Logit and fuzzy models in data analysis: estimation of risk in cardiac patients.

The aim of this study was a comparison of risk stratification for death in patients after myocardial infarction (MI) and of risk stratification for malignant arrhythmias in patients with implantable cardioverter-defibrillator (ICD). The individual risk factors and more complex approaches were used, which take into account that a borderline between a risky and non-risky value of each predictor is not clear-cut (fuzzification of a critical value) and that individual risk factors have different weight (area under receiver operating curve - AUC or Sommers' D - Dxy). The risk factors were baroreflex sensitivity, ejection fraction and the number of ventricular premature complexes/hour on Holter monitoring. Those factors were evaluated separately and they were involved into logit model and fuzzy models (Fuzzy, Fuzzy-AUC, and Fuzzy-Dxy). Two groups of patients were examined: a) 308 patients 7-21 days after MI (23 patients died within period of 24 month); b) 53 patients with left ventricular dysfunction examined before implantation of ICD (7 patients with malignant arrhythmia and electric discharge within 11 month after implantation). Our results obtained in MI patients demonstrated that the application of logit and fuzzy models was superior over the risk stratification based on algorithm where the decision making is dependent on one parameter. In patients with implanted defibrillator only logit method yielded statistically significant result, but its reliability was doubtful because all other tests were statistically insignificant. We recommend evaluating the data not only by tests based on logit model but also by tests based on fuzzy models.

Anthracycline therapy and 24-hour blood-pressure profile in long-term survivors of childhood cancer.

In our previous studies, a decreased blood pressure was reported in children treated by anthracycline (AC). The aim of this study was to assess the long-term effects of AC anticancer therapy in 45 subjects aged 13-22 years by repeated 24-hour Holter monitoring of blood pressure. Sixty four aged-matched subjects served as controls. The differences between mean values of systolic (SBP) and diastolic blood pressure (DBP) in each hour of both groups were evaluated by Mann-Whitney test. Also the parameters of the least-squares fit of the sinusoidal curve in each subject were estimated (M - mesor, midline-estimating, a mean value of sinusoidal curve corresponds to 24-hours mean pressure; A - amplitude, double amplitude corresponds to night-day difference; Acr - acrophase is a time of maximal value of a sinusoidal curve). SBP and DBP was significantly lower only during night hours in anthracycline patients 19-22 years old. Also M was lower in this age subgroup of patients comparing to age matched controls (SBP: 112+/-6 mm Hg versus 117+/-7 mm Hg, p<0.05; DBP: 67+/-3 mm Hg versus 69+/-6 mm Hg, p<0.05), A was not different, Acr in patients was shifted one hour earlier (SBP: 2.4 p.m. versus 3.6 p.m., p<0.05; DBP: 2.1 p.m. versus 3.3 p.m., p<0.01). This corresponds to the shift of the morning blood-pressure increase seen on 24-hours blood pressure profiles. M correlated with age in controls (SBP: r=0.374, p<0.01; regression coefficient b=1.34 mm Hg/1 year; DBP: r=0.365, p<0.01; b=0.95 mm Hg/1 year), but not in patients (SBP: r=0.182, DBP: r=0.064). A and Acr were age-independent in all subjects. It is concluded that blood pressure in 19-22 years old AC patients is lower during night hours, the age-dependent increase of blood pressure seen in healthy controls between 13 and 22 years of age does not occur in patients. This finding is consistent with the long-lasting impairment of the sympathetic nervous system caused by anthracyclines.

Overweight and decreased baroreflex sensitivity as independent risk factors for hypertension in children, adolescents, and young adults.

We studied the relationship between blood pressure (BP), body mass index (BMI, kg/m(2)) and baroreflex sensitivity (BRS, ms/mmHg) in adolescents. We examined 34 subjects aged 16.2+/-2.4 years who had repeatedly high causal BP (H) and 52 controls (C) aged 16.4+/-2.2 years. Forty-four C and 22 H were of normal weight (BMI between 19-23.9), and 8 C and 12 H were overweight (BMI between 24-30). Systolic BP was recorded beat-to-beat for 5 min (Finapres, controlled breathing 0.33 Hz). BRS was determined by the cross-spectral method. The predicting power of BMI and BRS for hypertension was evaluated by sensitivity, specificity, and receiver operating curve (ROC - plot of sensitivity versus specificity). H compared with C had lower BRS (p<0.01) and higher BMI (p<0.05). Multiple logistic regression analysis (p<0.001) revealed that a decreased BRS (p<0.05) and an increased BMI (p<0.01) were independently associated with an increased risk of hypertension. No correlation between BMI and BRS was found either in H or in C. Following optimal critical values by ROC, the sensitivity, specificity and area under ROC were determined for: BMI - 22.2 kg/m(2), 61.8 %, 69.2 %, 66.0 %; BRS - 7.1 ms/mmHg, 67.7 %, 69.2 %, 70.0 %; BMI and BRS - 0.439 a.u., 73.5 %, 82.7 %, and 77.3 %. Decreased BRS and overweight were found to be independent risk factors for hypertension.