Epidemiology Lectures 1-4 Objectives – Flashcards

In the 1900s, the leading causes of death were communicable diseases such as pneumonia, tuberculosis, diarrheal disease and enteritis, and non-communicable diseases such as heart disease, liver disease, injuries, cancer, and senility In the 2000s, the leading causes of death made an epidemiologic shift and are now heart disease, cancer, stroke, COPD, unintentional injuries, diabetes mellitus, influenza & pneumonia, Alzheimer's disease, and kidney disease Actual causes of death are tobacco use, poor diet and physical activity, microbial agents, toxic agents, motor vehicle accidents, firearms, sexual behavior, and illicit drug use

Epidemiology is the study of the occurrence, distribution, and determinants of health-related states or events (diseases, morbidity, injuries, disability, and mortality) in populations -Epidemiologic studies are applied to the control of health problems in populations -Key aspects of this definition are determinants, distribution, population, and health phenomena (morbidity and mortality) Epidemiology derives from the word epidemic -epi (upon) + demos (people) + olody (study of)

Epidemiology can be traced back to 2000 years ago: -Classical Greeks in 4th/5th century BC traced disease causality to environmental factors such as: wrath of gods, breakdown of mortality, weather, and bad air -400 B.C., Hippocrates: Lifestyle and environmental factors affect human disease - 1600s, Bacon: Principles of inductive logic developed forming philosophical basis for epidemiology -1662, Graunt: London births and deaths analyzed for population 1747, Lind: First experimental study - treatment of scurvy 1839, Farr: System developed for routine summary for causes of death -1849-1854, John Snow: Analytic epidemiology used to form and test hypothesis for origin of cholera 1912, Lane-Claypon: First historical cohort study was done on the benefits of breast feeding 1949, National Heart, Lung, and Blood Institute: First population-based cohort study was the Framingham Heart Study 1950s, Doll ; Hill: First case control study linked cigarette smoking to lung cancer

1. Study the history of the health of populations 2. Diagnose the health of the community 3. Examine the working of health services 4. Estimte individual risks and chances 5. Identify syndromes 6. Complete the clinical picture 7. Search for causes

Descriptive Epidemiology: describes the distribution of disease, health-related states, and risk factors (person, place, time) 1. Study history of health of populations 2. Diagnose the health of the community 3. Examine the working of health services Analytical Epidemiology: Seems to ultimately determine causal, aka etiological, relationships -Applications include: 4. Defines "individual risks" (i.e., measurements that are grouped by characteristics of individuals, like age, gender, etc) 5. Search for causes 6. Identify clinical concerns (i.e., syndromes) 7. Complete the clinical picture

- Human disease does not occur at random - Human disease has causal and preventive factors that can be identified through systematic investigation of different populations or subgroups of individuals within a population

Cause: An event, condition, or characteristic without which the disease would not have occurred Risk Factor: Exposure that is associated with a disease -Due to the uncertainty of "causal" factors the term risk factor is used -Example: smoking Criteria (3): 1. The frequency of the disease varies by category or value of the factor, e.g., light smokers vs. heavy smokers. 2. The risk factor precedes onset of the disease. 3. The observation must not be due to error.

Hill's Causal Criteria: *temporality (or time sequence): exposure precedes development of disease by a period consistent with proposed biologic mechanism - sufficient induction / latent period exists -ESSENTIAL *strength of association: the stronger the association the more likely the exposure-disease relationship is causal -a strong association is less likely than a weak association to be due to confounding bias -IMPORTANT *biologic gradient (or dose response): strength of association increases with intensity or duration of exposure, as predicted -example: examples: linear, threshold, curvilinear -IMPORTANT *consistency: other studies using different populations and methodology show similar results -IMPORTANT *plausibility: there are known or postulated biological mechanisms that help explain exposure-disease relationship -example: examples: contaminated water / cholera, cigarette smoke / lung cancer, promiscuity / AIDS -IMPORTANT coherence: association must not seriously conflict with what is already known about natural history or biology of disease specificity: a specific exposure is associated with only one disease -this may not be applicable to chronic diseases since a specific exposure may be associated with several diseases analogy: evidence exists that similar exposures may have similar effects -example: heavy metals and toxic effects (lead and mercury) experiment: natural experiments provide evidence for exposure-disease relationship -example: communities with naturally fluoridated water / dental caries -rarely available for human populations

Ratio: Value obtained by dividing one quantity by another -RATE, PROPORTION, and PERCENTAGE are types of ratios -consists of a numerator and a denominator -most general form has no specified relationship between numbers -example: x/y, female/male Proportion: A type of ratio in which the numerator is part of the denominator; states a count relative to a size of the group (%) -can demonstrate magnitude of a problem -10 sick students in a dorm of 20 students: 50% are sick Rate: Ratio that consists of a numerator and a denominator and in which time forms part of the denominator -contains: disease frequency, unit size of population, and time period

The population at risk for whatever disease is being measured

Crude birth rates=# of live births within a given period/population size at the middle of that period x1,000 population Infant mortality rate=number of deaths among infants aged 1-365 days during the year/number of live births during the year x1,000 live births Maternal Mortality Rate=# of deaths assigned to cases related to childbirth/# of live births x1,000 live births (during a year)

Prevalence: The # of existing cases of disease or health condition in a population at some designated time -provides an indication of the extent of a health problem -example: prevalence of diarrhea in children at camp on 7/13 was 15. -describes the burden of a health problem -estimating the frequency of an exposure -determining allocation of health resources such as facility and personnel Incidence rate: Describes the rate of development of a disease in a group over a certain period of time -Idea as that the group of people are under study for the same amount of time -contains 3 elements: numerator (# of new cases, denominator (# of new cases), time (period during which the cases occur) -calculation: (# of new cases over a time period)/(total population at risk during the same period) x multiplier -multiplier makes the # meaningful THIS IS A MEASURE OF RISK!!! Prevalence=IxD Therefore, if duration of a disease is long and the incidence is small, prevalence increases greatly relative to incidence

Crude rate=unadjusted rate -use crude rates with caution when comparing disease frequencies between populations -observed differences in crude rates may be the result of systematic factors (sex or age) within the population rather then the true variation rate -If there is an underlying difference in characteristic related to death, will try and adjust that. When there is an underlying difference regards to the rate, you need to adjust. Adjusted rates: Summary measures of the rate of morbidity and mortality in a population in which statistical procedures have been applied to remove the effect of differences in composition of various populations. Specific rate=refer to a particular subgroup of the population defined in terms of any characteristic such as race, age, or single cause of death/illness -stratum causes specific rate -any way of categorizing by specific shared characteristics -example: brown hair -Cause-specific rate= (mortalty or frequency of a given disease) / (population size at midpoint of time period) x 100,000 -example: cause-specific mortality rates in the age group of 25-34 due to HIV in 2004 was X PMR (proportional mortality ratio)= number of deaths in a population due to a specific disease or cause divided by the total number of deaths in the population -Age-specific rate= (Number of deaths among those aged 5-14 years) / (# of people who are aged 5-14 years) (per time period) -The ) of case/age group of a pop during a specific time period -Adjusted rates summary measures of the rate of morbidity and mortality in a population in which statistical procedures have been applied to remove the effect of difference in composition of various populations

Direct method: To remove bias due to age... 1. Select a reference population. 2. Multiply the age-specific disease rates of the comparison population by the age distribution of the reference population. 3. Sum the "expected" number of cases of disease for the comparison population. 4. Divide the total "expected" number of cases of disease for the comparison population by the total number of people in the reference population. Indirect Method: Observed/Expected x 100 -The standardized mortality ratio (SMR) can be used to evaluate the results of the indirect method.

Direct: The direct method may be used if age-specific death rates in a population to be standardized are known and a suitable standard population is available. Indirect: Indirect method may be used if age-specific death rates of the population for standardization are unknown or unstable, for example, because the rates to be standardized are based on a small population.

Descriptive studies: used to identify a health problem that may exist. Characterize the amount and distribution of disease Analytical studies: follow descriptive studies, and are used to identify the cause of the health problem

To evaluate and compare trends in health and disease To provide a basis for planning, provision, and evaluation of health services To identify problems for analytic studies (creation of hypotheses)

Can provide general characteristics of the frequency of disease in a population, particularly by ... Person: Who has the lowest / highest disease rates in the population? -gender -socioeconomic status -race Place: Where are the lowest / highest disease rates for a population? Where do they live? -population density -migration -latitude / longitude -geographic unit state county census tract Time: How does this change over time? -short-term trends -long-term or secular trends -cyclic trends age, period, and birth cohort effects

1. Nature of the data -Refers to the source of data, e.g., vital statistics, case registries, physicians' records, surveys of the general population, or hospital and clinic cases. Will affect the types of statistical analyses and inferences that are possible. 2. Availability of the data -Refers to investigator's access to data. For example, medical records and other data with personal identifiers may not be used without patients' consent. 3. Completeness of population coverage -Representativeness: the degree to which a sample resembles a parent population. -Generalizability (external validity): (external validity)— ability to apply findings to a population that did not participate in the study. -Thoroughness: the care taken to identify all cases of a given disease. 4. Strengths versus limitations The utility of the data for various types of epidemiologic research. -Factors inherent in the data may limit their usefulness. -Incomplete diagnostic information. Case duplication.

Confidentiality: 1. Privacy Act of 1974 Prohibits the release of confidential data without the consent of the individual 2l Freedom of Information Act Mandates the release of government information to the public, except for personal and medical files 3. The Public Health Service Act Protects confidentiality of information collected by some federal agencies, e.g., NCHS The HIPPA Privacy Rule 1. Refers to the Health Insurance Portability and Accountability Act of 1996 2. Sections of HIPAA "...require the 3. Secretary of HHS to publicize standards for the electronic exchange, privacy and security of health information..." 4. 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

Vital Registration System: -Mortality Statistics: 1. Mortality data are nearly complete, as most deaths in the U.S. and other developed countries are unlikely to be unreported. 2. Death certificates include demographic information about the deceased and cause of death (immediate cause and contributing factors). Limitations: 1. Certification of cause of death. For example, in an elderly person with chronic illness, exact cause of death may be unclear. 2. Lack of standardization of diagnostic criteria. 3. Stigma associated with certain diseases, e.g., AIDS, may lead to inaccurate reporting. 4. Errors in coding by nosologist 5. Changes in coding Revisions in the (ICD) International -Classification of Disease. -Sudden increases or decreases in a particular cause of death may be due to changes in coding. -Birth Statistics 1. Birth certificate includes information that may affect the neonate, such as congenital malformations, birth weight, and length of gestation. 2. Sources of unreliability: Mothers' recall of events during pregnancy may be inaccurate. Conditions that affect neonate may not be present at birth. 3. Varying state requirements for fetal death certificates. 4. Both types of certificates have been used in studies of environmental influences upon congenital malformations. 5. Both provide nearly complete data.

Surveillance: Public health surveillance is the ongoing collection and timely analysis, interpretation, and communication of health information for public health action

Passive: 1. voluntary or mandatory reporting of disease cases 2. no outreach to identify missed disease cases 3. provider-initiated -responsibility for reporting falls on health care provider, lab or district health officer -example: reporting of common infectious diseases Active: 1. actively seeking cases to ensure completeness 2. extensive outreach to identify missed disease cases 3. surveys (providers, hospitals or defined population) 4. provider- or health department-initiated

Planning Steps 1. identify a health problem 2. develop case definition 3. collect data 4. analyze data 5. interpret and disseminate results 6. evaluate and improve surveillance system

Population Sampling: Method by which subjects in a given population are selected as representatives of the entire population Types: 1. non-probability samples members of the population are selected in some non-random manner -convenience sampling method for selecting study subjects is convenient for study coordinator, e.g., shoppers at local mall -judgment sampling method for selecting study subjects based on investigator's perception that sampled persons will be representative of entire population, e.g., choosing one neighborhood even though target population includes all neighborhoods -quota sampling method for selecting subjects based on subgroups in the population, e.g., gender and age, and on proportions within each subgroup, e.g., 200 females and 300 males between ages 45 and 60 - snowball sampling referrals from initial subjects are used to identify additional subjects, e.g., illicit drug users, STD cases 2. probability samples every member of the population has an equal probability of being included in the sample - Simple random sampling: n subjects are randomly chosen without replacement from a population of N subjects, each subject in population has an equal chance of being chosen in sample, requires enumeration of all potential subjects before sampling -Systematic Sampling: often used instead of simple random sampling, called an nth name selection technique, as good as simple random sample as long as sampling frame does not contain any hidden order -Stratified random sampling: used to ensure representation of specific subgroups of the population in overall sample, often used to reduce sampling variation, individuals are chosen such that proportion of individuals from given stratum in a sample is the same as proportion of individuals in the population (referred to as proportional allocation), subgroups may be over-sampled to provide sufficient sample size for sub-analyses (referred to as disproportionate stratified sampling) -Cluster Sampling: used when sampling unit is a cluster of subjects, cluster may include individuals from families, schools, counties, census tracts, etc., all else being equal, a large number of small clusters is preferable to a small number of large clusters -Multistage Sampling: Multi-stage sampling is combination of above sampling schemes First stage define primary sampling units (PSU), e.g., counties -select probability sample of PSUs, using any sampling scheme Second Stage define secondary sampling units (SSU), e.g., census tracts -select probability sample of SSUs in each PSU, using any sampling scheme Third stage define tertiary sampling units (TSU), e.g., households -select a sample of TSUs within each SSU, using any sampling scheme Fourth Stage select all subjects in each TSU or select one subject in each TSU