Good or bad responders? A single-case statistical method to evaluate AChEI treatment