Introduction
Prosthetic valve dysfunction is a potentially critical complication of heart
valve replacement. An easy and quickly applicable diagnostic procedure is
required for recognizing the prosthetic valve dysfunction. The purpose of
this study was to prospectively define the diagnostic value of D-dimer and
INR level in predicting prosthetic valve dysfunction.
Methods
This cross-sectional study was performed in 70 patients suspected to have
prosthetic valve dysfunction admitted to Imam Ali Hospital, affiliated with
Kermanshah University of Medical Sciences (KUMS), Kermanshah Province, Iran.
Cinefluoroscopy, as the gold standard diagnostic test, was used for the
diagnosis of prosthetic valve dysfunction in enrolled patients. Two
milliliters of blood from each patient were taken into a tube containing
sodium citrate anticoagulant. To evaluate D-dimer, the cutoff value was set
at 500 ng/ml. Also, to evaluate international normalized ratio (INR), the
cutoff value was set at 2. Sensitivity, specificity, positive predictive
value (PPV), negative predictive value (NPV), positive likelihood ratio
(PLR), and negative likelihood ratio (NLR) of the serum markers were used to
describe predictive properties.
Results
Of 70 patients, 27 (38.6%) were male and 43 (61.4%) were female, and the mean
age was 54.67±15.11 years (range, 18 to 80 years). Of 70 patients, 27
(38.6%) had prosthetic heart valve malfunction demonstrable by fluoroscopy,
and 19 patients (27.1%) had D-dimer levels >500 ng/ml. Elevated D-dimer
levels (>500 ng/ml) have been indicated to have sensitivity of 70.4%, and
hence an NPV of 84.3%, specificity of 100%, PPV of 100%, NLR of 0.3, and the
infinity value of PLR for predicting prosthetic valve dysfunction. There was
a significant relationship between fluoroscopy and D-dimer test (P=0.001). A
kappa coefficient value of 0.745 indicated a substantial agreement between
D-dimer and fluoroscopy testing. Mixing test (combination of D-dimer and
INR) showed to have 100% sensitivity, and hence a NPV of 69.8%, specificity
of 69.8%, PPV of 51.8%, NLR of 1.41, and PLR of 1.44 for predicting
prosthetic valve dysfunction.
Conclusion
D-dimer with moderate sensitivity and high specificity is an ideal marker for
the diagnosis of prosthetic valve dysfunction in suspected patients.
Enhanced plasma D-dimer level is not by itself diagnostic of a prosthetic
valve dysfunction but may alert physicians to refer the patient for more
detailed examination, preferably by fluoroscopy. Mixing test with 100%
sensitivity can apply as a rule-out test.