Explaining Health Risk

1
Explaining Health Risk

• NAACCR\2009\AnnualMeeting\TenRules\TenRules.PPT

2
Guidelines

• How do we remain alert prudent
  
• Versus
• Jumping to Conclusions

3
Health Risk Reports are Important

• Public Preoccupation with Health Risk
• Media Reports
• Frequent
• Alarming
• Scientists are also Interested

4
Do migraines reduce risk of breast cancer?

• Cancer researchers in Seattle found women who
  had a history of migraines had a 30 percent lower
  risk of breast cancer compared to women who did
  not have a history of such headaches
• Li, et al. Fred Hutchinson Cancer Research
  Center. Cancer Epidemiology, Biomarkers and
  Prevention.
• San Diego Union-Tribune, Nov 6, 2008.

5
New Study Links Parkinsons and Pesticides

• Increasing levels of exposure to pesticides were
  associated with an increasing risk of
  Parkinsons
• Dana Hancock, PhD. Duke University. BMC
  Neurology.

6
Examples of Reported Health Risks

• Second hand tobacco smoke
• Radon within residences
• Cell phones-brain tumors
• DDT breast cancer
• Dump sites
• Electromagnetic fields-childhood leukemia
• Coffee drinking pancreatic cancer

7
Other Examples

• Silicone implants connective tissue cancer
• Agent Orange Gulf War Syndrome
• Charred Meat
• Neighborhood Clusters, Hot Spots
• Saccharin/sweeteners
• Oxidants
• Salt

8
Different Consumers ofHealth Risk Data

• Scientists (Epidemiologists Environmental
  Toxicologists)
• Looking for possibles/clues
• Multivariate analysis
• Public
• Common sense
• Not statistical
• Large, or interesting, risk

9

• Balancing learning about
• every possible risk versus
• Only being concerned with evidence proven risk.
• Statisticians might call this balancing
  sensitivity versus specificity.

10

• Possible (commonsense) Guidelines for
  Explaining Health Risk

11
Guidelines

• Agenda/Funding/Purpose of Presenter (Bias)
• Knowledge/Preconception of Listener/Reader (Bias)
• Characteristics/Biology of the Diagnosis
• Biologic Mechanism
• Risk Factor Mileau
• Absolute vs Relative Risk
• Other Supporting Evidence
• Association versus Causation
• Dosage
• Control Bias/Confounding (external influence)
• Nature/Reliability of Data and the Analysis
• Opposing Findings/Hypotheses

12

• What Have I Omitted?

13
Thank You

• Herman R. Menck, BS, MBA, CPhil, FACE
• Los Angeles
• Cancer Surveillance Program
• menckh_at_aol.com

14
Guidelines

• Agenda/Funding/Purpose of Presenter (Bias)
• Knowledge Preconception of Listener/Reader (Bias)
• Characteristics/Biology of the Diagnosis
• Biologic Mechanism
• Risk Factor Mileau
• Absolute vs Relative Risk

15
Guidelines

• 7. Other Supporting Evidence
• 8. Association versus Causation
• 9. Dosage
• 10. Control Bias/Confounding (external
  influence)
• 11. Nature/Reliability of Data and the Analysis
• 12. Opposing Findings/Hypotheses

16
1. Agenda/Funding/Purpose of Presenter (Bias)

• How Do they Get Their Funding?
• Tobacco Institute
• American Cancer Society
• Trucking Association
• Previous Work

17
2. Knowledge/Preconception of Listener (Bias)

• Cultural - Ethnic Geographic
• Media Gender Linguistic
• Political Corporate Advertising
• Sociologic Personal Gain Religious
• Sensationalist Anti-Scientific

18
3. Define Diagnosis

• Is the Diagnosis Understood and Well Defined?
• Organ/System Function
• Biology of organ
• Which Cancer/Disease (Agent Orange?)
• Are There Related/Confounding Diseases?

19
4. Biologic Mechanism

• How do headaches influence risk to breast cancer?
• How does it biologically work?

20
5. Risk Factor Mileau

• What are the Major and Minor Risk Factors for
  this diagnosis?

21
Lung Cancer Risk Factors

• Cigarettes
• Secondary smoke
• Chewing tobacco
• Radon
• Other

22
Breast Cancer Risk Factors

• Hormonal mileau
• Medication HRT
• Diet/Green Tea
• Exercise
• Other

23
6. Absolute vs Relative Risk

• Discuss Both
• Establish Magnitude of Absolute Risk
• Relative Risk can be misunderstood.

24
7. Other Supporting Evidence

• Natural Experiments/International Correlations
• High Low Risk Populations
• International Correlations
• Other Human Studies
• Animal Studies

25
8. Is it an Association vs Causation?
26
9. Dosage Can Be Important

• Could the dosage in animal studies be
  unrealistic?

27
9. Are Data Biases Controlled

• Healthy Worker Effect
• Married Men Live Longer
• Asians Have Higher SAT Scores
• Breast Cancer Patients Have Headaches

28
11. Evaluate Nature of the Data and Its Analysis

• Recall versus measurements
• Recall error
• Other Possible Data Bases

29
12. Review Opposing Evidence/Hypotheses

• Animal Studies
• Null Findings
• Opposing Findings/Hypotheses

30
Question for Scientists

• Which Role Do You Wish to Play?
• Can you be both an impartial scientist, and a
  health risk activist?
• Which role should the scientist play?

31
Guidelines

• Agenda/Funding/Purpose of Presenter (Bias)
• Knowledge/Preconception of Listener/Reader (Bias)
• Characteristics/Biology of the Diagnosis
• Biologic Mechanism
• Risk Factor Mileau
• Absolute vs Relative Risk
• Other Supporting Evidence
• Association versus Causation
• Dosage
• Control Bias/Confounding (external influence)
• Nature/Reliability of Data and the Analysis
• Opposing Findings/Hypotheses