Do Case-Control Studies Always Estimate Odds Ratios?