A team reports that the percentage of patients screened for nutrition risk rose after a workflow change. During the same period, the clinic began excluding several visit types from the denominator. What should the team do before concluding that screening improved?