Opportunities and challenges in using studies without a control group in comparative effectiveness reviews