State of GIS Activities Among NAACCR Member Registries

1
State of GIS Activities Among NAACCR Member
Registries

• NAACCR GIS Committee
• Recinda L Sherman, MPH, CTR
• NAACCR Annual Meeting,
• Detroit, Michigan
• June 7, 2007

2
State of GIS Activities Among NAACCR Member
Registries

• Two surveys were conducted
• 1999 by NPCR
• Ric Skinner, formerly with NJ Cancer Registry
• 2005 by NAACCR
• NAACCR GIS Committee

3
Rationale/Benefits 1999

• Rationale for survey
• Desire to support local cancer control activities
  at geographic levels lower than county
• To provide tools to improve understanding of
  effects of cancer on populations within state
• To support small area analysis
• Increase data quality
• QA/QC of address data and for geocoding

• Reported Benefits
• Forces QA/QC of demographic data when geocoding
• New ways of presenting data
• Decreases response time to data requests
• Supports cancer prevention and control activities
• Enables additional research
• Time distance studies for services
• Cluster investigation
• Hypothesis generation
• Merge with environmental and sociodemographic
  factors

4
Extent of Activities(1999 2005)

• 1999 Survey
• 44 currently geocoding
• 40 currently using GIS
• 68 Registries Contacted
• 44 responded (65)

• 2005 Survey
• 82 currently geocoding
• 62 currently using GIS
• 72 Registries Contacted
• 47 Responded (65)

5
Geocoding Data Quality(1999 2005)

• 2005 Survey
• 83 of those geocoding clean addresses
• Variety of methods of data cleaning
• None, parsing, USPS standardization
• Variety of methods used for geocoding
• No interactive, multiple street reference files,
  level of geocoding (zip code, street, city), hand
  geocoding
• Variety of software and underlying street
  reference files used

• 1999 Survey
• 70 of those geocoding clean addresses
• Correct misspellings, correct no matches after
  batch match, correct for NAACCR edits, in-house
  standards, addresses standardization
• Use phone books, paper maps, Zip code directories

6
Geocoding Results 2005

• On average, the number of geocodes attempted each
  year is 1.7 times the annual case load
• On average, 16 of attempted geocodes failed the
  batch match
• On average, 7 of attempted geocodes were
  ungeocodable

7
GIS Implementation Obstacles(1999 2005)

• Data quality
• Institutional support
• IT issues
• Qualified personnel
• Training

8
Summation 1999 to 2005

• Considerable growth in geocoding/GIS activities
  since 1999
• Current needs have also grown
• Standardization/best practices
• Each registry is a renegade
• Training
• Software, basic GIS concepts/techniques
• Applicability of data use
• Interpretability of data results
• Spatial Analysis methodology
• Institutional support
• Institutional investment

9
Training Priorities 2005

• Ranked priorities for training
• 1. Geocoding methodological issues
• 2. GIS applications in cancer control
• 3. Basic mapping methodological issues
• 4. Spatial analysis
• 5. Web-based GIS applications
• 6. Data/map resources
• 90 would benefit from on-line training

10
Future Priorities 2005

• Ranked priorities for future GIS activities
• 1. Geocoding, address cleaning,
  standardization
• 2. Cartography
• 3. Spatial statistical analysis
• 4. Application for GIS in healthcare cancer
  control
• 5. Internet mapping applications

11
Key Point 1

• Efforts needed to improve/standardize geocoding
• A variety of methods is used to prepare and
  process data for geocoding.
• A variety of software and reference files is used
  for geocoding and GIS.    
• This indicates a need for standardization of
  methods across registries to ensure comparable
  data (demonstrated need for Geocoding Guidelines).

12
Key Point 2

• Lack of institutional support
• Of the registries currently geocoding and using
  GIS
• 63 have staff with geocoding or GIS in their job
  descriptions
• Of these with GIS or geocoding in job
  descriptions, the average percent of time
  allocated is 28
• 58 feel the level of support for GIS activities
  is not sufficient
• This is for current activity level only
• This may indicate lack of institutional support.

13
Key Point 3

• Training needs are immediate and long-term
• The majority of registries have the data
  available in geocoded form but not all these
  registries are currently using the data in a GIS.
  
• The majority of registries not currently using
  GIS have implementation plans.
• This likely indicates evolving training needs.

14
Geographical Information Systems

• a system of hardware, software, data, people,
  organizations and institutional arrangements for
  collecting, storing, analyzing and disseminating
  information about areas of the earth
• Dueker and Kjerne, 1989

15
Acknowledgements

• Results from the 1999 GIS survey were provided
  by Ric Skinner, Baystate Health (previously with
  the NJ State Cancer Registry)
• NAACCR GIS Committee