Comparing Mortality Trends in Heart Disease, Cancer, and Stroke from 2000–2017

A Walkthrough of the Data and Findings

Brian Blandino

Audience and Background

  • Target Audience
    • Public health planners, clinicians, and medical decision-makers
  • Why this Matters
    • Supports prevention focus, research prioritization, and resource allocation
  • Project Background
    • Compares mortality trends in Heart disease, Cancer, and Stroke
    • Uses data from 2000 to 2017
    • Evaluates whether mortality burden differed meaningfully across causes

Objective

  • Primary Goal
    • Compare long-term mortality patterns in Heart disease, Cancer, and Stroke from 2000 to 2017
  • Analytical Focus
    • Compare overall mortality burden
    • Examine changes in age-adjusted death rate over time
    • Test whether the three causes followed different patterns
  • Practical Purpose
    • Identify which causes warrant the greatest sustained public health attention

Data

  • Dataset
    • NCHS_final_2000_2017_with_population_enriched.csv
    • 10,296 observations and 11 variables
  • Variables Used
    • Year, Cause.Name, Deaths, Age.adjusted.Death.Rate, and Deaths_per_100k
  • Scope
    • Focused on Heart disease, Cancer, and Stroke
    • Designed to compare long-term mortality burden across the three causes

Data Preparation

  • Cause Selection
    • Restricted to Heart disease, Cancer, and Stroke
  • Cleaning Step
    • Removed the United States row to keep the analysis at the state level
  • Result
    • Produced a balanced dataset for direct comparison across the three causes

Initial Summary of the Data

  • Overall Burden
    • Heart disease showed the highest average mortality burden
    • Cancer ranked second
    • Stroke remained much lower than the other two causes
  • Average Deaths
    • Heart disease: 12,524.453
    • Cancer: 11,213.296
    • Stroke: 2,787.753

Initial Summary of the Data (Cont)

  • Interpretation
    • Heart disease carried the greatest mortality burden in this comparison
    • Cancer also remained high enough to warrant strong attention
    • Stroke was meaningfully lower across the major summary measures

Trend in AADR

Trend Interpretation

  • Main Pattern
    • Age-adjusted death rates declined from 2000 to 2017 for all three causes
  • Relative Burden
    • Heart disease remained highest across most of the period
    • Cancer also declined steadily
    • Stroke remained much lower throughout
  • Takeaway
    • All three causes improved over time, but they did not carry the same mortality burden

Average Deaths by Cause

Average Deaths Interpretation

  • Main Findings
    • Heart disease had the highest average deaths
    • Cancer ranked second
    • Stroke remained much lower than the other two causes
  • Meaning
    • Heart disease and Cancer carried the greatest mortality burden in this comparison
  • Takeaway
    • These two causes appear to warrant the strongest sustained public health attention

Assumptions and Limitations

  • Assumptions
    • Mortality data were recorded consistently across years and states
    • The selected variables were appropriate for comparing long-term mortality patterns
    • Age-adjusted death rate was a reasonable measure for comparing trends over time
  • Limitations
    • The analysis focused only on Heart disease, Cancer, and Stroke
    • The study was observational, so it identifies patterns rather than causation
    • Other factors, such as demographics, healthcare access, and policy differences, were not included
  • Interpretive Caution
    • The findings are best used for comparison and prioritization, not as a complete decision framework

Hypothesis Test

  • Test Used
    • A one-way ANOVA was used to compare the mean age-adjusted death rate across Heart disease, Cancer, and Stroke
  • Result
    • The test produced an F-statistic of 7770
    • The p-value was less than 2e-16
  • Interpretation
    • The mean age-adjusted death rate was not the same across the three causes
    • The differences were statistically meaningful rather than due to random variation alone
    • This result supported the earlier visual evidence that the causes followed different mortality patterns

Regression Model

  • Model Used
    • A linear regression model estimated age-adjusted death rate using Year and Cause.Name
  • Key Results
    • Death rates declined over time (Year = -3.10813)
    • Heart disease had a higher rate than Cancer
    • Stroke had a much lower rate than Cancer
  • Interpretation
    • Mortality rates decreased over time, but meaningful differences across causes remained

Conclusions

  • Overall Pattern
    • Heart disease, Cancer, and Stroke all declined in age-adjusted death rate from 2000 to 2017
  • Relative Burden
    • Heart disease had the highest overall mortality burden
    • Cancer also remained high across the study period
    • Stroke was substantially lower than the other two causes
  • Statistical Support
    • The ANOVA and regression results both supported meaningful differences across the three causes

Recommendations

  • Primary Recommendation
    • Public health efforts should give the greatest sustained attention to Heart disease and Cancer
  • Reasoning
    • Heart disease had the highest average deaths and age-adjusted death rate
    • Cancer also remained high across the major mortality measures
    • Stroke was important, but meaningfully lower in this comparison
  • Caution
    • These findings support prioritization, but not causal conclusions or a complete decision framework