Find Paper, Faster
Example:10.1021/acsami.1c06204 or Chem. Rev., 2007, 107, 2411-2502
Assessment of the Change of a Continuous Variable as a Function of its Initial Value
JAMA Cardiology  (IF14.676),  Pub Date : 2021-11-01, DOI: 10.1001/jamacardio.2021.3025
Christian Funck-Brentano

To the Editor Marcusa et al1 have examined the influence of baseline low-density lipoprotein cholesterol (LDL-C) levels on percentage LDL-C reduction with statins, ezetimibe, and PCSK9 inhibition. Using this mode of calculation, they report a higher percentage reduction in LDL-C with evolocumab in patients with lower baseline LDL-C levels, compared with a less marked percentage LDL-C reduction with simvastatin at lower baseline LDL-C levels and no significant reduction of LDL-C levels with ezetimibe at lower baseline values.1 However, the use of percentage changes might be misleading since it is influenced by absolute baseline values, eg, a 30% reduction from 100 mg/dL (to convert to millimoles per liter, multiply by 0.0259) converts to 70 mg/dL, a 30 mg/dL absolute reduction, whereas a 30% reduction from 70 mg/dL converts to 49 mg/dL, a 21 mg/dL absolute reduction. More importantly, assessing the change of a parameter as a function of its baseline value is subject to bias because baseline values are used to calculate percentage changes.2,3 An analysis of the slope and intercept of the regression lines between LDL-C levels under treatment and baseline LDL-C levels using both linear and log scales, as promoted by MacGregor et al,2 seems more appropriate and might yield different results. This mode of analysis and of graphical display would allow a better comparison of evolocumab, simvastatin, and ezetimibe.