### INTRODUCTION

### METHODS

### Minimum P-value approach

_{min}) (Table 1). However, with such a procedure, the problem of multiple testing arises. This means that the significance of the chosen cutoff value tends to be overestimated [6-10]. Adjustment approaches have been developed to resolve the increase in false positive error rate.

### Adjustment methods to correct P-values

#### Bonferroni method

#### P_{bon}=P_{min}×the number of tests performed

_{min}, P

_{bon}, equals P

_{min}×9. This method is relatively easy for calculating the corrected P-value and can be applied to any type of outcome. However, the method is conservative, and with more candidate cutoff values it becomes more difficult to reach statistical significance.

#### Miller and Siegmund method

_{min}=0.05, z is the 97.5th percentile in the standard normal distribution, 1.96, and ϕ (1.96) is 0.975. This method is less conservative than the Bonferroni method because multiplication of the minimum P-value by the number of tests performed is not necessary, but this method can only apply to binary outcome and is inappropriate for small data sets. Additionally, the results of this method depend on the selection of the value of ε.

#### Lausen and Schumacher method 1

_{min}is computed using the following formula:

#### Lausen and Schumacher method 2

_{i}is the number of observations which is less than or equal to i

^{th}candidate cutoff value (i=1,…,k-1). The value of the function of D is calculated with the following formula:

^{(2)}and P

^{(3)}[16].

#### Contal and O’Quigley method

_{k}is the log-rank statistic computed at the k

^{th}candidate cutoff value, and s is calculated with

_{j}is the score for j

^{th}event and equals

#### Woo et al. method

_{k}is the Gray’s statistic calculated at the k

^{th}cutoff value and s is calculated with

### APPLICATIONS

### Leukocyte elastase and coronary artery disease

^{(1)}, P

^{(2)}, and P

^{(3)}were all significant. The results demonstrate that we can divide the subjects into two groups regarding CAD using a significant cutoff of 36 for leukocyte elastase.