Or you could consider if there’s reason to believe there’s a causal relationship, if there is you could change your protocol (offer it in the evening as an option), measure the improvement, publish the result and simultaneously improve your patient outcomes and move science forward.