Term
| Purpose of descriptive epidemiology |
|
Definition
Reveal patterns related to the distribution of disease that form the basis for hypothesis generation.
Identify problems to be studied by analytic methods and suggest areas which may be fruitful for investigation
Provide basis for public health service surveillance, prevention and intervention efforts. |
|
|
Term
| Disease patterns can be described according to: |
|
Definition
|
|
Term
| Disease distribution patterns according to: PERSON |
|
Definition
Individual characteristics that may determine or be associated with disease:
age, gender, ethnicity, occupation, SES, marital status, income
knowledge, attitudes, behaviors, beliefs. |
|
|
Term
| Disease Distribution patterns can be described according to: PLACE |
|
Definition
characteristics of place that could alter reported disease prevalence:
environmental exposures
medical diagnosis traditions
3rd party payer policies
neighborhood characteristics, cultural differences |
|
|
Term
| disease distribution patterns can be described according to: TIME |
|
Definition
cyclic and short term changes
secular trends: cohort effects vs. period effects |
|
|
Term
|
Definition
Change in disease occurrence over time:
cohort effect
vs.
period effect |
|
|
Term
|
Definition
change in disease occurrence over a long period of time due to changes (that are consistent in direction) in population exposure levels.
Example: trends in lung cancer mortality in the US |
|
|
Term
|
Definition
change in disease occurrence as a result of a temporary, unusual exposure.
Example: 1944-45 Dutch famine impact on birth weight |
|
|
Term
| The Descriptive Epidemiology Toolkit for summarizing and presenting data |
|
Definition
discrete vs. continuous data
interpreting histograms and distributions: measures of central tendency, measures of variability |
|
|
Term
|
Definition
binary (example: male/female)
ordinal (example: parity)
categorical (example: blood type)
Counts and proportions are used to present and summarize discrete variables |
|
|
Term
|
Definition
categorical or binary measures may start out as continuous variables
age in years --> age 0-19, age 20-59, age 60+
education in years --> less than high school, high school grad, some college, college grad, graduate degree |
|
|
Term
| presenting continuous data |
|
Definition
examples: body mass index, blood pressure
averages are used to present and summarize continuous data
requires a measure of central tendency for summarization (mean, median, mode)
helpful to present a measure of variability too |
|
|
Term
| measure of central tendency |
|
Definition
|
|
Term
|
Definition
| how tightly or loosely clustered are the data around a central point |
|
|
Term
| normal distribution or unimodal distribution |
|
Definition
|
|
Term
| skewed distribution measures of central tendency |
|
Definition
| median might be more informative; transformation allows for use of a transformed mean in stats |
|
|
Term
| bimodal distribution measures of central tendency |
|
Definition
| mode of each subgroup might be useful |
|
|
Term
|
Definition
range: lowest to highest observed value
interquartile range: 25th to 75th percentile
variance: dispersion of individual observations around the mean
standard deviation: square root of variance; used more commonly than variance and expressed in the same units as observations. |
|
|
Term
| Key Health Indicators from the IOM: health-related behaviors |
|
Definition
smoking
physical activity
excessive drinking
nutrition
obesity
condom use |
|
|
Term
| key health indicators: health outcomes |
|
Definition
| life expectancy at birth, infant mortality, life expectancy at age 65, injury-related mortality, self-reported health status, unhealthy days in last 30 days, chronic disease prevalence, serious psychological distress |
|
|
Term
| key health indicators: health systems |
|
Definition
per capita health expenditures
insurance coverage
unmet medical, dental, rx needs
preventive services
childhood immunization
preventable hospitalizations |
|
|
Term
| why these 20 health indicators? (IOM) |
|
Definition
reflect health of the nation and determinants of health (current and future)
reflect effectiveness and efficiency of US health care system
indicators are salient for audience
data available for US overall and for relevant subgroups
data are reliable and high quality
indicators are sensitive to changes in social, environmental, and policy domains |
|
|
Term
| what makes a good population health metric |
|
Definition
1. you can measure it well
2. you can move it.
3. it matters - relevance, salience |
|
|
Term
| self-reported health status as a health outcome indicator (aka self-rated health): what does it tell us? |
|
Definition
can be highly correlated with physical health as assessed by physicians
can be highly predictive of later mortality
captures subjective perceptions of health that biomarkers may not capture |
|
|
Term
|
Definition
SRH can be highly correlated with physical health as assessed by physicians
SRH can be highly predictive of later mortality
SRH captures subjective perceptions of health that biomarkers may not capture. |
|
|
Term
|
Definition
is SRH valid across cultures/ethnic groups/SES strata?
is it affected by access to care and utilization of services?
example: latinos in US report lower SRH even when physically healthier. why? hypotheses: somatization of discrimination; different weighting of past, resolved problems; language/translation issues |
|
|
Term
|
Definition
| Poor SRH may be less predictive of mortality for lower SES groups |
|
|
Term
|
Definition
| (a health-related behavior listed as a key health indicator on the IOM) |
|
|
Term
| measuring smoking/tobacco use: defining a smoker |
|
Definition
| The NHIS is consistent with the IOM Report's definition of current smoker: has smoked 100+ cigarettes in entire life and now smokes every day or some days. |
|
|
Term
|
Definition
| key health indicator, health-related behavior |
|
|
Term
| measuring overweight / obesity in adults |
|
Definition
most common measure is body mass index (bmi) = weight taking into account differences in height = weight in kg / height in meters squared
underweight <18.5
obese BMI> 30
overweight 25-30 |
|
|
Term
| Problems with BMI as a measure of healthy/unhealthy weight |
|
Definition
does not distinguish lean vs fat body mass
thresholds may vary by gender, race, activity level |
|
|
Term
| chronic disease prevalence |
|
Definition
| key health indicator - health outcome - iom report |
|
|
Term
| where do cancer prevalence and incidence statistics come from? |
|
Definition
| surveillance epidemiology and end results - cancer statistics |
|
|
Term
| where do immunization coverage statistics come from? |
|
Definition
| national immunization survey |
|
|
Term
| 5 questions to ask about health and disease burden statistics |
|
Definition
what metric is used in the statistic?
is this statistic a valid measure of population health?
how good are the input data and estimation methods?
do contextual factors have a role?
how is uncertainty addressed? |
|
|
Term
| 5 questions to ask about health and disease burden statistics: what metric is used in the statistic? |
|
Definition
| rate or a ratio? incidence or prevalence? crude or adjusted? observed or predicted? |
|
|
Term
| 5 questions to ask about health and disease burden statistics: is this statistic a valid measure of population health? |
|
Definition
| is it related to the disease or construct of interest? is it on the causal pathway from risk factors to disease? |
|
|
Term
| 5 questions to ask about health and disease burden statistics: how good are the input data and estimation methods? |
|
Definition
| are the data current? representative? are the data biased towards successful programs and easy sites? are the estimation methods transparent? credible? |
|
|
Term
| 5 questions to ask about health and disease burden statistics: do contextual factors have a role? |
|
Definition
| are exemplars or anecdotes used that are not typical? are the statistics spun to drive funding or attention? is this a worst case scenario? |
|
|
Term
| 5 questions to ask about health and disease burden statistics: how is uncertainty addressed |
|
Definition
| how are uncertainties in estimates compounded over multiple calculations? are uncertainties or bounds stated? are they reasonable? up to: watch out for reports using high estimates when median/middle might be more reasonable. |
|
|
