STAT 6395 Special Topic in Statistics: Epidemiology Spring, 2008 Filardo and Ng, 2008.

STAT 6395

Special Topic in Statistics:

Epidemiology

Spring , 2008

Filardo and Ng, 2008

I. Epidemiology

The study of the distribution and determinants of health-

related states or events in specified populations and the

translation of study results to control health problems

Distribution

• Persons affected

• Place

• Time

Epidemiology: The study of the distribution and determinants of health-related states or events in specified populations and the translation of study results to control of health problems

Determinants

All the physical, biological, social, cultural, and behavioral

factors that influence health


Health-related states or events• Diseases • Mortality (death)• Specific causes of death• Injuries• Disability• Health-related behaviors• Physiological measurements• Results of preventive regimens• Clinical outcomes• Provision and use of health services


Specified populations

• Residents of a defined geographic area

• Students who attend a specified school

• Persons who belong to a specified organization

• Workers at a specified workplace


Translation

Study Results


Scientific articles and presentations at meetings

Prevention programs

Quality of care improvement programs

Patient safety programs

Clinical guidelines

Control

Operations or programs aimed at reducing the adverse impact of the disease on the community

1) Prevention2) Cure3) Management


Primary prevention

An action taken to prevent the development of

a disease in a person who is well and does

not have the disease in question

Operations or programs aimed at reducing the adverse impact of the disease on the community: 1) Prevention; 2) Cure; 3) Management

Secondary prevention (Cure and management)

The identification and treatment of people

who have already developed a disease or

precursors of the disease, through screening,

at an early enough stage in the disease’s

natural history (early detection) such that

intervention will be more effective than if the

disease had been discovered later

Operations or programs aimed at reducing the adverse impact of the disease on the community: 1) Prevention; 2) Cure; 3) Management

II. Objectives of Epidemiology

• To describe the extent of disease in the community

• To identify risk factors (factors that influence a person’s risk of acquiring a disease) for disease and the etiology or cause of disease

• To study the natural history (course from onset to resolution) and prognosis of disease

• To evaluate both existing and new preventive and therapeutic measures (including health care delivery)

• To provide the foundation for developing public policy and regulatory decisions relating to environmental problems

…bottom line, Epidemiology research requires a multidisciplinary effort and statisticians play a key role in:

• Hypothesis development

• Study execution

Hypothesis development

Epidemiologists study the specific distribution and determinants of specific diseases

Development of hypotheses to test in an epidemiologic study requires in-depth knowledge of the disease and determinants under study

• Medical and biological sciences

• Social and behavioral sciences

Epidemiology Research hypothesis development and study execution

Study execution

• Statistics ------------>Biostatistics

• Medical and biological sciences

• Social and behavioral sciences

Epidemiology Research hypothesis development and study execution

Biostatistics

• Separate chance observations from meaningful observations

• Sampling

• Sophisticated statistical analyses

Epidemiology Research Biostatistics, medical and biological sciences, social and behavioral sciences

Medical and biological sciences

• Microbiology, e.g., to identify infectious agents

• Clinical medicine and pathology, (e.g., to identify cases of disease)

• Molecular biology, (e.g., to identify genotype of individuals)

• Biochemistry, (e.g., to measure serum hormone levels)


Social and behavioral science

• Design questionnaires for obtaining valid information

• Design effective interventions for lifestyle changes


Epidemiology Sub-disciplines

• Disease-specific subject matter

• Exposure-specific subject matter

• Health services research

• Infectious disease epidemiology• HIV/AIDS epidemiology• Malaria epidemiology

• Chronic disease epidemiology• Cancer epidemiology• Cardiovascular epidemiology• Perinatal epidemiology• Neuroepidemiology• Psychiatric epidemiology

Epidemiology disciplines Disease-specific subject matter, Exposure-specific subject matter, Health services research

Disease-specific subject matter

• Environmental epidemiology

• Occupational epidemiology

• Pyschosocial epidemiology

• Genetic epidemiology

• Nutritional epidemiology

• Pharmacoepidemiology


Determinant (exposure)-specific subject matter

• Operations research: the study of the placement of health services in the community and the optimum utilization of such services

• Program evaluation

• Clinical epidemiology or outcomes research: the study of illness outcomes in persons seen by providers of health care; evaluation of medical treatments


Health services research

Types of epidemiologic studies

• Experimental

• Observational (this course will focus on this second type of studies)

Experimental Studies

• Studies in which conditions are under the control of the investigator(s).

• The investigators assigns subjects to different study groups. The effect of the treatment is determined by comparing the outcome of interest in these groups.

Type of studies Experimental, Observational

Experimental studies (examples)

• Randomized clinical trial (unit of study is the individual)

• Community trial (unit of study is the community)


Observational studies

• Studies in which the investigators does not control conditions, but rather observe nature taking its course by gathering information, recording, classifying, counting, and analyzing the collected data.

• Exposure and disease outcomes would have occurred whether or not the studies have been performed because there were no ‘a priori’ intervention(s) on the part of the investigators.


Observational studies (examples)

• Descriptive

• Analytic


Descriptive studies

Studies aimed at describing the distribution of disease or other health-related variables with respect to person (age, gender, race, socioeconomic status), place (census tract, county, state, country, urban/rural), and time (season, year)

Type of studies Experimental, Observational (descriptive)

Descriptive studies

• Often use routinely-collected data

• Can define high-risk groups

• Can be used for hypothesis generation, but generally not hypothesis testing

Type of studies Experimental, Observational (descriptive)

Analytic studies

• Test specific etiologic hypotheses

• To generate new etiologic hypotheses

• To suggest mechanisms of causation

• To generate preventive hypotheses

• To suggest or identify potential methods for disease prevention

Type of studies Experimental, Observational (analytic)

In these studies, the epidemiologist observes the relationship between an exposure and a disease or other health outcome.

Definition: Exposure

A potential causal agent or characteristic, such as

infectious agent, behavior, dietary factor,

medication, medical treatment, genetic makeup,

environmental agent, or physiologic state (e.g.,

serum level of a hormone or nutrient; blood

pressure).

Type of studies Observational Analytic

An exposure may be harmful or beneficial

Types of analytic studies

• Cohort studies

• Case-control studies

Type of studies Observational

Cohort studies

A study in which a group of

persons exposed to a factor

of interest and a group of

persons not exposed are

followed

Type of studies Observational Cohort studies

and compared with respect

to the incidence rate of the

disease or other condition of

interest

Time

Cohort studies (study schema)

Type of studies Observational Cohort studies

Case-Control studies

Type of studies Observational Case-Control

Studies in which a group of persons with a disease

(cases) and a comparison group of persons without

the disease (controls) are compared with respect to

the history of past exposures to factors of interest

PresentPast

Case-Control studies (study schema)

Type of studies Observational Case-Control

Time

Either descriptive or analytic studies

Type of studies Observational

• Cross-sectional studies

• Ecologic studies

Cross-sectional studies

Studies of the distribution of exposures and/or disease in a defined population at one given point in time

Type of studies Observational Cross-sectional

Ecologic studies

Studies of the association between exposures and disease in which the units of analysis are populations or groups of people, rather than individuals.

This involves the assessment of the correlation of exposure rates and disease rates among different groups or populations.

Causality, though???

Type of studies Observational Ecologic

Types of Epidemiologic studies

Example involving several types of study designs

Ecologic study results

Case-Control study results

Cohort study results

Randomized controlled trial results

Which type of study is the ‘Gold Standard’ and/or more common/feasible?

Experimental

Efficacy

controlled setting

(difficult to reproduce in real life)

Observational

Effectiveness

real life setting

Focus of this course is on

observational Epidemiologic research(research regarding the direct study of disease in human populations)

Some triumphs of observational Epidemiology

• Smoking causes lung cancer

• Identification of cardiovascular disease risk factors

• Characterization of how HIV spreads through a population

• Identification of occupational hazards (e.g., asbestos)

Three ‘Eras’ of Epidemiology

• Sanitary (1800-1875)

• Infectious disease (1875-1950)

• Chronic disease (1950-present)

Sanitary ‘Era’ (1800-1875)• Miasma theory of disease – poisoning by foul emanations (miasma) from the

soil, water, and air.

• Created national vital statistics systems: much valuable descriptive epidemiology

• Demonstrated clustering of disease in slums and among the poor

• Solutions – sewage systems, drainage, clean water supplies, garbage collection, decent housing

• Incorrect miasma theory, but solutions were a major contribution to public health

Sanitary, Infectious disease, Chronic disease

Lesson: prevention doesn’t necessarily require understanding of cause

Infectious disease ‘Era’ (1875-1950)

• Germ theory: single microscopic agents relate one-to-one to specific diseases

• Epidemiology took a back seat to laboratory science, although in the1920s-30s, the germ theory was broadened to accommodate the interactive roles of host (immune and nutritional status), environment, and agent in infectious disease

• Other epidemiologic contributions:– Occupational exposures as causes of cancer– Specific vitamin deficiencies as causes of disease


Chronic Disease (Modern) ‘Era’

• By end of World War II, infectious diseases were under control in the developed countries; coronary heart disease and lung cancer were epidemic

• Multifactor causation of chronic disease

• Focus on identification and control of risk factors at the individual level

• Black box approach: emphasis on risk factor identification with only a secondary concern about mechanism or pathogenesis


Most of the methodology we will cover in this course was developed during this ‘Modern Era’.

New ‘Era’(???) • Emerging infectious diseases (e.g., HIV)

• Continued burden of infectious diseases in majority of world

• Traditional chronic disease epidemiology has hit a wall in its ability to discover important new risk factors

• Advances in molecular biology and genetics allow the study of pathogenesis and causality at the molecular and genetic levels using epidemiologic approaches

• The need to be concerned with causal pathways at multiple levels, including the societal level, as opposed to an exclusive focus on risk factors at the individual level, has become apparent to many

Sanitary, Infectious disease, Chronic disease, ???

Levels of causality

• Societal or population

• Individual

• Biochemical

• Cellular

• Molecular

What causes lung cancer?

Individual/societal: Cigarette smoking (nicotine addiction)

Biochemical: Specific chemicals in cigarette smoke that cause the mutations

Cellular: Specific phenotypic changes in the cells that result in loss of growth control

Molecular: Mutations in DNA

What causes AIDS?Societal:

PovertyProstitution

Individual/societal: a. Multiple sexual partners; b. Intravenous drug use

X Biochemical:

Cellular: Infection -the HIV viruses progressively destroy lymphocytes (a types of white blood cells)

Molecular: Mutations in DNA -the viral DNA is incorporated into the DNA of the infected lymphocyte

On the mode of communication of cholera

John Snow, M.D.London, 1855

John Snow 1857

Why study Snow?

• Appreciate those who came before us and paved the way

• Brilliant piece of work: lucid and thorough

• Snow’s work on cholera illustrates a key epidemiologic principle:

…the most important information to have about any communicable disease is its mode of communication

John Snow (1813-1858)

• Physician

• Pioneer in both epidemiology and anesthesiology

• Experiments in administration of anesthesia himself may have contributed to his early death

Snow administered chloroform to Queen Victoria for the birth of Prince Leopold and Princess Beatrice

Cholera: Acute Gastrointestinal Disease• Incubation period: 12-72 hours

• Sudden onset of severe vomiting

• Followed shortly by voluminous, watery, non-bloody diarrhea, described as rice water stool (white and opalescent)

• Abdominal cramps

• Severe cases: severe dehydration, circulatory collapse, renal failure (death may occur within a few hours of first symptoms)

• Case fatality rate may range from 1-50%, depending on strain of Vibrio cholerae and treatment

In 1817, four years after John Snow’s birth, cholera emerged from the Indian

subcontinent, where it had existed for centuries, to spread across the world.

Cholera Epidemics in Great Britain

• 1831-32– 56,000 deaths

• 1848-49– 125,000 deaths

• 1853-54– Cholera returns to England

Cholera: Apparently contradictory facts in 1854

• Local spread with evidence of direct communication from person to person

• Failure to spread to many in close contact with the sick

• Cases occur without traceable relation to prior cases

• Highest rates in low-lying areas and in filthy environments - Exceptions too numerous to be disregarded

• Unpredictability of its spread around the world and its geographic distribution

Cholera: Miasma Theory

Report of the London General Board of Health on cholera epidemic of 1848-49:

“…it appears as if some organic matter, which constitutes the essence of the epidemic, when brought in contact with other organic matter proceeding from living bodies, or from decomposition, has the power of so changing the condition of the latter as to impress it with poisonous qualities of a peculiar kind similar to its own.”

Cholera: Elaborations of Miasma theory

• Localizing influences

• Predisposition

• Spontaneous generation of “cholera poison”

• “Poison” spread by diffusion through the atmosphere vs. poison attached itself to solid bodies

• “Poison” communicated by an effluvium (contagion) given off by the sick

WilliamFarr

Farr’s elaboration of Miasma theory

• Soil at low elevations, especially near the banks of the Thames River, contained much organic matter that produced deadly miasmata.

• Miasmata diffused through the atmosphere in a cloud or mist

• Concentration of miasmata would be greater at lower elevations than in higher elevations, accounting for the geographic distribution in the London epidemic of 1849.

Farr’s theory did have some consistency with the facts.

John Snow’s Germ Theory

• Cholera caused by a germ cell, not yet identified

• 2 main modes of transmission of germ cell, which was found in the evacuations of cholera victims:

1. Drinking water contaminated with sewage2. Contaminated food, bedding, or clothing

Snow was firmly convinced of his theory by start of 1853-54 epidemic.

Snow’s ecologic observations prior to the 1853-54 epidemic

• Epidemics of cholera followed major routes of commerce and warfare.

• Cholera always appeared first at seaports, when extending to a new island or continent.

Cholera “has never appeared except where there has been ample opportunity for it to be conveyed by

human connections.”

Snow’s observations from ‘case’ histories

• Cholera can be communicated from the sick to the healthy.

• Persons attending those with cholera do not necessarily become afflicted.

• Close contact with a cholera patient is not necessary to become afflicted.

Snow’s conclusion: “…cholera is communicated from person to person,

but not through the air”

When cholera returned to London in August 1853, Snow had a definite hypothesis:

cholera was spread by contaminated water.

Water supply of south districts of London

• Until 1852, drew water from the Thames River in London, contaminated with London’s sewage

• In 1852, moved intake 22 miles up river and far from the contaminated water

• Continued to draw water from the contaminated Thames

Lambeth Water Company

Southwark and Vauxhall Water Company

Water Company

1849 Epidemic

1853-54 Epidemic

Lambeth Contaminated (London)

Pure (Up River)

Southwark & Vauxhall

Contaminated (London)

Contaminated (London)

Cholera Deaths in South Districts of London

Deaths/100,000

Water CompanyEpidemic

1849 1853

S&V, Lambeth 1290 61

S&V 1420 94

S&V, Kent 2050 107

Cholera deaths (per 100,000) in south districts of London

District Water supply 1849 1853

Bermondsey S&V 1610 150

St. Saviour S&V, Lam 1530 146

St. George S&V, Lam 1640 143

St. Olave S&V 1810 134

Rotherhithe S&V, Kent 2050 112

Newington S&V, Lam 1440 57

Wandsworth S&V, others 1000 51

Camberwell S&V, Lam 970 40

Lambeth Lam, S&V 1200 34

Notes: 1. Lambeth was supplied mostly by Lambeth Water Co. 2. Rotherhithe supplied partly by Kent in 1853

Cholera Deaths in south districts of London, 1853, sub-district analysis

Water Cholera Deaths/Supply Population Deaths 100,000

S&V 167,654 192 116

Lambeth 14,632 0 0

Both 301,149 182 60

How many cholera deaths would we expect in sub-districts supplied by Lambeth if they had the same death rate as those supplied by S&V?

Cholera Deaths in south districts of London, 1853, sub-district analysis

Water Cholera Deaths/Supply Population Deaths 100,000

S&V 167,654 192 116

Lambeth 14,632 0 0

Both 301,149 182 60

Expected deaths for Lambeth = (116/100,000)*14,632 = 16

Additional Observations on Sub-districts:

• Among the sub-districts supplied by both companies, those supplied mainly by Lambeth had a low death rate, while those supplied mainly by S&V had a high death rate.

• Two sub-districts supplied only by S&V also contained a number of pump-wells. These sub-districts had a low death rate.

By the return of cholera in July 1854, Snow recognized the full significance of the intermixing of the water supplies of the 2 companies

“…in the sub-districts … supplied by both Companies, the mixing of the supply is of the most intimate kind… A few houses are supplied by one Company and a few by the other, according to the decision of the owner or occupier at that time when the Water Companies were in active competition …”

By the return of cholera in July 1854, Snow recognized the full significance of the intermixing of the water supplies of the 2 companies

“…in many cases a single house has a supply different from that on either side. Each company supplies both rich and poor, both large houses and small; there is no difference either in the condition or occupation of the persons receiving the water of the different Companies.”

An Experiment on the Grandest Scale

“…no fewer than three hundred thousand people of both sexes, of every age and occupation, and of every rank and station, from gentlefolks down to the very poor, were divided into two groups without their choice …”

An Experiment on the Grandest Scale

“…one group being supplied with water containing the sewage of London,

…the other group having water quite free from such impurity”

What type of study John Snow conducted?

What type of study John Snow conducted?

Cohort Study

• A study in which a group of persons exposed to a factor of interest and a group of persons not exposed are followed and compared with respect to the incidence rate of the disease or other condition of interest.

• Exposed group: persons using S&V water supply

• Comparison group: persons using Lambeth water supply

Compared cholera mortality rates in the two groups

Calculation of mortality rates required numerators and denominators

Mortality rate in exposed group=

(Cholera deaths among persons supplied with S&V water) / (Number of persons supplied with S&V water)

Mortality rate in comparison group =

(Cholera deaths among persons supplied with Lambeth water) / (Number of persons supplied with Lambeth water)

To determine numerators and denominators, Snow needed a way to classify each death and every

person in the population by water supply

The Numerators

• For each cholera death in the relevant districts, Snow obtained information on the water supply

• “The inquiry was necessarily attended with a good deal of trouble.”

• Chemical test problematic because S&V water had 40 times more NaCl than Lambeth

The Denominators

• Therefore Snow had to include all the south districts of London in his study, not just the districts where the water supply was intermingled.

A daunting undertaking, so Snow obtained an assistant (Mr. Whiting)

“ …a return had been made to Parliament of the entire number of houses supplied with water by each of the Water Companies, but ... the number of houses which they supplied in particular districts was not stated…”

Cholera deaths in south London districts during first 4 weeks of 1854 Epidemic, by water supply

Deaths/Water Cholera 10,000Supply Deaths Houses Houses

S&V 286 40,046 71.4

Lambeth 14 26,107 5.4

Relative risk = 71.4/5.4 ≈ 13.2

Thames 22

Pump-wells 4

Ditches 4

Unknown 4

London 277 287,345 9.6

(-S&V)

Note: Houses, not persons, used in denominator

More data …

• Cholera likely was imported from Baltic Fleet to Thames River, which was initially the primary source of the epidemic

• Later, cholera was also spread by other means, diluting the water company effect

“ …as the epidemic advanced, the disproportion between the number of cases in houses supplied by the Southwark and Vauxhall Company and those supplied by the Lambeth Company, became not quite so great, although it continued very striking…”

Cholera deaths in south London districts during first 7 weeks of 1854 Epidemic, by water supply


S&V 1263 40,046 315

Lambeth 98 26,107 37

Relative risk = 315/37 ≈ 8.5

London 1422 256,423 59(-S&V, Lambeth)

Cholera deaths in south London districts during last 7 weeks of 1854 Epidemic, by water supply


S&V 2353 40,046 573

Lambeth 302 26,107 115

Relative risk = 573/115 ≈ 5

Cholera deaths in south London districts during the 1854 epidemic, by water supply

Water Cholera Deaths/Supply Deaths Population 10,000

S&V 4,093 266,516 153

Lambeth 461 173,748 26

Relative risk = 153/26 ≈ 5.9

London 10,367 2,362,236 43

Note: populations supplied by water companies estimated by

Registrar General.

Was the Use of Houses in the Denominators Valid?

Water Persons/Supply Population Houses Household

S&V 266,516 40,046 6.7

Lambeth 173,748 26,107 6.7

London 2,362,236 322,576 7.3

Cholera Deaths in South Districts of London, Sub-district Analysis, 1849 vs. 1854

Water Cholera DeathsSupply 1849 1854

S&V 2261 2458

Both 3905 2547

Lambeth 1644 89

Here we are back to ecologic analysis

Cholera Outbreak in the Golden Square Area of London, 8/31 - 9/9, 1854. Within 250 yards of the intersection of Cambridge and Broad Streets, there were more than 500 fatal cases

Snow immediately suspected contamination of the water of the much-used street pump on Broad Street near Cambridge Street.

Snow mapped the places of residence of cholera decedents from August 31 - September 2 in the broader neighborhood…

…and found that 73 of 83 deaths had taken place within a short distance of the pump.

Snow investigated the water source of the 73 decedents who lived near Broad Street pump

• 61 -- drank water from the pump

• 6 -- did not

• 6 -- could get no information

10 Deaths in houses located nearer to another street pump

• 5 always used the Broad Street pump, as they preferred its water

• 3 (children) went to school near the Broad Street pump

Handle of the pump was removed on September 8

• Legend has it that the removal of the pump handle

caused the end of the epidemic

• Snow himself wondered whether removing the pump

handle had a beneficial effect– Epidemic was already subsiding– Much of the population in the neighborhood had fled

“…it is impossible to decide whether the water from the pump still contained the cholera poison in an active state.”

Snow’s investigation of Golden Square Outbreak: workhouse near Broad St.

• Surrounded by houses in which deaths from cholera occurred

• Only 5 deaths among 535 inmates

• Workhouse had a pump-well on the premises

• Also received water from the Grand Junction Water Works

• Did not use Broad Street pump

• Would have expected more than 100 deaths based on mortality in surrounding streets

Snow’s investigation of Golden Square Outbreak: Brewery on Broad St.

• Located near the pump

• More than 70 workers

• None died of cholera

• Workers drank malt liquor, not water

• Deep well located in brewery

• Workers never obtained water from Broad Street Pump

Snow’s recommendations for prevention of cholera during an epidemic

• Observe strictest cleanliness around the sick.

• Wash linens of patients as soon as they are removed.

• Boil water for drinking and preparing food (unless known to come from clean source).

• Wash or heat to 212°F all food.

• Healthy should not live in same room as sick.

• Pit-men should work 4 hour shifts, and not eat in mines.

• Educate the people about communicability of cholera.

Snow’s recommendations for long-term prevention of cholera

• Effect good and perfect drainage.

• Provide water supply free from contamination with contents of sewers, cesspools, house-drains, and refuse of people who navigate the rivers.

• Provide model lodging-houses for the vagrant class and sufficient house room for the poor in general.

• Teach habits of personal and domestic cleanliness among the people.

• Screen persons arriving from infected places.

Medical establishment slow to catch on

• 1855 report of Scientific Committee for Scientific Enquiries in Relation to the Cholera Epidemic of 1854: “…on the whole of evidence, it seems impossible to doubt that the influences, which determine in mass the geographical distribution of cholera in London, belong less to the water than to the air.”

• 1856 Report on the last two cholera epidemics of London as affected by the consumption of impure water: “…under the specific influence which determines an epidemic period, fecalized drinking-water and fecalized air equally may breed and convey the poison.”

Sanitary ‘movement’ eventually succeeded in spite of its incorrect miasma theory

• Extensive improvements in several of London’s water supplies, including Southwark and Vauxhall, had already been ordered before the 1853-54 epidemic.

• In next London epidemic (1866), William Farr himself used epidemiology to show that the source of the epidemic was impure water from the East London Water Company.

What We Know Now?• Cholera caused by a bacterium (Vibrio cholerae; discovered

in 1883 by Robert Koch)

• Small intestine is primary site of infection

• Diarrhea caused by cholera toxin produced by Vibrio cholerae

• Treatment: intravenous or oral fluid and electrolytes, depending on severity of illness

• Environmental reservoir for Vibrio cholerae in the sea, where it lives on zooplankton and shellfish

Vibrio Cholerae

2005

STAT 6395 Special Topic in Statistics: Epidemiology Spring, 2008 Filardo and Ng, 2008.

Documents

health epidemiology

determinants of health

healthrelated states

translation of study

health services epidemiology

management slide

management epidemiology

control operations