Table 1.
Study population characteristics.
Fig 1.
Overall concordances of the different equations for LDL-C estimation.
Overall concordances of different equations for LDL-C estimation for each assay (i.e., Roche, Siemens and Beckman) are given in a clustered bar chart. Each bar indicates the concordance of estimating LDL-C levels by different formulas given that the LDL-C levels measured by Roche, Siemens and Beckman direct methods, accordingly.
Table 2.
Median statistics for the ratio of triglycerides to very low-density lipoprotein cholesterol by the cross table of non-high-density lipoprotein cholesterol and triglycerides calculated from the Turkish population (calculated for each direct assay method for 180-cell strata).
Fig 2.
Distribution density of LDL-C concentrations calculated by direct methods and different formulas.
Box-plots are also represented to compare the LDL-C levels measured by direct method with the LDL-C levels estimated by Friedewald, Sampson, Martin-Hopkins and extended Martin-Hopkins formulas. Red dash line is depicted to see the difference between direct method (Roche, Beckman or Siemens) and the formulas used to measure the LDL-C levels.
Fig 3.
Concordances of the different equations for LDL-C estimation by LDL-C strata.
Concordances of different equations for LDL-C estimation by LDL-C groups assuming different direct measures (i.e., Roche, Beckman and Siemens) are given in a clustered bar chart. Each bar indicates the concordance of estimating LDL-C levels by different formulas for each group of LDL-C levels given that the LDL-C levels measured by Roche, Beckman and Siemens direct methods, accordingly.
Fig 4.
Concordances of the different equations for LDL-C estimation by triglycerides strata.
Concordances of different equations for LDL-C estimation by triglycerides groups for different direct measures (i.e., Roche, Beckman and Siemens) are given in a clustered bar chart. Each bar indicates the concordance of estimating LDL-C levels by different formulas for each group of triglycerides concentration given that the LDL-C levels measured by Roche, Beckman and Siemens direct methods, accordingly.
Table 3.
The concordances of LDL-C estimation equations in patients with low LDL-C and/or higher TG levels.
Fig 5.
Concordances of the different equations for LDL-C estimation by non-HDL-C strata.
Concordances of different equations for LDL-C estimation by non-HDL-C groups assuming different direct measures (i.e., Roche, Beckman and Siemens) are given in a clustered bar chart. Each bar indicates the concordance of estimating LDL-C levels by different formulas for each group of non-HDL-C concentration given that the LDL-C levels measured by Roche, Beckman and Siemens direct methods, accordingly.
Fig 6.
Regression analysis between LDL-C levels estimated by formulas and directly measured LDL-C levels.
Correlations of estimated LDL-C levels by Friedewald, Sampson, Martin-Hopkins and extended Martin-Hopkins formulas with LDL-C levels directly measured by Roche, Beckman and Siemens.
Fig 7.
Residual error plots for LDL-C by different formulas concerning to different direct assay methods.
While the values on x-axis show TG levels, the values on y-axis shows the difference between estimated LDL-C (by Friedewald, Sampson, Martin-Hopkins or extended Martin-Hopkins) and direct LDL-C levels (calculated by Roche, Beckman or Siemens). The mean absolute deviation (MAD) for each possible case is also given in each panel for each dataset.