Latent multimorbidity classes and antithrombotic medication distributions among patients hospitalized for bleeding after dental surgery: a retrospective study.

OBJECTIVES

To characterize patients hospitalized for postoperative intraoral bleeding after dental surgery and to describe latent multimorbidity classes and their observed antithrombotic medication distributions.

MATERIALS AND

METHODS

This single-center retrospective study included 700 patients hospitalized between 2011 and 2021. Demographic characteristics, clinical measurements, comorbidities, and antithrombotic medication categories were extracted from medical records. Latent class analysis was performed in 458 patients with at least two documented comorbidities using 11 binary indicators. Models containing one to six classes were compared using information criteria, entropy, the bootstrap likelihood ratio test, class size, and clinical interpretability. Stability of the selected solution was assessed using 100 bootstrap resamples.

RESULTS

The median age was 74 years, 460 patients (65.7%) were male, and 581 (83.0%) were receiving at least one antithrombotic medication. A five-class solution was selected. Class 1 (n = 26, 5.7%) was characterized by deep venous thrombosis, hypertension, and cardiac disease; Class 2 (n = 97, 21.2%) by heart valve replacement and cardiac disease; Class 3 (n = 55, 12.0%) by stroke, cardiac disease, and hypertension; Class 4 (n = 151, 33.0%) by hypertension and cardiac disease; and Class 5 (n = 129, 28.2%) by a broader multisystem profile involving cardiac disease, hypertension, diabetes, renal disease, and gastrointestinal disease. Vitamin K antagonist (VKA) use was most frequent in Classes 1 and 2, whereas direct oral anticoagulant (DOAC) use was most frequent in Class 4. Bootstrap analysis indicated moderate overall stability.

CONCLUSION

The study describes clinically relevant heterogeneity among hospitalized bleeding cases.