Introduction to Biostatistics (PubHlth 540) Lecture 1: Overview Ed Stanek

1
Introduction to Biostatistics(PubHlth 540)
Lecture 1 OverviewEd Stanek
Acknowledgement Thanks to Professor
Balasubramanian and Professor Pagano for lecture
material
2
Topics

• Course Logistics
• Why Biostatistics?
• Course outline
• Data Presentation
• Inference
• Prediction

3
Topics

• Course Logistics
• Why Biostatistics?
• Course outline
• Data Presentation
• Inference
• Prediction

4
Course Logistics

• Instructor Ed Stanek
• Office Hours Tu/Th, 400 515
• Office 401 Arnold
• Email stanek_at_schoolph.umass.edu
• LocationMorrill III Room 212

5
Course Logistics

• Grading
• Homework (n10 assignmentsreports) 10
• Exam1 30
• Exam2 30
• Exam3 30
• Final 30

Best 2 of 3
6
Topics

• Course Logistics
• Why Biostatistics?
• Course outline
• Data Presentation
• Inference
• Prediction

7
Biostatistics

• There are three kind of lies lies, damn lies,
  and statistics
• Mark Twain (1835-1910)
• Bio --- bios
• Greek --- life
• Statistics
• Latin --- statisticum collegium
• (lecture about state affairs)
• A branch of applied mathematics concerned with
  the collection and interpretation of quantitative
  data and the use of probability theory to
  estimate population parameters --
  www.hyperdictionary.com

8
Course Outline

• Presentation
• Inference
• Prediction

9
Data presentation

• Data presentation techniques enable us to
  condense large amounts of information into a
  digestable form
• Examples
• Tables
• Bar graphs
• Histograms etc.

10
Data Presentation - example
11
Data Presentation - example
12
Data presentation - exampleAdults and children
estimated to be living with HIV as of end 2004
13
Election 2004 Results
http//www-personal.umich.edu/mejn/election/
14
Course Outline

• Presentation
• Inference
• Prediction

15
Inference

• But Do We Believe It?
• Inference
• Sample from a populationFrom the sample infer
  (guess) characteristics of the population.

16
Inference

• Population Entire group of interest
• Sample a small subset of population to be
  studied
• Parameter a summary measure or characteristic of
  a population (e.g. mean)
• Statistic summary measure or characteristic of a
  sample

17
Inference

• Theory and methodology for generalizing from a
  sample to a population

Population
Sample drawn from population
sample
guess
Inference regarding the population made from
sample
18
Mortality before and after the 2003 invasion of
Iraq cluster sample survey -- Lancet 2004 364
1857-64
Inference - example

• We estimate that there were 98000 extra deaths
  (95 CI 8000-194 000) during the post-war period
  in the 97 of Iraq represented by all the
  clusters except Falluja.

19
Meat Consumption and Risk of Colorectal Cancer -
JAMA 293 (2) Jan. 12, 2005
Inference - example

• In our analyses, the association between colon
  cancer risk and high intake of red (RR, 1.41 95
  CI, 1.121.78) and processed meat (RR, 1.33 95
  CI, 1.081.64) measured at a single time point is
  consistent with meta-analysis results, 50
  adjusting for age and energy intake. However, the
  association was substantially attenuated with
  further adjustment for educational attainment,
  cigarette smoking, physical activity, and other
  lifestyle factors associated with red meat
  intake.

20
Course Outline

• Presentation
• Inference
• Prediction

21
Prediction

• A diagnosis of diabetes can be suspected in the
  presence of the following signs and symptoms of
  hyperglycemia
• Polydipsia (increased thirst)
• Polyuria (increased urinary frequency
• with increased volume)
• Fatigue
• Polyphagia (increased appetite)
• Weight loss
• Abnormal healing
• Blurred vision
• Increased occurrence of infections,
• particularly those caused by yeast.

22
Prediction

• The risk of diabetes is increased in asymptomatic
  individuals if any of the following risk factors
  are present
• A strong family history of diabetes (parents or
  sibling)
• Obesity (20 above ideal body weight)
• Certain races (American Indian, Hispanic,
  African, or Pacific Islander ancestry)
• Women with previous gestational diabetes or
  history of babies of 9 pounds (4Kg) or more at
  birth
• Previously identified impaired glucose tolerance
  (IGT)
• Hypertension or significant hypertriglyceridemia
  (gt 250 mg/dL)
• 40 years of age with any of the preceding
  factors.

23
Prediction

• Could you have diabetes and not know it?
• There are 18.2 million Americans with diabetes --
  and nearly one-third of them (or 5.2 million
  people) do not know it! Take this test to see if
  you are at risk for having diabetes. Diabetes is
  more common in African Americans, Latinos, Native
  Americans, Asian Americans and Pacific Islanders.
  If you are a member of one of these ethnic
  groups, you need to pay special attention to this
  test.
• http//www.diabetes.org/diabetes-basics/prevention
  /diabetes-risk-test/
• To find out if you are at risk, answer the
  following questions as they apply to you, then
  click the "CALCULATE" button to run the test and
  view your score.
• Please select your age category. 0-4445-6465
  or Older
• Please select your height.
• Please enter your weight in pounds.
• I am a woman who has had a baby weighing more
  than nine pounds at birth.True or False

24
Prediction

• I have a sister or brother with
  diabetes.TrueFalse
• I have a parent with diabetes.TrueFalse
• I am under 65 years of age and I get little or no
  exercise.TrueFalse
• CALCULATE your score
• Read our Frequently Asked Questions regarding
  this Risk Test.
• The information contained in this American
  Diabetes Association (ADA) Web site is not a
  substitute for medical advice or treatment, and
  the ADA recommends consultation with your doctor
  or health care professional.  

25
Summary

• Website
• Syllabus
• Lecture Notes
• Homework Problems
• Course Outline
• Data Presentation
• Inference
• Prediction