How to construct a nomogram for hypertension using complex sampling data from Korean adults

The purpose of this study is to propose a novel nomogram that predicts the incidence of hypertension with the data surveyed by the two-stage stratified cluster sampling design method, which is one of the complex sampling design method. Complex sample is a sampling of large-scale population by reflecting the design effects (stratification, clustering, sample weight, random sampling etc.) to increase the efficiency of the survey and to represent the population well. When the complex sample is analyzed as simple random sample (SRS), it is possible to obtain biased results from the variance estimates. We used the Rao-Scott chi-squared test to identify the association between two categorical variables by adjusting the Pearson chi-squared statistic. In addition, logistic regression analysis considering design effects was performed in estimating the coefficients to reduce the bias. Based on these results, we constructed a novel nomogram that predicts the probability of the incidence of hypertension using the Korean national health and nutrition examination survey (KNHANES). The constructed nomogram shows that age is the strongest effect on hypertension incidence, followed by BMI, stroke, and family history of hypertension. Finally, we verified the nomogram by the Receiver operating characteristic (ROC) curve and calibration plot.