# Limited Predictive Performance of the SAMe-TT₂R₂ Score for Suboptimal Time in Therapeutic Range (TTR) in Thai Patients With Atrial Fibrillation: A Retrospective Analysis

## TL;DR

The SAMe-TT₂R₂ score is not effective at predicting poor anticoagulation control in Thai patients with atrial fibrillation, suggesting it should not be used for clinical decisions in this population.

## Contribution

This study evaluates and challenges the clinical utility of the SAMe-TT₂R₂ score in a Thai atrial fibrillation population.

## Key findings

- The SAMe-TT₂R₂ score had poor predictive performance (AUC = 0.55) for suboptimal TTR in Thai patients.
- High sensitivity but low specificity at the standard cutoff limits the score's clinical usefulness.
- Age <60 years and longer warfarin use were associated with better anticoagulation control.

## Abstract

Background: The SAMe-TT₂R₂ score is used to predict poor anticoagulation control (time in therapeutic range (TTR)) in warfarin users. However, its performance varies across ethnicities, and its validity in the Thai population is uncertain. This study aimed to evaluate the score's predictive performance and redefine the optimal cutoff for Thai patients with atrial fibrillation (AF).

Methodology: This retrospective analysis included 140 patients with non-valvular AF on warfarin at a single University Hospital. TTR was calculated using the Rosendaal method, with suboptimal control defined as TTR <70%. The score's overall predictive performance was assessed using the area under the ROC curve (AUC).

Results: The mean of TTR was 59.89% ± 28.56%, and 83 patients (59.3%) had suboptimal TTR. The SAMe-TT₂R₂ score ≥3 demonstrated poor predictive ability for suboptimal TTR (AUC = 0.55; 95% confidence interval (CI), 0.50-0.60). At this cutoff, key diagnostic metrics included high sensitivity (95.2%) but extremely low specificity (14.0%), a modest positive predictive value (PPV) of 61.7% and a low negative predictive value (NPV) of 66.7%. Notably, in multivariable analysis, age <60 years, a risk factor in the original score, was independently associated with a lower risk of suboptimal TTR. A longer duration of warfarin use was also associated with a lower risk of suboptimal TTR.

Conclusions: The SAMe-TT₂R₂ score showed poor performance in predicting suboptimal TTR in our Thai cohort and has limited clinical utility. The score cannot reliably rule in or rule out patients at risk for poor control. Clinical decisions, such as intensive monitoring or switching to a direct oral anticoagulant (DOAC), should be guided by a patient's actual TTR rather than the SAMe-TT₂R₂ score. The single-center, retrospective design and modest sample size limit generalizability and statistical power, necessitating a further prospective trial for validation.