16
Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Embed Size (px)

Citation preview

Page 1: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Spatial Statistics for Cancer Surveillance

Martin Kulldorff

Harvard Medical School and

Harvard Pilgrim Health Care

Page 2: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Two Applications of Spatial Data and GIS in Cancer Research

Studies of Specific Hypotheses: Evaluate the relationship between cancer and geographical variables of interest such as radon, pesticide use or income levels, adjusting for geographical variation.

Surveillance: Evaluate the geographical variation of cancer, adjusting for known or suspected variables such as age, gender or income.

Page 3: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Reasons for Geographical Cancer Surveillance

• Disease Etiology• Known Etiology but

Unknown Presence• Health Services

• Public Education• Outbreak Detection• New Diseases

Page 4: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Cancer Prevention and Control

• Are people in some geographical area at higher risk of brain cancer? This could be due to environmental, socio-economical, behavioral or genetic risk factors.

Page 5: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Cancer Prevention and Control

• Are there geographical differences in the access to and/or use of early detection programs, such as mammography screening?

Page 6: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Cancer Prevention and Control

• Are there geographical differences in the access to and/or use of state-of-the-art breast cancer treatment?

Page 7: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Different Types of Cancer Data

• Count Data: Incidence, Mortality, Prevalence

• Categorical Data: Stage, Histology, Treatment

• Continuous Data: Survival

Page 8: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

For Incidence and Mortality

Poisson Data

Numerator: Number of CasesDenominator: Person-years at risk

Page 9: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

For Prevalence

Bernoulli Data (0/1 Data)

Numerator: People with Thyroid CancerDenominator: Those without Thyroid Cancer

Note: When prevalence is low, a Poisson model is a very good approximation for Bernoulli data.

Page 10: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

For Stage, Histology and Treatment

Bernoulli Data (0/1 Data)Numerator: Cases of a specific type, e.g. late stage.Denominator: All cases.

Ordinal Data

For example: Stage 1, 2, 3, 4

Page 11: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

For Survival

Survival Data

Length of Survival

(Censored Data is Common)

Page 12: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Data Aggregation (spatial resolution)

• Exact Location

• Census Block Group

• Zip Code

• Census Tract

• County

• State

Page 13: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Data Aggregation

• Same level of aggregation usually needed due to data availability.

• Less aggregation is typically better as more information is retained.

• Many statistical methods can be used irrespectively of aggregation level.

Page 14: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Geographical Cancer Surveillance1. Mapping Rates and Proportions2. Smoothed Maps3. Tests for Spatial Randomness4. Spatial Scan Statistic5. Global Clustering Tests6. Brain Cancer Mortality7. Survival Data

Course Outline

Page 15: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

Space-Time Cancer Surveillance8. Space-Time Scan Statistic for the Early

Detection of Disease Outbreaks

Statistical Software9. SaTScan Demonstration

Course Outline

Page 16: Spatial Statistics for Cancer Surveillance Martin Kulldorff Harvard Medical School and Harvard Pilgrim Health Care

WELCOME AT ANY TIME

Comments and Questions

Software and Slide Presentation

AVAILABLE FROM THE WEB