The Web system for coronary disease risk estimation by logistic regression (“CHD risk”)

Abstract

Aim Coronary and heart diseases (CHD) represent one of thegreatest medical problems in the developed world. To facilitatethe CHD probability estimation procedure, as well as to speed upthe procedure of making and issuing patient’s final diagnosis, wedeveloped the Risk estimation application (‘’CHD Risk’’).Methods Risk estimation is based upon a multivariate analysis ofstatistical data by using logistic regression as a method for probabilityestimation. The method estimates how the final outcome isinfluenced by every single factor. Risk factors represent independentvariables of the model, while a coronary disease risk indicatoris dependent variable.Results The ‘’CHD Risk” was tested for three cases and showedapplication credibility. The system provides coronary disease riskprobability estimation for the given risk factor values, as well asadvice for factors whose values exceed the range of normal values.The system allows input of additional statistical data, whichimproves its learning properties.Conclusions Although the “CHD Risk” system can neither makefinal decisions nor replace the medical professionals themselves, itfulfills the aim to develop the web based tool to help the physiciansto monitor their patients’ health condition, as well as to suggestpreventive measures and therapy.

Keywords

• coronary diseases
• logistic regression
• multivariate analysis
• risk estimation
• web application