B

Sample size for estimating prevalence is calculated as  
n = Z² × p × q / L², where  
• Z = 1.96 for α = 5% (two-sided)  
• p = 0.10 and q = 0.90  
• L = allowable absolute error = relative precision × p = 0.20 × 0.10 = 0.02  

n = (1.96)² × 0.10 × 0.90 / (0.02)² = 3.84 × 0.09 / 0.0004 ≈ 864 → about 900.  

Power is not required for a simple prevalence study, so the closest option is 900. [Naing, 2006, PMID 16984641]