Using social contact data to improve the overall effect estimate of a cluster‐randomized influenza vaccination program in Senegal