Construction and evaluation of an aspirin resistance risk prediction model for ischemic stroke

doi:10.21203/rs.3.rs-5280647/v1

Download PDF

Research Article

Construction and evaluation of an aspirin resistance risk prediction model for ischemic stroke

https://doi.org/10.21203/rs.3.rs-5280647/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background Aspirin has become the drug of choice for the prevention and treatment of ischemic stroke, but approximately a quarter of patients may be resistant to its effects and have an increased risk of recurrent ischemic events while also developing aspirin resistance. This study aimed to build a risk prediction model for AR in IS patients, predicts the likelihood of IS patients developing AR.

Methods The retrospective research study included the clinical data of patients with ischemic stroke were retrospectively collected from January 2021 to January 2023 at the Affiliated Hospital of Beihua University in the Jilin Province. Univariate and logistic regression analyses were used to construct a risk prediction model. The Hosmer–Lemeshow χ2 test and a receiver operating characteristic (ROC) curve were used to check the differential validity and calibration of the risk prediction model. The AR risk assessment criteria for ischemic stroke were established based on the β values of each risk factor and its variable types in the prediction model. The two evaluation criteria were compared and analyzed to determine the best criteria.

Results Seven risk factors were included in the prediction model. Sex (female), age (≥ 60 years), smoking, diabetes, hyperlipidemia, platelets > 350 × 10⁹ g/L, and glycosylated hemoglobin > 6.5% were independent influencing factors for the occurrence of AR in ischemic stroke. The area under the ROC curve (AUC) of the risk score model in the training group was 0.834 (0.772–0.896, P < 0.001). The Hosmer–Lemeshow test predicted the model fit effect χ²= 9.979, P = 0.267 > 0.05. In the validation group, the AUC was 0.819 (0.715–0.922, P < 0.001). Using the β value × 4 partial regression coefficient method, the scores and stratification of the AR risk prediction model were divided into three groups: no risk (0–3 points), low risk (4–15 points), and high risk (16–36 points).

Conclusions The AR ischemic stroke risk prediction model has strong prediction and assessment capabilities, enabling the precise identification of patients at risk of AR.

Ischemic stroke

aspirin resistance

risk prediction model

risk factors

Ischemic stroke, also known as cerebral infarction, is responsible for 60–82% of all strokes [1]. Ischemic stroke is a common type of stroke and a leading cause of long-term disability worldwide [2]. Patients with ischemic stroke typically experience relapse within 12 months [3]. Secondary prevention is an effective method for reducing recurrence and death from ischemic stroke [4]. Aspirin is an economical and effective antiplatelet agent, and it is the only antiplatelet agent of choice for the secondary prevention and acute treatment of ischemic stroke. This has been proven by evidence-based medicine, suggesting that aspirin can significantly reduce the incidence of ischemic stroke [5]. Additionally, aspirin is widely used in clinical practice [6, 7].

Aspirin can effectively prevent almost 75% of recurrent vascular events [4]. More than 80% of patients who have had a stroke in China use aspirin [8]; however, up to about 40% of patients have varying degrees of aspirin resistance (AR, also known as aspirin non-response) [9]. This means that even after consistent aspirin treatment, patients are still unable to successfully stop the development of malignant cerebrovascular events and the recurrence of thrombotic diseases [10]. AR (ineffectiveness) is as common in patients who have experienced ischemic stroke and recurrence as in patients with vascular malignant events (AR), with an incidence of up to 61.73% [11]. The incidence of cerebrovascular events is three times higher in patients with AR than in individuals with a normal aspirin response [12]. Thrombosis, fatal events, and poor clinical prognosis are more common in patients with AR [13]. Furthermore, patients often do not recognize the existence of AR, which can lead to the failure of ischemic stroke-aspirin secondary prevention, the occurrence of malignant cerebrovascular events, and even death [14]. This is because the clinical manifestations of AR are not always obvious; its onset is hidden, and its mechanism of occurrence is unclear [10]. Therefore, the diagnosis of AR depends on clinical laboratory evidence (aspirin reactivity unit ≥ 550 or arachidonic acid [AA]-induced platelet aggregation rate > 30%) [15]. Therefore, it is possible to reduce the incidence of malignant cerebrovascular events and the recurrence rate of ischemic stroke, improve patients’ quality of life, reduce the burden on families and society, and effectively prevent AR if detected early [16]. This also allows for the timely correction of ineffective secondary prevention of aspirin. Risk factors and clinical treatments are the main areas of current research on AR in ischemic stroke [17, 18]. AR has not been accurately assessed, and inadequate early detection is one of the problems.

Disease prediction models are an essential tool for assessing the risk of chronic diseases [19]. The primary purposes of these models are to identify high-risk populations, determine risk variables, and improve disease prevention [20]. In developed countries, disease prediction models are useful for preventing and treating numerous diseases [21]. The risk-critical value of objective clinical risk indicators is fairly removed from clinical practice applications and clinical prediction rates. This value is missing in traditional risk prediction models that primarily focus on subjective risk factors and are derived through literature reviews or meta-analyses [22]. Few studies exist on objective clinical treatment indicators, such as biochemical indicators, behavioral indicators, and pharmacotherapeutics [23]. Further research is needed to determine the risk-critical value of clinical indicators. To increase the clinical prediction rate of the disease risk prediction model, employing a series of data mining techniques and clinical data evaluation focused on deep clinical objective risk factors is imperative.

To provide patients with a reliable tool to predict the risk of AR, we collected the medical records of patients with ischemic stroke who took aspirin, examined the risk factors for AR in ischemic stroke, and introduced a model that provides an AR risk score for ischemic stroke.

Aim

To provide patients with a reliable tool to predict the risk of AR based on clinical data, and a score model for predicting AR risk was constructed.

Study design and participants

Data selection and analysis were performed using a retrospective cross-sectional design. The data were collected between January 2021 and April 2023.

The study participants were patients who experienced ischemic stroke and were admitted to the Department of Neurology and Stroke at the Affiliated Hospital of Beihua University in the Jilin Province between January 2021 and April 2023. The governing principle involved at least 10 times the number of logistically independent variables [24]. The incidence of AR in ischemic stroke can reach 65% [25], so the sample size should be increased by 20% and not less than (number of independent variables×10)/incidence. This study included a retrospective analysis of the demographic and laboratory examination characteristics of 285 patients. The hospital ethics committee reviewed and approved this study (BHFS-2021-009). Patients who met the following criteria were eligible for inclusion: (1) had an ischemic stroke diagnosed by head magnetic resonance imaging or computed tomography [26]; (2) had taken aspirin regularly, with a daily loading dose of 100 mg and a continuous duration of at least 7 days; and (3) had undergone thromboelastography [27]. The exclusion criteria were as follows: (1) use of nonsteroidal anti-inflammatory drugs, anticoagulants, or additional antiplatelet medications within 4 weeks of enrollment; (2) history of surgery within 1 week before admission; and (3) blood disease or bleeding tendency.

Data grouping

AR group: According to the inpatient medical records, the medication needed to be changed, the results of the AR risk genetic test were attenuated, or the AA inhibition rate was above 50%. Aspirin non-resistant group: Inpatient records demonstrated a moderate or good response to AR risk genetic testing and aspirin intolerance or AA inhibition rate of 50% ≤ AA ≤ 100%. For 199 cases (70%) in the training group and 86 cases (30%) in the validation group, a random division into the prediction model and verification model effects was made based on a ratio of 7:3.

Candidate predictor variables

Clinical and laboratory variables such as sex, age, body mass index, smoking, hypertension (HTN), coronary artery disease (CAD), diabetes mellitus (DM), hyperlipidemia (HLP), history of stroke, and fibrinogen (FBI) were included in the data collection. Other variables included platelet count (PLT), fasting blood glucose (FBG), low-density lipoprotein cholesterol (LDH), homocysteine (HCY), and glycosylated hemoglobin (HbA1c). Information collected at the time of the initial diagnosis from computerized and printed medical records served as the baseline. Using standardized data collection and quality control procedures, trained staff retrospectively collected relevant information from the medical data platform of the Affiliated Hospital of Beihua University. General data and risk factor indicators of the participants were collected through a two-person verification process and arranged based on the contents of the collection table.

Risk scoring and stratification

The AR ischemic stroke risk prediction model was evaluated based on the value of the partial regression coefficient (β) of each variable in the binary logistic regression analysis. Two methods were used to establish the risk assessment criteria:

(1) The basic score × grade score method was used to determine the scoring criteria for AR risk based on the β value of each risk factor and its variable types. For a given risk factor, a point was assigned to the lowest β value, and the division value (integer) of the factor was used to determine the baseline score of other factors. If a risk factor was a dichotomous variable, it was assigned a value of 0 or 1.

(2) The risk factors were ranked based on the partial regression coefficient (β value) in the logistic regression model using the partial regression coefficient method β × 4 (rounded to an integer). The risk factor score was determined using β × 4 (rounded to an integer), with the reference category for each variable being 0. This was accomplished by considering previous research and the special circumstances of this study.

(3) Prediction and evaluation of the AR risk prediction model for ischemic stroke. The prediction accuracy = (true positive + true negative)/total number of cases was used to evaluate the predictive capacity of the model in the validation set. Two evaluation criteria were used to compare and examine the frequency of AR prediction.

Statistical analysis

SPSS software (version 24.0; IBM; Armonk, NY, USA) was used to analyze all data. The frequency, rate, and percentage were used to represent the counting data. The risk factors were analyzed using the chi-square test or Fisher’s exact probability method to compare the resistant and non-resistant groups. P-values less than 0.05 indicated statistical significance.

Through systematic sampling, the selected patients with ischemic stroke were randomly assigned to training and validation groups, with 70% (199/285) of patients in the training group and 30% (86/285) in the validation group. The training group data were selected to build a prediction model, and single-factor analysis was used to eliminate the risk factors associated with AR in patients with ischemic stroke. AR was considered the dependent variable, and the risk factors associated with AR were considered independent variables. This analysis focused on determining the likelihood of AR when multiple independent variables were present. The backward likelihood ratio method is a special method of logistic regression. After determining the β value of the partial regression coefficient of each risk factor, the inclusion criteria were set at 0 points05, the exclusion criteria were set at 0 points1, and the prediction model was constructed.

The receiver operating characteristic (ROC) curve was drawn. The validity of the identification was better when the area under the line was greater than 0.7. The calibration level of the model was tested using the Hosmer–Lemeshow (H–L) test, and P > 0.05 was considered a good calibration level.

Demographic and clinical characteristics of the two groups

Demographic and clinical characteristics of the 285 participants are shown in Table 1: 206 patients (72.3%) did not have AR, and 79 patients (27.7%) developed AR. They were randomly divided into training (199 cases, 70%) and validation groups (86 cases, 30%) based on a ratio of 7:3. The training group comprised 55 cases, and the aspirin non-resistance group had 144 cases. The validation group consisted of 24 patients with AR (27.9% of the total) and 62 patients without AR (72.1%). Regarding the general demographic data, no significant differences were observed between the validation and training groups (P > 0.05).

Table 1

Demographics and clinical characteristics of the two groups.
Variables		The modeling group				χ²	P
Variables		(N = 199, %)		(N = 86, %)		χ²	P
Gender	Male		136(68.3%)	53(61.6%)	1.212		0.271
	Female		63(31.7%)	33(38.4%)
Age	≤ 60 Y		125(62.8%)	53(61.6%)	0.036		0.849
	>60Y		74(37.2%)	33(38.4%)
BMI	≤ 24		63(31.7%)	36(41.9%)	2.757		0.097
	>24		136(68.3%)	50(58.1%)
Smoking	No		105(52.8%)	53(61.6%)	1.910		0.167
	Yes		94(47.2%)	33(38.4%)
HTN	No		61(30.7%)	29(33.7%)	0.262		0.609
	Yes		138(69.3%)	57(66.3%)
CHD	No		122(61.3%)	61(70.9%)	2.420		0.120
	Yes		77(38.7%)	25(29.1%)
DM	No		128(64.3%)	59(68.6%)	0.488		0.485
	Yes		71(35.7%)	27(31.4%)
HLP	No		87(43.7%)	47(54.7%)	2.881		0.090
	Yes		112(56.3%)	39(45.3%)
Stroke history	No		131(65.8%)	63(73.3%)	1.524		0.217
	Yes		68(34.2%)	23(26.7%)
FBI (g/L)	≤ 4.98		167(83.9%)	70(81.4%)	0.273		0.601
	> 4.98		32(16.1%)	16(18.6%)
PLT (g/L)	≤ 350		183(92.0%)	76(88.4%)	0.932		0.334
	> 350		16(8.0%)	10(11.6%)
FBG (mmol/L)	≤ 6.1		112(56.3%)	50(58.1%)	0.085		0.771
	> 6.1		87(43.7%)	36(41.9%)
LDH (mmol/L)	≤ 4.11		185(93.0%)	82(95.3%)	0.577		0.448
	> 4.11		14(7.0%)	4(4.7%)
HCY (µmol/L)	≤ 15		118(59.3%)	49(57.0%)	0.133		0.715
	> 15		81(40.7%)	37(43.0%)
HbA1C (%)	≤ 6.5		69(34.7%)	40(46.5%)	3.563		0.059
	> 6.5		130(65.3%)	46(53.5%)

Risk factors for AR in ischemic stroke

For the single-factor analysis, 199 patients with ischemic stroke in the training group were divided into resistance and non-resistance groups based on whether they developed AR. The results are shown in Table 2, where it was found that AR was not associated with body mass index, stroke history, FBI, LDH, or HCY (P > 0.05). AR was found to be associated with sex (χ² = 6.687, P = 0.010), age (χ² = 7.859, P = 0.005), smoking (χ² = 12.243, P < 0.001), HTN (χ² = 4.457, P = 0.035), CAD (χ² = 4.179, P = 0.041), DM (χ² = 14.171, P < 0.001), HLP (χ² = 8.355, P = 0.004), PLT (χ² = 17.024, P < 0.001), FBG (χ² = 4.939, P = 0.026), and HbA1c (χ² = 7.225, P = 0.007). These differences were significant (P < 0.05).

Table 2

Univariate analysis of risk factors for AR in ischemic stroke (modeling group).
Variables		Non-AR	AR	χ²	P
Variables		(N = 144,%)	(N = 55,%)	χ²	P
Gender	Male	106(73.6%)	30(54.5%)	6.687	0.010
	Female	38(26.4%)	25(45.5%)
Age (yr)	≤ 60	99(68.8%)	26(47.3%)	7.859	0.005
	>60	45(31.3%)	29(52.7%)
BMI	≤ 24	41(28.5%)	22(40.0%)	2.444	0.118
	>24	103(71.5%)	33(60.0%)
Smoking	No	87(60.4%)	18(32.7%)	12.243	<0.001
	Yes	57(39.6%)	37(67.3%)
HTN	No	38(26.4%)	23(41.8%)	4.457	0.035
	Yes	106(73.6%)	32(58.2%)
CHD	No	82(56.9%)	40(72.7%)	4.179	0.041
	Yes	62(43.1%)	15(27.3%)
DM	No	104(72.2%)	24(43.6%)	14.171	<0.001
	Yes	40(27.8%)	31(56.4%)
HLP	No	72(50.0%)	15(27.3%)	8.355	0.004
	Yes	72(50.0%)	40(72.7%)
Stroke history	No	92(63.9%)	39(70.9%)	0.872	0.350
	Yes	52(36.1%)	16(29.1%)
FBI (g/L)	≤ 4.98	118(81.9%)	49(89.1%)	1.506	0.220
	>4.98	26(18.1%)	6(10.9%)
PLT (g/L)	≤ 350	140(97.2%)	43(78.2%)	17.024	<0.001
	>350	4(2.8%)	12(21.8%)
FBG (mmol/L)	≤ 6.1	88(61.6%)	24(43.6%)	4.939	0.026
	>6.1	56(38.9%)	31(56.4%)
LDH (mmol/L)	≤ 4.11	133(92.4%)	52(94.5%)	0.290	0.590
	>4.11	11(7.6%)	3(5.5%)
HCY (µmol/L)	≤ 15	83(57.6%)	35(63.6%)	0.593	0.441
	>15	61(42.4%)	20(36.4%)
HbA1C (%)	≤ 6.5	58(40.3%)	11(20.0%)	7.225	0.007
	>6.5	86(59.7%)	44(80.0%)

The dependent variable was the occurrence of AR (0 = non-occurrence, 1 = occurrence), and the entry and removal criteria were 0.05 and 0.10, respectively. The stepwise backward method estimated by the maximum likelihood ratio was used to include the 10 significant factors in the univariate analysis: sex (P = 0.010), age (P = 0.005), smoking (P < 0.001), HTN (P = 0.035), CAD (P = 0.041), DM (P < 0.001), HLP (P = 0.004), PLT (P < 0.001), FBG (P = 0.026), and HBA1c (P = 0.007).

The results of the binary logistic regression analyses are presented in Table 3. Seven risk factors for AR in ischemic stroke were included in the logistic regression analysis: female sex (odds ratio [OR] = 4.352, 95% confidence interval [CI]: 1.798–10.537), age ≥ 60 years (OR = 2.594, 95% CI: 1.191–5.651), smoking (OR = 3.515, 95% CI: 1.535–8.049), DM (OR = 3.784, 95% CI: 1.689–8.476), HLP (OR = 2.797, 95% CI: 1.228–6.368), PLT > 350 × 10⁹ g/L (OR = 8.319, 95% CI: 2.004–34.539), and HbA1c > 6.5% (OR = 2.875, 95% CI: 1.167–7.087).

Table 3

Multivariate analysis on risk factors for AR in ischemic stroke (modeling group).
Variables	β	S.E	Wald	P	OR	95%CI
Female	1.471	0.451	10.629	0.001	4.352	1.798 ~ 10.537
Age (>60years)	0.953	0.397	5.761	0.016	2.594	1.191 ~ 5.651
Smonking (yes)	1.257	0.423	8.843	0.003	3.515	1.535 ~ 8.049
Diabetes (yes)	1.331	0.412	10.455	0.001	3.784	1.689 ~ 8.476
HLP (yes)	1.029	0.42	6.003	0.014	2.797	1.228 ~ 6.368
PLT>350×10⁹g/L	2.118	0.726	8.507	0.004	8.319	2.004 ~ 34.539
HbA1C>6.5%	1.056	0.46	5.266	0.022	2.875	1.167 ~ 7.087
Constant	-6.160	1.030	35.745	0.000	0.002

Derivation of the risk prediction model

Independent risk factors associated with AR in patients with ischemic stroke were identified using multivariate logistic regression. The constant term in the logistic regression was − 6.16, and the β value in Table 3 represents the β value of the partial regression coefficient for each variable in the equation. Logit P = -6.16 + sex (female) × 1.471 + age × 0.953 (< 60 = 0; ≥ 60 = 1) + smoking × 1.257 + DM × 1.331 + HLP × 1.029 + PLT × 2.118 + HBA1c × 1.056.

Validation of the risk prediction model

The H–L chi-square test was used to evaluate the goodness-of-fit of the predictive model and compare the consistency between the observed probability of the event and the predicted probability of the model. If P > 0.05, the training group data were considered fully extracted, indicating a high goodness-of-fit for the model. The predicted probability of AR in patients with ischemic stroke was higher than the actual probability and fit well according to the H–L test results, which showed that P = 0.267 > 0.05, x² = 9.979, and the difference was not significant.

To evaluate the discriminant validity of the prediction model, an ROC curve was created, and the area under the curve (AUC) was computed. The maximum approximate entry index was 0.487, the corresponding optimal critical value was 0.249, or the threshold value was 0.249. The AUC of the training group was 0.834, 95% CI (0.772–0.896). The model had good prediction efficiency, as demonstrated in Fig. 1, with a sensitivity of 70.9% and a specificity of 77.8%. The model formula classified patients with AR as high risk when P_prediction ≥ 0.249. P_prediction < 0.249 was used to identify patients with low-risk AR. The AUC value in the validation group was 0.819 (95% CI: 0.715–0.922), demonstrating a high degree of model differentiation and good predictive ability. This indicates that the model can be used to predict the risk of AR in patients who have had ischemic strokes.

Scoring and stratification of AR risk scoring model for ischemic stroke

The base score × grade score method and the partial regression coefficient β × 4 (rounded to an integer) method were used to establish the scoring criteria. The results are summarized in Table 4. In the base score × grade score method, the scoring standard of AR occurrence risk was established according to the β value of each risk factor and its variable type in the established prediction model. A total score of 0 indicates that the patient is not at risk of AR. In this score, 15 patients scored 0 points, and the remaining 270 patients were divided into risk categories. All patients were scored according to the set scoring criteria. Risk stratification was performed based on the segmentation method of the area truncation value under the curve, and patients were divided into high- and low-risk groups. The maximum threshold was 0.46, and the corresponding score was 4 points. The no AR group had 0 points, the low-risk group had 1–4 points, and the high-risk group had 5–8 points for a total of 8 points. The scoring criteria for AR occurrence risk were determined by considering the variable types and β values of each risk factor in the established prediction model. The risk factors were scored as follows: 6 points for sex (female), 4 points for age (≥ 60 years), 4 points for HLP and HbA1c > 6.5 mmol/L, 5 points for DM and smoking, and 8 points for PLT > 350 × 10⁹ g/L. This resulted in an overall score of 36 points. All patients were scored according to preset scoring criteria using the tricolor method. Three risk groups were formed from the patient pool: no risk (0–3 points), low risk (4–15 points), and high risk (16–36 points).

Table 4

Scoring criteria for risk prediction of aspirin resistance.
	Base score × grade score	β × 4
	Points	Points
Female	1	6
Age (>60 years)	1	4
Smonking (yes)	1	5
Diabetes (yes)	1	5
HLP (yes)	1	4
PLT>350×10⁹g/L	2	8
HbA1C>6.5%	1	4
Total	8	36

Using the two scoring criteria, we compared 206 non-AR and 79 AR cases to predict the incidence of AR in patients with ischemic stroke. The results are shown in Table 5, with a total score of 8. The AR scoring standard established by the basic score × grade scoring method revealed that 31 patients had developed AR and 201 patients had not, yielding an 81-point prediction accuracy. Following the establishment of the AR scoring standard using the partial regression coefficient β × 4 (rounded to an integer) method, 40 actual patients with AR and 196 actual patients with AR with a total score of 36 points were identified. The forecast accuracy was 82.8%. The prediction accuracy of the partial regression coefficient method was higher than that of the basic score × grade method.

Table 5

Comparison of the two scoring criteria's effects on the aspirin resistance screening accuracy in clinical cases of ischemic stroke risk prediction.
Base score × grade score						β × 4
		Non-AR		AR	Total	Non-AR	AR	Total
Actual	Non-AR	201	5		206	196	10	206
	AR	48	31		79	39	40	79
		249	36		285	235	50	285
Accuracy					81.4%			82.8%

Stroke has become a major public health concern owing to its high frequency and high rates of disability, death, and recurrence [8]. Of all stroke cases, 60–80% are caused by an ischemic stroke. Aspirin can stop nearly 75% of recurrent vascular events. Aspirin is an inexpensive and effective antiplatelet drug that can be used to treat and prevent thrombotic events and cardiovascular diseases [28]. Nonetheless, research indicates that AR impacts roughly 40% of patients to differing extents [9]. Even with consistent aspirin treatment, patients are still unable to successfully prevent the development of malignant cerebrovascular events and the recurrence of thrombotic diseases [29]. It is easy to ignore the existence of AR because of its unclear mechanism of occurrence and hidden onset, which frequently happens without obvious clinical symptoms. This results in the failure of secondary ischemic stroke prevention with aspirin [30]. Because AR increases the incidence of malignant cerebrovascular events in patients with ischemic stroke to 25%, patients experience permanent harm when AR occurs [31]. Thus, clinical significance arises from the early identification of patients with ischemic stroke who are at a high risk of AR. In addition to screening for risk factors for ischemic stroke AR and developing a risk score model, this study involved the retrospective collection of the medical records of patients with ischemic stroke who took aspirin. The results showed that sex (female), age (above 60 years), smoking, DM, HLP, PLT > 350 × 10⁹ g/L, and HBA1c > 6.5% were independent risk factors for AR in patients with ischemic stroke. Previous studies have shown that the incidence of AR in patients with ischemic stroke is highly correlated with age (over 60), sex (female), DM, and smoking [32]. In agreement with the findings of Paven et al., platelet reactivity and 11-DH-TXB2 levels in urine were higher in female patients than in male patients [33]. As a result, COX-1 inhibition was reduced, and aspirin insensitivity was more likely to occur [34]. Diminished disease tolerance, noncompliance with medication, and markedly diminished physiological function are all independent risk factors for ischemic stroke and can lead to an elevated risk of recurrence [35].

Smoking and other poor lifestyle choices can damage the intima of blood vessels, leading to vasoconstriction, lumen narrowing, increased risk of drug absorption, and vascular blockage [36]. Patients with ischemic stroke and DM may experience increased blood viscosity, platelet adhesion to blood vessel walls, red blood cell aggregation, or platelet agglutination, leading to the enlargement of atherosclerotic plaques and the development of thrombosis [37]. Diseases and other factors decrease the body’s platelet survival rate while increasing platelet turnover rates. In other words, the platelet effect of AR in patients with thrombus is diminished, increasing PLT. Studies have demonstrated a positive association between elevated HBA1c and platelet hyperresponse [38]. Elevations in HbA1c are associated with improved platelet aggregation capacity and a higher risk of acute respiratory distress (AR) [39, 40]. An effective measure of the body’s ability to regulate blood glucose levels is the HbA1c [41]. The β values of each risk factor and their corresponding variable types were calculated to establish the AR risk scoring criteria, considering the strong correlation between the aforementioned variables and AR in patients with ischemic stroke [42].

The ROC curve and H–L test (P > 0.05) were used to evaluate the degree of calibration and differentiation of the risk scoring models. A prediction effect with an AUC greater than 0.7 is generally considered good. The higher the AUC, the better the risk prediction. Our results showed that the AUCs for the scoring model for the training and validation groups were 0.834 (0.772–0.896, P < 0.001) and 0.819 (0.715–0.922, P < 0.001), respectively, and the H–L test results were both P > 0.05. This shows that in the future, this model can be used to screen patients with ischemic stroke receiving aspirin therapy for AR and has good clinical application value in identifying high-risk AR groups. The best scoring standard was determined by comparing and analyzing the two risk scoring techniques. In the first-grade scoring method, the lowest value factor was given a base score of 1, and the scores of the other factors were calculated by dividing each of the lowest values by 1. Each risk factor in the second quadruple scoring method was determined by the integer β value × 4. Based on the incidence of AR in patients matching each risk score, the quantile method was used to divide the patients into three risk groups: low, medium, and high.

Based on the AR evaluation standard set with a total score of 36, comparative analysis results showed that 40 patients with AR developed AR, and 196 did not meet the study evaluation standard for the occurrence of AR, with a prediction accuracy rate of up to 82.8%. The risk was determined using a four-way scoring method. Clinicians can regularly assess patients’ AR risk scores, determine whether patients with ischemic stroke are likely to be aspirin-resistant, and use a risk prediction score model to provide targeted interventions to reduce the risk of early AR.

In this study, we investigated the risk factors for AR in patients with ischemic stroke based on clinical medical records and established the evaluation criteria of the AR risk prediction model for ischemic stroke. Furthermore, an AR risk prediction model for patients with ischemic stroke was constructed using logistic regression analysis, and its assessment, prediction, and early detection sensitivities were validated. The AR ischemic stroke risk prediction model has strong predictive, scoring, and sensitivity functions. It can reliably identify AR in patients with ischemic stroke and provide clinical staff with a dependable risk prediction tool to help them identify high-risk populations and promptly implement prevention and intervention measures.

Author contributions

TYM and MZ conceived and designed the study idea. TYM and YS performed the search, quality assessment, and extraction of the data. TYM and XW contributed to the analysis and interpretation of data. All authors contributed to the writing and editing of the manuscript and agreed to the final manuscript.

Funding

This work was partially supported by the Government Technology Department of the Jilin Province (grant numbers: 20210203072SF and 20230203054SF) and the Education Department of the Jilin Province, China (grant number: JJKH20210068KJ).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

The study was approved by the Medical Ethics Committee (ethics review number: 2023 (25)).

Clinical trial number

Not applicable

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Liu L, Chen W, Zhou H, Duan W, Li S, Huo X, et al. Chinese Stroke Association guidelines for clinical management of cerebrovascular disorders: executive summary and 2019 update of clinical management of ischaemic cerebrovascular diseases[J]. Stroke Vasc Neurol. 2020;5(2):159–76.
Jayaraj RL, Azimullah S, Beiram R, Jalal FY, Rosen GA. Neuroinflammation: friend and foe for ischemic stroke[J]. J Neuroinflammation. 2019;16(1):142.
XU J, Zhang X, Jin A, Pan Y, Li Z, Meng X, et al. Trends and Risk Factors Associated With Stroke Recurrence in China, 2007–2018[J]. JAMA Netw Open. 2022;5(6):e2216341.
Flach C, Muruet W, Wolfe CDA, Bhalla A, Douiri A. Risk and Secondary Prevention of Stroke Recurrence: A Population-Base Cohort Study[J]. Stroke. 2020;51(8):2435–44.
Fintel DJ. Oral antiplatelet therapy for atherothrombotic disease: overview of current and emerging treatment options[J]. Vasc Health Risk Manag. 2012;8:77–89.
Kamarova M, Baig S, Patel H, Monks K, Wasay M, et al. Antiplatelet Use in Ischemic Stroke[J]. Ann Pharmacother. 2022;56(10):1159–73.
Mekaj YH, Daci FT, Mekaj AY. New insights into the mechanisms of action of aspirin and its use in the prevention and treatment of arterial and venous thromboembolism[J]. Ther Clin Risk Manag. 2015;11:1449–56.
Tu WJ, Wang LD. China stroke surveillance report 2021[J]. Mil Med Res. 2023;10(1):33.
Macchi L, Sorel N, Christiaens L. Aspirin resistance: definitions, mechanisms, prevalence, and clinical significance[J]. Curr Pharm Des. 2006;12(2):251–8.
Floyd CN, Ferro A. Mechanisms of aspirin resistance[J]. Pharmacol Ther. 2014;141(1):69–78.
Sambu N, Radhakrishnan A, Englyst N, Weir N, Curze N. Aspirin resistance in ischemic stroke: insights using short thrombelastography[J]. J Stroke Cerebrovasc Dis. 2013;22(8):1412–9.
Gum PA, Kottke-marchant K, Welsh PA, White J, Topol EJ. A prospective, blinded determination of the natural history of aspirin resistance among stable patients with cardiovascular disease[J]. J Am Coll Cardiol. 2003;41(6):961–5.
Krasopoulos G, Brister SJ, Beattie WS, Buchanan MR. Aspirin resistance and risk of cardiovascular morbidity: systematic review and meta-analysis[J]. BMJ. 2008;336(7637):195–8.
Yi X, Wang C, Liu P, Fu C, Lin J. Antiplatelet drug resistance is associated with early neurological deterioration in acute minor ischemic stroke in the Chinese population[J]. J Neurol. 2016;263(8):1612–9.
Lev EI, Solodky A, Harel N, Mager A, Brosh D, et al. Treatment of aspirin-resistant patients with omega-3 fatty acids versus aspirin dose escalation[J]. J Am Coll Cardiol. 2010;55(2):114–21.
Kolmos M, Christofeersen L, Kruuse C. Recurrent Ischemic Stroke - A Systematic Review and Meta-Analysis[J]. J Stroke Cerebrovasc Dis. 2021;30(8):105935.
Roman-Gonzalez A, Naranjo CA, Cardona-maya WD, Vallejo D, Garcia F et al. (2021). Frequency of Aspirin Resistance in Ischemic Stroke Patients and Healthy Controls from Colombia[J]. Stroke Res Treat,2021: 9924710.
Thomsno RM, Anderson DC. Aspirin and clopidogrel for prevention of ischemic stroke[J]. Curr Neurol Neurosci Rep. 2013;13(2):327.
Oh SM, Stefanl KM, Kim HC. Development and application of chronic disease risk prediction models[J]. Yonsei Med J. 2014;55(4):853–60.
Moons KG, De Groot JA, Bouwmeester W, et al. Critical appraisal and data extraction for systematic reviews of prediction modelling studies: the CHARMS checklist[J]. PLoS Med. 2014;11(10):e1001744.
Farzadfar F. Cardiovascular disease risk prediction models: challenges and perspectives[J]. Lancet Glob Health. 2019;7(10):e1288–9.
Staszewski J, Piusinska-macoch R, Skrobowska E, Brodacki B, Macek K, et al. Aspirin Resistance: Risk Factors and Prognostic Significance in Patients with Cerebral Small Vessel Disease[J]. Ann Clin Lab Sci. 2018;48(1):45–54.
Stuckey TD, Kirtane AJ, Brodie BR, Witzenbichler B, Litherland C, et al. Impact of Aspirin and Clopidogrel Hyporesponsiveness in Patients Treated With Drug-Eluting Stents: 2-Year Results of a Prospective, Multicenter Registry Study[J]. JACC Cardiovasc Interv. 2017;10(16):1607–17.
Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis[J]. J Clin Epidemiol. 1996;49(12):1373–9.
Zytkiewicz M, Gielwanowska L, Wojtasinska E, Psuja P, Zawilska K. Resistance to acetylsalicylic acid in patients after ischemic stroke[J]. Pol Arch Med Wewn. 2008;118(12):727–33.
Jing Y, Yue X, Yang S, Li S. Association of Aspirin Resistance with Increased Mortality in Ischemic Stroke[J]. J Nutr Health Aging. 2019;23(3):266–70.
Tantry US, Bliden KP, Gurbel PA. Overestimation of platelet aspirin resistance detection by thrombelastograph platelet mapping and validation by conventional aggregometry using arachidonic acid stimulation[J]. J Am Coll Cardiol. 2005;46(9):1705–9.
Khan H, Kanny O, Syed MH, Qadura M. (2022). Aspirin Resistance in Vascular Disease: A Review Highlighting the Critical Need for Improved Point-of-Care Testing and Personalized Therapy[J]. Int J Mol Sci,23(19).
Lusk JB, Xu H, Petersin ED, Bhatt DL, Fonarow G. Antithrombotic Therapy for Stroke Prevention in Patients With Ischemic Stroke With Aspirin Treatment Failure[J]. Stroke. 2021;52(12):e777–81.
Ansa BE, Hoffman Z, Lewis N, Savoy C, Hickson A. (2019). Aspirin Use among Adults with Cardiovascular Disease in the United States: Implications for an Intervention Approach[J]. J Clin Med, 8(2).
Poulsen TS, Kristensen SR, Korsholm L, Haghfelt T, Jorgensen B, et al. Variation and importance of aspirin resistance in patients with known cardiovascular disease[J]. Thromb Res. 2007;120(4):477–84.
Pu L, Wang L, Zhang R, Zhao T, Jiang Y, et al. Projected Global Trends in Ischemic Stroke Incidence, Deaths and Disability-Adjusted Life Years From 2020 to 2030[J]. Stroke. 2023;54(5):1330–9.
Paven E, Dillinger JG, Bal Dit Sollier C, Vidal-trecan T, Berge N, et al. Determinants of aspirin resistance in patients with type 2 diabetes[J]. Diabetes Metab. 2020;46(5):370–6.
Qayyum R, Becker DM, Yanek LR, Moy TF, Becker LC, et al. Platelet inhibition by aspirin 81 and 325 mg/day in men versus women without clinically apparent cardiovascular disease[J]. Am J Cardiol. 2008;101(9):1359–63.
Aremu TO, Oluwolf OE, Adeyinka KO, Schommer JC. (2022). Medication Adherence and Compliance: Recipe for Improving Patient Outcomes[J]. Pharm (Basel), 10(5).
Salehi N, Janjani P, Tadbiri H, Rozbahani M, Jalilian M. Effect of cigarette smoking on coronary arteries and pattern and severity of coronary artery disease: a review[J]. J Int Med Res. 2021;49(12):3000605211059893.
Maida CD, Daidone M, Pacinella G, Norrito RL, Pinto A et al. (2022). Diabetes and Ischemic Stroke: An Old and New Relationship an Overview of the Close Interaction between These Diseases[J]. Int J Mol Sci, 23(4).
Ertugrul DT, Tutal E, Yildiz M, Akin O, Yalcin AA, et al. Aspirin resistance is associated with glycemic control, the dose of aspirin, and obesity in type 2 diabetes mellitus[J]. J Clin Endocrinol Metab. 2021;95(6):2897–901.
Lee S, Liu T, Zhou J, Zhang Q, Wong WT, et al. Predictions of diabetes complications and mortality using hba1c variability: a 10-year observational cohort study[J]. Acta Diabetol. 2021;58(2):171–80.
Neergaard-petersen S, Hvas AM, Grove EL, Larsen SB, Gregersen S, et al. The Influence of Haemoglobin A1c Levels on Platelet Aggregation and Platelet Turnover in Patients with Coronary Artery Disease Treated with Aspirin[J]. PLoS ONE. 2015;10(7):e0132629.
Weinstock RS, Aleppo G, Bailey TS, Bergenstal RM, Fisher WA, Arlington et al. (VA); American Diabetes Association.
Hu L, Zhu Y, Chen M, Li X, Lu X, et al. Development and Validation of a Disease Severity Scoring Model for Pediatric Sepsis[J]. Iran J public health. 2016;45(7):875–84.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Construction and evaluation of an aspirin resistance risk prediction model for ischemic stroke

Status:

Version 1

Abstract

Figures

Background

Methods

Aim

Study design and participants

Data grouping

Candidate predictor variables

Risk scoring and stratification

Statistical analysis

Results

Demographic and clinical characteristics of the two groups

Risk factors for AR in ischemic stroke

Derivation of the risk prediction model

Validation of the risk prediction model

Scoring and stratification of AR risk scoring model for ischemic stroke

Discussion

Conlusions

Declarations

References

Additional Declarations

Status:

Version 1