Objectives: This study aimed to develop a high-risk drinking scorecard using cross-sectional data from the 2014 Korea Community Health Survey.
Methods: Data were collected from records for 149,592 subjects who had participated in the Korea Community Health Survey conducted from 2014. The scorecard model was developed using data mining, a scorecard and points to double the odds approach for weighted multiple logistic regression.
Results: This study found that there were many major influencing factors for high-risk drinkers which included gender, age, educational level, occupation, whether they received health check-ups, depressive symptoms, over-moderate physical activity, mental stress, smoking status, obese status, and regular breakfast. Men in their thirties to fifties had a high risk of being a drinker and the risks in office workers and sales workers were high. Those individuals who were current smokers had a higher risk of drinking. In the scorecard results, the highest score range was observed for gender, age, educational level, and smoking status, suggesting that these were the most important risk factors.
Conclusion: A credit risk scorecard system can be applied to quantify the scoring method, not only to help the medical service provider to understand the meaning, but also to help the general public to understand the danger of high-risk drinking more easily.
Keywords: Korea Community Health Survey; data mining; high-risk drinking; scorecard; weighted multiple logistic regression.