Record linkage: making maximum use of the discriminating power of identifying information
Communications of the ACM
Efficient Transitive Closure Algorithms
VLDB '88 Proceedings of the 14th International Conference on Very Large Data Bases
Transitive Closure and the LOGA+-Strategy for its Efficient Evaluation
MFDBS '89 Proceedings of the 2nd Symposium on Mathematical Fundamentals of Database Systems
Febrl: a freely available record linkage system with a graphical user interface
HDKM '08 Proceedings of the second Australasian workshop on Health data and knowledge management - Volume 80
Automatic cervical cell segmentation and classification in Pap smears
Computer Methods and Programs in Biomedicine
Hi-index | 0.00 |
This paper aims at to present the integration of the files of the Brazilian Cervical Cancer Information System (SISCOLO) in order to identify all women in the system. SISCOLO has the exam as the unit of observation and the women are not uniquely identified. It has two main tables: histology and cytology, containing the histological and cytological examinations of women, respectively. In this study, data from June 2006 to December 2009 were used. Each table was linked with itself and with the other through record linkage methods. The integration identified 6236 women in the histology table and 1,678,993 in the cytology table. 5324 women from the histology table had records in the cytology table. The sensitivities were above 90% and the specificities and precisions near 100%. This study showed that it is possible to integrate SISCOLO to produce indicators for the evaluation of the cervical cancer screening programme taking the woman as the unit of observation.