# Latent profile analysis of medication adherence in lower extremity deep venous thrombosis-cross-sectional study

## TL;DR

This study identifies three distinct medication adherence patterns in patients with lower extremity deep venous thrombosis and finds factors influencing these patterns to guide personalized interventions.

## Contribution

The study introduces latent profile analysis to categorize medication adherence in LEDVT patients into distinct classes and identifies influencing factors for targeted interventions.

## Key findings

- Three distinct medication adherence classes were identified: poorest adherence (44.99%), moderate adherence (19.83%), and good adherence (35.18%).
- Factors like perceived health competence, hope, activation, and medication beliefs significantly influence adherence profiles (p < 0.05).
- Personalized interventions based on adherence classes can improve medication adherence and patient outcomes.

## Abstract

The cornerstone of treating lower extremity deep venous thrombosis (LEDVT) lies in anticoagulation therapy to prevent thrombus progression and recurrence. However, patient adherence to medication is a critical factor influencing treatment efficacy. Traditional research often simplifies adherence into binary categories of “adherent” and “non-adherent,” which fails to comprehensively reflect the complex behavioral patterns. Based on latent profile analysis (LPA), medication adherence in LEDVT patients can be categorized into distinct classes, enabling more precise identification of their characteristics. Therefore, exploring these latent classes and their influencing factors holds significant importance for optimizing intervention strategies and improving prognosis.

A cross-sectional survey was used to study LEDVT. From March 14, 2024 to September 20, 2024, a random sampling method was used to recruit 469 patients with LEDVT from four grade-A tertiary hospitals in Urumqi, China. Participants completed questionnaires on general demographic information, the Medication Adherence Scale, the Perceived Health Competence Scale, the Herth Hope Index, the Patient Activation Measure, the Beliefs about Medicines Questionnaire-Specific. LPA was conducted to analyze the medication adherence characteristics of patients with LEDVT. Univariate analysis and multivariate logistic regression were used to identify the influencing factors of their latent profiles. Data analysis was performed using Mplus 8.3 and SPSS 25.0 software.

LPA was employed to investigate medication adherence in LEDVT patients, revealing three distinct latent classes: the poorest adherence group (44.99%), the moderate adherence group (19.83%), and the good adherence group (35.18%). The logistic regression results demonstrated that, perceived health competence, hope, activation, beliefs about medication necessity, and concerns about medication were influential factors affecting the potential profile of medication adherence (all p < 0.05).

LEDVT patients exhibit significant individual differences in medication adherence. Personalized intervention strategies can be designed based on different adherence classes to enhance medication adherence. Additionally, targeted interventions addressing perceived health competence, hope, positive affect, and medication beliefs can effectively improve adherence.