3C SEER Moderator Guide for Public

Participant # ____

Attachment C:

SEER Statistics

Initial Interviews With End Users

Interviewer’s Guide - Public

[Note: The purpose of this document is to guide the interviewer.  The questions and tasks contained herein may not be asked as written.  The facilitator often draws upon participant comments and the natural flow of the interview process.  While the facilitator will try to follow the order of the guide, many times questions will come up ahead of time or in a different order.  The facilitator may allow the order of the questions to change in order to let the process flow naturally.]

OMB No.: 0925-0642-03

Expiration Date: 9/30/2014

PRIVACY ACT NOTIFICATION STATEMENT

The National Cancer Program—Sec. 411 [285a] provides authority for collection of information. (For details about the authority see http://codes.lp.findlaw.com/uscode/42/6A/III/C/1.) Personally identifying information (name, address, phone number, and email) are collected to contact participants and arrange a time of participation and to provide an incentive as a thank you for their time. This information will be shared only with those who need to contact participants about the time or with those who will compensate participants for their time. Providing this information is voluntary although without this information, the participant cannot be scheduled or receive an incentive to participate. This information is not shared further and it is destroyed after it has been used for these purposes.

NOTIFICATION TO RESPONDENT OF ESTIMATED BURDEN

Public reporting burden for this collection of information is estimated to average 60 minutes per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB control number. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to: NIH, Project Clearance Branch, 6705 Rockledge Drive, MSC 7974, Bethesda, MD 20892-7974, ATTN: PRA (0925-0642). Do not return the completed form to this address.

Introduction

Before we start, let me tell you about what we’re going to be doing today. My colleagues and I are working on redesigning the Cancer Statistics features on the NCI Surveillance Epidemiology and End Results (SEER) website, but before we do that we need a better understanding of who the users are and what they need. In order to get a better contextual understanding of whom the data is appropriate for, I’ll be asking you questions about how and why you would use cancer statistics. Then I’m going to show you some examples of statistics so we can discuss different types of formats and what works or doesn’t work for you.

You may or may not have had experience with SEER before. If you have, I want you to ignore what you currently can and cannot do with it. I’d like to focus on cancer statistics at a more general level.

[Basic information:]

1. Why don’t we start with you telling me a little about your interest in cancer information?

   1. What types of cancer information would you be interested in looking for?

   2. If I were to say “cancer information” versus “cancer data”, what does that mean to you?

   3. Has there ever been an instance where you’ve ever had a need to look for cancer statistics?

[If YES:]

2. What kind of cancer statistics did you look for?

[Probe on which type of statistic or specific cancer site was searched for; how the information related to them (looking for a friend, personal diagnosis, statistics about an area, etc)]

3. What resources did you use to find these statistics?

(Look at examples if possible.)
[Probe on what kind of keywords were used on an Internet search]

4. How did you weed through information until you found what you were looking for?

[Probe on what made it {hard/easy} to understand the information]

5. How was the information provided to you? What format was it in?

[Probe on whether it was a line chart, bar graph, or text format; how would information shown in a summary format change the way they used it; What they do when they don’t understand the information]

6. After you got the information, what did you do with it?

[Probe on how information would be used differently if it were shared with a doctor, or a friend/family member; how they would modify the information (or prefer the information to be modified) to make it easier to share it with someone]

[If NO:]

2. Do you think there is a situation in which you would ever want to look for cancer data for cancer statistics data?

[Probe on what they imagine they would want to know]

Examples

Ok, now I’m going to show you some examples of various outputs of cancer statistics and I’d like you to tell me what you think of them.

[Survival]

One statistic estimates the probability that a person will still be alive a certain number of years after being diagnosed with cancer. For example, approximately 65.7% of men between ages 60 and 64 when diagnosed with colorectal cancer are still alive 5 years later.

How do you describe this type of statistic? Do you know a specific name for it? [Probe: Survival rate, prognosis, probability]

Are you interested in this type of statistic? [If no, show the next statistic type. If yes, continue.]

If you were looking for this type of statistic, what specifically would you look for? [Probe on options like age, gender, etc.]

[Show SEER options for Survival]

There is a system I know of that has this data and you can get it using these options:

• Age: 0-44 50 and Older All Ages

0-49 65 and Older

0-64 75 and Older

45-54

55-64

65-74

• Sex: Male Female Both
  
  

• Race: White Black

Asian/Pacific Islander American Indian/Alaska Native Hispanic All

• Cancer stage: Localized Regional Distant Unstaged
  
  

• Survival term: 1-year

2-year
3-year
4-year
5-year
10-year

• The year of the diagnosis: 1975-1977 1981-1983 1990-1992 2001-2007

1978-1980 1984-1986 1993-1995

1987-1989 1996-1998

• Type of survival: Relative Overall

Which of these factors are of interest to you and why? Which factors seem unclear to you? What other factors would you be interested in that are not in this list? [Probe: Geographical location, trends in data, uncertainty data]

[If participant does not understand a certain term, provide additional information and discuss further:

Age: Explain that it is the age at diagnosis, discuss age ranges, span of ages available
Race: Differentiation between “race” and “ethnicity”
Cancer stage: Explain the categories of cancer diagnosis stages, other terms besides “stage,” alternative stage labeling outside of the typical Localized/Regional/Distant classification, what “unstaged” might imply
Length of survival term: How long a survival term or what options would be useful to them
Year of diagnosis: How far back is useful for them, how recent should the data be
Type of survival: Explain the difference between relative survival and overall survival, discuss other types they may be interested in]

Sometimes cancer data is shown as an estimate of how data for the current year might look (as a projection). How would you use these data?

Here is an example of data output that you might generate on the SEER website.

Stage Distribution and 5-year Relative Survival by Stage at Diagnosis for 2001-2007, All Races, Both Sexes
Stage at Diagnosis	Stage Distribution (%)	5-year Relative Survival (%)
Localized (confined to primary site)	39	90.1
Regional (spread to regional lymphnodes)	37	69.2
Distant (cancer has metastasized)	20	11.7
Unknown (unstaged)	5	33.3

1. Is it clear what it’s trying to tell you? What specifically seems [clear/unclear] to you?

2. What do you like about it, and what don’t you like about it?

   • Do you like the table, the graphs, both, or neither?

3. What would you change about the format, if anything?
   
   

[Prevalence]

Another statistic estimates the number of people in the US population who were alive on a specific date after having been diagnosed with cancer. For example, 168,326 men were alive on Jan. 1 2008 who had a prior diagnosis of lung or bronchus cancer.

How do you describe this type of statistic? Do you know a specific name for it? [Probe: Most common cancer, most prevalent cancer, most frequent cancer]

Are you interested in this type of statistic? [If no, show the next statistic type. If yes, continue.]

If you were looking for this type of statistic, what specifically would you look for? [Probe on options like age, gender, etc.]

[Show SEER options for Prevalence]

There is a system I know of that has this data and you can get it using these options:

• Age: 1-4 45-49

5-10 50-54

11-14 55-59

15-19 60-64

20-24 65-69

25-29 70-74

30-34 75-79

35-39 80-84

40-44 85 and Older

• Sex: Male Female Both
  
  

• Race: White Black Asian/Pacific Islander Hispanic All
  
  

• Format: Count Percentage
  
  

• Type of prevalence: Limited duration Complete
  
  

• Years since diagnosis: Less than 5 5-Less than 10 More than 33

Less than 18 10-Less than 20

Less than 33 20-Less than 25

25-Less than 30

Which of these factors are of interest to you and why? Which factors seem unclear to you? What other factors would you be interested in that are not in this list? [Probe: Geographical location, trends in data, uncertainty data]

[If participant does not understand a certain term, provide additional information and discuss further:

Age: Explain that it is the age at diagnosis, discuss age ranges, span of ages available
Race: Differentiation between “race” and “ethnicity”
Format: Explain the difference between the two formats. Does having a specific number of individuals diagnosed with cancer, or a percentage of people diagnosed with cancer more useful to you?
Type of prevalence: Explain the difference between the two types of prevalence. Discuss the range of years that would be useful to the participant.
Years since diagnosis: How far back is useful for them, discuss ranges, how recent should the data be]

Sometimes cancer data is shown as an estimate of how data for the current year might look (as a projection). How would you use these data?

Here is an example of data output that you might generate on the SEER website.

[Show output for prevalence]

1. Is it clear what it’s trying to tell you? What specifically seems [clear/unclear] to you?

2. What do you like about it, and what don’t you like about it?

   • Do you like the table, the graphs, both, or neither?

3. What would you change about the format, if anything?
   
   

[Incidence]

Another statistic estimates the number of new cases of cancer. For example, about 21,520 new stomach cancer cases are expected to be diagnosed in 2011.

How do you describe this type of statistic? Do you know a specific name for it? [Probe: Incidence, additional cases, new cases]

Are you interested in this type of statistic? [If no, show the next statistic type. If yes, continue.]

If you were looking for this type of statistic, what specifically would you look for? [Probe on options like age, gender, etc.]

[Show SEER options for Incidence]

There is a system I know of that has this data and you can get it using these options:

• Age: 0-14 15-34 30-39 55-64 40 and Older

0-19 15-44 35-44 60-69 50 and Older
0-39 20-29 40-49 65-74 60 and Older

0-49 20-44 45-54 70-79 65 and Older

0-54 20-54 50-59 75-84 75 and Older

0-64 80 and Older

• Sex: Male Female Both
  
  

• Race: White Total Hispanic

Black White Hispanic

Asian/Pacific Islander White Non-Hispanic

American Indian/Alaska Native All

• Geographic location: San Francisco

Connecticut
Detroit
Hawaii
Iowa
New Mexico
Seattle
Utah
Atlanta

• Population: 2000 U.S.

1970 U.S.

World

1991 Canadian

1996 Canadian

European

• Year of diagnosis: Each single year from 1973 to 2008

1973-2008 1990-2008 2000-2008

1975-2008 1992-2008 2004-2008

1999-2008

Which of these factors are of interest to you and why? Which factors seem unclear to you? What other factors would you be interested in that are not in this list? [Probe: Trends in data, uncertainty data, understanding of age adjusted rates]

[If participant does not understand a certain term, provide additional information and discuss further:

Age: Explain that it is the age at diagnosis. Discuss age ranges and/or span of ages available.
Race: Differentiation between “race” and “ethnicity.”
Geographic location: Explain that the data was only collected in these areas. Discuss the fact that locations include cities, regions and states. [Probe: is this a limitation that affects the participant’s understanding of this data?]
Population: Explain the comparison to different populations and how that affects the incidence rate. [Probe: are the population options relevant to the participant?]
Year of diagnosis: Discuss ranges and length of time necessary for data to be useful. [Probe: How recent should the data be?]

Sometimes cancer data is shown as an estimate of how data for the current year might look (as a projection). How would you use these data?

Here is an example of data output that you might generate on the SEER website.

[Show output for incidence]

Trends in Rates

Trends in rates can be described in many ways. Trends over a fixed period of time can be evaluated by the annual percentage change (APC)X Close
The average annual percent change over several years. The APC is used to measure trends or the change in rates over time. For information on how this is calculated, go to Trend Algortihms in the SEER*Stat Help system. The calculation involves fitting a straight line to the natural logarithm of the data when it is displayed by calendar year.. If the number is negative, the trend is a decrease; otherwise it is an increase. An asterisk after the number indicates the trend was significant--that one believes with a certain level of confidence (usually 95%) that the increase or decrease is beyond chance. If the trend is not significant, the trend is usually reported as stable or level. Joinpoint analysesX Close
A statistical model for characterizing cancer trends which uses statistical criteria to determine how many times and when the trends in incidence or mortality rates have changed. The results of joinpoint are given as calendar year ranges, and the annual percent change (APC) in the rates over each period can be used over a long period of time to evaluate when changes in the trend have occurred. The APC then shows how much the trend has changed between each of the joinpoints.

The joinpoint trend in SEER cancer incidence with associated APC(%) for cancer of the stomach between 1975-2008, All Races
Male and Female		Male		Female
Trend	Period	Trend	Period	Trend	Period
-1.6*	1975-2008	-1.2*	1975-1988	-1.6*	1975-2008
		-2.0*	1988-2008

1. Is it clear what it’s trying to tell you? What specifically seems [clear/unclear] to you [Probe: Joinpoint understanding]?

2. What do you like about it, and what don’t you like about it?

   • Do you like the table, the graphs, both, or neither?

3. What would you change about the format, if anything?

[Mortality]

Another statistic estimates the number of people who will die from cancer in a given year. For example, in 2011, about 571,950 Americans are expected to die of cancer.

How do you describe this type of statistic? Do you know a specific name for it? [Probe: Death rate, death count, cancer patients who did not survive]

Are you interested in this type of statistic? [If no, show the next statistic type. If yes, continue.]

If you were looking for this type of statistic, what specifically would you look for? [Probe on options like age, gender, rate vs. count, rate per 100,000 vs. percent etc.]

[Show SEER options for Mortality]

There is a system I know of that has this data and you can get it using these options:

• Age: 0-14 15-34 30-39 55-64 40 and Older

0-19 15-44 35-44 60-69 50 and Older
0-39 20-29 40-49 65-74 60 and Older

0-49 20-44 45-54 70-79 65 and Older

0-54 20-54 50-59 75-84 75 and Older

0-64 80 and Older

• Sex: Male Female Both
  
  

• Race: White Total Hispanic

Black White Hispanic

Asian/Pacific Islander White Non-Hispanic

American Indian/Alaska Native All

• Population: 2000 U.S.

1970 U.S.

World

1991 Canadian

1996 Canadian

European

• Cause of death: All causes of death

All malignant cancers

A specific type of cancer

• Year of death: Each single year from 1969 to 2008

1969-2008 1975-2008 1989-2008 1999-2008 2000-2008

1969-1978 1975-1998 1989-1999 2004-2008

1975-1998

1979-1998

Which of these factors are of interest to you and why? Which factors seem unclear to you? What other factors would you be interested in that are not in this list? [Probe: Geographical location, trends in data, uncertainty data]

[If participant does not understand a certain term, provide additional information and discuss further:

Age: Explain that it is the age at diagnosis, discuss age ranges, span of ages available
Race: Differentiation between “race” and “ethnicity”
Population: Explain the comparison to different populations and how that affects the incidence rate, are these options relevant to the participant?
Cause of death: Explain “malignant cancers,” Are these distinctions useful for the participant?
Year of death: How far back is useful for them, discuss ranges, how recent should the data be]

Sometimes cancer data is shown as an estimate of how data for the current year might look (as a projection). How would you use these data?

Here is an example of data output that you might generate on the SEER website.

[Show output for mortality]

1. Is it clear what it’s trying to tell you? What specifically seems [clear/unclear] to you?

2. What do you like about it, and what don’t you like about it?

   • Do you like the table, the graphs, both, or neither?

   • Would you prefer to see the data as a percent or out of 100,000 people?

3. What would you change about the format, if anything?

[Risk]

Here is another statistic for you to consider:

Risk of developing stomach cancer over an entire lifespan in 2005-2007:

Approximately 0.88%

If you were looking for this type of statistic, what specifically would you look for? [Probe on options like age, gender, etc.]

Are you interested in this type of statistic? [If no, wrap up the interview. If yes, continue.]

[Show SEER options for Risk]

There is a system I know of that has this data and you can get it using these options:

• Starting age: 0 40 80

5 45 85

10 50 90

15 55 95

20 60

25 65

30 70

35 75

• Ending age: 5 45 85

10 50 90

15 55 95

20 60 95 and Older

25 65

30 70

35 75

40 80

• Sex: Male Female Both
  
  

• Race: White Total Hispanic

Black White Hispanic

Asian/Pacific Islander White Non-Hispanic

American Indian/Alaska Native All

• Risk type: Developing cancer

Dying from cancer

• Year of diagnosis: 2005-2007

2003-2005

2000-2002

Which of these factors are of interest to you and why? Which factors seem unclear to you? What other factors would you be interested in that are not in this list? [Probe: Geographical location, trends in data, uncertainty data]

Sometimes cancer data is shown as an estimate of how data for the current year might look (as a projection). How would you use these data?

Here is an example of data output that you might generate on the SEER website.

[Show output for risk]

1. Is it clear what it’s trying to tell you? What specifically seems [clear/unclear] to you?

2. What do you like about it, and what don’t you like about it?

   • Do you like the table format?

3. What would you change about the format, if anything?

13