Page 10 - BH-1-2
P. 10

Brain & Heart                                                  Post-stroke atrial fibrillation and predictive scores





                                   Individual patient data from three   Investigating the detection of AF  with prolonged Holter monitoring,  analyzed using multivariate   Age (0.76×year), NIHSS ≤5 (9), and   The threshold of 67.5 serves as  the delineation between low- and   CHADS 2  0.61 (NR), P=0.00032  (Cont’d...)





                                Uphaus et al. [10]  prospective studies  72-h NIEM  NIHSS >5 (21)  high-risk categories.


                             AS5F             regression  191  69.7±13.4  4.9%  >30 s   0.78















                             HAVOC  Kwong et al. [9]  Retrospective  Multivariate regression  9,589  68.1±13.4/67.5±13.4  ICD-9-CM data bank during follow-up   revealing AF diagnosis  5%  Any  Age ≥75 years (2), BMI≥30 (1), blood  hypertension (2), congestive HF (4), CAD  (2), PAD (1), and cardiac valve disease (2)  0 – 14  0.77 (NR)  Significant better accuracy, sensitivity,  and specificity (P<0.001) compared with   CHA 2 DS 2 -VASc (AUC NR)







                          Table 1. Synoptic table for predictive scores of post‑stroke atrial fibrillation detection



                                Sudacevschi et al. [8]  Retrospective  Multivariate regression  63.2±16.0 (with AF: 72.9±9.7; without AF:   21-day NIEM  Age>70 years (1), premature atrial  complex on 12-lead ECG (1), left ventricle  hypertrophy on echocardiography (1),  previous white matter lesions on brain   magnetic resonance (1)  NR, 4.8 (1.5 – 13.8) for Score 1, 23 (6.2 –  86.4) for Score 2, and 110 (15.5 – 778.5)




                             NR                 171  61.4±16.4)  15%  >30 s      0 – 4      for Score 3  /







                                Bugnicourt et al. [7]  Prospective  Multivariate regression  ECG and/or 24-h Holter-ECG  performed for any reason during a   1-year follow-up  Age≥72 years (2), history of CAD  (1), history of stroke (1), and LAE   NR, 0% Score 0 – 1, 7% Score 2,  14% Score 3, 32% Score 4, 67%




                             NR                 164  65.4±15.1  13%  Any  (2)    0 – 6      Score 5 – 6  /






                             Score acronym  Study  Study design  Statistic method for defining   AF predictors  Sample size (patient number)  Age (mean±SD or median   [IQR])  Methodology for AF   diagnosis  AF detection rate  Episode duration for AF   diagnosis  Score variables (point)  Score range  Score AUC (95% CI)  Comparator AUC (95% CI)







            Volume 1 Issue 2 (2023)                         4                         https://doi.org/10.36922/bh.0955
   5   6   7   8   9   10   11   12   13   14   15