2. CONTENTS
1. Introduction
2. Person Variables
3. Place Variables
4. Time Variables
5. Types of Descriptive Epidemiologic Studies
6. Uses of Descriptive Epidemiologic Studies
7. Summary
2
3. EPIDEMIOLOGY
The study of the distribution and determinants of health related states or events
in specified populations and the application of this study to the control of health
problems*
*John M Last (1988)
3
4. OBJECTIVES OF EPIDEMIOLOGY
1. Determine extent of disease
2. Identify etiology or cause of disease
3. Study natural history and prognosis of disease
4. Evaluate both existing and newly developed preventive and therapeutic measures
5. Provide foundation for developing public policy
4
5. TYPES OF EPIDEMIOLOGICAL STUDIES
1. Observational
a. Descriptive
b. Analytical i. Case-control
ii. Cohort
2. Experimental
a. Clinical trial(or Randomized controlled trial)
b. Field trial(or community intervention studies)
c. Community trial
5
7. CHARACTERISTICS OF DESCRIPTIVE
EPIDEMIOLOGY
• Relatively inexpensive
• Less time consuming
• First toe in water1
• Document the health of population
• Prompt more rigorous studies
• Describe the existing distribution of variables, without regard to causal or
other hypothesis
1. Hulley SB, Cummings SR, Browner WS, Grady D, Hearst N, Newman RB, eds. Designing clinical research: an epidemiologic approach, 2nd
edn. Baltimore: Lippincott Williams and Wilkins, 2001.
7
8. Descriptive Analytical
Used when little is known about the disease Used when insight about various aspects of
disease is available
Rely on pre existing data Rely on development of new data
Start - no explicit hypothesis regarding
cause-effect relationship
Start - definite hypothesis regarding an
exposure possibly causing an outcome
End - development of possible hypotheses
regarding cause and effect relationship but
does not confirm or reject such hypothesis
End - confirms or rejects the hypothesis with
which it started
8
10. PERSON
• AGE
• SEX
• RACE
• SOCIO ECONOMIC STATUS
• RESIDENCE
• MARITAL STATUS
• OCCUPATION
10
11. AGE
Death rate tend to be higher in
infant,pre-school age and in elderly
RTA have high mortality in
young/early middle age Bimodality in Hodgkins lymphoma
11
12. SEX, RACE, SOCIOECONOMIC STATUS
Thyroid and gall bladder
disease common in females
CHD and lung cancer in
males
Tay – sachs common in Jews
Indoor air pollution
Overcrowding leading to spread
of disease
12
16. GDM
Seshiah V, Balaji V, Balaji MS, et al. Prevalence of gestational diabetes mellitus in south India (Tamil Nadu): a community
based study. J Assoc Physicians India. 2008;56:329–333. 16
17. spot map of fatal cases, John Snow
Kangri cancer
Localised pattern of disease
17
18. TIME
• Common source epidemic
• Point source or single exposure
• Continuous or multiple exposure
• Propagated epidemic
• Seasonal or cyclical trend
• Secular trend
18
26. Case report
• Detailed description of disease occurrence in a single person
• Lowest level of evidence and considered base of pyramid
• Should explore and infer; not conform, deduce or prove
Advantages: Disadvantages
First line of evidence where new ideas
and issues emerge
Cases may not be generalizable
Identification of new trends of disease Cause and association may have other
explanations
Educational purpose Can be seen as focusing on misleading elements
Identifies rare presentation of cases No comparison made
Cannot demonstrate causality or argue for
adoption of a new treatment approach 26
27. E.g. Curious Case of Accidental Kerosene Aspiration in an Adolescent Girl and
Blue Colored Carbonated Soft Drinks
Babu TA. Curious case of accidental kerosene aspiration in an adolescent girl and blue colored carbonated soft drinks. Indian J
Community Med. 2015 Apr-Jun;40(2):139-40
27
28. Case series
• Describe a set of patients having similar symptoms and outcome
• Low level of evidence
Advantages Disadvantages
Identify new clinical issues
If retrospective the availability and accuracy of
the data records
Lead to development of hypothesis Selection bias
Feasibility, easy to conduct, less expensive No comparison group
Less time consuming The results not generalizable
28
30. Cross-sectional
• Relationship between diseases (or other health-related characteristics) and other
variables of interest as they exist in a defined population at a particular point in time
• Done on a sample of the total population
• Community or Hospital based
• Mainly directed to work out the :
i) Prevalence
ii) Mean of a factor of interest
iii) Pattern
iv) Surrogate for longitudinal descriptive studies
* textbook of public health and community medicine by dept. of community medicine,AFMC,Pune,in collaboration with
WHO,India office,New Delhi 30
32. Longitudinal Descriptive Studies
• Follow up a single group of subjects over a defined period of time
• More scientific but more costly and time consuming
• Objectives:
i) Incidence of a disease
ii) Natural history of a disease
iii) Describe a health related natural phenomena
iv) Trend of a disease
v) Trend of a health - related phenomena
* textbook of public health and community medicine by dept. of community medicine,AFMC,Pune,in collaboration with
WHO,India office,New Delhi
32
34. • Example of Cross-sectional study followed by longitudinal study
• There are three major types of longitudinal studies:
• Panel Study: Involves sampling a cross-section of individuals.
• Retrospective Study: Involves looking to the past by looking at historical information
such as medical records.
• Cohort Study: Involves selecting a group based on a specific event such as birth,
geographic location or historical experience.
34
35. Surveillance studies
• The ongoing systematic collection, analysis and interpretation of health data essential to
the planning, implementation and evaluation of public health practice, closely integrated
with the timely dissemination of these data to those who need to know.(CDC)
• Provide information to guide interventions
• Watchfulness over the community
• Key feature is feedback
* textbook of public health and community medicine by dept. of community medicine,AFMC,Pune,in collaboration with
WHO,India office,New Delhi
35
36. Types of surveillance:
• Passive
e.g. Malaria
• Active
e.g. Malaria
• Sentinel
e.g. HIV, Polio
36
37. • Uses of surveillance studies:
• Estimating the magnitude of problem
• Determining the geographic distribution of illness
• Portraying the natural history of disease
• Detecting epidemics or defining a problem
• Generating hypothesis, stimulating research
• Evaluating control measures
• Monitoring changes in infectious agents
• Detecting changes in health practices
• Facilitating planning
37
41. Ecological studies
It is an investigation of health and its determinants between group of
individuals
Unit of observation: group of people
Advantages:
• Data required is readily available
• Correlation coefficient:
Disadvantages:
• Ecological fallacy
• Effect of other factors on outcome cannot be determined
• Dose response relationships are masked
41
42. Rationale for ecologic studies
• Low cost and convenience
• Measurement limitations of individual level studies
• Design limitations of individual level studies
• Interest in ecologic effects
• Simplicity of analysis and presentation
42
43. The Correlation and Association between Short-term Exposure to Ambient Air Pollution and Case Fatality of SARS in People's
Republic of China.
*Cui Y et al.Environmental Health :A Global Access Science Source .2003;2:15 43
47. Types of ecologic studies
1. Purely ecological: no data on individual
2. Partially ecological: some individual data available
Types of Study designs:
1. Multiple group studies
2. Time trend studies
3. Mixed designs
47
48. The relation between male obesity and income inequalities in 21 rich countries
Pickett K.E. et al. Wider income gaps, wider waistbands? An ecological study of obesity and income inequality.
J Epidemiol Community Health. 2005 August; 59(8): 670–674
1.Multiple Group Study
48
49. Percentage of Children Receiving
Measles-Mumps-Rubella (MMR)
Immunization in Second Year of Life and
Caseload of Children With Autism, by
Year of Birth, California, 1980-1994
Dales L.et al. Time trends in autism and in MMR immunization
coverage in California.JAMA.March2001;285(9):1183-1185
2.Time Trend Study
49
50. Examples of ecological study
• Cholera study
(by John Snow,Golden Square,1854)
• UV radiation and cancer
• Diet and Alzheimer’s disease
(Total Fat and total Caloric supply were strongly correlated with
prevalence,while fish and cereals/grains were inversely correlated)
• UV radiation and Influenza
(A randomized controlled trial involving Japanese school children found that
taking 1000 IU per day vitamin D3 reduced the risk of type A influenza by
twothirds.)
50
51. Uses of descriptive studies
• Monitoring and reporting on the health status and health related behaviours in populations
• Identifying emerging health problems
• Alerting us to potential threats from bioterrorism
• Establishing public health priorities for a population
• Evaluating the effectiveness of intervention programmes
• Trend analysis
• Health care planning
• Hypothesis generation
51
52. Hypothesis formulation
• By examining characteristics of person, place and time
• Should specify:
• Population
• Specific cause
• Expected outcome
• Dose response relationship
• Time response relationship
• Differences
• Similarities
• Correlations
52
53. Advantages of descriptive studies:
• Data are already available
• Inexpensive
• Efficient to use
• Few ethical problems
Disadvantages:
• Absence of a clear, specific, and reproducible case definition
• Overstepping the data
53
54. SUMMARY
• Descriptive studies often represent the “first scientific toe in the water” in
new areas of inquiry.
• It answers: who , what,when and where.
• Case reports, case-series reports, cross-sectional studies, and surveillance
studies deal with individuals, whereas ecological studies examine
populations.
• Three important uses of descriptive studies include trend analysis, health-
care planning, and hypothesis generation.
54
Distribution:How is the disease?people,geographical area;increased/decreased
Disease frequency
Disease distribution
Determinants of disease
To describe the distribution and magnitude of health disease and disease problems in human and populations
To identify etiological factors/risk factors in the pathogenesis of disease
To provide the data essential to the planning, implementation and evaluation of services for the prevention, control and treatment of diseases and to the setting up of priorities among those services.
Descriptive studies are concerned with observing the distribution of disease or health-related characteristics in human populations and identifying the characteristics with which the disease in question seems to be associated.
Goal:
DS: describes a health related event
AS: Explain why a health related event occurs
Who- age,sex, race, occupation, or recreational activities
What- is the condition or disease being studied? Development of a clear, specific, and measurable case definition is an essential step in descriptive epidemiology.
When- Time
Where-Place
Why-cause
How-cause
They do not proceed to test a “pre - formed hypothesis”
The study is done only on one group of subjects; there is no comparison group
The main objective of a descriptive study is to
“describe” the “mean value” of a health related condition, or
“proportion” (i.e. incidence or prevalence of a disease/ health - related condition)or
natural history of disease/ a health - related phenomena; and
describe the “distribution” according to “person”, “Place” and “time”
Then it culminates by making suggestions / hypotheses about certain “cause and effect” relationships which can be further tested by analytical studies.
Age is strongly related to disease than any other single host factor.
Certain diseases are more frequent in certain age groups than in others, e.g., measles in childhood, cancer in middle age and atherosclerosis in old age
Kangri cancer is a type of squamous-cell carcinoma of the skin. It is found only in Kashmir in the northwest of the Indian subcontinent. It occurs on the lower abdomen and inner thighs and is due to the use of a kangri - a baked clay pot covered in wicker-work, used as a source for warmth by people in Kashmir during cold weather.
the pattern of disease may be described by the time of its occurrence,
It raises questions whether the disease is seasonal in occurrence; whether it shows periodic increase or decrease; or whether it follows a consistent time trend.
The exposure to the disease agent is brief and essentially simultaneous, the resultant cases all develop within one incubation period of the disease
the epidemic curve rises and falls rapidly, with no secondary waves
the epidemic tends to be explosive, there is clustering of cases within a narrow interval of time
all the cases develop within one incubation period of disease.
The Great Smog of 1952 or Big Smoke[1] was a severe air-pollution event that affected London during December 1952. A period of cold weather, combined with an anticyclone and windless conditions, collected airborne pollutants—mostly arising from the use of coal—to form a thick layer of smog over the city. It lasted from Friday 5 December to Tuesday 9 December 1952, and then dispersed quickly when the weather changed.
Although it caused major disruption due to the effect on visibility, and even penetrated indoor areas, it was not thought to be a significant event at the time, with London having experienced many smog events in the past, so-called "pea soupers". Government medical reports in the following weeks estimated that up until 8 December 4,000 people had died prematurely and 100,000 more were made ill because of the smog's effects on the human respiratory tract. More recent research suggests that the total number of fatalities was considerably greater, at about 12,000
Sometimes the exposure from the same source may be prolonged - continuous, repeated or intermittent - not necessarily at the same time or place.
The outbreak continued beyond the range of one incubation period.
From 1900 to 1907,mary mallon worked as a cook in New york city area of 7 families and presumed to have infected 51 individuals.
The Broad Street cholera outbreak was a severe outbreak of cholera that occurred near Broad Street in the Soho district of London, England in 1854 (in the area now known as Carnaby Street). This outbreak is best known for the physician John Snow's study of the outbreak and his hypothesis that contaminated water, not air, spread cholera.[1] This discovery came to influence public health and the construction of improved sanitation facilities beginning in the 19th century
The epidemic usually shows a gradual rise and tails off over a much longer period of time.
This is a propagated (progressive source) epidemic in which one or more of the first wave of cases serves as a source of infection for subsequent cases and those subsequent cases, in turn, serve as sources for later cases. The shape of the curve usually contains a series of successively larger peaks, which are one incubation apart in which an increasingly large number of cases caused by person-to-person contact, until the pool of susceptibles is exhausted or control measures are implemented.
Measles is usually at its height in early spring.
Bacterial Gastrointestinal infection are prominent during summer months
The seasonal variations of disease occurrence may be related to environmental conditions (e.g., temperature, humidity, rainfall, overcrowding, life cycle of vectors, etc.) which directly or indirectly favour disease transmission.
H1n1 greatest medical holocaust of history
Secular Trend: Gradual change in frequency of disease over period of time.
Other E.g.
Varied Presentation of Complicated Falciparum Malaria in a Family1
A Case Report of Plasmodium Vivax, Plasmodium Falciparum and Dengue Co-Infection in a 6 Months Pregnancy2
Cutaneous findings in five cases of malaria3
Sanklecha M, Mehta N, Bagban H. Varied presentation of complicated falciparum malaria in a family. Indian Pediatr. 2012 May;49(5):413-4
A Case Report of Plasmodium Vivax, Plasmodium Falciparum and Dengue Co-Infection in a 6 Months Pregnancy
Vaishnani JB. Cutaneous findings in five cases of malaria. Indian J Dermatol Venereol Leprol. 2011 Jan-Feb;77(1)
In Colombia, P. vivax as responsible for . of cases in the last fie years. This study pretends to describe the clinical and paraclinical profile of seere ia malaria cases hospitalied is a descriptie, retrospectie case-series study of seere malaria ia cases, hospitalied beteen and. Descriptie analyses of reason for consultation, sins and symptoms, dianosis, treatment, paraclinical characteristics, complications, and time hospitalied, ere conducted. Siteen cases of seere P. vivax ere analysed. Feer, chills and headache ere shon to be the main admission symptoms. Eleation of total bilirubin leels in ., and seere thrombocytopaenia in of cases ere the main complications presented durin hospitaliation. All cases responded to treatment, there ere no deaths. This is a descriptie, retrospectie case-series study of seere malaria ia cases, hospitalied beteen and . The data ere mined from medical records reised at four hospitals located in endemic areas of the Colombian states of Nario and Valle del Cauca, coerin the municipalities of Tumaco, Buenaentura and Cali.
Definition: cross-sectional studies examine the relationship between diseases(or other health-related characteristics) and other variables of interest as they exist in a defined population at a particular point in time(Last 2001).
They could be defined as ‘studies taking a snapshot of a society’.
Synonyms : prevalence, disease-frequency studies.
Such descriptive studies are done on a sample of the total population and may be community based or hospital based. They are mainly directed to work out the :
“Prevalence” of a factor of interest, e.g. prevalence of leprosy among general population, i.e. community based, or prevalence of nosocomial Pseudomonas infection in renal units, i.e. hospital based
“Mean” of a factor of interest (e.g. Hb% level in ladies in the community; or serum protein levels among patients admitted with open pulmonary TB)
Description of a “Pattern” (e.g. pattern of antibiotic prescription in an Acute Medical Ward; or knowledge, attitudes and practices regarding contraceptives in a community)
As a surrogate for longitudinal descriptive studies
Elimination of leprosy as a public health pThe National Leprosy Eradication Program (NLEP) was launched in 1983 with the World Health Organization (WHO) goal of eliminating the disease by 2000. Through a series of aggressive NLEP activities, Government of India declared elimination of leprosy on 31 December, 2005 as a public health problem with a national prevalence rate (PR) of 0.96.roblem was reached at the global level in the year 2000 and by India on 31 December, 2005
i) To see the incidence of a disease (e.g. the incidence of poliomyelitis among children in the community or the incidence of acute glomerulonephritis among children admitted in hospital, with acute sore throat)
ii) To describe the ‘natural history of a disease’ (e.g. the clinical progression of cases of AIDS in hospital; or the clinical progression of cases of viral hepatitis in a community)
iii) To describe a health related natural phenomena (e.g. the examples of weight charting of children given earlier)
iv) To study the ‘trend’ of a disease (e.g. to see whether the incidence of TB is rising or falling in a community)
v) To study the ‘trend of a health - related phenomena (e.g. to see whether the blood pressure of children from well - to - do families rises progressively; or whether the level of knowledge regarding prevention of AIDS is rising, falling or remaining stationary
Self-reported abnormal sensitivity, qualitative distortions and phantom sensations with respect to smell and taste was assessed with a longitudinal design, based on questions referring to gestational weeks 13-16 and 31-34 of pregnancy in comparison with 9-12 weeks post partum and with non-pregnant women with corresponding time durations and intervals. The results show that abnormal smell and/or taste perception was reported by 76% of the pregnant women, typically believed to be caused by their pregnancy. Increased smell sensitivity was found to be very common at the early stage of pregnancy (67% of all pregnant respondents) and occasionally accompanied by qualitative smell distortions (17%) and phantom smells (14%). The smell abnormalities were less common at the late pregnancy stage and almost absent post partum. Abnormal taste sensitivity was fairly commonly reported (26%), often described as increased bitter sensitivity and decreased salt sensitivity.
Instituto Nacional de Rehabilitacin Rehabilitation National Institute INR deeloped a prototype heelchair cushion INR cushion desined to adust to the anthropometry of the users ischioluteal area and preent pressure ulcer formation hile maintainin or promotin functionality. A prospectie, lonitudinal, descriptie study as conducted from February to February to ealuate the effect of usin the INR cushion on clinical ariables, functionality, and user satisfaction Siteen patients ere recruited To patients ithdre due to appointment conflicts of the remainin , sinificant differences beteen the users eperience ith other products and the INR ere found ith reard to pressure redistribution all participants but 1 raded the INR as Good in all interie cateories. No participants deeloped a pressure ulcer durin the study.
One example of this progression can be found in an Institute for Work & Health (IWH) project on the links between computer work and musculoskeletal disorders (MSDs) at a large newspaper. This project began with a cross-sectional study, aimed at exploring the links between injuries and different characteristics of the job (e.g. work stress) or of the worker (e.g. the social support he or she had at work). Knowing which links were strongest helped the researchers develop theories to test. In the next study, a longitudinal one, they studied changes in workers’ MSD symptoms over time. That study gave the researchers a better understanding of the cause-and-effect relationship between MSD symptoms and work/worker characteristics, which in turn lay the groundwork for intervention studies down the line.
The key objective of surveillance is to provide information to guide interventions.
Public health surveillance provides the scientific and factual database essential to informed decision making and appropriate public health action.
The fundamental principle of public health surveillance is that the surveillance should be designed and implemented to provide valid (true) information to decision makers in a timely manner at the lowest possible cost.
CDC has been doing active surveillance of abortion related deaths using overlapping sources- state maternal mortality study committees, professional organisations, newpapers and colleagues in the speciality. It identifies twice as many deaths
Most important contribution of surveillance has been to eradication of smallpox. Mass immunisation of the world’s population had failed, the approach of identification of cases through surveillance and then immunisaiton of susceptible persons in the surrounding communities stopped transmission. Without a non-human vector the virus died out.
Passive surveillance often gathers disease data from all potential reporting health care workers. Health authorities do not stimulate reporting by reminding health care workers to report disease nor providing feedback to individual health workers. Passive surveillance is the most common type of surveillance in humanitarian emergencies. Most surveillance for communicable diseases is passive. The surveillance coordinator may provide training to health workers in how to complete the surveillance forms, and may even send someone to periodically collect forms from health facilities. But little attention is given to individual health workers who report the information. The data requested of each health worker is minimal. Nonetheless, passive surveillance is often incomplete because there are few incentives for health workers to report. A system by which a health jurisdiction receives reports submitted from hospitals, clinics,
public health units, or other sources. Passive surveillance is a relatively inexpensive strategy to cover large areas,
and it provides critical information for monitoring a community's health. However, because passive surveillance
depends on people in different institutions to provide data, data quality and timeliness are difficult to control.
Eg. Notifiable diseases, hospital and billing data, vital statistics,
An active surveillance system provides stimulus to health care workers in the form of individual feedback or other incentives. Often reporting frequency by individual health workers is monitored; health workers who consistently fail to report or complete the forms incorrectly are provided specific feedback to improve their performance. There may also be incentives provided for complete reporting. Active surveillance requires substantially more time and resources and is therefore less commonly used in emergencies. But it is often more complete than passive surveillance. It is often used if an outbreak has begun or is suspected to keep close track of the number of cases. Community health workers may be asked to do active case finding in the community in order to detect those patients who may not come to health facilities for treatment. Active surveillance: a system employing staff members to regularly contact heath care providers or the population to seek information about health conditions. Active surveillance provides the most accurate and timely information, but it is also expensive. Eg. health surveys, The census
Instead of attempting to gather surveillance data from all health care workers, a sentinel surveillance system selects, either randomly or intentionally, a small group of health workers from whom to gather data. These health workers then receive greater attention from health authorities than would be possible with universal surveillance.
Sentinel surveillance also requires more time and resources, but can often produce more detailed data on cases of illness because the health care workers have agreed to participate and may receive incentives. It may be the best type of surveillance if more intensive investigation of each case is necessary to collect the necessary data. For example, sentinel influenza surveillance in the United States collects nasopharyngeal swabs from each patient at selected sites to identify the type of influenza virus. Collection of such data from all health workers would not be possible. In a sentinel surveillance system, a prearranged sample of reporting sources agrees to report all cases of defined
conditions, which might indicate trends in the entire target population (Birkhead and Maylahn 2000). When properly
implemented, these systems offer an effective method of using limited resources and enable prompt and flexible
monitoring and investigation of suspected public health problems. Examples of sentinel surveillance are networks of
private practitioners reporting cases of influenza or a laboratory based sentinel system reporting cases of certain bacterial infections among children. Sentinel surveillance is excellent for detecting large public health problems, but it may be insensitive to rare events, such as the early emergence of a new disease, because these infections may emerge anywhere in the population.
Advantages:
Data required is readily available
Correlation coefficient: provides a measure of how closely the observed data points confirm to a straight line
Disadvantages:
We cannot link the risk factor to a disease-ecological fallacy
Effect of other factors on outcome cannot be determined
Dose response relationships are masked
1. cost and convenience Ecologic studies are inexpensive and take little time because various secondary data sources, each involving different
information needed for the analysis, can easily be linked at the aggregate level. For example, data obtained from population registries, vital records, large surveys, and the census are often linked at the state, county, or census-tract level.
2. Measurement limitations of individual-level studies In environmental epidemiology and other research areas, we often cannot accurately measure
relevant exposures or doses at the individual level for large numbers of subjects-at least not with available time and resources. Thus, the only
practical way to measure the exposure may be ecologically (45,46). This advantage is especially true when investigating apparent clusters of disease
in small areas (66). Sometimes individual-level exposures, such as dietary factors, cannot be measured accurately because of substantial within-person
variability; yet ecologic measures might accurately reflect group averages
3. Design limitations a/individual-level studies Individual-level studies may not be practical for estimating exposure effects if the exposure varies little
within the study area. However, ecologic studies covering a much wider area might be able to achieve substantial variation in mean exposure across
Groups
4. Interest in ecologic effects As noted above, the stated purpose of a study may be to assess an ecologic effect, i.e. the target level of inference may
be ecologic rather than biologic. Ecologic effects are particularly relevant when evaluating the impacts of population interventions such as new programs, policies, or legislation
5. Simplicity of analysis and presentation In large, complex studies conducted at the individual level, it may be conceptually and statistically
simpler to perform ecologic analyses and to present ecologic results than to do individual-level analyses. For example, data from large, periodic
surveys, such as the National Health Interview Survey, are often analysed ecologically by treating some combination of year, region, and demographic
group as the unit of analysis.
Individual level data not available
Individual exposure measurements not available but grouped level data are (eg. mean radon gas levels from county wide surveys)
Adequate design of an individual-level study is not possible
The hypothesis is relatively new
Comparison between large jurisdictional unit (eg. comparison of breast cancer rates with mean daily fat intake between countries)
Easy, quick and inexpensive
Interest in effect from ecological variables, for which there is no correlate at the individual level (eg. does increasing taxes on tobacco reduce consumption in different regions)
Causal inferences from ecological studies can be made at three different levels
Biological: causation is only through a biological pathway
Among unvaccinated persons, low income people have increased risk of hepatitis B because having a low income increases a person’s stress which increases a person’s risk for Hep. B
Biologic effect of wearing a motorcycle helmet on the risk of motorcycle related mortality among motorcycle riders
Ecological: causation is only through a group characteristic
Among unvaccinated persons, people living in a low income community have increased risk for Hep. B infection because low income communities have high levels of exposure to Hep. B
Ecologic effect of helmet-use laws on the motorcycle related mortality rate of riders in different states , target level is ecologic
Contextual: causation is through both biologic and group characteristics
Among unvaccinated persons, both high income and low income people who live in a low income community have increased risk for Hep.B infection because of increased stress and increased exposure
Contextual effect of living in a poor area on the risk of disease, controlling for poor individual poverty
Contextual effect of living in a state that mandates helmet use on the risk of motorcycle related mortality in riders, controlling for individual helmet use
Aggregate Variables: a summary or composite measure derived from values collected from individuals/summaries of attributes calculated from data on individuals for whole populations in well-defined geographic regions
E.g. mean income, percentage of families below the poverty line, mean number of household members
Environmental/Group level measures: a measure of the physical characteristics of the environment in which people reside, work, recreate or attend school OR estimates of (environmental) attributes that have individual analogues, usually obtained from surveys.
E.g. maximum daily exposure to ozone, daily mean levels of environmental tobacco smoke in public buildings
Global measures (contextual): attributes that pertain to groups and do not have analogues at the individual level. Measure exposures, not outcomes.
E.g. total area of green space, number of private medical clinics. Population density
Purely ecological: no data on individual
E.g. age-sex-race standardised mortality rates of oral cancer, comparisons at the country level
Partially ecological: some individual data available
E.g. study of low birth weight and environmental exposures to biogas from a landfill site
Individual data: age of mother, sex, birth, weight, gestational age of baby, and geographic area or residence
Ecological: geographic region of residence as a surrogate for exposure to biogas in the ambient air
Multiple group study
–
Compares disease rates among many regions during the same period
•
Time-trend studies
–
Comparison of disease rates over time in one population
•
Mixed designs
–
Multiple groups + multiple time periods
The study by John Snow regarding a cholera outbreak in London is considered the first ecological study to solve a health issue. He used a map of deaths from cholera to determine that the source of the cholera was a pump on Broad Street. He had the pump handle removed in 1854 and people stopped dying there.[1] It was only when Robert Koch discovered bacteria years later that the mechanism of cholera transmission was understood
An important advancement in the understanding of risk modifying factors for cancer was made by examining maps of cancer mortality rates. The map of colon cancer mortality rates in the United States was used by the brothers Cedric and Frank C. Garland to propose the hypothesis that solar ultraviolet B (UVB) radiation, through vitamin D production, reduced the risk of cancer (the UVBvitamin D cancer hypothesis).[6] Since then many ecological studies have been performed relating the reduction of incidence or mortality rates of over 20 types of cancer to higher solar UVB doses.[7]
Links between diet and Alzheimer’s disease have been studied using both geographical and temporal ecological studies. The first paper linking diet to risk of Alzheimer’s disease was a multicountry ecological study published in 1997.[8] It used prevalence of Alzheimer’s disease in 11 countries along with dietary supply factors, finding that total fat and total energy (caloric) supply were strongly correlated with prevalence, while fish and cereals/grains were inversely correlated (i.e., protective). Diet is now considered an important riskmodifying factor for Alzheimer’s disease.[9] Recently it was reported that the rapid rise of Alzheimer’s disease in Japan between 1985 and 2007 was likely due to the nutrition transition from the traditional Japanese diet to the Western diet.[10]
Another example of the use of temporal ecological studies relates to influenza. John Cannell and associates hypothesized that the seasonality of influenza was largely driven by seasonal variations in solar UVB doses and calcidiol levels.[11] A randomized controlled trial involving Japanese school children found that taking 1000 IU per day vitamin D3 reduced the risk of type A influenza by twothirds.
By examining characteristics of person, place and time
Should specify:
The population- the characteristics of the persons to whom the hypothesis applies
The specific cause being considered
The expected outcome i.e. the disease
The dose response relationship- the amount of the cause needed to lead to a stated incidence of the effect
The time-response relationship- the time period that will elapse between exposure to the cause and observations of the effect
Differences- if the frequency of diseases differs in two circumstances, it may be caused by a factor that differs between the two circumstances
Similarities- if a high frequency of disease is found in several different circumstances and one can identify a common factor, then the common factor may be responsible
Correlations: if the frequency of disease varies in relation to some factors that that factor may be a cause of that disease
Overstepping the data- temporal association is incorrectly inferred to be as causal