8/13/2019 Extra Disease stuff

1/5

1. Describe the following tests and under what circumstances you might usethem over the others.

a. Chi-squaredb. Fishers exact test of independencec. CochranMantelHaenszel test

It is early February 2002, and news media around the country arereporting that Senator John McCain has been diagnosed with a

recurring melanoma. Successful surgery follows, but the report hascome on the heels of the death from melanoma of former President

Reagan's daughter, Maureen Reagan.

These high-profile cases focus attention on skin cancer, and you, asChief Disease Detective for the State of Delaware, are asked toprepare a briefing for the governor on the problem nationally and in

your state, and how to prevent it.

A quick search of the data tells you that in 2001 doctors diagnosedapproximately 51,400 new cases of melanoma, and about 7,800

people died of the disease. You locate data indicating that theincidence of melanoma has been on the rise since the early 1970s.

The Surveillance, Epidemiology, and End Results (SEER) Program ofthe National Cancer Institute collects and publishes cancer incidence

and survival data from registries covering approximately 14% of theU.S. population. Using the SEER data in the table, answer the

questions that follow.

Figure 1. Age-adjusted incidence of melanoma by race andgender; SEER data, 19731998.

(Source: National Cancer Institute)

8/13/2019 Extra Disease stuff

2/5

2. Give the definition of incidence rate.

3. For some subgroups, the incidence rates appear to haveincreased overall in the past 20 years. Give two reasons for thisapparent increase.

8/13/2019 Extra Disease stuff

3/5

4. Referring to Figure 1, give the 1998 incidence rate for thefollowing groups:

White Men ______________

White Women ______________

Black Men ______________

Black Women ______________

5. Give two possible reasons for the observed large differences inincidence rates between races.

6. The rates in Figure 1 are age-adjusted.a. Define age adjustment and

b. explain why Disease Detectives would use age-adjustedrates.

ANSWERS:

Answer:

1. Describe the following tests and under what circumstances you might usethem over the others.

a) Chi-square is a statistical test commonly used to compare observeddata with data we would expect to obtain according to a specific

hypothesis. Usually used to determine goodness of fit of expected

values, and must have numerical values for variables having choices

(usually yes or no). Better than b and c in cases with large numbers.

b) Fishers exact test of independence is used when you have twonominal variables and is similar to the above except is more accurate

in cases where the expected value is small.

8/13/2019 Extra Disease stuff

4/5

c) Used when there are three nominal variables; you want to know whether twoof the variables are independent of each other, and the third variableidentifies the repeats, i.e. when you have the above conditions but want tocontrol for a third variable (ex: time of year) on the correlation betweenthe other two.

2. Give the definition of incidence rate. (3 points 1 each forexplanation of numerator, denominator, and time)

Answer: Incidence rate is a measure of the frequency with

which a health problem or health event (such as a new injury or

case of illness) occurs in a population. In calculating incidence,the numerator is the number of new cases occurring in the

population during a given period of time, and the denominator isthe total population at risk during that time.

3. For some subgroups, the incidence rates appear to haveincreased overall in the past 20 years. Give two reasons for thisapparent increase. (2 points 1 for each valid reason)

Answer: Factors that could contribute to the rise in rates include

a. increased exposure to UV radiation as a result of morepeople being exposed to the sun

(Note: This is the most important reason.);b. ozone depletion;c. heightened awareness among health care providers;d. increased reporting (improved surveillance); ande. people living longer increased survival.

4. Referring to Figure 1, give the 1998 incidence rate for thefollowing groups: (8 points 2 for each rate. Estimates areacceptable; however, units must be cited. Take off 1 point for

each rate without units.)

Answer:a. White Men 19.3 per 100,000 personsb. White Women 13.6 per 100,000 personsc. Black Men 1.1 per 100,000 personsd. Black Women 1.0 per 100,000 persons

8/13/2019 Extra Disease stuff

5/5