Page 15 - Read Online

Hsu et al. Vessel Plus 2021;5:2 I http://dx.doi.org/10.20517/2574-1209.2020.45 Page 9 of 12

Table 3. Comparison of variables selected by different prognostic models in the literature
Performance
Author (year) Sample size Outcome assessed Variables included in the model
(AUC)
Counsell et al. [45] (2002) 530 30-day mortality and age, living alone, independence before stroke, 0.84-0.88
six-month independent verbal component of GCS, arm strength, ability to
survival walk
Muscari et al. [46] (2011) 211 9-month mRS NIHSS, need of urinary catheter, oxygen 0.84
administration, upper limb paralysis
Teale et al. (2012) 27-8964 30-180 days functional 2-11 variables (age, NIHSS, limb weakness, 0.75-0.88
[8]
review 17 models assessment dysarthria, conscious, diabetes, previous stroke,
fever, mRS, etc.)
Wouters et al. [27] (2018) 369 90 days mRS Baseline-NIHSS, age, ischemic heart disease 0.86
Jampathong et al. [25] 75-4441 90-365 days functional 1-11 variables (NIHSS, age, infarct volume, 0.73-0.84
(2018) review 23 models assessment diabetes, previous stroke, pre-stroke disability,
small-vessel stroke, t-PA use, preadmission mRS,
sex, atrial fibrillation,..,etc.)
Proposed model 37,094 90-day mRS age, discharge mRS, discharge NIHSS, recurrent 0.95-0.96
by LR method ischemia, previous stroke, Barthel index (BI)-
(This study) grooming, BI-dressing, aspirin use

NIHSS: National Institutes of Health Stroke Scale; AUCs: area under the curves

smoking and drinking were not selected in our final prediction models [Figure 3]. The relationship between
stroke outcomes and hospital distances, socioeconomic status, and timely treatment has been previously
discussed in the literature . Our study suggested that patients treated at close by medical centers had
[40]
good outcomes, and so were those in the east part of Taiwan (rural countryside) with farther distance to
the hospital, which may be associated with a younger population and smaller sample size, although further
studies may be required to explain this effect.

In our study, age of onset and previous CVA were found to be the most frequently selected predictors
[Figure 2], which were consistent with previously reported studies, however, these two variables were
non-modifiable factors which make them unusable for triage or treatment. Nevertheless, some variables
selected in our models might provide useful guidance during the triage and for the treatment plan. For
example, according to the coefficients of LR selected variables in different patient subgroups [Figure 3], the
models for all patients at admission required fewer variables to achieve similar performance as those of
all 140 variables used. In the case of the hemorrhagic patients, only two variables were selected that might
occur due to the different natural courses and pathology between hemorrhagic and ischemic patients [34,35] .
The negative coefficients of albumin and hemoglobin found in our study indicated higher values might
improve stroke outcomes. On the contrary, the positive coefficients of white blood cells (WBC), fasting
sugar, and heart rate provided warning signs to clinicians that these variables might be prone to poor
outcomes. The negative coefficient of Aspirin prescribed as the discharge medication was also shown a
positive effect in our discharge model of ischemic patients. The associations between stroke outcomes and
[42]
[41]
albumin , hemoglobin , and WBC have been reported, but Aspirin prescription has not been shown
[43]
[44]
as beneficial to stroke outcomes as found in our current population study. For the potential optimal
options of treatment, further evaluations on Aspirin were required for targeted interventions to prove its
positive effort on the improvement of stroke outcomes. Furthermore, several imaging variables as shown in
Supplemental Figure S2 including MRI no Finding (OR = 0.37), CT no Finding (OR = 0.62), MRI Lesion:
Left subcortical MCA (OR = 1.58), MRI: Left brainstem (OR = 1.85), and MRI Lesion: Right brainstem (OR
= 2.03), were selected in our admission prediction models. These clinical imaging findings can be used as
early predictors and indicators for predicting stroke outcomes to alert and assist clinicians during triages of
stroke patients.

Several studies have tried to build different prognostic models aiming for stroke outcome predictions [Table 3]
[45]
using various sample sizes in different populations. For example, Counsell et al. utilized six variables