Biometric blood and primary care examinations associated with carotid plaques: CHAID decision trees

Abstract

Introduction: Carotid plaques increase the risk of cerebrovascular accident (CVA). Primary care testing could predict its presence. Objective: To determine the biochemical factors and primary care measurements associated with the presence of carotid plaques. Material and methods: Analytical and crosssectional study of a secondary database of 1380 adults. The variables were: score and number of carotid plaques, glycosylated hemoglobin (HbA1c), LDL, HDL, triglycerides (TG), basal glucose, body mass index (BMI), systolic blood pressure (SBP) and diastolic blood pressure (DBP). Decisions were made through chi square automatic interaction detection (CHAID), Cramer's V, Odds Ratio, prevalence ratio, sensitivity, specificity, positive and negative predictive value. Results: The decision tree had a percentage of correct predictions of 70% and 79% in men and women, respectively. The characteristic associated with risk carotid plaques was HbA1c>5.80% in men; in women: HbA1c >5.70% with SBP >117 mmHg. Men with HbA1c>5.80% had a 2.85- and 1.90-times greater probability and frequency, respectively, of carotid plaque scores >1.20 mm than men with HbA1c≤5.80%. In women, HbA1c >5.70% and SBP >117 mmHg were associated with a probability and frequency 3.71 and 2.60 times higher, respectively. There was high specificity (men: 87%; women: 84%). Conclusions: HbA1c >5.80% in men and HbA1c >5.70% with SBP >117 mmHg in women are associated with a greater frequency and probability of carotid plaques at risk for stroke. SBP limits lower than 130mmHg could be insufficient in preventing stroke. The inclusion of HbA1c as a risk marker for carotid plaques in primary care needs to be evaluated.