ACS Conference Presentation
Download
Report
Transcript ACS Conference Presentation
Assessing Cancer Registry Systems:
Challenges and Opportunities for Improving
Outreach in Rhode Island Minority Populations
The Rhode Island Latino Cancer
Control Task Force (LCCTF)
Bridging the Health Care Divide: Research
and Programs to Eliminate Cancer Disparities
American Cancer Society Conference
New Orleans, Louisiana
April 18-20, 2007
Building a Coalition
2000 General Assembly urges Rhode
Island cities and towns to create
municipal Cancer Control Task Forces.
Established partnership among American
Cancer Society, RIDOH, RICAN
13 coalitions were formed.
2
Building a Coalition
Neither one focused on dispair
populations.
Hispanic/Latinos are the majority of the
minority groups in RI, representing almost
12% of the overall population.
Among the Hispanic/Latino population,
Cancer was the leading cost of death
during the period of 1999-2004 in the
State of Rhode Island.
3
Building a Coalition
The Formation of the Latino Cancer Control
Task Force will help:
to determine the nature of cancer challenge in
our community,
to examine its causes, and
to identify resources, gaps and opportunities,
in order to develop prioritized strategies that
will diminish the cancer deaths among the
Latino community in our state.
4
Building a Coalition
October 20, 2005,
The Latino Cancer Control Task Force
(LCCTF) was created as a
volunteer-based Community group formed
in response to the need to reduce the
leading causes of cancer morbidity and
mortality among Hispanic/Latinos in
Rhode Island.
5
The Latino Cancer Control Task Force
Mission Statement
The Latino Cancer Control Task Force aims to
reduce the impact of cancer in the Latino
community in Rhode Island
by facilitating cancer prevention through
outreach, public education, awareness, early
detection, access to care and advocacy,
and improving care and quality of life for cancer
patients, their families, and caregivers.
6
The Latino Cancer Control Task Force
Goals and Objectives
Advocacy. Being vigorous advocates for cancer
control, supporting public policy initiatives and
research, and developing strategic alliances with
governmental agencies, hospitals, nongovernmental
health promotion organizations, health
professionals, the community, and the media.
Outreach. Providing outreach targeting the different
populations, institutions, providers, policy makers,
advocates, community leaders and organizations to
increase knowledge and awareness of cancer
issues affecting Latinos.
7
The Latino Cancer Control Task Force
Goals and Objectives
Education. Preventing cancer by increasing
knowledge and awareness of cancer and by
encouraging healthier behaviors.
Services. Promoting the early detection of
cancer through public education and access
to screenings and treatment.
Quality of life. Improving patients' quality of
life and minimizing their suffering through
education, support, and palliative care.
8
The Latino Cancer Control Structure
Co-Chairs
(2)
Secretary
Data
Analysis
Workgroup
Outreach
Workgroup
Treasurer
Policy &
Advocacy
Workgroup
Media
Workgroup
9
LCCTF Projects
Conduct an assessment on what resources are
available to the Hispanic/Latino population.
Assess how data is collected and interpreted by
regulatory offices.
Bring together Latino health professionals to educate
and create community awareness on the issues of
cancer prevention and early detection.
Trough our Advocacy workgroup support American
Cancer Society Legislative Policy Priorities.
Ensure fair access to screenings and treatment.
10
LCCTF Accomplishments
Membership: LCCTF membership represents the vast
diversity of Latinos in RI.
Data Analysis: Data Analysis workgroup conducted a
survey to all RI Hospitals.
1st. Annual Conference “Mujeres Hablemos
Claro! La Salud del Seno” 183 registered, 126
attended, 60 new volunteers, 13 “De Amiga a Amiga” volunteers
trained.
Legislation Introduced by one of LCCTF members
Senator Juan Pichardo, asking to increase the funds for the
WCSP.
11
LCCTF Data Systems Project
Assess how data is collected and
interpreted by regulatory offices
The Rhode Island Cancer Registry
System Study
12
The RICR System Study -Objectives
Conduct an assessment of the Rhode Island
Cancer Registry (RICR) System
Determine how race and ethnicity information
of cancer patients is collected and reported
Identify strengths, limitations and
opportunities for improving CR information
quality and data utilization for population
targeting and outreach
13
Main Questions
What are the start and end-points of the
RI CR?
What is the guideline?
Is there compliance with OMB 15?
What are the implications of non-uniform
reporting?
Other study findings
Can under-reporting be corrected?
14
Start and End-Points - RICR
COLLECTION
AGGREGATION
LOCAL, REGIONAL AND NATIONAL REPORTING
DATA SYSTEM
Cancer
Patients
Cancer
Patients/Cancer
Families
15
What is the Guideline?
The Rhode Island Department of Health requires health
providers to collect the following demographic
information about their patients:
What is your ethnic background?
Hispanic/Latino
Not Hispanic/Latino
What is your race? (please indicate all races that apply)
American Indian or Alaskan Native
Asian
Black or African American
Native Hawaiian or other Pacific Islander
White
16
OMB 15
Is there compliance in the collection and
reporting of race and ethnicity?
17
RICRS Assessment - Methods
Hospital Survey
14 acute care hospitals
13 cancer registries
hospital cancer registry data not a start or
end-point
Key Informant Interviews
3 key informants
public health, health care industry and
national voluntary health organization
18
Hospital Cancer Registry Survey
Unit of Analysis: hospitals (information collected
via questionnaire)
Population of Interest: Rhode Islanders with
a diagnosis of primary cancer
Main Topics: forms used for cancer patient
intake, collection of cancer risk factors and
race/ethnicity information, content of database
and access to cancer registry data, feasibility of
CR and MR linkage, frequency of reports and
organizations receiving reports (22 questions)
19
Participating Hospitals (N=14)
Kent Hospital
Landmark Medical Center
Memorial Hospital of Rhode Island
Miriam Hospital
Naval Health Care New England
Newport Hospital
Rhode Island Hospital
Roger Williams Medical Center
South County Hospital
St. Joseph Hospital
Our Lady of Fatima Hospital
VA Medical Center
Westerly Hospital
Women & Infants Hospital
20
Key Informant Interviews
Unit of Analysis: individual expertise in cancer
data systems (information collected via questionnaire)
Population of Interest: US population with a
diagnosis of primary cancer
Main Topics: reporting organizations, types/
format of information reported, risk factor and
race/ethnicity information quality, methods for
adjusting under-reporting, OMB compliance,
database content and access to cancer registry
data, feasibility of CR and MR linkage, frequency
of reports and agencies targeted (26 questions)
21
Key Informants (N=3)
John Fulton, PhD
Director, Rhode Island Cancer Registry
Associate Director of Health
Rhode Island Department of Health
David Rousseau, PhD
Director, Cancer Information Systems
Hospital Association of Rhode Island
Jimmy Parson, PhD
Research and Analysis Manager
American Cancer Society, NHO
22
Analysis
Hospital Survey
Since sample size was small, results are
presented as frequencies and percents
Because some questions asked for more than
1 response, some figures exceed the total
number of expected responses
Responses were converted into ordinal scales
(i.e.; all=100%; most=70-99%; some=40-69%;
a few=1-39%; none=0%)
Ordinal scales were grouped by topic in order
to summarize findings
23
Analysis
Key Informant Interviews
Emphasis set on the qualitative aspects of
information
Results not presented in frequencies or
percents
Information used to gain a better insight into
hospital survey results and inform discussion
of findings and recommendations to 1) help
improve the quality of cancer registries and 2)
translate findings into improved outreach
practices
24
Results Hospital Survey
Summary
14 Rhode Island hospitals have cancer registries
Requested participation of hospital cancer
registrars
Respondents signed consent form
1 initial mailing and two respondent follow-ups
11 hospital cancer registries responded (79%
response rate)
Findings presented here exclude non-respondents
Respondent follow-up ended 04-16-07
2 hospitals share the same reporting system
25
Results Hospital Survey
Forms
Do hospitals with CRs have intake forms (IF)
specific for cancer?
Is information on cancer risk factors obtained via
cancer IF?
5 hospitals (45%) have an IF for cancer; 5 hospitals
(45%) use a general IF
3 hospitals (30%) collect cancer risk factor
information on these forms
Is other form used to obtain cancer risk factor
information?
6 hospitals (56%) use other sources
26
Results Hospital Survey (cont.)
Forms
What form is used to determine the race of cancer
patients?
10% (1 out of 4 with cancer intake forms) ask for race
5 hospitals (50%) ask for race thru general intake
forms
4 hospitals (40%) use other forms (medical records,
physical exam, medical history)
27
Results Hospital Survey (cont.)
Forms
What form is used to determine the ethnicity of
cancer patients?
11% (1 out of 4 with cancer intake forms)
4 hospitals (44%) use general intake forms
4 hospitals (44%) use other forms
2 hospitals did not respond to this question
28
Results Hospital Survey (cont.)
Forms
From findings presented so far: race (50%) and
ethnicity (56%) are obtained via intake forms by
40% of hospitals with cancer registries
How many races can be checked in the intake form?
1 race: 1 hospital (9%)
2 races: 2 hospitals (18%)
2 or more races: 4 hospitals (36%)
4 hospitals (36%) do not collect this information in the
intake form
29
Results Hospital Survey (cont.)
Forms
What form is used to determine the country of
origin of cancer patients?
1 hospital (11%) uses the cancer intake form to
obtain information on country of origin
8 hospitals (88%) reported not collecting information
on country of origin
30
Results Hospital Survey (cont.)
Race/Ethnicity/Country of Origin
How is the race of a cancer patient determined?
(multiple response question)
a. In 4 hospitals (36%) patient self-reports race
b. In 5 hospitals (45%) person filling form asks patient
c. In 1 hospital (9%) person filling form makes a
guess
d. In 4 hospitals (36%) information obtained from MR
e. 1 hospital (9%) uses medical history/physical exam
2 hospitals use a combination of methods b and d
1 hospital use a combination of methods b and c
31
Results Hospital Survey (cont.)
Race/Ethnicity/Country of Origin
How is the ethnicity of a cancer patient
determined? (multiple response question)
a. 3 hospitals (27%) present the patient with a
standard form and the question “Are you from
Hispanic or Latino origin”
b. 6 hospitals (56%) ask patients about their
ethnicity in a non-standard way
c. 2 hospitals (18%) reported “unknown”
32
Results Hospital Survey (cont.)
Access to Cancer Data
Does the intake form have a number that
can be used to link with the patient medical
record?
All (7 out of 7 hospitals with intake forms) can link
to MRs
Do hospitals aggregate information of cancer
cases (from intake and other forms) into a database?
All (11 out of 11 hospitals)
33
Results Hospital Survey (cont.)
Access to Cancer Data
Is cancer data available to researchers from
academic institutions?
10 hospitals (91%) provide access to cancer data
Is there an internal procedure to obtain access?
10 hospitals (91%) have an IRB process in place
34
Results Hospital Survey (cont.)
Reporting
To what organizations do you report cancer data?
5 hospitals (45%) to NCBD (National CAHPS
Benchmarking Database); yearly
1 hospital (9%) to CIRF (Cancer Information Reference
File); yearly
4 hospitals (36%) to the National Cancer Database
of the American College of Surgeons (ACS); yearly
All to the Rhode Island Cancer Registry (RICR);
monthly
All to the Hospital Association of Rhode Island
(HARI); monthly
35
Results Hospital Survey (cont.)
Reporting
Is the data you report linked to medical
records?
4 hospitals (37%) have MR linkage capacity
36
Key Informant Results
TOPIC
HARI
RI CR
ACS
Who reports; 13 hospital CR; 21 Century
13 hospital CR; 6 state CR; Americans for Non-Smokers
how often?
Oncology Services (monthly) pathology labs; surgical
Rights; National Breast and
centers (monthly)
Cervical Cancer Early
Detection Program; Tobacco
Free Kids; National Committee
for Quality Assurance (NCQA),
Colorectal Screenings in
Health Plans; UC Berkeley,
Smoking Cessation for
Medicaid; Smoke Free
Grade, American Lung
Association; Cancer Incidence
Rates, National Program of
Cancer Registries; Behavioral
Risk Factor Surveillance
System (BRFSS) and Youth
Risk Factor Surveillance
System (YRBFS) (all yearly)
37
Key Informant Results (cont.)
TOPIC
Data format
Database
Specifications
HARI
single format: North
American Association
of Central Cancer
Registries (NAACCR)
RI CR
single format (NAACCR)
ACS
multiple formats (excel,
word, pdf and text)
www.naaccr.org
www.naaccr.org
Cancer risk
Factors (RF)
smoking status and
occupation
no cancer RF data
smoking; Thompson Medstat
PULSE Survey
Screening
no screening data
no screening data
breast and colorectal screen.
yes
yes
most report 1 race
most report 1 race (“can
report up to six”)
Race/ethnicity
# of races
Reported
Information on
how is race
determined?
no
some; “this is a variable in
the NCAACCR data set, but
HEALTH does not require it”
n/a
yes
most report 1 race
no
38
Key Informant Results (cont.)
TOPIC
Information on
how ethnicity is
determined?
Do you provide
guidance to
comply with
OMB 15?
HARI
RI CR
ACS
no
no
a few
“Race and ethnicity treated When directed by feds
like all other data items
When required by state
required by law”
When requested by hospitals
When required by RI HARI
Target year for
helping partners
achieve OMB 15
compliance?
no
Organizations
you report data
to?
-NAACCR
-National Program of
Cancer Registries (CDC)
-RI Department of Health
-Central Brain Tumor
Registry of United
Sates (CBTRUS); (yearly)
“It’s being discussed but
no date yet”
“We only look at race/
ethnicity data for
screening data and data
coming from CDC”
no response
-NAACCR
-ACS Divisions (monthly)
-CDC
-ACS Management (as
-International Association
needed)
of Cancer Registries (IACR)
(yearly)
39
Key Informant Results (cont.)
TOPIC
HARI
RI CR
ACS
Uses Spanish
Uses NAACCR algorithm “We can do this, but as a “We don’t go into that level
surname method (last name + place of birth) separate variable…we
of detail in our reporting.
to adjust
to determine Hispanic
don’t report it to CDC,
We rely on external
under-reporting?
sources”
How do you
report data on
race/ethnicity?
1 race + ethnicity (“2nd
racial group almost never
recorded in medical
Do you keep
identifiers to link
cancer registry
and MR data?
yes
Is your data
available to
researchers?
yes
“Several formats per
request“
yes
yes
IRB approval may be
necessary
1 race + ethnicity
“Not to my knowledge. We
have worked with ACoS to
determine underserved in
hospitals but it is not
possible to link up”
“We dot provide this data
at this time”
40
Key Informant Results (cont.)
TOPIC
HARI
How many data
inquiries and
research
proposals do
receive per year?
blank
Is there an IRB
procedure?
yes
What cancer
reports does your
organization
produce?
Do you report
cancer data
within your
organization?
Contact Dr. John Fulton
yes
RI CR
6 data inquiries
[Note: research proposals
are handled by the IRB
Committee]
yes
-Annual Report
-RI Cancer Burden
(annual update)
-Special reports (ongoing)
yes
ACS
0
0
blank
-Mission Scorecard (yearly)
-Mission Dashboard (yearly)
-Best Practices Reports
(ongoing)
yes
41
RI CR Assessment: Start/End-Points
Cancer
Patients
Hospitals
CRs
Kent H.
1
Landmark MC
2
Memorial H RI
3
Miriam H.
4
Naval HC NE
5
Newport H.
6
Rhode Island H.
7
R. Williams MC
8
South County H.
9
Colorectal Screenings
in Health Plans
NCQA
21st Century
Oncology Services
ALA
BRFSS
YBRFS
CDC
NAACCR
NPCR
HARI
CBTRUS
ACS
CDC
RI CR
NAACCR
IACR
St .Joseph H.
10
ACS
Divisions
ACS
Management
6 states
Our Lady of Fatima
VA MC
11
Westerly H.
12
Women & Infants.
13
pathology labs,
surgical centers
Cancer
Patients/Cancer
Families/Communities
42
Compliance with OMB 15?
Findings from the hospital cancer registry survey
indicated lack of uniform collection of
information on race and ethnicity.
Only a few hospitals ask patients to self-report
their race (27%) and ethnicity (27%) according to
the guideline
Only a few hospitals (36%) collect information on 2
or more races
The Key Informant Interview revealed that most
cancer data sources only report 1 race
There is no target date yet in RI for achieving
OMB 15 compliance with cancer data reporting
43
Summary of Findings
1. Use of Intake Forms (IFs) specific to cancer
2. Use of IFs (cancer-specific and general)
3. Use of IFs to obtain cancer risk factor (CRF) information
4. Use of other forms to obtain CRF information
5. Collection of CRF information from all sources
6. Is reporting of CRF information a RI CR requirement?
7. Race information obtained at point of entry (POE)
8. Race information obtained from other sources
9. Ethnicity information obtained at POE
10. Ethnicity information obtained from other sources
11. RI CR mechanism to monitor OMB 15 compliance?
12. Collection of information on country of origin
13. IF/MR linkage
14. RI CR/MR linkage
15. Patient self-reporting of race and ethnicity
16. Collection of information of two or more races
17. Reporting of two or more races
18. Target date for OMB 15 compliance?
19. Cancer data available/IRB process in place
44
Summary of Findings
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
45%
91%
30%
60%
90%
NO
60%
40%
56%
44%
NO
12%
100%
??
37%
27%
36%
9%
NO
91%
0%
10%
20%
Hospital Cancer Registry Survey
Key Informant Interview
30%
40%
50%
60%
70%
80%
90%
100%
45
Implications of Non-Compliance
Undercount of multi-racial ethnic groups;
over count of non multi-racial ethnic groups
Underestimates in cancer prevalence,
incidence and mortality
Problem may not be limited to the RI CRS;
other RI/national data sets may be affected
Difficult targeting and outreach of multiracial ethnic groups
Misdirecting resources for targeting/outreach
Potential for unsound public health practice
46
Other Study Conclusions
All = 100%; Most = 70-99%; Some = 40-69%; A few = 1-39%; None = 0%
The form closest to the information entry point is
the IF
Most hospitals (91%) reported use of IFs
A few hospitals (30%) use IFs to collect information
on cancer risk factors (RFs)
Some hospitals (60%) use other forms
While some cancer RF information is obtained at
the point of entry, the Key Informant Survey
revealed no reporting requirement
47
Other Study Conclusions (cont.)
All = 100%; Most = 70-99%; Some = 40-69%; A few = 1-39%; None = 0%
About
half of the information on race and ethnicity
obtained at the point of entry (IF)
Some hospitals (60%) obtain race information thru
IFs; some hospitals (56%) use IFs for ethnicity
A few hospitals (40%) use other forms to determine
race and some (44%) for ethnicity
The Key Informant Survey determined that:
There is no mechanism in the RI CR system to
monitor OMB 15 compliance
No information is reported on how race and
ethnicity data are collected
48
Other Study Conclusions (cont.)
All = 100%; Most = 70-99%; Some = 40-69%; A few = 1-39%; None = 0%
Most hospitals (88%) do not collect information
on country of origin
1 asks for country of origin
1 asks for main language
All hospitals with IFs can link back to MRs; only
a few hospitals (37%) can link back to MRs data
reported to the RI CR
However, HARI and the RI CR require a field in
reported databases that can be used to link to MRs
49
Can the Problem be Fixed?
Our recommendations:
Improve data quality at the point of entry
and monitor data quality collection and
reporting on key variables including race and
ethnicity
In the interim, promote use and refinement of
the Spanish Surname Algorithm to determine
the extent of the undercount of Hispanic
cancer cases and institutionalize dissemination
of more reliable cancer statistics
50
Apply the Guideline Please
51
Contact Information
Alvaro M Tinajero, MD, MPH, ScM
[email protected]
[email protected]
[email protected]
Mercedes (Betty) Bernal
[email protected]
52