Meghan A Baker1, Michael Nguyen2, David V Cole3, Grace M Lee4, Tracy A Lieu5. 1. Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School, Boston, MA, United States; Division of Infectious Diseases, Department of Medicine, Brigham and Women's Hospital, Boston, MA, United States. Electronic address: meghan_baker@harvardpilgrim.org. 2. US Food and Drug Administration Center for Biologics Evaluation and Research, Rockville, MD, United States. Electronic address: Michael.Nguyen@fda.hhs.gov. 3. Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School, Boston, MA, United States. Electronic address: david_cole@harvardpilgrim.org. 4. Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School, Boston, MA, United States; Division of Infectious Diseases, Department of Medicine, Boston Children's Hospital, Boston, MA, United States. Electronic address: Grace.Lee@childrens.harvard.edu. 5. Department of Population Medicine, Harvard Pilgrim Health Care Institute and Harvard Medical School, Boston, MA, United States. Electronic address: tracy.lieu@kp.org.
Abstract
BACKGROUND: The Post-Licensure Rapid Immunization Safety Monitoring (PRISM) program is the immunization safety monitoring component of FDA's Mini-Sentinel project, a program to actively monitor the safety of medical products using electronic health information. FDA sought to assess the surveillance capabilities of this large claims-based distributed database for vaccine safety surveillance by characterizing the underlying data. METHODS: We characterized data available on vaccine exposures in PRISM, estimated how much additional data was gained by matching with select state and local immunization registries, and compared vaccination coverage estimates based on PRISM data with other available data sources. We generated rates of computerized codes representing potential health outcomes relevant to vaccine safety monitoring. Standardized algorithms including ICD-9 codes, number of codes required, exclusion criteria and location of the encounter were used to obtain the background rates. RESULTS: The majority of the vaccines routinely administered to infants, children, adolescents and adults were well captured by claims data. Immunization registry data in up to seven states comprised between 5% and 9% of data for all vaccine categories with the exception of 10% for hepatitis B and 3% and 4% for rotavirus and zoster respectively. Vaccination coverage estimates based on PRISM's computerized data were similar to but lower than coverage estimates from the National Immunization Survey and Healthcare Effectiveness Data and Information Set. For the 25 health outcomes of interest studied, the rates of potential outcomes based on ICD-9 codes were generally higher than rates described in the literature, which are typically clinically confirmed cases. CONCLUSION: PRISM program's data on vaccine exposures and health outcomes appear complete enough to support robust safety monitoring.
BACKGROUND: The Post-Licensure Rapid Immunization Safety Monitoring (PRISM) program is the immunization safety monitoring component of FDA's Mini-Sentinel project, a program to actively monitor the safety of medical products using electronic health information. FDA sought to assess the surveillance capabilities of this large claims-based distributed database for vaccine safety surveillance by characterizing the underlying data. METHODS: We characterized data available on vaccine exposures in PRISM, estimated how much additional data was gained by matching with select state and local immunization registries, and compared vaccination coverage estimates based on PRISM data with other available data sources. We generated rates of computerized codes representing potential health outcomes relevant to vaccine safety monitoring. Standardized algorithms including ICD-9 codes, number of codes required, exclusion criteria and location of the encounter were used to obtain the background rates. RESULTS: The majority of the vaccines routinely administered to infants, children, adolescents and adults were well captured by claims data. Immunization registry data in up to seven states comprised between 5% and 9% of data for all vaccine categories with the exception of 10% for hepatitis B and 3% and 4% for rotavirus and zoster respectively. Vaccination coverage estimates based on PRISM's computerized data were similar to but lower than coverage estimates from the National Immunization Survey and Healthcare Effectiveness Data and Information Set. For the 25 health outcomes of interest studied, the rates of potential outcomes based on ICD-9 codes were generally higher than rates described in the literature, which are typically clinically confirmed cases. CONCLUSION: PRISM program's data on vaccine exposures and health outcomes appear complete enough to support robust safety monitoring.
Keywords:
CDC; CPT; CVX; Centers for Disease Control and Prevention; Codes for Vaccine Administered; Common Procedural Terminology; DTaP; Data characterization; ED; FDA; Food and Drug Administration; HEDIS; HPV; Health outcome; Healthcare Effectiveness Data and Information Set; Hib; ICD; IPV; International Classification of Diseases; MCV4; MMR; Mini-Sentinel; NIS; National Immunization Survey; PCV; PPV; PRISM; Post-Licensure Rapid Immunization Safety Monitoring; Surveillance; Td; Tdap; Vaccine; combined tetanus, diphtheria and pertussis; diphtheria, tetanus and pertussis; emergency department; haemophilus influenzae type B; human papillomavirus (HPV); inactivated polio vaccine; measles, mumps and rubella; meningococcal conjugate vaccine; pneumococcal conjugate vaccine; pneumococcal polysaccharide vaccine; tetanus and diphtheria
Authors: Christian Hoppe; Patrick Obermeier; Susann Muehlhans; Maren Alchikh; Lea Seeber; Franziska Tief; Katharina Karsch; Xi Chen; Sindy Boettcher; Sabine Diedrich; Tim Conrad; Bron Kisler; Barbara Rath Journal: Drug Saf Date: 2016-10 Impact factor: 5.606
Authors: Sophia R Newcomer; Martin Kulldorff; Stan Xu; Matthew F Daley; Bruce Fireman; Edwin Lewis; Jason M Glanz Journal: Pharmacoepidemiol Drug Saf Date: 2018-01-02 Impact factor: 2.890
Authors: Xintong Li; Anna Ostropolets; Rupa Makadia; Azza Shoaibi; Gowtham Rao; Anthony G Sena; Eugenia Martinez-Hernandez; Antonella Delmestri; Katia Verhamme; Peter R Rijnbeek; Talita Duarte-Salles; Marc A Suchard; Patrick B Ryan; George Hripcsak; Daniel Prieto-Alhambra Journal: BMJ Date: 2021-06-14
Authors: Benedikt F H Becker; Paul Avillach; Silvana Romio; Erik M van Mulligen; Daniel Weibel; Miriam C J M Sturkenboom; Jan A Kors Journal: Pharmacoepidemiol Drug Saf Date: 2017-06-28 Impact factor: 2.890
Authors: Sonali Kochhar; Jean-Louis Excler; Karin Bok; Marc Gurwith; Michael M McNeil; Stephen J Seligman; Najwa Khuri-Bulos; Bettina Klug; Marian Laderoute; James S Robertson; Vidisha Singh; Robert T Chen Journal: Vaccine Date: 2018-11-26 Impact factor: 3.641