Scheduled maintenance: Thursday, December 8 from 5PM to 6PM PST
Only $35.99/year

Terms in this set (194)

was an English physician and anesthesiologist.
Called "The Father of Epidemiology"
He investigated a cholera outbreak that occurred during the mid-19th century in Broad Street, Golden Square, London.
natural experiments

Linked the cholera epidemic to contaminated water supplies
Groundbreaking because he used many features of epidemiologic inquiry:
He used powers of observation
He used a spot map (Cluster map) of cases
He tabulated (counted) the number of fatal attacks and deaths

At this time (1849), people didn't know what caused cholera.....they didn't know it was caused by water contaminated by sewage!

So John Snow's natural experiment was groundbreaking—because he used the spot map (cluster map) and counted the number of deaths and those sick—these are epidemiologic methods!

Through his observations and inference, he was the first to hypothesize that water contaminated with sewage was the cause of cholera.

In 1849, ALL residents in this area of London received contaminated water from 2 water companies, the Lambeth Company and the Southwark and Vauxhall Company, which both got their water from the Thames River at a point heavily contaminated with sewage.

Snow's Nat Experiment:
Two different water companies supplied water from the Thames River to houses in the same area.
The Lambeth Company moved its source of water to a less polluted portion of the river.
Snow noted that during the next cholera outbreak those served by the Lambeth Company had fewer cases of cholera.

Then, after the cholera outbreak in 1849, one of the companies, the Lambeth company, relocated their water plant to a less contaminated part of the Thames river.

When another cholera epidemic happened in 1854, two-thirds of London's resident population south of the Thames River was still being served by both water companies—their water lines were laid out in an interpenetrating manner, so that houses on the same street were receiving their water from different sources.

This time, Snow was able to demonstrate that a disproportionate number of residents who contracted cholera in the 1854 outbreak used water from the company that was still using the highly contaminated water, in comparison with the other company, which was using relatively unpolluted water!
Infectious diseases
SARS, pandemic influenza 2009 H1N1, use of epidemiologic methods to attempt to eradicate, when possible, polio, measles, smallpox, and other communicable diseases. As mentioned earlier, the ONLY one that has been ERADICATED has been smallpox. The rest may have been somewhat eliminated in certain areas of the world, but they are still around. Outbreaks of infectious diseases, such as H5N1, Ebola, etc. will also be investigated.
Environmental health: human health may be affected by toxic chemicals used in industry, air pollutants, contaminated drinking water, unsafe homes, etc. Both occupational and environmental epidemiology address this.
Chronic Disease, lifestyle, and health promotion: Epidemiology in this area has looked at connections between Lifestyle choices and poor health—exercise, diet, smoking, ETOH consumption—can play a part in obesity, coronary heart disease, arthritis, diabetes, and cancer. Epidemiology has also looked at connections between the environment and exercise—does the environment dissuade exercise? How much do cultural practices affect behaviors linked to health and disease?
Psychological and Social Factors: Stress, social support, socioeconomic status affect the occurrence and outcomes of mental and physical health. What is the relationship between psychological and illnesses such as arthritis, some GI conditions, and essential hypertension?! What about personality factors and disease? Are some personalities more prone to heart disease? Psychosocial determinants: social, cultural, demographic factors (socioeconomic status, gender, employment, marital status, and race) are demonstrated correlates of mental and physical health status. This type of epidemiology looks at all of this.
Molecular and genetic epidemiology: use of DNA typing (genetic and molecular markers) to determine behavioral outcomes and host susceptibility to disease. Genetic epidemiology looks at things such as inherited susceptibility to breast and ovarian cancer as well as to alcohol use disorders!
Chronic diseases have replaced acute infectious diseases as the major causes of morbidity and mortality.
In 2009, the leading causes of U.S. deaths were heart disease, cancer, and chronic lower respiratory disease.
Increases were reported for Alzheimer's disease, kidney disease, and hypertension.

Dramatic changes have occurred:
Chronic conditions have replaced acute infectious diseases as the major causes of morbidity & mortality
The next slide will compare mortality causes in 1900 with those in 2009—leading causes of deaths in the U.S. in 2009 were heart disease, cancer, and chronic lower respiratory disease
Recently, death rates for Alzheimer's disease, kidney disease, and hypertension have increased while death rates for heart disease, cancer, and stroke have been declining since the 1960's.Disappearing: Conditions that were once common but are no longer present in epidemic form.
Examples include smallpox, poliomyelitis, and measles.
Brought under control by immunizations, improvement in sanitary conditions, and the use of antibiotics and other medications; led to eradication of smallpox and the elimination of the other 2 diseases in certain areas of the world.
Residual: Contributing factors are largely known, but methods of control have not been implemented effectively
Examples:
STDs
Perinatal and infant mortality among low SES persons
Problems associated with alcohol and tobacco use
Persisting: No effective method of prevention or cure has been discovered so they remain common.
Examples: certain types of cancer and mental disorders
New epidemic: diseases that are increasing markedly in frequency in comparison with previous time periods.
Examples: Lung cancer, AIDS, Obesity, Type 2 diabetes
The emergence of new epidemics of diseases may be a result of increased life expectancy of the population, new environmental exposures, or changes in lifestyle, diet, and other practices.
Denotes changes in the demographic structure of populations associated with such factors as births and deaths and immigration and emigration
Two types of populations
Fixed populations
Dynamic populations

Fixed population: one distinguished by a specific happening and consequently adds no new members; therefore, the population decreases in size as a result of deaths only. No change in this population except through death of members!

Dynamic population: This population CHANGES—with the addition of new members through immigration and births or the loss of members through emigration and deaths.

Three major factors affect the sizes of population births, deaths, and migration:
Population in equilibrium or a steady state
Population increasing in size
Population decreasing in size

Three major factors affect the sizes of population births, deaths, and migration:
Population in equilibrium or a steady state
The 3 factors do not contribute to net increases or decreases in the number
of persons—meaning that the number of members exiting for various
reasons equals the number entering.
Population increasing in size
The number of persons immigrating plus the number of births exceeds the
number of persons emigrating plus the number of deaths.
Population decreasing in size
The number of persons emigrating plus the number of deaths exceeds the
number of persons immigrating plus the number of births.

Demographic Transition
Shift from high birth and death rates found in agrarian societies to lower birth and death rates found in developed countries.
Decline in death rate: attributed to improvement in general hygienic & social conditions
Decline in birth rate: attributed to industrialization and urbanization
Industrialization and urbanization, while they have led to a decline in the birth rate, both have led to environmental contamination, concentration of social and health problems in the urban core areas of the US, and out-migration of inner city residents to the suburbs.

Demographic and Social Variables:
Age and sex distribution: How old is the population? Young? Old? If old, you'll see health problems r/t aging—chronic diseases, heart attacks, CVAs. Also, in an older community, women tend to live longer than men, so you'll see more of women's issues: osteoporosis, screenings needed for various things.
With younger people, the focus will be on immunizations for children and teens, education on STDs, etc. for Teens as well as health promotional education on smoking, substance use, and prevention of unintentional deaths/injuries.
Socioeconomic status (SES): income level, educational level, types of occupation—insurance? Health care? Educational level influences diet and exercise—people with lower educational levels may be less aware of importance.
Family structure: Marital status, single moms,/dads, number of children in single parent homes, number of single parent families.
Racial, ethnic, and religious composition: Some health outcomes are more common in one race or ethic group than another. Example: Sickle cell anemia, diabetes, Tay-Sachs.
Some religions restrict dietary items, may abstain from ETOH, may not smoke, may eat foods low in fat and avoid high fat foods!

They should be REALLY healthy, right?!!
1964 Surgeon General's Report
1. Strength of association: Rates of morbidity and mortality must be higher for the exposed group than for the non-exposed group (risk of heart disease
is higher in smokers than in non-smokers)
2. Time Sequence: Demonstration of correct temporal sequence: Exposure to the causal factor must occur before the effect, or the disease.
3. Consistency with other studies: Varying types of studies in other populations must observe similar associations.
4. Specificity of the association: The exposure variable must be necessary and sufficient to cause disease; there is only one causal factor. **This one is not as important today since diseases have multifactorial origins.
5. Coherence of explanation: the association must not seriously conflict with what is already known about the natural history and biology of the disease.

Sir Austin Bradford Hill
1. Biological gradient (Dose-response relationship): An increased exposure to the risk
factor causes a concomitant increase in disease (risk for heart disease is higher in
heavy smokers compared to light smokers)
2. Biological plausibility: The data must make biological sense and represent a coherent
explanation for the relationship. It must be credible on the basis of existing biomedical
knowledge—but what if the biologic knowledge of the day is wrong, not up-to-date,
etc.?
3. Experimental evidence: experimental designs provide the strongest epidemiologic
evidence for causal associations, but they are not feasible or ethical to conduct for
many risk factors—disease association. However, "natural experiments" that happen
(remember Snow) may shed light on the subject! Think about the observation that
dental caries decreased in populations with fluoride added to the water supply!
4. Analogy: Thalidomide and Rubella in pregnancy—caused great problems with the
baby during pregnancy (birth defects [both], stillbirths, miscarriages), so considering
this, "we would surely be ready to accept slighter but similar evidence with another
drug or viral infection during pregnancy."

Absolute causality is rarely established.... Some non-smokers get heart disease but most all smokers have heart disease....Smoking doesn't give you heart disease but puts you at greater risk for it.
***Keep in mind that there are multiple causes of disease in most cases—this is referred to as multiple causality.
Case control studies:
A type of design that compares persons who have a disease (cases) with those who are free from the disease (controls).
This design explores whether differences between cases and controls result from exposures to risk factors.
the study begins after the health outcome has already occurred.
People are selected from a population because they are known to have the outcome of interest (cases) or they are known NOT to have the outcome of interest (controls).
Do the people with the outcome of interest (cases) have the exposure characteristic (or history of exposure) more frequently that those without the outcome (controls)?

Cohort Design
A group of people free from a disease is assembled according to a variety of exposures.
The group (cohort) is followed over a period of time for development of disease.

A cohort is a group of people who share a common characteristic or experience within a defined period and are followed over time to observe some health outcome (e.g., are born, leave school, lose their virginity, are exposed to a drug or a vaccine)
Example: the reports of a relationship with autistic children and the MMR immunization
Parents noted there were behavioral problems after their child was immunized with the second MMR. They stated that 8 out 12 children receiving the MMR vaccination displayed early signs of autism which sparked a detailed analytical study by the government to disprove the theory.

Eventually, it was a combination of a case-control and cohort study that proved the MMR does not cause autism!
Prepathogenesis—before agent reacts with host.
Susceptibility Stage: disease is not present and individuals have not been exposed; Primary Prevention stage

Pathogenesis—after agent reacts with host
Subclinical Stage—NO symptoms—individuals are exposed but have not started showing symptoms; this is where incubation period takes place. The organism multiples and grows to sufficient numbers to produce a host reaction and clinical symptoms. Secondary Prevention Stage
Clinical Stage- The beginning of symptoms—early stages of clinical stage may show signs in lab tests or other exams. Like TB lesions on a chest x-ray or cervical diseases in a PAP smear. Late stages acute symptoms are clearly visible like with food poisoning.
Later stages include development of active signs and symptoms. Clinical end points are: recovery, disability, or death. Resolution Stage- Symptoms may vary from mild to severe, residual or chronic forms of the disease that ends either with disabling limitations or death. The condition or disease has caused enough bodily harm/changes that recognizable S/S occur.
Also known as the advanced disease stage because it may have completed its course, or it may even conclude with a return to health. Tertiary Prevention Stage
*The Goal of epidemiology is to identify and understand the causes and mechanism of disease, disability, and injury so that effective interventions can be created & implemented to prevent these things before they occur or before they get any worse.
**Recognizing these different stages of the evolution of disease, one can determine the most effective application of the Levels of Prevention—a framework commonly used in public health.
A ratio that consists of a numerator and a denominator but is different from a proportion because the denominator involves a measure of time.
Contains the following elements:
disease frequency
unit size of population
time period during which an event occurs

When you think rate, think heart rate, pulse rate, respiratory rate.....what is this telling you? The number of beats or breaths per a unit of time!

The numerator consists of the frequency of a disease over a specified period of time, and the denominator is a unit size of population.

CRITICAL: to calculate a rate, remember that TWO periods of time are involved—the beginning of the period and the end of the period.

Rates improve the ability to make comparisons, but they also have limitations: rates of mortality and morbidity for a specific disease reduce that standard of comparison to a common denominator, the unit size of population.
So if the crude death rates for diseases of the heart is 188.9 per 100,000 in Texas and 288.0 per 100,000 in New York, it looks like the death rate is higher in New York but there is no way to tell if there are important differences in population composition (such as age differences between populations) that would affect mortality experience.

Crude death rate = [Number of deaths in a given year/
Reference population
(during the midpoint of the year] x 100,000

Example:
Number of deaths in the United States during 2007 = 2,423,712
Population of the U.S. as of July 1, 2007 = 301,621,157

2,423,712
301,621,157

Rates can be expressed in any form that is convenient—per 1,000, per 100,000, or per 1,000,000.

Many of the most commonly used rates are expressed in particular conventions: cancer rates are typically expressed per 100,000 population, infant mortality is expressed per 1,000 live births.

This is done to prevent the answer from being a very small number—for example, rather than saying 0.04 per 1,000, say 4.2 per 100,000
Definition: The number of new cases of a disease that occur in a group during a certain time period.
Example of Incidence measured as frequency:
Number of new cases of HIV infection diagnosed in a population in a given year: a total of 164 HIV diagnoses were reported among American Indians or Alaska natives in the U.S. during 2009.

The number of new health-related events in a defined population within a specified time period.
It can be measured as a frequency count, a rate, or a proportion.
It is a measure of the risk of a specified health-related event.

Incidence rate: Describes the rate of development of a disease in a group over a certain time period.
Contains three elements:
Numerator = the number of new cases.
Denominator = the population at risk.
Time = the period during which the cases occur.
Here it is used as a rate! It describes the rate of development of a disease in a group over a certain time period—this time period is included in the denominator.
The number of new cases: uses the frequency of new cases in the numerator—so individuals who have a history of the disease are not included (they are existing cases—prevalence!)
The population at risk: people who have already developed the disease or who are not capable of developing it should be excluded. For example, if looking at the risk for ovarian cancer, you would not include women who have already developed the disease and women who have had their ovaries removed—they can't develop it.
Specification of a time period: a date of onset for the condition or disease during the time period—this is easily done with acute illnesses (stroke, MI) but other illnesses, such as cancer, may be defined by the date of diagnosis.

Here it is used as a rate! It describes the rate of development of a disease in a group over a certain time period—this time period is included in the denominator.
The number of new cases: uses the frequency of new cases in the numerator—so individuals who have a history of the disease are not included (they are existing cases—prevalence!)
The population at risk: people who have already developed the disease or who are not capable of developing it should be excluded. For example, if looking at the risk for ovarian cancer, you would not include women who have already developed the disease and women who have had their ovaries removed—they can't develop it.
Specification of a time period: a date of onset for the condition or disease during the time period—this is easily done with acute illnesses (stroke, MI) but other illnesses, such as cancer, may be defined by the date of diagnosis.

Here it is used as a rate! It describes the rate of development of a disease in a group over a certain time period—this time period is included in the denominator.
The number of new cases: uses the frequency of new cases in the numerator—so individuals who have a history of the disease are not included (they are existing cases—prevalence!)
The population at risk: people who have already developed the disease or who are not capable of developing it should be excluded. For example, if looking at the risk for ovarian cancer, you would not include women who have already developed the disease and women who have had their ovaries removed—they can't develop it.
Specification of a time period: a date of onset for the condition or disease during the time period—this is easily done with acute illnesses (stroke, MI) but other illnesses, such as cancer, may be defined by the date of diagnosis.

Incidence rate= [# of new cases over a period of time/ total population at risk during the same time period] x multi[lier (e.g. 100,000)

Number of new cases= 1085
pop at risk= 37,105
Incidence rate= 1085/37105= = 0.02924/8 (8 year total)—divide this number
by 8 for an annual total of 0.003655; multiply by 100,000 = = 365.5 cases per 100,000 women per year
follow descriptive studies, and are used to identify the cause of the health problem
Analytic studies follow the descriptive studies—they explore the determinants of disease—variables such as infectious agents, environmental exposures, and risky behaviors. They ask How and Why! How is the disease/condition being transmitted or spread and Why are some people getting sick and others not? They take the hypothesis/hypotheses developed by the descriptive studies and use them to find the cause of the health problem.

Person
Place
Time
Secular trends
Cyclical patterns
Event-related clusters

Descriptive epidemiology describes the distribution of disease, death, and other health outcomes in the population according to person, place, and time, providing a picture of how things are or have been—the who, where, and when of disease patterns.
The variables of person, place, and time directly or indirectly relate to the occurrence of illnesses by affecting a wide range of exposures associated with lifestyle, behavioral patterns, healthcare access, and exposure to environmental hazards....just to name a few.
In this chapter, we will identify descriptive characteristics (age, sex, race, etc—all Person characteristics) that help to delineate patterns of disease and generate hypotheses regarding their underlying causes.
We will also look at Place and Time characteristics.
To get your thinking going in the right direction, let's look at a few characteristics of Person, Place, and Time!
Personal characteristics of interest in epidemiology include race, ethnicity, sex, age, education, occupation, income (and related socioeconomic status), and marital status. The most important predictor of overall mortality is age. The combination of variables, such as age and sex ,is noteworthy, too.
When we think of the distribution of a disease, geographical patterns come to mind: does the rate of disease differ from place to place (e.g., with local environment)?
In relation to time, epidemiologists ask these questions: Is there an increase or decrease in the frequency of the disease over time? Are other temporal (and spatial) patterns evident? Temporal patterns of interest to epidemiologists include secular trends, point epidemic, cyclical patterns, and event-related clusters.

3 Broad Objectives of Descriptive Epidemiology:
To evaluate and compare trends in health and disease: this includes monitoring known diseases as well as identifying emerging problems
To provide a basis for planning, provision, and evaluation of health services: data needed for efficient allocation of resources often come from descriptive epidemiologic studies
To identify problems for analytic studies (creation of hypotheses) and suggest areas that may be fruitful for investigation.

Hypotheses—remember high school science?!!

3 types:
Positive declaration (research hypothesis): The infant mortality rate is higher in one area than another.
Negative declaration (null hypothesis): there is no difference between the infant mortality rates of two regions
Implicit question: To study the association between infant mortality and geographic region of residence.

Hypotheses should be made as explicit as possible and not left as implicit!
Case reports: counts—helpful for describing diseases. Astute observations of unusual cases could spur more investigation to determine whether larger numbers of these types of cases exist. Case reports are a starting point for exploring the causes of diseases and for introducing preventive interventions. Example: a single case of methylene chloride poisoning.
Case series: One case report may not be enough to draw firm conclusions, so an observation may need to be expanded to a series of cases. Typical features are generated from a set of observations. An example from CDC is one that involved 5 cases of Hantavirus pulmonary syndrome in 5 pediatric patients. These 5 cases happened from May 2009 until November 2009.
Cross-sectional studies: a cross-section of a population or group; information is collected on current health status, personal characteristics, and potential risk factors/exposures—and it is collected all at once. It is a FAST study! This study is efficient at identifying association, but may have trouble deciding cause and effect. With data at only one time point, you don't know whether the chicken or the egg came first, so these are often conducted at different points in time. This way, they are effective in detecting trends in prevalence of disease or risk factors.
Example: National Health Interview Survey

Persons:
Age- One of the most important factors to consider when describing the occurrence of any disease or illness
Age-specific disease rates usually show greater variation than rates defined by almost any other personal attribute.
This is why age-specific rates are used when comparing the disease burden among populations.
Childhood to early adolescence
Leading cause of death, ages 1-14 years—unintentional injuries
Infants—mortality from developmental problems, e.g., congenital birth defects
Childhood—occurrence of infectious diseases & communicable diseases
Meningococcal disease, otitis media, measles, mumps, and chicken pox.
Teenage years (15-24 years)
Leading causes of death—unintentional injuries, homicide, and suicide
Other issues—unplanned pregnancy, tobacco use, substance abuse, and binge drinking
Alcohol, marijuana, tobacco are the drugs of choice among 12-17 y.o.—also Rx. drugs.
Lifestyle choices: obesity d/t sedentary hobbies (video games, computers); sugary & high-fat foods too available.

Adults: also homicide and HIV disease (even though death rates have declined by 50% since the 1990's). Lung, brain, and colon cancers were the most common causes of cancer deaths among men; leading among women were breast, lung, and cervical cancers.
Older adults: Age-specific rates of cancer incidence increase with age
deaths from chronic diseases (e.g., cancer & heart disease) dominate after age 45. leading causes of death
Unintentional injuries
Cancer
Heart disease
Suicide (3X higher in men than women)


Elderly: (65+) 5 leading causes of death in 2007 were heart, malignant neoplasms, Cerebrovascular disease, chronic lower respiratory disease, and Alzheimer's. Some of the elderly have problems caring for themselves and may not be able to live on their own.
deaths from chronic diseases and limitations in activities of daily living

A few reasons for age affects on mortality are:
Validity of diagnoses across the life span: exact cause of death of elderly people may be hard to determine as many have multiple co-morbidities
Latency effects: long latency period between environmental exposures and the development of certain diseases—example: the passage of many years between the initial exposure to a potential carcinogen and the subsequent appearance of cancer later in life.
Action of the "human biologic clock": as we age, we get more vulnerable to disease—the immune system may wane, producing increased tissue susceptibility to disease.
Life cycle and behavioral phenomena: causes of death can be influenced by factors such as personal behavior and risk-taking; lifestyle influences the occurrence of diabetes and other chronic diseases.
Age-specific rates of cancer incidence increase with age with apparent declines late in life—this is deceptive, though, because the 80-84 age group is few in number when compared to other age groups, hence the numerators and denominators for the very elderly categories are smaller. The data is not reliable due to this.



Sex
Marital Status
Race and ethnicity
Nativity and migration
Religion
Socioeconomic status
Men- All-cause age-specific mortality rates is higher for men than for women.
May be due to social factors
May have biological basis
Men often develop severe forms of chronic disease.
Hearing impairment, smoking-associated conditions, & cardiovascular disease are more common among men.
Men affected by the same chronic diseases as women (lung cancer, cardiovascular disease, and diabetes) are more likely to develop severe forms of these conditions and die from them.

female paradox- Reports from the 1970s indicated female age-standardized morbidity rates for many acute and chronic conditions were higher than rates for males, even though mortality was higher among males.
Higher female rates for:
Pain
Asthma
Some lung difficulties
Depression
This phenomenon is known as the Female Paradox—females have higher morbidity rates than males for acute and chronic conditions even though the mortality rate was higher among males.
Cancer- Cancer of the lung and bronchus is leading cause of cancer death for both men and women in the U.S.
Increases among women are related to changes in lifestyle and risk behavior, e.g., smoking. With women adopting more and more of the same behaviors and habits as men (smoking, etc.), diseases such as lung cancer have increased among women much faster than men. Males have a greater frequency of smoking, a greater prevalence of the coronary-prone behavior patterns, higher suicide and motor vehicle accident rates, as well as risky behavioral patterns that are expected of and condoned among men.

While women don't have as high of a mortality rate for lung & bronchus cancer YET, women's roles in society have changed and many have been smoking more and more since the late 1960's.....Consequently, when women started smoking more, low and behold, lung cancer increased much faster in women than in men between the years of 1975 and 1990—"you've come a long way baby"!! This supports the view that certain behavioral and lifestyle variables (smoking behavior, etc.) may relate to male/female lung cancer morality differences.
CHD is the leading cause of mortality among women and men, but sex differences in mortality rates of CHD exist between women and men even when both have high-risk factor status for serum cholesterol, blood pressure, and smoking.

Women tend to be protected from CHD in the pre-menopause stage but cardiac disease increases in the post-menopausal phase (after age 60).

Minority Women in Economically Disadvantaged U.S. Areas


In Los Angeles County, some have higher rates of diabetes and hypertension than men.
A large percentage are physically inactive.
High rates of obesity among Latinas and African Americans.
Some women may be unaware of the fact that they may be at high risk of cardiac disease—consequently, they may not be alert for symptoms of CHD—this can cause a delay in seeking treatment. Many women resist the lifestyle changes (increased activity level and consumption of low-fat food).

Minority women who live in Los Angeles County have higher rates of diabetes, HTN, and elevated cholesterol. In 2005, about ½ of these women reported little physical activity and the frequency of obesity was high, especially among Latinas (25% or 1 out of 4 women) & African Americans (33% or 1 out of 3 women).

Marital Status- In general, married people, especially men, tend to have lower rates of morbidity and mortality and tend to be healthier than adults in other marital status categories—REGARDLESS of population subgroup (age, sex, race, education, income, or nativity) or health indicator (fair or poor health, limitations in activities, low back pain, headaches, serious psychological distress, smoking, or leisure-time physical inactivity). This is true for chronic diseases (CHD & cancer), some infectious diseases, suicides, and accidents.
Among older women, divorce and separation are linked with adverse health outcomes such as physical impairments. Men had higher mortality risks than women if divorced or widowed.
Married women have been found to have a reduced risk of breast cancer mortality when compared with single women.
Never married adults (especially men) less likely to be overweight whereas being married was associated with obesity, especially among men (this was the one exception to being healthier for married people).
Single people (40-79 years) experienced a higher mortality risk than married persons, according to a 2007 Japanese study.
Protective hypothesis: marriage makes a positive impact to health by influencing lifestyle factors, providing mutual psychological and social support, and increasing available financial resources. The marital environment and factors associated with marriage apparently reduce the risk of death and therefore, should be considered as a possible source of differences in disease rates.

Marital Selection Model: physically attractive people are more likely to be successful in competing for a mate and are healthier than those persons who never marry; also, this model proposes that less healthy individuals, if married, are more predisposed than healthy persons to gravitate to non-married status. Suicide tends to be higher in widowed young men in comparison with married young men in the same age group (20-34 years). Depressed mood more common among widowed individuals than among persons who were married, living together, never married, or divorced/separated. Married people had the lowest percentage of depressed mood.
Race—refers to a person's physical characteristics (bone structure, skin, hair, eye color)
Ethnicity—refers to cultural factors, including nationality, regional culture, ancestry, and language.
However, this being said, some scientists propose that race is primarily a social and cultural construct—take a look at the 2000 Census.
Census 2000 used five categories of race—white, black/AA, American Indian & Alaska Native, Asian, Native Hawaiian & Other Pacific Islander.
Also, Census 2000 changed the race category by allowing respondents to choose one or more race categories.
The 2 words, race and ethnicity, tend to be used interchangeably because using physical characteristics to assign someone to a particular race may be difficult at best, as in the case of persons of mixed race parentage.

So as far as the "Who" (Person) goes in descriptive epidemiology, we will be briefly discussing the characteristics of 4 races to demonstrate how the prevalence and the causes of morbidity/mortality may vary from one race to another.
Heart disease is the leading cause of death in white, black, and American Indians/Alaskan Natives.
Stroke (CVA) is the 3rd most common cause of death in blacks and Asian/Pacific Islanders
AA: In 2007, age-adjusted death rate for African Americans was 1.3 times rate for whites. Included in the 10 leading causes of death in non-Hispanic blacks were homicide, HIV, and septicemia—these are not among the top 10 causes of death in non-Hispanic whites.
AAs have more problems with HTN: causes? Possible influence of stress or diet (decreased fruits/veggies), decreased social support, increased rates of obesity, lack of participation in cardiovascular risk reduction programs.
Cancer is high—breast CA higher in black women than in 4 of the 5 races. Are any of these things due to lack of access? Are services being offered but not utilized? Lack of education on risky behaviors, diet, resources available, etc.?
Differences in life expectancy—black men don't live as long as white men; black women don't live as long as white women. Why? Is lack of access the ONLY factor? What if we compared blacks with other races?! What would the life expectancy be when comparing blacks with Native Americans?!
Notice that most of these statistics compared blacks to ONLY whites—the next slide on American Indians compares them to the general US population.
Heart disease is the leading cause of death in white, black, and American Indians/Alaskan Natives.

AI: High rates of chronic diseases, adverse birth outcomes, and some infectious diseases (TB & Hepatitis A)
Diabetes—11.9 times the rate for all races in US
Cirrhosis—6.5 times the rate for all races in US
Prevalence of TB is 2 times that of US Population—7 times that of non-Hispanic white population.
Accidents—5.9 times the rate for all races in US
Homicide—7.4 times the rate for all races in US
Suicide—4.3 times the rate for all races in US

Life expectancy—decreased compared to general US Population

Notice that this group of individuals was compared to the general US population and/or to all races and not to just whites.
Asian
Japanese: lower mortality rates than whites
Lower rates of CHD and cancer
Attributed to low-fat diet & stress-reducing strategies
Some Asian groups (Cambodians) have high smoking rates
70% smoking rates
TB rates highest among Asian/Pacific Islander group
Almost 25 times higher than Non-Hispanic whites.
Stroke (CVA) is the 3rd most common cause of death in blacks and Asian/Pacific Islanders
Cancer is the leading cause of death in Asian/Pacific Islanders.

Japanese culture seems to protect them—low rates of CHD d/t diet and stress-reducing strategies; however, Japanese moving to the US and becoming acculturated lost this protection over time. Studies show a link to environmental and behavioral factors on chronic diseases.

Acculturation—modifications that individuals or groups undergo when they come into contact with another country (migrate, move, etc.). Studying those who have immigrated provides evidence of the influence of environmental and behavioral factors on chronic disease. Example: Japanese migrants experience a shift in rates of chronic disease toward those of the host country.

Stroke (CVA) is the 3rd most common cause of death in blacks and Asian/Pacific Islanders
Cancer is the leading cause of death in Asian/Pacific Islanders.

HHANES—found out that there is much diversity between Hispanic groups (Cubans, Puerto Ricans, etc.) and that they need to be studied individually.
Low rates of CHD in Mexican Americans—this may be d/t culture—diet preferences, social support in large and extended families; Puerto Ricans also had a low prevalence of CHD.

San Antonio Heart Study

Hispanic Mortality Paradox: Despite having a high prevalence of Diabetes and other risk factors for chronic disease, Hispanics/Latinos in the US have a lower mortality rate than non-Hispanic whites (28.5% lower--and 44.7% lower than non-Hispanic blacks!) Could this be d/t underreporting of deaths among Hispanics? Do many of them return back to their native country to die (salmon bias effect)? It was found that Cubans and Puerto Ricans do NOT return home to die......and this paradox occurs with them, too. This difference has not been fully explained.
Hispanic/Latino
HHANES—Hispanic Health and Nutrition Examination Survey
1st special population survey of Hispanics in the US
Examined health & nutrition status of major Hispanic/Latino population in the US
San Antonio Heart Study
Found high rates of obesity and diabetes among Mexican Americans
Hispanic mortality paradox (text box)

Nativity—where they or parents were born.

Categories are foreign born and native born—these are common subdivisions used in epidemiology. Nativity is tied to migration because foreign-born persons have immigrated to their host country.

Natural experiment—migration meets the criteria for this, in which the effects of change from one environment can be studied—various health dimensions (stress, acculturation, chronic and infectious diseases) can be studied as far as the effect of moving from one environment to another (remember the Japanese?)

Studies on admission to mental hospitals in New York—admission rates were higher among foreign-born than native-born persons, suggesting that foreign-born individuals may experience stresses associated with migration to a new environment.....but WHAT else may it suggest?

Impact of migration: Importation of "Third World" disease by immigrants from developing countries
Leprosy during 1980s
Programmatic needs resulting from migration:
Specialized screening programs (tuberculosis and nutrition)
Familiarization of doctors with formerly uncommon (in U.S.) tropical diseases
In the 1980's, it was leprosy.
Currently, with people coming across our southern border unchecked, who knows what they are carrying with them?! Some of the diseases they bring with them are diseases that people in the U.S. have never been exposed to—think about the diseases that are "endemic" to their country! These are things to which they may have natural immunity but we don't because we've never seen them before.

Also, some migrants are inadequately immunized with respect to measles and other vaccine-preventable diseases (remember that not all nations have the immunization program that the U.S. does)—this has hampered efforts of health officials to eliminate these diseases from the U.S.
Observation that healthier, younger persons usually form the majority of migrants
Often difficult to separate environmental influences in the host country from selective factors operative among those who choose to migrate
Certain religions prescribe lifestyles that may influence rates of morbidity and mortality.
Example: Seventh Day Adventists
Follow vegetarian diet and abstain from alcohol and tobacco use
Have lower rates of CHD, reduced cancer risk, and lower blood pressure
Similar findings for Mormons
SES: Low social class is related to excess mortality, morbidity, and disability rates.
Factors that negatively effect health include:
Poor housing
Crowded conditions
Racial disadvantage
Low income
Poor education
Unemployment
Relationship between SES and health has been demonstrated for a wide range of health outcomes and confirmed by a massive body of evidence.


Variables include:
Prestige of occupation or social position
Educational attainment
Income
Combined indices of two or more of the above variables
But there are some problems with each of these.....

Many epidemiologic terms come from Sociology, such as the measurement of social class, which is a means to measure the economic position.
The measurement of social class includes all of these variables, but it is also related to ethnicity, race, religion, & nativity.
Occupation/social position: the problem with relying on this one is that there may be 2 workers in the same family who are at different levels of occupational prestige, but even those with the same level of occupational prestige may or may not make the same money.
Education: higher levels of education, in contrast to income or occupation, appear to be the strongest and most important predictor of positive health status! (1992 study cited)
Income: problem with using income as a measure of social class:
1. Some people may not know the income of the entire family or may not
disclose it.
2. Two workers in a lower SES family may actually make more combined than 1
worker in a higher SES family.
Sample of subjects (N) is selected first, then exposure and level of disease is determined—however, some studies may focus on just the level of disease while others focus on just the level/distribution of exposure.
Individual level
Single period of observation—both histories (exposure & history of disease) collected simultaneously.
They are basically descriptive in nature—they yield quantitative estimates of the magnitude of a problem but do not measure cause and effect.
Sample design: Probability and non-probability sampling is used—
Probability sample: every element in the population has a nonzero probability of being included in the sample. Ex. Simple random samples (everyone has same chance of being selected), systematic samples ( some simple, systematic rule—all first names start with C), stratified samples (population divided by age, sex, race, etc.)
Non-probability sample: based on a sampling plan that does not have this feature. Ex. Quota samples (certain number of samples must be completed) and Judgmental samples (subjects selected on the basis of investigator's perception that they represent the population as a whole).

Surveys of smokeless tobacco use among high school students
Prevalence surveys of the number of vasectomies performed
Prevalence surveys of cigarette smoking among Cambodian Americans in Long Beach, California
Prevalence surveys are helpful for identifying resource needs for health interventions.
Read about these studies and more on Pages 296-300

Hypothesis generation
Intervention planning
Planning health services and administering medical care facilities
Estimation of the magnitude and distribution of a health problem
Examine trends in disease or risk factors that can vary over time
Limited usefulness for inferring disease etiology
Do not provide incidence data
Cannot study low prevalence diseases
Cannot determine temporality of exposure and disease
Cannot study low prevalence diseases—prevalence is proportional to the incidence of the disease times its duration. A large survey may have many cases of diseases with short duration.
Which came first, the chicken or the egg? They are done so quickly, with exposure and disease histories being taken at the same time, it is hard to determine whether the exposure came first or the disease. This makes it hard to determine cause and effect.
Five key questions to be asked:
Could the association have been observed by chance?
Determined through the use of statistical tests.
Could the association be due to bias?
Bias refers to systematic errors, i.e., how samples were selected or how data was analyzed.
Could the association have been observed by chance? The P-value indicates the probability that the findings observed could have occurred by chance alone. A small P-value (a highly significant result) for an observed association should provide some assurance that the results were not obtained simply by chance. However, a very small P-value doesn't imply that the association is real.

Could association be d/t bias? Errors at any of the following stages may lead to results that are not valid:
+How study groups selected
+How information about exposure and disease was collected
+How data were analyzed

Confounding—refers to the masking of an association between an exposure and an outcome because of the influence of a 3rd variable that wasn't considered in the design or analysis.

To whom does this apply? Population-based samples are important and the sampling procedures used enhance the likelihood of generalizability (ability to apply results to other populations not involved), but this doesn't mean that you'll get this outcome (generalizability).
Representativeness of sample: a lot can be learned from an unrepresentative study sample! It's easier to generalize from one group to another if groups are similar! (Ex. White women aged 55 to 69
who live in Iowa could be generalized to other white women of the same age who live in the Midwest).
Participation Rates: % of a sample that completes the data collection phase of a study. Some people think this needs to be high to generalizability, but high participation rates don't necessarily ensure
generalizability. Generalizability can be high even with low participation rates.

Cause-and-effect relationship? Strength of association---more on this later in the presentation.
Establishing causality:
Diseases have not one, but multiple causes
Determining the relationships between the different causes of disease is important to PH practitioners who seek to prevent, diagnose, and treat disease.
Screening:
Surveillance: Passive vs. Active surveillance

Also referred to as multifactorial etiology--"...requirement that more than one factor be present for disease to develop..."
It is often discovered that not one but many things cause a disease or event or a chain of causes occur....producing a "Chain of Causality"

Screening: Key component in many secondary prevention interventions
Used to identify risk factors and diseases in their earliest stages.

Surveillance: the systematic collection, analysis, and interpretation of data related to the occurrence of disease and the health status of a given population.

Passive surveillance: health care providers in the community
report cases of notifiable diseases to public health authorities through the
use of standardized reports. Inexpensive, but limited—information
depends on providers reporting practices.

Active surveillance: purposeful, ongoing search for new cases of
disease by public health personnel, through personal or telephone
contacts or review of laboratory reports or hospital and/or clinic
records. Costly—so it's only used in cases of emergence of new
diseases, severe diseases, or re-emergence of a previously eradicated
disease.

Epidemiologic triangle
Web of causation, e.g., in avian influenza
A two-dimensional causal web that considers multiple levels of factors that affect health and disease
Looks like a spider web
What might be the "spider"?
This concept is based upon the fact that there are various factors interacting, sometimes in subtle ways, to increase the risk for illness or to decrease it.

Krieger (1994) has suggested that in addition to research on the relationships within the web, we need to look for "the spider," that is, focus on those larger factors and contexts that influence or create the causal web itself.
More on this later.....

Krieger (1994) has suggested that in addition to research on the relationships within the web, we need to look for "the spider," that is, focus on those larger factors and contexts that influence or create the causal web itself.

The concept of multiple causation is comparable to that of the chain of disease transmission. It reflects the complex relationships between many different factors that interact to either increase or decrease the risk of disease.

This model helps to explain how diseases have multifactorial etiology. Some of these may be both positive factors (or protective factors) that may protect someone from diabetes and negative (predictive) factors, depending upon existence.

-Think of this as a concept map! Just as we know that the body is made up of many systems, and we need to address them all in care, there is no single cause of disease;
-Most of the causes of diseases interact like this, too, & this illustrates the interconnectedness of possible contributing factors--


**Also, remember that not everyone who smokes gets lung cancer and not everyone who has an MI has high cholesterol or ate high-fat foods all of their lives!


Wheel model, e.g., childhood lead poisoning
Pie model, e.g., lung cancer
The relationship between an event (the cause) and a second event (the effect), where the second event is a consequence of the first.
Why is an event or a disease occurring? What aspect of "environment" (broadly defined) if removed/reduced/controlled would reduce outcome/burden of disease?

Is there an association between a disease and other variables?
Statistical Associations—Does a statistical association exist between some factor and a health outcome? If the probability of disease seems unaffected by the presence or level of the factor, then there is no association. But if probability of disease DOES vary according to whether the factor is present, then there is a statistical association.

Bias is a systematic error resulting from the study, design, execution, or confounding; information (classification) bias: r/t how information is collected, including information that subjects supply or how subjects are classified. Selection bias: bias attributable to the way subjects enter a study.

Assessing for Causality: just because statistical association exists doesn't mean that there is a causal relationship or that causality is present. The association could be d/t bias (flaws in study design or execution) or may be a chance event.

Some epidemiologists use criteria for causality to evaluate the link between an infectious agent and disease—the next slide will cover these criteria
Used to identify etiological factors of diseases; it is a very useful tool in epidemiology to determine the most effective primary prevention activities and develop treatment modalities.
Strength of association: Rates of morbidity and mortality must be higher for the exposed group than for the non-exposed group (risk of heart disease is higher in smokers than in non-smokers)
Consistency with other studies: Varying types of studies in other populations must observe similar associations.
Biological plausibility: The data must make biological sense and represent a coherent explanation for the relationship.
Demonstration of correct temporal sequence: Exposure to the causal factor must occur before the effect, or the disease.
Dose-response relationship: An increased exposure to the risk factor causes a concomitant increase in disease (risk for heart disease is higher in heavy smokers compared to light smokers)
Specificity of the association: The exposure variable must be necessary and sufficient to cause disease; there is only one causal factor. This one is not as important today since diseases have multifactorial origins.
Experimental evidence: experimental designs provide the strongest epidemiologic evidence for causal associations, but they are not feasible or ethical to conduct for many risk factors—disease association.
Absolute causality is rarely established.... Some non-smokers get heart disease but most all smokers have heart disease....Smoking doesn't give you heart disease but puts you at greater risk for it.
***Keep in mind that there are multiple causes of disease in most cases—this is referred to as multiple causality or causation.
Multiple causality (or causation) results in what is known as "web of causation" or "chain of causation" which is very common for noncommunicable/chronic diseases (cancer, cardiovascular, etc.)
random: Reflect fluctuations around a true value of a parameter because of sampling variability.
Factors That Contribute to Random Error- Poor precision
Occurs when the factor being measured is not measured sharply.
Analogous to aiming a rifle at a target that is not in focus.
Precision can be increased by increasing sample size or the number of measurements.
Example: Bogalusa Heart Study

Sampling error
Arises when obtained sample values (statistics) differ from the values (parameters) of the parent population.
Although there is no way to prevent a non-representative sample from occurring, increasing the sample size can reduce the likelihood of its happening.

Variability in measurement
The lack of agreement in results from time to time reflects random error inherent in the type of measurement procedure employed.

Bias (Systematic Errors)
"Deviation of results or inferences from the truth, or processes leading to such deviation. Any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth."
Selection bias-
Refers to distortions that result from procedures used to select subjects and from factors that influence participation in the study.
Arises when the relation between exposure and disease is different for those who participate and those who theoretically would be eligible for study but do not participate.
Example: Respondents to the Iowa Women's Health Study were younger, weighed less, and were more likely to live in rural, less affluent counties than non-respondents.

Information bias-
Can be introduced as a result of measurement error in assessment of both exposure and disease.
Types of information bias:
Recall bias: better recall among cases than among controls.
Example: Family recall bias
Interviewer/abstractor bias--occurs when interviewers probe more thoroughly for an exposure in a case than in a control.
Prevarication (lying) bias--occurs when participants have ulterior motives for answering a question and thus may underestimate or exaggerate an exposure

Confounding
The distortion of the estimate of the effect of an exposure of interest because it is mixed with the effect of an extraneous factor.
Occurs when the crude and adjusted measures of effect are not equal (difference of at least 10%).
Can be controlled for in the data analysis.
To be a confounder, an extraneous factor must satisfy the following criteria:
Be a risk factor for the disease.
Be associated with the exposure.
Not be an intermediate step in the causal path between exposure and disease.
Simpson's paradox: means that an association in observed subgroups of a population may be reversed in the entire population.
Illustrated by examining the data (% of black and gray hats) first according to two individual tables and then by combining all the hats on a single table.
When the hats are on separate tables, a greater proportion of black hats than gray hats on each table fit.
On table 1:
90% of black hats fit
85% of gray hats fit
On table 2:
15% of black hats fit
10% of gray hats fit
When the man returns the next day and all of the hats are on one table:
60% of gray hats fit (18 of 30)
40% of black hats fit (12 of 30)
Note that combining all of the hats on one table is analogous to confounding.
ex. Air pollution and bronchitis are positively associated. Both are influenced by crowding, a confounding variable.
The association between high altitude and lower heart disease mortality also may be linked to the ethnic composition of the people in these regions.

Techniques to Reduce Selection Bias-
Develop an explicit (objective) case definition.
Enroll all cases in a defined time and region.
Strive for high participation rates.
Take precautions to ensure representativeness.
Ensure that all medical facilities are thoroughly canvassed.
Develop an effective system for case ascertainment.
Consider whether all cases require medical attention; consider possible strategies to identify where else the cases might be ascertained.
Compare the prevalence of the exposure with other sources to evaluate credibility.
Attempt to draw controls from a variety of sources.
Use memory aids; validate exposures.
Blind interviewers as to subjects' study status.
Provide standardized training sessions and protocols.
Use standardized data collection forms.
Blind participants as to study goals and classification status.
Try to ensure that questions are clearly understood through careful wording and pretesting.


Prevention strategies--attempt to control confounding through the study design itself.
Three types of prevention strategies:
Randomization
Restriction
Matching
Two types of analysis strategies:
Stratification
Multivariate techniques
Screening--the presumptive identification of unrecognized disease or defects by the application of tests, examinations, or other procedures that can be applied rapidly.
Positive screening results are followed by diagnostic tests to confirm actual disease.
Example: phenylalanine loading test in children positive on PKU screening

Surveys
Conducted on an ad hoc basis to identify individuals who may have infectious or chronic diseases.
Examples: breast cancer screenings, health fairs.
Clientele are highly selected.
Individuals who participate are concerned about the particular health issue.

Ad hoc—a particular group of people

Test and procedures to identify unrecognized/suspected cases of disease.

Results need to be sent to PCP for follow-up!!




multiphasic- Defined as the use of two or more screening tests together among large groups of people.
Information obtained on risk factor status, history of illness, and physiologic and health measurements.
Commonly used by employers and health maintenance organizations.
Administration of 2 or more screening tests during a single screening program
Ongoing screening programs often are carried out at worksites.
Potential biases from worker attrition
Data can be useful for research on occupational health problems.
Data may not contain etiologic information.


Mass screening--screening on a large scale of total population groups regardless of risk status.
Selective screening--screens subsets of the population at high risk for disease.
More economical, and likely to yield more true cases.
Example: Screening high-risk persons for Tay-Sachs disease.

Mass Health Examinations- Population or epidemiologic surveys--purpose is to gain knowledge regarding the distribution and determinants of diseases in selected populations.
No benefit to the participant is implied.
Epidemiologic surveillance--aims at the protection of community health through case detection and intervention (e.g., tuberculosis control).
Case finding (opportunistic screening)--the utilization of screening tests for detection of conditions unrelated to the patient's chief complaint.

social- The health problem should be important for the individual and the community.
Diagnostic follow-up and intervention should be available to all who require them.
There should be a favorable cost-benefit ratio.
Public acceptance must be high.

scientific- Natural history of the condition should be adequately understood.
This knowledge permits identification of early stages of disease and appropriate biologic markers of progression.
A knowledge base exists for the efficacy of prevention and the occurrence of side effects.
Prevalence of the disease or condition is high.
Ethical-
The program can alter the natural history of the condition in a significant proportion of those screened.
Suitable, acceptable tests for screening and diagnosis of the condition as well as acceptable, effective methods of prevention are available.

Characteristics of a Good Screening Test- Simple--easy to learn and perform.
Rapid--quick to administer; results available rapidly.
Inexpensive--good cost-benefit ratio.
Safe--no harm to participants.
Acceptable--to target group.

Evaluation of Screening Tests-
Reliability types
Repeated measurements
Internal consistency
Interjudge

Validity types
Content
Criterion-referenced
Predictive
Concurrent
Construct
In 1537, Ambroise Paré applied experimental treatment for battlefield wounds.
East India Shipping Company (1600) found that lemon juice protected against scurvy.
James Lind (1747) used the concurrently treated control group method.

Pare—turpentine, rose oil, egg yolks concoction to treat battlefield wounds—was better than boiling oil at treating wounds!

Lemon juice saved sailors in 1600—this was discovered when the East India Shipping Company compared sailors from ships with lemons and sailors from ships without.

1747—citrus fruits discovered to be the treatment for scurvy. 12 sailors suffering from scurvy were fed 6 different types of diets—those who ate citrus fruits recovered the best.

Edward Jenner's efforts to develop a smallpox vaccine in the late 18th century
Most recent historical developments include the use of multicenter trials.
Instrumental in the development of treatments for infectious diseases and recently in chronic diseases that are of noninfectious origin

Jenner—while the early experiments were carried out without a control group or a comparison group, subsequent studies contributed to the development of control treatments and randomization.

Multicenter trials—recruitment of participants is extended across several to hundreds of accrual sites with data sent to a coordinating center for analysis.


A research activity that involves the administration of a test regimen to humans to evaluate its efficacy and safety
Wide variation in usage:
The first use of the term was for studies in humans without any control treatment
Now denotes a rigorously designed and executed experiment involving RANDOM ALLOCATION of test and control treatments

Pg. 368—NIH Definition: A prospective biomedical or behavioral research study of human subjects that is designed to answer specific questions about biomedical or behavioral interventions (drugs, treatments, devices, or new ways of using known drugs, treatments, or devices).

Clinical Trials are used to determine whether new biomedical or behavioral interventions are safe, efficacious, and effective.

The key point is that clinical trials enroll individual subjects and enable randomization of subjects to either receive or not receive the intervention.

Carefully designed and rigidly enforced protocol
Tightly controlled in terms of eligibility, delivery of the intervention, and monitoring of outcomes
Duration ranges from days to years
Participation is generally restricted to a highly selected group of individuals.


Participation is generally restricted to a highly selected group of individuals—mainly people who have been diagnosed with a disease, who are screened subjects at high risk for disease, or just interested people.

Once subjects agree to participate, they are randomly assigned to one of the study groups, e.g., intervention or control (placebo)
Eligibility of potential subjects is determined first. Eligibility rules must be carefully defined and rigidly enforced. Criteria for inclusion will vary by the type and nature of the intervention proposed.

Once eligible subjects agree to participate, they are randomly assigned to a study group—either intervention or control

Strengths:
Provide the greatest control over:
the amount of exposure (drug dosage)
the timing and frequency of exposure
the period of observation for end points
Ability to randomize reduces the likelihood that groups will differ significantly.
Less likelihood of variables influencing the outcome

Limitations:
Artificial setting
Limited scope of potential impact
Adherence to protocol is difficult to enforce
Especially if treatment produces undesirable side effects &/or a significant burden to the subjects.
Ethical dilemmas
Withholding a potentially beneficial treatment from the control group
Artificial setting—treatments may not work as well in a setting other than the controlled clinical area.
Community intervention trials determine the potential benefit of new policies and programs
Intervention: Any program or other planned effort designed to produce changes in a target population
Community refers to a defined unit, e.g., a county, state, or school district
trial—an "experiment in which the unit of allocation to receive a preventive, therapeutic, or social intervention is an entire community or political subdivision." Example—fluoridation of drinking water.

Start by determining eligible communities and their willingness to participate
Collect baseline measures of the problem to be addressed in the intervention and control communities
Use a variety of measures, e.g., disease rates, knowledge, attitudes, and practices
Permission to enroll the community is usually given by someone in charge—the mayor, governor, school board, etc.

Measures to use as baselines may include: disease prevalence or incidence, knowledge/attitudes/practices, purchase of lean relative to fatty cuts of meat, etc.

Communities are randomized and followed over time
Outcomes of interest are measured
After baseline information is obtained, communities are randomized to receive or not receive the intervention.

These randomized community trials are also called cluster randomized trials.

North Karelia Project
Minnesota Heart Health Program
Stanford Five-City Project
Pawtucket Heart Health Program
Community Intervention Trial for Smoking Cessation (COMMIT)
Project Respect

They represent the only way to estimate directly the impact of change in behavior or modifiable exposure on the incidence of disease.

They are inferior to clinical trials with respect to:
ability to control entrance into study, delivery of the intervention, and monitoring of outcomes.
Fewer study units are capable of being randomized, which affects comparability.
They are affected by population dynamics, secular trends, and nonintervention influences.
Affected by population dynamics, secular trends, and nonintervention influences. For example, the fluoridation of the community water—these studies take place over long periods of time, so people come into the community and move out of the community. This will have some effect on the results—it can negatively impact the results by making them look worse than they really would have been or it can positively impact the results by making them look better than they would have been.
A cohort is defined as a population group, or subset thereof, that is followed over a period of time.
The term cohort is said to originate from the Latin cohors, which referred to one of ten divisions of an ancient Roman legion.
Cohort group members experience a common exposure associated with a specific setting (e.g., an occupational cohort or a school cohort) or they share a non-specific exposure associated with a general classification (e.g., a birth cohort—being born in the same year or era).
The tabulation and analysis of morbidity or mortality rates in relationship to the ages of a specific group of people (cohort) identified at a particular period of time and followed as they pass through different ages during part or all of their life span.

cohort effect- The influence of membership in a particular cohort.
Example: Tobacco use in the U.S.
Fewer than 5% of population smoked around the early 1900s.
Free cigarettes for WWI troops increased prevalence of smoking in the population.
During WWI, age of onset varied greatly; then people began smoking earlier in life.
One net effect was a shift in the distribution of the age of onset of lung cancer.
Cohort Studies-
Include at least two observation points: one to determine exposure status and eligibility and a second (or more) to determine the number of incident cases
This permits the calculation of disease/incidence rates.

Cohort studies measure incidence directly. Not only is incidence measured, but mortality, health status (morbidity), and certain biological parameters as well.

Going from cause to effect: Exposure of interest (cause) is determined for each member of the cohort at the start of the study—group is followed through time to document the incidence of an outcome (effect) among exposed and non-exposed members.
The individual forms the unit of observation and the unit of analysis.
Involve the collection of primary data, although secondary data sources are used sometimes for both exposure and disease assessment

Population-Based Cohort Studies-
The cohort includes either an entire population or a representative sample of the population.
Population-based cohorts have been used in studies of coronary heart disease.
Exposures unknown until the first period of observation when exposure information is collected
Examples: After administration of questionnaires, collection of biologic samples, and clinical examinations, there can be two or more levels of exposure.

The Alameda County Study
Studied factors associated with health and mortality
Involved residents of Alameda County, CA, ages 16-94 years
Data collected through mailed questionnaires; telephone interviews or home interviews of non-respondents
Follow-up with same procedures at years 9, 18, and 29
Honolulu Heart Program
Studied coronary heart disease and stroke in men of Japanese ancestry
Involved men of Japanese ancestry living on Oahu, HI, ages 45-65 years
Data were collected through mailed questionnaires, interviews, and clinic examinations.
Nurses' Health Study
Originally studied oral contraceptive use; expanded to women's health
Married female R.N.s ages 30-55 years
Data collected through mailed questionnaires
Follow-up every 2 years; toenail sample at year 6 and blood sample at year 13

Permit direct determination of risk.
Because you start with disease-free subjects, it permits direct determination of risk.

Time sequencing of exposure and outcome.
These studies provide evidence about lag time between exposure and disease occurrence (the time from exposure to development of the disease).

Can study multiple outcomes.
If they are properly designed and executed, they allow examination of multiple outcomes.

Can study rare exposures.
Cohort studies can increase the efficiency for rare exposure studies through selection of cohorts with known exposures (such as certain occupational groups).

Subjects lost to follow-up—either because they dropped out of the study or died. Can be a significant problem if the loss is too high. Questions concerning the reliability and validity of the results can arise due to this.
A nested case-control study is defined as a type of case-control study ". . . in which cases and controls are drawn from the population in a cohort study."
Example: nested case-control breast cancer study
Controls are a subset of the source population for the cohort study of breast cancer.
Cases of breast cancer identified from the cohort study would comprise the cases.

Example: a nested case-control breast cancer study—the population of this cohort would involve both exposed (women who used birth control pills) and the non-exposed (women who do not use birth control pills)---just like the cohort study of children with autism. That group would have contained children who had been exposed to immunizations and children who had NOT been exposed to immunizations!

Provide a degree of control over confounding factors.
Reduce cost because exposure information is collected from a subset of the cohort only.
An example is an investigation of suicide among electric utility workers.
Suicide among electric utility workers—this study examined the association between exposure to very low-frequency magnetic fields and suicide.

Cases (536 suicides) and controls (n=5248) were selected from a cohort of 138,905 male utility workers. Findings supported the hypothesis—there was association found.

Permit direct determination of risk.
Because you start with disease-free subjects, it permits direct determination of risk.

Time sequencing of exposure and outcome.
These studies provide evidence about lag time between exposure and disease occurrence (the time from exposure to development of the disease).

Can study multiple outcomes.
If they are properly designed and executed, they allow examination of multiple outcomes.

Can study rare exposures.
Cohort studies can increase the efficiency for rare exposure studies through selection of cohorts with known exposures (such as certain occupational groups).
The easiest studies to start with when not much is known about the topic of study are studies that can use existing data, that are quick and easy to conduct, and that are economical. As knowledge increases, the complexity of the research questions increases....this will require more rigorous studies at that time.
Number of observations made: some studies require that observations be made at just one time while others may do observations at 2 or more points in time.
Directionality of exposure: varies because it is relative to disease; the investigator may go back in time (retrospective) and ask subjects about the disease they currently have or may elect to start with a group who does NOT have the disease and follow them prospectively for the development of the disease.
Data collection methods: Some methods require use of existing data while some require collection of new data.
Timing of data collection: data should be collected and used quickly or questions might be raised about the quality and applicability of the data. Ex. if there was a long time between the measurement of exposure and disease.
Unit of observation: Some studies use groups while others use individuals
Availability of subjects: certain classes of subjects may be off-limits due to a number of considerations, including ethical issues.


Experimental Studies: maintain the greatest control over the research setting; investigator both manipulates variables in the study (drug or placebo—blind, double-blind to eliminate bias) and randomly assigns subjects to exposed and non-exposed groups. One example of an experimental study is a clinical trial—usually a new medication, surgical procedure, etc. They may be used to test hypotheses developed from the observational studies. Usually participants are randomly assigned.
Quasi-experimental studies: Manipulation of variables in the study but no randomization of participants—could be thought of as a natural experiment. Ex. Before federal seat belt laws, some states had laws and some did not. Residents didn't determine their "exposure" (seat belt), their politicians did (manipulation)—so a study of the traffic fatalities in states with laws vs. those without is an example of a quasi-experimental study. Another type of quasi-experimental study is a Community trial (or intervention)—most are oriented towards education and behavior modification. Ex. Fluoridation of community's water supply, smoking cessation, control of ETOH use, weight loss, establishing healthy eating behaviors, encouraging increased activity, etc. May also focus on people at high risk for a certain disease within the community and may even help shape public health policy, such as mandatory seat belt usage. Data from these studies can also be used to evaluate programs, to compare programs to determine why something worked or why it failed, to compare costs, and to suggest changes in current health policies.
Observational: Descriptive: Person, Place, & Time studies—case reports, case series, cross-sectional surveys. Also, estimates disease frequency and time trends. Mostly used to generate hypotheses for further studies. Analytic: ecologic studies, case-control studies, and cohort studies. Used to test the hypotheses generated by descriptive studies, to generate more hypotheses, and to suggest mechanisms of causation.
The unit of analysis is the group, not the individual.
They can be used for generating hypotheses.
The level of exposure for each individual in the unit being studied is unknown.
Generally makes use of secondary data.
Advantageous with cost and duration.
With ecologic studies, the marginal numbers are usually known or inferred; the interior cells are unknown.
Cheaper because they make use of available data.

Ecologic comparison study—involves an assessment of the correlation between exposure rates and disease rates among different groups over the same time period.
Ecologic trend study—involves correlation of changes in exposure with changes in disease within the same community, country, or other aggregate unit. used to ascertain trends—Ex. the consistent downward trend in the incidence of and mortality from CHD. Is it really happening? What are the reasons? If it is due to doctors working hard to achieve this result, then they need to claim responsibility for it....using this type of study, ecologic correlation data could be generated that can support the claim that there was an increase in BP meds being prescribed or an increase in CABG surgeries.
Example of an Ecologic Correlation study.
The association between breast cancer and dietary fat for 39 countries.
High intakes of dietary fats associated with high rates of breast cancer mortality.
Another example is the study of childhood lead poisoning in Massachusetts. More than 200,000 children screened in doctors' offices, hospitals, state-funded screening sites, through nutritional programs, and door-to-door screening in high-risk areas (houses, daycares, schools older than 1978).
There are a few more examples of ecologic studies in your textbook....read about them from page291-292.
Is the ranking of cities by air pollution levels associated with the ranking of cities by mortality from cardiovascular disease, adjusting for differences in average age, percent of the population below poverty level, and occupational structure?
What are long-term trends (1950-1995) for mortality from the major cancers in the US, Canada, and Mexico?
The effect of fluoridation of the water supply on hip fractures
The association of naturally occurring fluoride levels and cancer incidence rates
The relationship between neighborhood or local area social characteristics and health outcomes

Advantages—When little is known about the association between an exposure and a disease, an ecologic study is a good place to start—it's quick, simple, and inexpensive. When a disease etiology is unknown, ecologic studies are a good approach for generating hypotheses.
Disadvantages—
ecological fallacy and imprecise measurement of exposure and disease makes accurate quantification of this exposure/disease relationship difficult.
In a case-control study with two groups, one group has the disease of interest (cases) and a comparable group is free from the disease (controls).
The case-control study identifies possible causes of disease by finding out how the two groups differ with respect to exposure to some factor.

A single point of observation
Unit of observation and the unit of analysis are the individual
Exposure is determined retrospectively
Does not directly provide incidence data
Data collection typically involves a combination of both primary and secondary sources.

A mainstay of epidemiologic research!
Usually going from the "effect" to the "cause" as these studies are done retrospectively—by looking into the history of the subject's exposure after the disease has already occurred.
Data is usually collected by researchers, but valuable information can also be obtained from medical, school, and employee records.

Selection of Cases
Two tasks are involved in case selection:
Defining a case conceptually
Identifying a case operationally

Sources
Need to define a case conceptually
Ideally, identify and enroll all incident cases in a defined population in a specified time period
A tumor registry or vital statistics bureau may provide a complete listing of all cases
Medical facilities also may be a source of cases, but not always incident cases
Lack of suitable disease registries may mean that you're left with only medical facilities that may receive only the most severe cases, so this can skew your data. Also, medical facilities don't always treat just incident cases.....

Selection of controls- The ideal controls should have the same characteristics as the cases (except for the exposure of interest).
If the controls were equal to the cases in all respects other than disease and the hypothesized risk factor, one would be in a stronger position to ascribe differences in disease status to the exposure of interest.

If the controls differ in demographic factors such as age or socioeconomic status, these factors could operate as a rival explanation to account for the observed outcomes.
Number of controls—usually it is a one-to-one ratio—one control for each case, but researchers can also accept up to a 4:1 controls to cases ratio. This is the maximum.

Population-based controls—Obtain a list that contains names and addresses of most residents in the same geographic area as the cases.
A driver's license list would include most people between the ages of 16 and 65.
Tax lists, voting lists, and telephone directories
Patients from the same hospital as the cases
Relatives of cases

They should come from the same population at risk for the disease or condition as the cases being studied. Ex. In a study of ovarian cancer, women who have had their ovaries removed would not be eligible to be controls as they no longer have the risk of contracting ovarian cancer.

Measures of Association
This just shows you how to link association with an illness....

The OR represents the odds that an outcome will occur given a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure. Odds ratios are most commonly used in case-control studies, however they can also be used in cross-sectional

On the association between chili pepper consumption and gastric cancer risk: a population-based case-control study conducted in Mexico City
Source: Lopez-Carillo, et al. Am J Epidemiol. 1994;139:263-71.

Calc. Odds Among Cases = 204/9 = 22.67
Odds Among Controls = 552/145 = 3.81
The OR (unadjusted for age and sex) is:

AD = (204)(145) = 5.95
BC (552)(9)

The study involved 220 incident cases and 752 controls who were randomly selected from the general population. The CASES had or have had gastric cancer; the CONTROLS do NOT or have NOT had gastric cancer.
So using the data from this study, we will plug in the numbers to calculate the OR.

Out of the Cases, 204 had been diagnosed with gastric cancer who had a history of eating chili peppers (this is A); 9 got gastric cancer who did NOT eat chili peppers (this is C).

Out of the Controls (People alike in age, gender, and location but without Cancer!), 552 had a history of pepper intake (this is B) and 145 did NOT (this is D). So setting up the calculation as shown......we get 5.95 for our OR

The OR = 5.95 suggests that cases were nearly six times more likely to eat chili peppers than controls.
• The OR = 5.95 also suggests that the odds of gastric cancer is nearly six-fold higher in those consuming chili peppers than non-consumers.
• These results suggest that eating chili peppers may be a risk factor for the development of gastric cancer.
• Chili peppers contain the hot spice capsaicin which irritates the gastric mucosa.
• Chronic consumption of chili peppers may therefore increase the risk of developing gastric carcinoma.

Young women's cancers resulting from utero exposure to diethylstilbestrol
Green tea consumption and lung cancer
Maternal anesthesia and development of fetal birth defects
Passive smoking at home and risk of acute myocardial infarction
Household antibiotic use and antibiotic resistant pneumococcal infection

Tend to use smaller sample sizes than surveys or prospective studies
Quick and easy to complete
Cost effective
Useful for studies of rare diseases

Smaller—hundreds to thousands vs. thousands to tens of thousands! Increases the likelihood that a case-control study will be repeated—this is desirable, as it provides consistency in epidemiological research.
Much better for studying rare diseases than cohorts.

Unclear temporal relationships between exposures and diseases
Use of indirect estimate of risk
Representativeness of cases and controls often unknown
Refers to the source of data, e.g., vital statistics, case registries, physicians' records, surveys of the general population, or hospital and clinic cases.
The nature of the data will affect the types of statistical analyses and inferences that are possible—so it will affect the usefulness.

Are the sources vital statistics (death, birth, etc.), case registries, doctors' records, general population surveys, or cases from hospitals/clinics?

Refers to investigator's access to data.
For example, medical records and other data with personal identifiers may not be used without patients' consent. (HIPAA)
Data Perturbation

If the data has been stripped of all identifying characteristics, it can be used.

The process of modifying the identifying characteristics is called "data perturbation".

Insurance Data
Sources include:
Social Security--provides data on disability benefits and Medicare.
Health insurance--provides data on those who receive care through a prepaid medical program.
Life insurance--provides information on causes of mortality; also provides results of physical examinations.
All of these are examples of insurance data that have been widely used in epidemiologic studies.
Limitation of Insurance Data: Data may not be representative of entire population, as the uninsured are excluded.

Clinical Data Sources
Hospital data
Diseases treated in special clinics and hospitals
Data from physicians' practices
Results from clinical laboratories affiliated with clinical sites.
Remember that hospital cancer registries forward information to the SEER database (Conducted by the National Cancer Institute (NCI)). Using clinical data presents challenges because of the confidential nature of most of it, a lack of standardized recordkeeping and diagnostic procedures, etc.

Hospital Data
Consists of both inpatient and outpatient data
Deficiencies of data:
Not representative of any specific population
Different information collected on each patient
Settings may differ according to social class of patients; e.g., specialized clinics, emergency rooms

Deficiencies: Individuals do not represent any specific population (the denominator is undefined)—they may come from all over a large metro area or even from other countries. The lack of standardization in recordkeeping and in diagnostic procedures can cause problem with this data as well.

Socioeconomic data may be useless, too, because you may have different levels of socioeconomic status being seen in different areas—specialized clinics and renowned hospitals in urban areas vs. hospital ERs and outpatient clinics (low SES will use these more often as their primary source of medical care).

A Positive: Electronic Health Records should facilitate the sharing of information among providers and enhance data standardization.

Data from Physicians' Practices
Limited application due to:
Confidentiality of patient data
Highly selected group of patients
Lack of standardization of information collected
Useful for the purposes of:
Verification of self-reports
Source of exposure data
Need written informed consent from patients to release confidential information.

Highly selected group of patients: Most patients at private doctor's offices are unrepresentative of the entire population because they generally have insurance or the means by which to pay for the services.

Because of the lack of standardization, the records are likely to be highly idiosyncratic documents that cannot be linked readily to other data sources!

Data from doctors' offices may be useful in analytic studies—such as finding a population of women at risk for breast cancer.....you could verify self-reports on questionnaires by using the medical records to help you exclude the women who had developed cancer previously. These records could also be used as a source of data for exposure to breast cancer agents, such as birth control pills—more complete data on age at first use, duration of use, and strength, etc. can be obtained reliably from medical records.

Absenteeism Data
Records of absenteeism from work or school
Possible deficiencies:
Data omits people who neither work nor attend school.
Not all people who are ill take time off.
Those absent are not necessarily ill.
Useful for the study of rapidly spreading conditions
Subject to a host of possible deficiencies!

Useful for the study of rapidly spreading conditions, such as respiratory disease outbreaks & epidemics of flus, etc.

Morbidity Data from the Armed Forces
Reports from physicals, hospitalizations, and selective service examinations
Data have been used for:
Studies of disease etiology.
Study of twins serving in Korean War or WWII to determine influence of "nature and nurture" on cause of disease.
Studies investigating genetic factors in obesity
Information on reported morbidity from active armed forces personnel and veterans.

Since the draft has been abolished, there are no more "selective service exams"---these physical exams are now done selectively to volunteers for military service—this makes the information from these exams less useful than before.

Overall, this type of data is useful for studying disease etiology.

With the study of obesity in twins, results suggested that the identical twins were more alike in various measures of obesity than the non-identical twins. This expectation was consistent with genetic influences.
Data cannot be generalized because patients are a highly selected group.
Case-control studies can be done with unusual and rare diseases.
However, it is not possible to determine incidence and prevalence rates without knowing the size of the denominator.

Data from these sources are not very useful because of these problems—however.......
An exception to the rule about special clinics and hospitals is......The Mayo Clinic in Rochester, Minnesota.

The Rochester Epidemiology Project, housed at the Mayo Clinic, uses a medical records-linkage system that has afforded access to details of medical care provided to the residents of Rochester and Olmsted County, Minnesota, since the early 1900's!! It has worked so well because this area is geographically isolated from other urban centers. This Project links all data about a specific patient to a unique Mayo identification number. The records of more than 5 million patients are kept in a central repository and tracked by computer bar codes. This project has successfully provided data and facilities to complete over 1,000 reports on the epidemiology of acute and chronic diseases.

The unique traits that make the Mayo Clinic a natural laboratory for population-based studies are the following:
Mayo has always provided primary, secondary, and tertiary care for the residents in Rochester, Minnesota—
They also offer care in every medical and surgical specialty and sub-specialty, so local residents don't have to leave the area for those services.
This has created a unique environment from which to study population-based disease causes and outcomes unlike any other area in the U.S. or world!
Privacy Act of 1974
Prohibits the release of confidential data (by a fed govt. agency or its contractor) without the consent of the individual
Freedom of Information Act
Mandates the release of government information to the public, except for personal and medical files
The Public Health Service Act
Protects confidentiality of information collected by some federal agencies, such as the National Center for Health Statistics (NCHS)

Privacy and confidentiality of certain information is legally protected.

Personally identifiable information includes information that a person has not given permission for release to public. Also covers information on whether a person participated in a study and any information that may identify a deceased person.


The HIPAA Privacy Rule

Refers to the Health Insurance Portability and Accountability Act of 1996
Sections of HIPAA "...require the Secretary of HHS to publicize standards for the electronic exchange, privacy and security of health information..."
Categories of protected health information pertain to individually identifiable data re:
The individual's physical and mental health
Provision of health care to the individual
Payment for provision of health care
HIPAA.......NOT HIPPA!!
Protects individually identifiable health information, among other things (insurance portability—transfer of insurance from one job to another, etc.).

PHI—Protected Health Information pertains to individually identifiable data. Many common identifiers such as name, address, date of birth, social security number. This information is typically demographic data related to:
-individual's past, present, or future mental or physical health
-provision of health care to the individual
-past, present, or future payment for provision of health care.


Data Sharing
The voluntary release of information by one investigator or institution to another for the purpose of scientific research.
Can enhance data quality and increase knowledge from research.
Key issue
The primary investigator's potential loss of control over information

The linkage of large data sets and the pooling of multiple studies that combine results from several research projects in order to enhance the quality of data and to increase knowledge from research.

Key issue: The primary investigator's potential loss of control over information—but because of the benefit to society, many researchers are willing to make their non-confidential data available to the research community.


Record Linkage

Joining data from two or more sources,
Employment records and mortality data.
Applications include genetic research, planning of health services, and chronic disease tracking.
Sources of Epidemiologic data
Table 5-1, pages 248 - 250
Sources include a range of information from vital statistics to absenteeism from work and/or school
Morbidity Surveys of the General Population
Morbidity surveys collect data on the health status of a population group.
Ones authorized by US govt. obtain more comprehensive information than would be available from routinely collected data
Example: National Health Interview Survey

Remember that Morbidity means illness/condition

These surveys use a scientifically designed representative sample of the population.

Their purpose: to determine the frequency of chronic and acute diseases and disability, collect measurements of bodily characteristics, conduct physical exams and laboratory tests, and explore other health-related problems that may be of concern to whoever is sponsoring the survey.

National Health Survey—established by The National Health Survey Act of 1956 in order to obtain information about the overall health status of the US population.


National Health Survey
Refers generically to a group of surveys and not a single survey.
In response to the Act, the National Center for Health Statistics (NCHS) conducts three separate and distinct programs.
National Health Interview Survey (NHIS)
Health Examination Survey (HES)
Various surveys of health resources
National Hospital Discharge Survey
National Ambulatory Medical Care Survey

National Health Interview Survey (NHIS)

General household health survey of the U.S. civilian non-institutionalized population
Studies a comprehensive range of conditions such as diseases, injuries, disabilities, and impairments

Health Examination Survey (HES)
Provides direct information about morbidity through examinations, measurements, and clinical tests
Identifies conditions previously unreported or undiagnosed
Provides information not previously available for a defined population
Now known as the Health and Nutrition Examination Survey (HANES)
http://wonder.cdc.gov/datasets.html
Other healthcare surveys conducted as part of the National Health Survey include: National Hospital Discharge Survey, National Nursing Home Survey, National Ambulatory Medical Care Survey. In addition, there are also these vital statistics surveys: National Natality Survey, National Fetal Mortality Survey, National Mortality Followback Survey.

The CDC is the center for the data collected from these surveys and other sources—CDC uses the CDC WONDER online databases to release this data. The link is listed above.

Behavioral Risk Factor Surveillance System (BRFSS)
Collects data on behaviorally related phenomena
Behavioral risks for chronic diseases
Preventive activities
Healthcare utilization
The largest telephone survey in the world

Established in 1984 by CDC, BRFSS collects data from all 50 states plus DC, Puerto Rico, U.S. Virgin Islands, and Guam.

Used for evaluating public health policies/activities, and for providing quantitative support for state-based legislative initiatives.

California Health Interview Survey (CHIS)
Provides information on the health and demographic characteristics of California residents
Uses telephone survey methods
Topics include
Physical and mental health conditions
Health behaviors
Health insurance coverage and utilization
Conducted on a continuing basis
http://www.chis.ucla.edu/about.html
Begun in 2001, CHIS has surveyed 42,000 to 56,000 randomly selected households in each wave of collection. A full data collection cycle takes 2 years to complete! It is conducted on a continuing basis so that annual estimates can be made.

Data is available for public use by accessing the AskCHIS option on the website.
Mortality data are nearly complete
Most deaths in the U.S. and other developed countries are unlikely to be unreported.
Death certificates include demographic information about the deceased, cause of death (immediate cause & contributing factors), as well as attending doctor & date of death.
Medical Examiner or Coroner will complete if death is due to accident, suicide, or homicide or if attending is unavailable.

Certification of cause of death: Diabetes may not be given as the cause of death—may be heart failure or pneumonia, which could be complications of Diabetes.
There can be errors in coding the deaths—a nosologist (a person who classifies diseases) reviews the death certificate and codes it. Errors are minimized by standardized training and routine audits.
Codes change over time—we are using ICD-10, but it was the 6th edition in 1948 when WHO took over coding. ICD-10 is the International Statistical Classification of Diseases & Related Health Problems.

Sudden increases or decreases in a particular cause of death may be due to changes in coding. **To avoid this error, be very careful when using data that spans more than one version of the ICD—interpret results carefully because codes and groupings of diseases may have changed.

Doctors may be reluctant to put AIDS or Alcoholism as the cause of death, especially if they are friends of the deceased.

Lack of standardization of diagnostic criteria.
Stigma associated with certain diseases
AIDS or alcoholism, may lead to inaccurate reporting.


Birth Statistics: Certificates of Birth and of Fetal Death
Birth certificate includes information that may affect the neonate, such as congenital malformations, birth weight, and length of gestation.
Sources of unreliability:
Mothers' recall of events during pregnancy may be inaccurate.
Conditions that affect neonate may not be present at birth.
Needed to calculate birth rates

Also collects information about the conditions listed on the slide.

Keep in mind that some of this data collected may not be reliable due to the mom's recall of events....plus some conditions may not be present at birth.
Varying state requirements for fetal death certificates.
Both types of certificates have been used in studies of environmental influences upon congenital malformations.
Have also been used successfully to search for cluster of birth defects where chemical/biological exposures have occurred.
Varying state requirements can also reduce the usefulness of fetal death certificates.

The same things that affect accuracy of birth certificates (mom's recall, etc.) can affect accuracy of certificates of fetal death certificates.


Reportable Disease Statistics
Federal and state statutes require health care providers to report those cases of diseases classified as reportable and notifiable.
Include infectious and communicable diseases that endanger a population, e.g., STDs, measles, foodborne illness.
Certain diseases and conditions are notifiable/reportable to health agencies at the local, state, and federal level. This is public health surveillance!!



Notice the difference in time frames of required reporting from Class A to Class B, then from Class B to Class C, etc.

Reporting chain is Local health department, State, CDC, then WHO (for international diseases for which quarantine is needed).

CDC collects information from 4 sources: labs across the US, flu data from 121 cities, sentinel physicians (150 Family Practice Docs), and state epidemiologists.

Sentinel physicians—doctors who raise the alarm about unusual or untimely cases of illnesses, disease, disability, or deaths whose occurrences may be a warning of potential outbreaks.
Limitations of Reportable Disease Statistics
Possible incompleteness of population coverage.
For example, not everyone will seek treatment, esp. if little to no symptoms.
Failure of physician to fill out required forms.
Unwillingness to report cases that carry a social stigma (HIV, AIDS, etc.).
Failure of physician to fill out required forms.
Docs, NPs, have to keep current with reporting requirements.

Also, things that are widespread but less dramatic than some of the others on the list sometimes go unreported. The more severe—at least with symptoms that are severe—will usually be reported.