Monday, October 27, 2003 - 2:15 PM
3901

This presentation is part of A2: Development and Implementation of Data Quality Monitoring Processes

Data Quality for 23 Million Immunizations; Is it Possible?

Marlene Lugg, Kaiser Permanente, 13652 Cantara St, South 1 Bldg, Panorama City, CA, USA, Constance Vadheim, Research and Education Institute, Center for Vaccine Studies, Harbor/UCLA Bldg. RB3, 1124 West Carson St, Torrance, CA, USA, Allan S. Lieberthal, Panorama City, Kaiser Permanente, 13652 Cantara St. South 1 Bldg, Panorama City, CA, USA, and Stephen Tannenbaum, Pediatrics, Southern California Permanente Medical Group, 4700 Sunset Blvd, Los Angeles, CA, USA.


KEYWORDS:
Automated Data Quality
Large Registries

BACKGROUND:
As data begins to flow in to a new registry, quality often reverts to a lower priority. Hopefully, it soon becomes an on-going part of the registry.
The Kaiser Permanente Immunization Tracking System (KITS) management team began working on quality issues as the system was being implemented in the 11 Southern California Medical Centers, which serve over 3 million members.

OBJECTIVE:
The Kaiser Permanente Immunization Tracking System (KITS) management team began working on quality issues as the system was being implemented in the 11 Southern California Medical Centers, which serve over 3 million members.
The development of data quality measures included routine automated data analysis and special analysis carried out by immunization coordinators at medical center and office practice levels.

METHOD:
Data was collected by programming for automated monthly/quarterly printouts which were then analyzed by Immunization COordinators centrally and/or at each Medical Center.

RESULT:
Routine monthly or quarterly data printouts include: (1) immunizations given in the arm to infants under 1 year old, (2) expired lot numbers, (3) lot numbers whose "masks" do not match appropriate manufacturers, (4) duplicate immunizations recorded for the same date, (5) wrong agent abbreviations, (6) children with more than the recommended number of immunizations [i.e., 4 MMR's, etc.], (7) persons not up-to-date for age, and (8) possible duplication of members. (De-duplication is probably the biggest quality issue facing most registries that receive data from various sources.)
Some of these data reports require further investigation locally, using patient or provider records. Working with personnel involved and educating staff, as well as modifying KITS data entry screens and procedures has improved data quality as it is an often invisible on-going part of the process.

CONCLUSION:
This paper dicusses the usefulness of each of the quality items noted above, and shows how other registries can make use of similar data quality measures.


LEARNING OBJECTIVES:
How qaulity assurance procedures can be used in small and large immunization registries

Back to Development and Implementation of Data Quality Monitoring Processes
Back to The 2003 Immunization Registry Conference (October 27-29, 2003)