Skip to main content
Download PDF

Predictive models for delay in medical decision-making among older patients with acute ischemic stroke: a comparative study using logistic regression analysis and lightGBM algorithm

BMC Public Health volume 24, Article number: 1413 (2024) Cite this article

Abstract

Objective

To explore the factors affecting delayed medical decision-making in older patients with acute ischemic stroke (AIS) using logistic regression analysis and the Light Gradient Boosting Machine (LightGBM) algorithm, and compare the two predictive models.

Methods

A cross-sectional study was conducted among 309 older patients aged ≥ 60 who underwent AIS. Demographic characteristics, stroke onset characteristics, previous stroke knowledge level, health literacy, and social network were recorded. These data were separately inputted into logistic regression analysis and the LightGBM algorithm to build the predictive models for delay in medical decision-making among older patients with AIS. Five parameters of Accuracy, Recall, F1 Score, AUC and Precision were compared between the two models.

Results

The medical decision-making delay rate in older patients with AIS was 74.76%. The factors affecting medical decision-making delay, identified through logistic regression and LightGBM algorithm, were as follows: stroke severity, stroke recognition, previous stroke knowledge, health literacy, social network (common factors), mode of onset (logistic regression model only), and reaction from others (LightGBM algorithm only). The LightGBM model demonstrated the more superior performance, achieving the higher AUC of 0.909.

Conclusions

This study used advanced LightGBM algorithm to enable early identification of delay in medical decision-making groups in the older patients with AIS. The identified influencing factors can provide critical insights for the development of early prevention and intervention strategies to reduce delay in medical decisions-making among older patients with AIS and promote patients’ health. The LightGBM algorithm is the optimal model for predicting the delay in medical decision-making among older patients with AIS.

Peer Review reports

Background

An aging society has become a common concern in both developed and developing countries. During the last 200 years, the average human life expectancy has doubled in most developed countries. It is predicted that nearly one-fifth of the population in the world will be aged ≥ 65 by 2030 [1,2,3], leading to an increase in the number of patients with chronic diseases [4,5,6]. Stroke is considered as one of the most important chronic diseases as it is the second largest cause of death and the third largest cause of disability worldwide [7]. In 2019, the global incidence, prevalence, and mortality of stroke was 12 million, 101 million, and 7 million, respectively [8, 9]. The disease burden caused by stroke can also lead to significant losses in productivity [10], and the negative consequences of stroke are expected to increase in further years [10]. Due to the aging population, a majority of stroke patients in China are older; thus, the burden of stroke predominantly impacts this age group [11].

Acute ischemic stroke (AIS) comprises 60–70% of the total number of stroke patients [12] and has various causes, AIS can lead to blood supply disorders in the brain tissue, which can result in ischaemic necrosis, hypoxic necrosis, and brain dysfunction. The most recommended and effective treatment approved for AIS is a thrombolytic agent with recombinant tissue plasminogen activator (rt-PA) [13]. Intravenous thrombolysis with rt-PA can effectively restore cerebral tissue blood supply, rescue corresponding neurological function, and improve prognosis in patients with AIS. However, this treatment measure must be used within the optimal treatment time window of 3–4.5 h after the onset of stroke symptoms. Rt-PA treatment outside the treatment time window has a reduced thrombolytic effect and can lead to haemorrhagic transformation, thereby causing additional damage to the brain [12]. Therefore, early medical care is crucial in patients with AIS.

However, older patients with AIS do not always follow the best treatment plan [14] and often face delays in medical decision-making. This delay is characterised as a duration of more 1 h between the onset of discomfort symptoms to the patient’s initial decision to seek medical help [15, 16]. This delay represents over 50% of the total time delay and is a stage that has not yet seen effective improvement. Understanding the risks of delayed medical decision-making in older patients with AIS and performing primary prevention are critically important. However, research on factors influencing delayed medical decision-making in older patients are still lacking. In existing studies, there is still controversy over whether the level of stroke knowledge has an impact on delayed medical decision-making in patients with AIS [17, 18]. Researchers consider health literacy as an important prerequisite for health, healthy behavior, and health-related decision-making [19], however, the predictive role of health literacy on delayed medical decision-making in older patients with AIS is not yet clear. In addition, social networks are conceptualized as the group of close social relationships one has, serving as sources of advice, help, support, and companionship [20], these networks gain particular importance in older age [21]. Social support resources carried by social network have an impact on cognitive function, happiness, health management behavior and mortality of older people [22,23,24]. Despite this, few studies have examined the impact of social network factors delayed medical decision-making in older patients with AIS. Therefore, it is urgent to verify the impact of these variables on delay in medical decision-making among older patients with AIS; this information can aid in developing more effective methods for identifying older patients at risk of delay in medical decision-making.

Logistic regression is a statistical method used to establish predictive models and is primarily used to solve binary classification problems. It is based on the concept of a linear regression model that maps the output of the linear regression to an S-shaped curve for probability estimation and classification prediction. In practical applications, logistic regression is commonly used in fields such as marketing, medical research, and risk assessment to predict the probability of an event occurring or to make classification predictions. The LightGBM is a machine learning algorithm based on a gradient-lifting tree developed by Microsoft Research Asia. It combines the advantages of extreme gradient boosting (XGBoost) and Gradient Boosting. Notably, it shows high efficiency and accuracy when processing large-scale datasets. This alleviates the limitations of decision tree algorithms and is considered superior to machine tree algorithms, thus becoming increasingly popular among scholars for developing risk prediction models. To data, the performance of logistic regression and lightGBM algorithm in predicting risk factors for delay in medical decision-making among older patients with AIS remains still unclear.

This study aimed to establish and evaluate two mathematical models for predicting the delay in medical decision-making among high-risk older patients with AIS by analyzing demographic information, disease occurrence characteristics, stroke knowledge levels, health literacy, and social networks. To identify the high-risk factors for delayed medical decision-making among older patients with AIS in the early stages and provide a theoretical basis for specifying relevant intervention measures, reducing medical delays, and promoting patient health. Simultaneously applying LightGBM to the stroke delayed medical decision-making provides new directions for future research.

Materials and methods

Participants

Data were obtained between October 2021 and June 2022 from the neurology wards of four tertiary hospitals in Qingdao, China. Convenience sampling was used to assess the physical and mental health of older patients with AIS prior to stroke onset. According to the formula, \(Z\alpha ?\pi ?1-\pi ?/\delta ?\) [25], the allowable error between sample rate and unknown population rate was 5%, \(\alpha\) was 0.05 and\(Z\alpha\)was 1.96. Considering the high delay rate of medical decision-making among older patients with AIS, the rate was taken as the highest value, which was 0.85 in precious studies [26]. A total of 309 patients were included in the study.

The inclusion criteria were patients who were (1) aged over 60 years, (2) confirmed to have AIS, and (3) agreed to participate in the study. The exclusion criteria were patiens with (1) severe organ diseases (severe myocardial infarction, renal failure, hepatitis, cancer, etc.) and mental illnesses (alzheimer’s disease, senile dementia, schizophrenia, or serious mental disorders caused by physical infections, endocrine disorders, nutritional and metabolic disorders, etc.), (2) unclear consciousness and inability to communicate, (3) could not perceived symptom time and medical decision-making time (nor could their family memvers), or (4) experienced cognitive impairment symptoms during seizures, which caosued them to be unable to make decisions themselves.

Measurements

Demographic characteristics

Demographic data, including age, sex, education level, and marital status, were obtained through questionnaires and confirmed by checking the patients’ medical records. To gather disease information, we collected information including data on participants’ symptoms, National Institutes of Health Stroke Scale (NIHSS) score, onset time, location, first decision to seek medical attention. Whether a patient presents a medical decision-making delay was determined based on whether the time interval between the onset of symptoms and the initial decision to seek medical exceeded 1 h [15, 16].

Previous stroke knowledge

The 40- question Stroke-related Knowledge Questionnaire (SKQ), created by Yao Qiping in 2016, was used to evaluate patients’ previous level of stroke knowledge [27]. The questions focused on stroke symptoms, first-aid measures, risk factors, safe medication, healthy behaviours, and knowledge of rehabilitation. Each question is assigned a score of either 0 or 1, with 1 point given for correct answers and 0 foe incorrect ones. The maximum total score is 40 and Cronbach’s α is 0.858.

Health literacy

The Health Literacy Management Scale (HeLMS) for chronic disease patients was rendered into the Chinese language by Sun Haolin in 2012 [28]. This scale consists of 24 items, and it reflects an individual’s health literacy status on four aspects: health information ability, health information assistance ability, health willingness, and economic support. The Likert 5-point scoring method is employed, with values ranging from ‘absolutely impossible’ (1 point) to ‘no difficulties’ (5 points). The total score ranges from 24 to 120 points. Higher scores indicated higher levels of health literacy. Cronbach’s α is 0.977.

Social network

The Lubben Social Network Scale (LBSNS) has been used previously to assess family, friends, and social networks among older patients with AIS [29]. It is currently the most widely used tool for assessing social isolation among older adults. The scale comprises 11 questions, each with a specific score. Lower scores indicate a higher degree of social isolation. If the total score is less than or equal to 19 points, there may be a risk of isolation. Cronbach’s α is 0.92.

Statistical analyses

Data were analyzed using IBM SPSS. The Harman single factor test was used to test for common method bias. Econometric data that followed a normal distribution were listed as mean ± standard deviation, and the t-test was used for comparative analysis. Data not adhering to a normal distribution were described using the median and interquartile range, and comparisons and analyses are conducted using the rank sum test for two independent samples in non-parametric testing. Qualitative data were listed as frequency and percentage, and either the chi-square test or Fisher’s exact test was used foe comparison and analysis. These statistical methods were employed to investigate potential differences in each risk factor between two patient groups: those making timely decisions (≤ 1 h) and those with delayed decision-making (> 1 h). Correlation analysis was conducted and VIF values were calculated to determine collinearity between variables. A P-value < 0.05 obtained using logistic regression analysis was deemed statistically significant. The LightGBM algorithm analysis was performed using Pycharm 2023.3.

Results

Sample characteristics and clinical data

A total of 309 valid questionnaires were collected. As shown in Tables 1 and 58.6% of patients were males, and 60.2% of patients were between 60 and 70 years of age. The proportion of patients in rural and urban was equivalent; 45% were retired, 33.3% had a per capita monthly income ranging from 3000 to 5000 yuan, and only 1.3% had received higher education. Overall, most people lived with their children or spouses, and approximately half of the respondents underwent physical examination once a year.

Table 1 Characteristics of the study population and clinical data
Full size table

Delayed medical decision-making among older patients with AIS

In this study, the median delay in medical decision-making in older patients with AIS was 8 h. Of the 309 patients, 74.76% presented medical decision-making delay (> 1 h), and only 78 patients presented early decision delay (≤ 1 h). Overall, 110 (35.60%) and 137 (44.34%) patients had a decision delay of 3 h or less and between 3 and 6 h, respectively, whereas 17 patients (15.5%) had a decision delay of > 24 h.

Univariate analysis of decision-making delay

The results of the normality test indicate that the previous knowledge level and social networks follow a normal distribution, while the level of health literacy follows a skewed distribution. There were statistically significant differences in the medical decision-making time of older patients with AIS in terms of age, per capita monthly income, physical examinations, first stroke, stroke severity, reaction from others, stroke recognition, previous stroke knowledge, health literacy, and social networks (P < 0.05). This is shown in Table 1.

Multicollinearity test

We conducted a correlation analysis of the variables to avoid collinearity. The correlation heat map shows the correlation between the variables and combination of statistically significant variables (Fig. 1). This indicates that a correlation exists between some of the variables; however, the degree of correlation was low (< 0.5). The variance inflation factor (VIF) was used to test variable multicollinearity. A VIF value of < 5 indicates weak multicollinearity. Table 2 shows that the VIF values of each variable were close to 1, indicating weak collinearity of the variables and, consequently, demonstrating that the selected variables could effectively prevent the negative impact of feature collinearity on the classification performance of the model.

Fig. 1
figure 1

Heatmap of the correlation coefficient of variables

Full size image
Table 2 VIF test values
Full size table

Prediction model based on logistic regressive analysis

The dependent variable was whether older patients with AIS experienced delay in medical decision-making, and the independent variables were selected for inclusion in the model based on literature review and univariate analysis results. The variables underwent forward (conditional) entry regression analysis, with a criterion for variable inclusion set at 0.05 and a criterion for exclusion set at 0.10. The result of the Hosmer-Lemeshow test indicated that the p-value was greater than 0.05, suggesting that the current data were been adequately captured, and thus the model exhibited good fit.

The results of the binary logistic regression analysis are presented in Table 3. The possibility of delayed decision-making in patients with moderate-to-severe stroke (odds ratio [OR]: 0.301; confidence interval [CI]: 0.139–0.653), sudden-onset stroke (OR: 2.648; CI: 1.034–6.779), or recognition of symptoms as stroke symptoms by patients or their families was lower (OR: 0.275; CI: 0.123–0.616) compared to other patients. Moreover, patients with higher levels of stroke knowledge (OR: 0.882; CI: 0.832–0.935) and health literacy (OR: 0.972; CI: 0.950–0.995), as well as denser social networks (OR: 0.899; CI: 0.842–0.961), were less likely to make delayed medical decisions compared to other patients.

Table 3 Logistic regressive analysis in older patients with AIS
Full size table

Prediction model based on the LightGBM algorithm

To identify key parameters influencing delayed medical decision-making in older patients with AIS, the importance scores of features in the LightGBM model were generated. Previous stroke knowledge, health literacy, social networks, stroke recognition, reactions from others, and stroke severity were the most important parameters for predicting the delay in medical decision-making. (Fig. 2).

Fig. 2
figure 2

Importance scores of features presented by LightGBM model

Full size image

Evaluation of the prediction performance of the two models

The performance of the two prediction models was evaluated using five parameters including Accuracy, Recall, F1 Score, Area Under Curve (AUC) and Precision. The results are shown in Table 4. The LightGBM model demonstrated superior performance, achieving the higher AUC of 0.909. Given that the LightGBM model exhibited the ideal performance across the two algorithms, it was deemed the best model.

Table 4 The result of comparison between two prediction models
Full size table

Discussion

Our study included 309 participants aged ≥ 60 years and the results showed that approximately three-quarters of the patients experienced delayed medical decisions-making. Logistic regression analysis and LigitGBM algorithm were used to establish predictive models and screen for stroke-related risk factors. The two predictive models jointly reported that disease severity, stroke recognition, previous stroke knowledge, health literacy, and social networks were factors influencing delayed medical decision-making in older patients with AIS. In the logistic regression model, ‘stroke onset characteristics’ was statistically significant in determining whether there was a delay in medical decision-making. In the GBM algorithm, the ‘response of others’ was of high importance. We also evaluated the performance of both models and found that GBM algorithm outperformed logistic regression.

Research reports have shown that the delay rate in medical decision-making ranges from 40.9 to 71.5%. A study conducted in Malaysia showed that the delay rate in medical decision-making for patients with AIS in hospitals was 54.9% [15]. Zhang et al. investigated the delay status of patients with AIS in rural areas and found that the delay rate for this group of patients was 71.5% [16]. However, in this study, the delay rate for medical decision-making in older patients with AIS was 74.76%, which is high. The median decision delay (480 min) in seeking treatment for older patients with AIS in this study is much higher than previous research results [15,16,17, 30, 31]. The main reason is the difference in the study population, which means that the delay in medical decision-making among older AIS patients is more severe than the median delay in decision-making among AIS patients in adult studies [15, 17, 30, 31] or in rural areas [16]. Older patients are less sensitive to physical symptoms and have higher pain thresholds, and most older patients fear hospitalization and are unwilling to inconvenience their families, leading to delays in seeking medical care [32]. Additionally, longitudinal research reports indicate that with increasing age, older adults experience declines in financial skills [33], health literacy [34], and medical decision-making abilities [35, 36], leading them to exhibit a greater tendency for decision delay compared to younger adults [37]. However, the physical function of the older gradually declines with age [38], and the damage caused by delayed medical treatment is more serious; therefore, it is necessary to manage and regulate the intervenable factors that affect medical decision-making delays to reduce treatment delays for older patients who have undergone a stroke.

Our results show that stroke severity and stroke recognition are factors influencing the delay in medical decision-making among older patients with AIS, which is consistent with the findings of Lim et al. [15]. Mild symptoms, such as mild dizziness and headaches, confuse older people with their pre-existing symptoms of other common chronic diseases, including high blood pressure. Older people may rest, take medication, and observe themselves at home before going to the hosipital, leading to delayed decision-making in seeking medical attention. In contrast, if the symptoms are severe or older patients perceive them as having a stroke, they may seek medical attention promptly [17]. As bodily sensations beyond the ‘normal’ range may trigger a self-evaluation process, unfamiliar or severe symptoms can attract the patient’s attention and lead to timely seeking of professional medical assistance [39]. Ivynian et al. [39] believed that the inability to correctly recognize stroke symptoms indicated cognitive dissonance, which may lead to emotional responses of avoidance and denial.

The level of previous stroke knowledge was an important factor affecting delayed medical decision-making in older patients with AIS, ranking first in importance in the decision tree model. However, studies have suggested no correlation between high levels of stroke knowledge and timely decision-making [18, 40, 41], possibly because they only use a single dimension to evaluate stroke knowledge levels [17]. However, in our study, patients’ knowledge was evaluated by the symptoms of stroke, first-aid measures, risk factors, safe medication, healthy behaviour, and rehabilitation knowledge. Higher knowledge scores indicated that older patients were better at recognizing stroke abilities and emergency awareness, ensuring that they could make timely medical decisions when symptoms appear. Moreover, research has found that the positive effect of publishing stroke knowledge through public health campaigns has given unexpected results. Therefore, more in-depth research is needed to formulate more sustainable, multilevel, practicable, and effective health education strategies for older patients with AIS [15]. Additionally, broad, multi-levelled, practicable, and effective health education strategies, such as geriatric health related lectures and simulation, are required instead of focusing too narrowly on stroke symptoms alone [42]. These strategies are warranted to ensure that improved knowledge can be translated into symptom recognition and correct seeking behaviour when symptoms appear [18].

This study found a negative correlation between the health literacy level of older patients with AIS and delays in medical decision-making. Health literacy is a complex concept that involves many aspects of individual skills [43]. Patients with high levels of health literacy demonstrate rich motivation, ability, cognition, and social skills to acquire, understand, and apply information to promote and maintain good health [44, 45], actively manage their own health, and making timely medical decisions. Therefore, improving the health literacy level of the older population at high-risk for stroke can be an effective strategy to reduce delays in medical decision-making for stroke patients. Community workers should use scientific tools to assess the health literacy levels of older high-risk stroke populations and develop personalized intervention measures based on patient characteristics and health literacy levels to improve individual health literacy levels and reduce delays in medical decision-making.

Previous studies have focused on the positive effects of social support on timely medical treatment [16, 46,47,48]. One Chinese study demonstrated that social support was the only feature in the prediction model that affected decision-making delays in rural stroke patients [16], which is consistent with our studies. A correlation exists between social networks and delayed medical decision-making in older patients with AIS. Gao constructed a simulation model and found that the effective scale of social networks was the strongest indicator affecting the number of prehospital delays [24]. Older patients with large, effective social networks can access diverse health information and support resources. Others in the social network can participate in health activities with patients and help the older patient identify their symptoms and take urgent action more quickly. Moreover, socialising can maintain the activity-related performance of the older brain and improve emergency response speed. Therefore, expanding the social network of the older population at high-risk for stroke can help to reduce delay in medical decision-making. Community health organisations should encourage this population to actively participate in social activities, regularly perform physical exercise, engage in activities that involve communication. This can help create a positive social support environment, and establish stable social networks for older adults.

Our study had several limitations. Firstly, cross-sectional studies cannot determine causal relationships. Longitudinal or intervention studies should be conducted to further determine the influencing factors of delayed medical decision-making in older patients with AIS. Secondly, the main variables were dependent on the description provided by patients, which may result in biases related to social expectations and reporting. Lastly, there may be potential mutual influences between variables, and further in-depth research on the underlying mechanisms of influencing factors is needed in the future.

Conclusion

In this study, logistic regression analysis and the LightGBM algorithm were applied to screen the relevant risk factors in older patients with AIS and establish prediction models. We discovered that stroke severity, stroke recognition, stroke onset characteristics, response of others, previous stroke knowledge, health literacy, and social networks were factors impacting delayed decision-making in older patients with AIS. This finding demonstrates the need for targeted prevention interventions that improve the stroke knowledge and health literacy among the older patients. It also emphasizes the need for expanding social networks to reduce the occurrence of delayed medical decision-making for older patients with AIS, with potential implications for future public health efforts. And the prediction effect of the LightGBM algorithm for Delayed medical decision-making among older patients with AIS was more accurate than that of logistic regressive analysis.

Data availability

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

Abbreviations

AIS:

Acute ischemic stroke

LightGBM:

Light Gradient Boosting Machine algorithm

rt-PA:

Recombinant tissue plasminogen activator

VIF:

Variance inflation factor

References

  1. World Health Organization. Ageing and Health. https://www.who.int/news-room/fact-sheets/detail/ageing-and-health#:~:text=At%20this%20time%20the%20share. Accessed 9 Jan 2024.

  2. Fang EF, Xie C, Schenkel JA, Wu C, Long Q, Cui H et al. A research agenda for ageing in China in the 21st century (2nd edition): Focusing on basic and translational research, long-term care, policy and social networks. Ageing Res Rev. 2020;64:101174. https://doi.org/10.1016/j.arr.2020.101174.

  3. Ho IS, McGill K, Malden S, Wilson C, Pearce C, Kaner E, et al. Examining the social networks of older adults receiving informal or formal care: a systematic review. BMC Geriatr. 2023;23(1):531. https://doi.org/10.1186/s12877-023-04190-9.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Lopreite M, Zhu Z. The effects of ageing population on health expenditure and economic growth in China: a Bayesian-VAR approach. Soc Sci Med. 2020;265:113513. https://doi.org/10.1016/j.socscimed.2020.113513.

    Article  PubMed  Google Scholar 

  5. Yang Y, Zheng R, Zhao L, Population Aging, Health Investment and Economic Growth. Based on a Cross-country Panel Data Analysis. Int J Environ Res Pub Health. 2021;18(4). https://doi.org/10.3390/ijerph18041801.

  6. Foreman KJ, Marquez N, Dolgert A, Fukutaki K, Fullman N, McGaughey M, et al. Forecasting life expectancy, years of life lost, and all-cause and cause-specific mortality for 250 causes of death: reference and alternative scenarios for 2016-40 for 195 countries and territories. Lancet. 2018;392(10159):2052–90. https://doi.org/10.1016/S0140-6736(18)31694-5.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Shuqi H, Siqin L, Xiaoyan W, Rong Y, Lihong Z. The risk factors of self-management behavior among Chinese stroke patients. Int J Clin Pract. 2023;4308517. https://doi.org/10.1155/2023/4308517.

  8. Global regional. Lancet Neurol. 2021;20(10):795–820. https://doi.org/10.1016/S1474-4422(21)00252-0. and national burden of stroke and its risk factors, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019.

  9. Gerstl JVE, Blitz SE, Qu QR, Yearley AG, Lassarén P, Lindberg R, et al. Global, Regional, and National Economic consequences of Stroke. Stroke. 2023;54(9):2380–9. https://doi.org/10.1161/STROKEAHA.123.043131.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Rajsic S, Gothe H, Borba HH, Sroczynski G, Vujicic J, Toell T, et al. Economic burden of stroke: a systematic review on post-stroke care. Eur J Health Econ. 2019;20(1):107–34. https://doi.org/10.1007/s10198-018-0984-0.

    Article  CAS  PubMed  Google Scholar 

  11. Wang XL. Study on Satus and influencing factors of Benefit Finding among Elderly Stroke patients. He Nan: Xinxiang Medical University; 2021.

    Google Scholar 

  12. Barthels D, Das H. Current advances in ischemic stroke research and therapies. Biochim Biophy Acta Mol Basis Dis. 2020;1866(4):165260. https://doi.org/10.1016/j.bbadis.2018.09.012.

    Article  CAS  Google Scholar 

  13. Powers WJ, Rabinstein AA, Ackerson T, Adeoye OM, Bambakidis NC, Becker K, Guidelines for the Early Management of Patients With Acute Ischemic Stroke, et al. Stroke. 2019;50(12):e344–418. https://doi.org/10.1161/STR.0000000000000211. 2019 Update to the 2018 Guidelines for the Early Management of Acute Ischemic Stroke: A Guideline for Healthcare Professionals From the American Heart Association/American Stroke Association.

  14. Sharrief A, Grotta JC. Stroke in the elderly. Handb Clin Neurol. 2019;167:393–418. https://doi.org/10.1016/B978-0-12-804766-8.00021-2.

    Article  PubMed  Google Scholar 

  15. Lim SH, Tan TL, Ngo PW, Lee LY, Ting SY, Tan HJ. Factors related to prehospital delay and decision delay among acute stroke patients in a district hospital, Malaysia. Med J Malaysia. 2023;78(2):241–9.

    CAS  PubMed  Google Scholar 

  16. Zhang B, Sun Q, Lv Y, Sun T, Zhao W, Yan R, Guo Y. Influencing factors for decision-making delay in seeking medical care among acute ischemic stroke patients in rural areas. Patient Educ Couns. 2023;108:107614. https://doi.org/10.1016/j.pec.2022.107614.

    Article  PubMed  Google Scholar 

  17. Potisopha W, Vuckovic KM, DeVon HA, Park CG, Phutthikhamin N, Hershberger PE. Decision Delay is a significant contributor to Prehospital Delay for stroke symptoms. West J Nurs Res. 2023;45(1):55–66. https://doi.org/10.1177/01939459221105827.

    Article  PubMed  Google Scholar 

  18. Iversen AB, Blauenfeldt RA, Johnsen SP, Sandal BF, Christensen B, Andersen G, et al. Understanding the seriousness of a stroke is essential for appropriate help-seeking and early arrival at a stroke centre: a cross-sectional study of stroke patients and their bystanders. Eur Stroke J. 2020;5(4):351–61. https://doi.org/10.1177/2396987320945834.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Rüegg R, Decision-Making Ability. A Missing Link between Health Literacy, Contextual Factors, and Health. Health Lit Res Pract. 2022;6(3):e213–23. https://doi.org/10.3928/24748307-20220718-01.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Ye L, Zhang X. Social Network Types and Health among older adults in Rural China: the mediating role of Social Support. Int J Environ Res Public Health. 2019;16(3):410. https://doi.org/10.3390/ijerph16030410.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Litwin H, Levinsky M, Schwartz E. Network type, transition patterns and well-being among older europeans. Eur J Ageing. 2019;17(2):241–50. https://doi.org/10.1007/s10433-019-00545-7.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Dodds L, Brayne C, Siette J. Associations between social networks, cognitive function, and quality of life among older adults in long-term care. BMC Geriatr. 2024;24(1):221. https://doi.org/10.1186/s12877-024-04794-9.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Fletcher CME, Woolford D, Gladigau J, Gunn KM. A ‘Vocal locals’ social network campaign is associated with increased frequency of conversations about mental health and improved engagement in wellbeing-promoting activities in an Australian farming community. BMC Public Health. 2024;24(1):673. https://doi.org/10.1186/s12889-024-18193-7.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Gao ZH. Modeling and Simulation Research on the Influence of Social Network on Prehospital Delay in patients with Acute ischemic stroke. Shan Dong: Qing Dao University; 2023.

    Google Scholar 

  25. Bolarinwa OA. Sample size estimation for health and social science researchers: the principles and considerations for different study designs. Niger Postgrad MedJ. 2020;27(2):67–75. https://doi.org/10.4103/npmj.npmj_19_20.

    Article  Google Scholar 

  26. García Ruiz R, Silva Fernández J, García Ruiz RM, Recio Bermejo M, Arias Arias Á, Santos Pinto A, et al. Factors related to immediate response to symptoms in patients with stroke or transient ischaemic attack. Neurologia. 2020;35(8):551–5. https://doi.org/10.1016/j.nrl.2017.09.013.

    Article  PubMed  Google Scholar 

  27. Yao Q. The effect of self-management intervention on knowledge, belief, behavior, and subjective well-being of stroke recovery patients. In. Nanjing: Southeast University; 2016.

    Google Scholar 

  28. Sun H. The study and Preliminary Application of the Health Literacy Scale for Chronic Disease patients. In. Shanghai: Fudan University; 2012.

    Google Scholar 

  29. Assessing social networks among elderly populations. L JE. Family Community Health. 1988;11(3):42–52.

  30. Gonzalez-Aquines A, Cordero-Pérez AC, Cristobal-Niño M, Pérez-Vázquez G, Góngora-Rivera F. Contribution of onset-to-alarm time to Prehospital Delay in patients with ischemic stroke. J Stroke Cerebrovasc. 2019;28(11):104331. https://doi.org/10.1016/j.jstrokecerebrovasdis.2019.104331.

    Article  Google Scholar 

  31. Soto-Cámara R, González-Santos J, González-Berna J, Trejo-Gabriel-Galán JM. Factors associated with a rapid call for assistance for patients with ischemic stroke. Emergencias. 2020;32(1):33–9.

    PubMed  Google Scholar 

  32. Wang X, Xu L, Yin T, Zhang R. Analysis on treatment-seeking delay status and related factors for patients with acute myocardial infarction[J]. Nurs J Chin People’s Liberation Army. 2010;27(11):801–4.

    Google Scholar 

  33. Martin RC, Gerstenecker A, Triebel KL, Falola M, McPherson T, Cutter G, et al. Declining financial capacity in mild cognitive impairment: a six-year longitudinal study. Arch Clin Neuropsychol. 2019;34(2):152–61. https://doi.org/10.1093/arclin/acy030.

    Article  PubMed  Google Scholar 

  34. Yu L, Wilson RS, Han SD, Leurgans S, Bennett DA, Boyle PA. Decline in literacy and Incident AD Dementia among Community-Dwelling older persons. J Aging Health. 2018;30(9):1389–405. https://doi.org/10.1177/0898264317716361.

    Article  PubMed  Google Scholar 

  35. Okonkwo OC, Griffith HR, Copeland JN, Belue K, Lanza S, Zamrini EY, et al. Medical decision-making capacity in mild cognitive impairment: a 3-year longitudinal study. Neurology. 2008;71(19):1474–80. https://doi.org/10.1212/01.wnl.0000334301.32358.4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Wilson RS, Yu L, Stewart CC, Bennett DA, Boyle PA. Change in decision-making analysis and preferences in Old Age. J Gerontol B Psychol Sci Soc Sci. 2023;78(10):1659–67. https://doi.org/10.1093/geronb/gbad037.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Kim YS, Park SS, Bae HJ, Cho AH, Cho YJ, Han MK, et al. Stroke awareness decreases prehospital delay after acute ischemic stroke in Korea. BMC Neurol. 2011;11:2. https://doi.org/10.1186/1471-2377-11-2.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Fan YX, Zeng HJ, Xie Y, Luo M, Zhou JF, Wu XF. Research progress on mutual support for the elderly under the background of active aging. J Nurs. 2023;38(21):122–5.

    Google Scholar 

  39. Ivynian SE, Newton PJ, DiGiacomo M. Patient preferences for heart failure education and perceptions of patient-provider communication. Scand J Caring Sci. 2020;34(4):1094–101. https://doi.org/10.1111/scs.12820.

    Article  PubMed  Google Scholar 

  40. Faiz KW, Sundseth A, Thommessen B, Rønning OM. Factors related to decision delay in acute stroke. J Stroke Cerebrovasc Dis. 2014;23(3):534–9.

    Article  PubMed  Google Scholar 

  41. Faiz KW, Sundseth A, Thommessen B, Rønning OM. Prehospital delay in acute stroke and TIA. Emerg Med J. 2013;30(8):669–74. https://doi.org/10.1136/emermed-2012-201543.

    Article  PubMed  Google Scholar 

  42. Lecouturier J, Rodgers H, Murtagh MJ, White M, Ford GA, Thomson RG. Systematic review of mass media interventions designed to improve public recognition of stroke symptoms, emergency response and early treatment. BMC Public Health. 2010;10:784. https://doi.org/10.1186/1471-2458-10-784.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Fan ZY, Yang Y, Yin RY, Tang L, Zhang F. Effect of health literacy on decision Delay in patients with Acute myocardial infarction. Front Cardiovasc Med. 2021;8:754321. https://doi.org/10.3389/fcvm.2021.754321.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Posawang P, Vatcharavongvan P. Development of health literacy assessment scale for Thai stroke patients. Top Stroke Rehabil. 2023;1–9. https://doi.org/10.1080/10749357.2023.2275091.

  45. Health Promotion Glossary. https://www.who.int/publications/i/item/WHO-HPR-HEP -98.1. Accessed 12 Jan. 2024.

  46. Kuang JK, Zhu XM, Yang L, Gao ZH, Wei X, Zhou KX, Xu MF. Factors influencing alertness to premonitory symptoms in stroke patients with pre-hospital delay. Public Health Nurs. 2022;39(6):1204–12. https://doi.org/10.1111/phn.13108.

    Article  PubMed  Google Scholar 

  47. An QP, Gao F, Liu D, Luo Y, Li N, Yang ZY. Causes of delay in seeking medical attention in patients with acute ischemic stroke: a Meta-integration of qualitative research. Evid Based Nurs. 2023;9(9):1531–7.

    Google Scholar 

  48. Yang JY, Liu KQ, Zhang LZ. Influence of health literacy and social support on the intention of prehospital delay in young and middle-aged patients with first-episode stroke. Practical Prev Med. 2023;30(5):607–10.

    Google Scholar 

Download references

Acknowledgements

We appreciate the contributions of all participants and their families, as well as the valuable work done by Qingdao University Affiliated Hospital in collecting and preparing the data.

Funding

This work was funded by the Natural Science Foundation of Shandong Province, China, ZR2021MG031.

Author information

Author notes

  1. Zhenwen Sheng and Jinke Kuang contributed equally to this work.

Authors and Affiliations

  1. Shandong Xiehe University, Jinan City, Shandong Province, China

    Zhenwen Sheng, Jinke Kuang, Guiyun Wang, Cuihong Gu, Yanxia Qi, Ruowei Wang, Yuehua Han & Jiaojiao Li

  2. Qingdao University, Qingdao City, Shandong Province, China

    Li Yang

  3. Qilu Hospital of Shandong University, Jinan City, Shandong Province, China

    Xia Wang

Authors
  1. Zhenwen Sheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jinke Kuang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Li Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Guiyun Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Cuihong Gu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Yanxia Qi
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Ruowei Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Yuehua Han
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Jiaojiao Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Xia Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

S.ZW. and K.JK. wrote the main manuscript text, Y.L. and W.GY. provided guidance on the research design and writing ideas of the article, G.CH. completed statistical analysis of the lightGBM model, W.RW., H.YH., and L.JJ. provided literature support for literature review, and W.X. provided guidance on the expression and grammar of the article. All authors reviewed the manuscript.

Corresponding author

Correspondence to Li Yang.

Ethics declarations

Ethical approval and consent to participate

Ethical approval for this cross-sectional study was obtained from the Ethics Subcommittee of the Affiliated Hospital of Qingdao University (approval number: QDU-HEC-2021180), and each participant provided written informed consent before the survey.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sheng, Z., Kuang, J., Yang, L. et al. Predictive models for delay in medical decision-making among older patients with acute ischemic stroke: a comparative study using logistic regression analysis and lightGBM algorithm. BMC Public Health 24, 1413 (2024). https://doi.org/10.1186/s12889-024-18855-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12889-024-18855-6

Keywords

BMC Public Health

ISSN: 1471-2458

Contact us