Reducing the cutoff of hemoglobin A1C to determine if a patient has prediabetes could be cost effective, according to a new study.