Results of logistic regression analysis
| Independent variable | Cut off concentration (ng/l) | Odds ratio (95% CI, p value) | Diagnostic accuracy* | Incorrect classification | |
|---|---|---|---|---|---|
| False positive | False negative | ||||
| *Diagnostic accuracy of the statistical model including all the given covariates simultaneously. | |||||
| Odds ratios (95% CI, p value) are related to the two analytes. Diagnostic accuracy and false positive and false negative results of classification are derived from the respective whole statistical model. Observed frequencies: patients with dyspnoea caused by CHF, n = 137; patients with dyspnoea attributable to other causes, n = 114. | |||||
| Prediction of symptomatic CHF, unadjusted model | |||||
| BNP | 295 | 25 (13 to 49, <0.001) | 83% | n = 16 | n = 27 |
| NT-proBNP | 825 | 28 (14 to 55, <0.001) | 84% | n = 22 | n = 18 |
| Prediction of symptomatic CHF, adjusted model 1 (controlling for age, sex, and eGFR) | |||||
| BNP | 295 | 27 (13 to 56, <0.001) | 83% | n = 16 | n = 27 |
| NT-proBNP | 825 | 29 (14 to 61, <0.001) | 84% | n = 22 | n = 18 |
| Prediction of symptomatic CHF, adjusted model 2 (controlling for age, sex, eGFR, history of CHF, and the presence of orthopnoea, PND, nocturnal cough, JVD, pulmonary rales, third heart sound, and peripheral oedema) | |||||
| BNP | 295 | 10 (4 to 24, <0.001) | 87% | n = 16 | n = 18 |
| NT-proBNP | 825 | 13 (6 to 32, <0.001) | 87% | n = 19 | n = 15 |