_3 WTP_4 WTP_5  
  WTP Threshold -->    
 
  Z (alpha/2)= Z value for probability <-- Prob of type I error (alpha)or  
  Z (beta)= Z value for probability of Type II error <-- Prob of type II error (beta)or  
 
  sdet= standard deviation of the effect in treatment group  
  sdec= standard deviation of the effect in control group  
  sdct= standard deviation of the cost in treatment group  
  sdcc= standard deviation of the cost in control group  
  dC= mean difference in cost  
  dE= mean difference in effect  
  Corr_dEdC correlation between difference in cost and difference in effect  
     
     
  Sample Size Calculator - Results - number needed in each arm based on data entered above.  
    WTP_1 WTP_2 WTP_3 WTP_4 WTP_5  
  N =  
 
                 
 
  Power Analysis  
  If the study has already been completed, this section below calculates the power to detect a difference, if a difference actually exists,  
  based on the numbers entered in the first group of yellow boxes above.  
    N_1 N_2 N_3 N_4 N_5  
  Study Sample Size    
  1-Z_beta  
  Power = (prob(1-Z_beta))  
 




Figure 1. Below is a graph example illustrating the data results generated by the HDS sample size calculator.This shows the sample size by willingness to pay and by different values of the correlation between cost and effect. Note the "effect only" line which is constant.
 
Figure 2. Below is a graph example illustrating the data results generated by the HDS sample size calculator.This shows the sample size by willingness to pay and by different values of power. Note the "effect only" line which is constant.
 
   
  Figure 3. Below is a graph example illustrating the data results generated by the HDS power calculator.This shows the power by sample size for different values of the correlation between cost and effect.
 
   
 




 
References:  
   
Briggs A and Gray A. Power and sample size calculations for stochastic  
cost-effectiveness analysis. Med Decis Making 1998: 18 suppl:S81-S92.  
 
Briggs A and Tambour M. The design and analysis of stochastic  
cost-effectiveness studies for the evaluation of health care interventionsl.  
Stockholm School of Economics. Working paper #234. April 1998.  
 
Walter SD et al. Estimation, power and sample size calculations for  
stochastic cost and effectiveness analysis. Pharmacoeconomics  
2007; 25(6):455-466.  
 
Glick H et al. Economic evaluation in clinical trials. Oxford University  
Press, 2007.  




Health Decision Strategies, LLC