Economic of Data Analysis

Health

Health • Infant Mortality

– CDC – CIA Factbook – Measured as infant deaths prior to first year after birth per 1,000 live births. – Often used to evaluate various health care systems and technologies.

• Perceived by many to be highly responsive to improvements and failures in health care.

– Differences exist across states, countries, race, and time? – Data Considerations when comparing countries:

• Some countries count early births as “miscarriages” and do not include loss in infant mortality statistics, lowering their infant mortality rate.

• Countries with a higher ratio of pre-term births, like the U.S., register higher infant mortality.

• Abortion – Data: Guttmacher Foundation – Opinions: Gallup – Measuring

• Abortion Rate per 1000 Women, percent of pregnancies ending in abortion, demographic characteristics, etc..

• Influencing Factors: income, race, age, etc..

Health • Life Expectancy

– CIA FactBook

– World Health Organization

– SSA

– Often used to compare various health care systems and technologies.

– U.S has lower life expectancy than many European nations.

• U.S. has higher deaths unrelated to the health care system such as homicides, suicides, and accidents, skewing the average lower.

• Critics point out these factors are not completely unrelated to the health care system.

– U.S. Hispanics live longer (especially immigrants from Mexico)

• Ironically, Hispanics have characteristics correlated with lower life expectancy (i.e. higher poverty, less education, and health insurance).

• Explained by “return migration” and lower smoking rates.

– Blacks have lower life expectancy than Whites (≈4 years difference)

• Gap increased in 1980s due to higher homicide rates in the inner cities.

• After age 75 the gap diminishes to <1 year.

• Black youth have twice the probability of death compared to whites.

Health • Life Expectancy

– CDC

– Very Elderly data may be unreliable

• <21 and >70 years of age tend to exaggerate their age causing an upward bias.

– Measuring Longevity: Mean vs. Median statistics.

• Longevity is skewed left due to high infant mortality rates.

• Median may be a better statistic when measuring the “typical” life span.

• Leading Causes of Death – CDC

Health • Cancer

– National Cancer Institute – American Cancer Society – Type of cancer matters for rates

• Lung cancer has highest mortality and contributes to the increasing death rates seen post 1950s.

• When lung cancer is excluded, death rate falls for cancer. • “War on cancer” may be difficult to track if lung cancer is included since it is

primarily caused by smoking rather than insufficient health care advancements.

– Survival Rates • Skewed upward: the half that survive the median time-frame can expect to

live many years longer than the median would suggest.

– Incidence rate • Changes with both actual cases and rates of detection. • 1974 breast examination campaign led to higher incidence as women rushed

out to get checked. • 1982-1992 the PSA blood count test for prostate cancer caused the

incidence of prostate cancer in men to rise as more got tested.

Health • Cost/Benefit Analysis:

– Valuing human life • Life-income-lost earnings approach

– Problem: certain categories of people are valued differently (i.e. children and high income earners are valued higher)

• Willingness to pay approach – Greater salary for higher risk job. – Problem: large range from $300K-$8 million.

• Cost effectiveness approach – No value placed on life. – Analyzes the cost per life saved of a particular policy.

• Traffic Fatalities – Insurance Institute for Highway Safety (IIHS) – Overall rates have declined but important trends are masked.

• #1 cause of death among teenagers (ages 16-19). • 3x the fatal crash rate of adults. • Alcohol is present in about one-third of all traffic fatalities. • Motorcycle deaths are about 15% of vehicular fatalities but constitute 28x the death rate

per mile driven in automobiles. • Increased motorcycle deaths due to greater middle-aged/inexperienced riders. • >90% of all motorcycle fatalities are males.