Spatio-temporal Analysis of Ischemic Heart Disease Mortality in US Counties (1999-2021) Using a Scalable Bayesian Approach

Kernkonzepte

While the overall age-adjusted mortality rate due to ischemic heart disease (IHD) in the US has declined, this trend has slowed down, and disparities exist at the county level, with some counties even experiencing increases, highlighting the need for targeted interventions.

Zusammenfassung

• Bibliographic Information: Urdangarin, A., Goicoa, T., Congdon, P., & Ugarte, M.D. (2024). A fast approach for analyzing spatio-temporal patterns in ischemic heart disease mortality across US counties (1999-2021). arXiv preprint arXiv:2411.14849v1.
• Research Objective: To analyze the spatio-temporal patterns of IHD mortality in US counties from 1999 to 2021, identify disparities in trends, and investigate potential change points in the decline of IHD mortality.
• Methodology: The study utilizes a scalable Bayesian spatio-temporal modeling approach with a "divide and conquer" strategy to analyze IHD mortality data from CDC WONDER. Missing data (counts less than 10) were imputed using spatial models. The selected model incorporated an ICAR spatial prior, a RW1 temporal prior, and a Type II spatio-temporal interaction.
• Key Findings:
  • The overall IHD mortality risk in the US declined until 2014, followed by a flattening of the trend and a potential increase after 2019.
  • Significant disparities exist in IHD mortality trends at the county level, with some counties mirroring the national trend while others show increasing trends or different temporal patterns.
  • Rural areas in the West, Midwest, and South regions generally exhibit higher IHD mortality risks compared to urban areas.
  • The "divide and conquer" approach allows for adaptive smoothing and captures local heterogeneities in spatial patterns.
• Main Conclusions:
  • While the overall IHD mortality risk in the US has decreased, the slowdown in decline and disparities at the county level are concerning.
  • The study highlights the importance of examining IHD mortality trends at a granular level to identify areas requiring targeted interventions.
  • The scalable Bayesian approach provides an efficient method for analyzing large spatio-temporal datasets and revealing local variations in trends.
• Significance: This research contributes to a deeper understanding of the evolving patterns of IHD mortality in the US, emphasizing the need for public health interventions tailored to specific geographic areas and demographic groups.
• Limitations and Future Research:
  • The study relies on imputed data for censored counts, which may introduce some bias.
  • Further investigation is needed to confirm the increase in IHD mortality risk after 2019 and its potential link to the COVID-19 pandemic.
  • Future research could explore the relationship between IHD mortality trends and potential risk factors, such as obesity and diabetes, at the county level.

Statistiken

12.5% of total observations are missing due to confidentiality issues.
3.7% of counties exhibit missing counts for each year in the dataset.
Approximately 6.3% of counties have between 15 and 22 years with missing counts.
Around 15.2% of counties have less than 15 years with missing data.
Large and medium metros constitute 1.3% and 29.9% of the counties, respectively.
Rural metros account for 68.8% of the counties.

Zitate

Wichtige Erkenntnisse aus

A fast approach for analyzing spatio-temporal patterns in ischemic heart disease mortality across US counties (1999-2021)

by A. Urdangari... um arxiv.org 11-25-2024

https://arxiv.org/pdf/2411.14849.pdf

Tiefere Fragen

How might social determinants of health, such as access to healthcare and socioeconomic factors, contribute to the observed disparities in IHD mortality across US counties?

Social determinants of health (SDOH) play a significant role in the observed disparities in IHD mortality across US counties. These factors, including access to healthcare, socioeconomic status, education, and neighborhood environment, can profoundly impact health outcomes. Here's how:

Access to Healthcare: Limited access to quality healthcare services in rural areas or among socioeconomically disadvantaged populations can lead to delayed diagnosis, inadequate management of risk factors (hypertension, diabetes, high cholesterol), and reduced adherence to treatment, ultimately increasing IHD mortality risk.
Socioeconomic Factors: Lower socioeconomic status is often associated with increased exposure to stressors like job insecurity, financial strain, and food insecurity. These stressors can trigger unhealthy coping mechanisms (smoking, poor diet, lack of physical activity) and contribute to the development of IHD risk factors.
Education: Lower levels of education can limit health literacy, hindering individuals' ability to make informed decisions about their health. This can lead to poor health behaviors and reduced engagement with preventive healthcare measures.
Neighborhood Environment: Living in neighborhoods with limited access to healthy food options, safe spaces for physical activity, and high exposure to environmental hazards (pollution, crime) can contribute to IHD risk. These environments often foster unhealthy behaviors and limit opportunities for positive health outcomes.
The study's findings, particularly the higher IHD mortality risks observed in rural areas compared to urban areas in certain regions, highlight the influence of SDOH. Rural communities often face greater challenges in accessing healthcare services, experience higher poverty rates, and have limited access to resources that promote healthy living. These disparities underscore the need for public health interventions that address the root causes of health inequities and improve access to resources that support cardiovascular health across all communities.

Could the observed slowdown and potential increase in IHD mortality be attributed to factors other than changes in individual-level risk factors, such as shifts in disease classification or coding practices?

Yes, the observed slowdown and potential increase in IHD mortality could be influenced by factors beyond changes in individual-level risk factors. Shifts in disease classification or coding practices can impact mortality trends and should be considered when interpreting the results:

Changes in ICD Coding: The International Classification of Diseases (ICD) coding system, used to classify death certificates, undergoes periodic revisions. While these revisions aim to improve accuracy and reflect medical advancements, they can lead to changes in how deaths are categorized. A shift in coding practices, even if subtle, could result in more deaths being classified as IHD, potentially contributing to the observed trends.
Diagnostic Practices: Advancements in diagnostic technologies and increased awareness of certain conditions can lead to changes in diagnostic practices. For instance, improved detection of heart failure, a common comorbidity with IHD, could lead to more individuals being diagnosed with both conditions, potentially influencing mortality statistics.
Misclassifications: While efforts are made to ensure accurate coding, misclassification of deaths on death certificates can occur. This is particularly relevant for conditions with overlapping symptoms, such as heart disease and respiratory illnesses. The COVID-19 pandemic further highlighted the potential for misclassification, as some deaths attributed to COVID-19 might have been primarily caused by underlying cardiovascular conditions.
It's crucial to acknowledge that attributing mortality trends solely to changes in individual-level risk factors might not provide a complete picture. Investigating the potential impact of shifts in disease classification, coding practices, and diagnostic criteria is essential to ensure accurate interpretation of IHD mortality trends and inform public health interventions effectively.

What are the ethical implications of using imputed data in public health research, and how can these concerns be addressed while maintaining data privacy and confidentiality?

Using imputed data in public health research presents ethical considerations, particularly regarding data privacy and the potential for biased results if imputation methods are flawed. Here's a breakdown of the concerns and how to address them:
Ethical Implications:

Data Privacy: Even though imputed data replaces actual values with statistically derived estimates, it's crucial to ensure that the imputation process doesn't inadvertently reveal sensitive information about individuals whose data was suppressed.
Transparency and Trust: Researchers must be transparent about the use of imputed data and the methods employed. Failing to disclose this information can erode public trust in research findings.
Potential for Bias: If imputation models are not carefully constructed and validated, they can introduce bias into the analysis, leading to inaccurate conclusions and potentially misinforming public health policies.
Addressing Concerns:

Robust Imputation Methods: Employing statistically sound and validated imputation techniques that preserve data confidentiality is paramount. The study's use of spatial models for imputation, a method that borrows strength from neighboring areas, is a step in the right direction.
Sensitivity Analyses: Conducting sensitivity analyses to assess the robustness of findings to different imputation methods and assumptions helps gauge the potential impact of imputation on the results.
Data Governance and Access: Adhering to strict data governance protocols and ensuring that researchers only access the minimum data necessary for analysis helps minimize privacy risks.
Ethical Review: Submitting research proposals involving imputed data to ethical review boards ensures that privacy concerns are adequately addressed and that the benefits of using imputed data outweigh potential risks.
Balancing the need for complete data to conduct meaningful public health research with the ethical imperative to protect individual privacy is crucial. By adopting robust imputation methods, being transparent about data handling procedures, and engaging in rigorous ethical review processes, researchers can utilize imputed data responsibly while upholding data privacy and confidentiality.