Two physicians in a department of emergency medicine reviewed the medical records of all patients admitted for diabetic crisis to the hospital in the last 5 years. They correlated several different variables, such as length of hospital stay, survival, cost of care, re[1]hospitalizations, age, race, family history, and so forth. If they found any statistically significant trends (p-values of 0.05 or less), they planned to publish those results, and they planned to ask a statistician to help them determine if any of their correlations are significant.

Do you see any potential methodological or ethical problems with this study design?

Would it be ethical for the researchers to publish their results if they do not make it clear in the paper that they have mined the data for specific trends?

Would you be more confident in their research if they had a hypothesis prior to conduction the study? Why?