Injury Epidemiology, Prevention and Control

1
SOM 218 Epidemiology Review
1) What is this class about? 2) What is on the
exam? 3) A brief review of old information 4)
New skeletons for new information
Joe Hyder Family and Preventive Medicine MS
III-To-Be
2
SOM 218 Epidemiology is

• 1) All the previous material
• 2) Occupational Health
• 3) Mental Health
• 4) Nutritional Epidemiology
• 5) SES and ethnicity
• 6) Tobacco and cancer
• 7) Genetics

Common health topics where information
often comes from epidemiologic studies. Common
concepts from the course unite these lectures.
3
What the exam is like
No one will tell me either Only multiple
choice and True/False questions Emphases on
both old and new lectures General emphasis
on concepts over details Study like it is SBS
Add details to conceptual framework
4
Previous Material Briefly Reviewed
Two commonly missed items from the midterm
Bias can be found anywhere Sensitivity,
Specificity, PPV, NPV, Prevalence
MY U.S.M.L.E. STEP I QUESTION
5
Previous Material Review
Be sensitive to people with disease
Truth - TP FP Test - FN TN
PPV TP/(TPFP) NPV TN(TNFN)
Measures that cross the D D cells change with
prevalence
Sensitivity Specificity TP/(TPFN) TN/(TNFP)
Of all the people who denominator, the index
is the fraction who are/ will insert TFNP
6
Occupational Health
Apply concepts learned previously Ascertainment
of D and E biases causality Learn the
definitions in the handout Understand the
examples in the handout What happens when the
exposure is unknown How to measure an
exposure? Inhaled vs On Skin vs Eaten vs In
the Eyes vs
7
Occupational Health (II)
Precautionary Principle Preventive action given
uncertainty Burden of proof to proponents of
activity/product Explore alternatives to
activity/product Increase public participation
in decision-making Role of the Physician and of
the Epidemiologist Potpourri Vinyl Chloride
Angiosarcoma of the Liver Other items from
handout Answers to the FOUR QUESTIONS in the
lecture
8
Epidemiology of Mental Health
Epidemiology can inform the understanding of
MH Epidemiology as a tool requires measurement
of E and D In mental health, Sx, E and D (or Dx)
are less certain Regional (and temporal)
variation in Dx prevalence and incidence Often
rely on ecological studies PTSD is one mental
illness with an exposure component Evidence
for other exposures (gene environment) in MH
9
Nutritional Epidemiology
Historically, a study of deficiency vs
excess Epidemiology as a tool requires
measurement of E and D How to measure diet? How
long to measure diet? How long to wait for
outcome? Difficult to assign (or validate) diet
for an RCT Surrogate markers for E and D can be
useful What are the main tools to measure
diet? How are they similar and different? What
are their advantages and weaknesses? Beware of
fishing expeditions, especially with
nutritiongenetics
10
(No Transcript)
11
Eating fish may fight depression in pregnant
women - CNN.com Wednesday, May 21, 2003
http//www.cnn.com/2003/HEALTH/conditions/05/21/fi
sh.depression.ap/
12
SES, Ethnicity and Diabetes
Epidemiology as a tool requires measurement of E
and D What is socioeconomic status? What is
race or ethnicity? Does the theoretical
definition match the empirical one? What about
genetics, diet, social status, income, education,
disability, geography, access to/use of care,
(Are these confounders?) Study of Race/ethnic
differences in D can inform all of the
above Other illustrations from the
lecture Leading causes of death by place and
ethnicity Filipina and NHW women, fat
deposition, DM/metabolic syndrome Pima Indians
(San Diego County) and DM
13
Tobacco
Previous concepts Theoretical vs empirical
definitions and smoking machines Incubation
period Causal standards If RD is constant, RR
is greater when AR is lower High RR is
indicative of causation PAR, PAR for smoking
and CHD Life History of a Smoker CHD, CA,
COPD Quitting helps CHD almost right away (esp
after 5 years) helps COPD less so
because Broad historical trends in the
consumption of tobacco Cigarettes vs pipes vs
snuff vs chew vs roll-your-own vs Lites vs
14
Genetic Epidemiology
This is just one lecture of many What is
it? How is it done? Why do it? What are the
epidemiology design tie-ins? Phenotype studies
(no DNA involved)
Is there a _________ component?
familial genetic OK, but how
so? Familial aggregation Heritability Segregatio
n Analysis G vs E vs alpha Genetic
component not emphasized Diseased relatives
Twins Matched CC C-C Odds ratios
pairwise vs proband Recurrence risk ratio
concordance (?) h2 p(Drelative)
Other relatives covariance p(Dpopulation)
15
Genetic Epidemiology (II)
Disease-Gene Discovery Genome-wide scan vs
Candidate gene approach No assumptions Use
biological guessing Look for LD Identify
and test variants Non-causal relationship Assume
d to be causative Gene Identification Linkage
analyses vs Association Studies Loci
cosegregate w/trait Genetic frequencies Within
families Across families (populations) Mendelia
n diseases Polygenic/complex diseases
16
Genetic Epidemiology (III)
Epi-gene identification design can be C-C or
cohort design Suggestions for case selection,
ascertainment of E, etc There are many models
of gene-environment interaction
17
Summary
Re-learn old lectures Integrate new material
into old concepts Be familiar with some new
illustrations of old concepts Study like it is
SBS except that some people might fail GOOD
LUCK AND SEE YOU THIS SUMMER!