Introduction:

the lack of early preventive measures in the management of Chronic Venous Disease (CVD) has burdened Brazilian public health.

Objectives:

to develop a mathematical pattern recognition model in the relationship between saphenofemoral junction (SFJ) insufficiency and the clinical picture of CVD, and to outline patients' sociodemographic profile.

Methods:

a quantitative, inductive and descriptive research was carried out, in which participants were asked objective questions. The CVD of 91 patients was clinically classified into group 0: mild disease (CEAP C2 and C3), and group 1: severe disease (CEAP C4, C5 and C6). All limbs underwent Doppler ultrasonography. A model for the evolution of CVD was developed based on 16 sociodemographic, clinical and imaging predictor variables, including reflux time, peak reflux velocity and venous reflux volume. Two methods of pattern recognition and classification were used: logistic regression and discriminant analysis with quadratic score. The participants' profile was analyzed by relative and absolute frequency.

Results:

the outlined profile was that of a female patient, multiparous, aged around 50 years, with high body mass index, low income and advanced disease (C4,5,6). Concerning the mathematical relationship of the variables, all the models presented satisfactory results, with average hit rates higher than 80 %: 87.06 % for prediction of mild disease (G0) and 71.88 % for severe disease (G1).

Conclusion:

a mathematical model for pattern recognition was established in the relationship between degree of reflux in the SFJ and the clinical picture of CVD. The model predicts CVD progression with a good accuracy.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )