Congenital Malformations Registry - Summary Report
Section VI: Current Topics
Introduction
There are no national data on birth defects despite it being a leading cause of infant mortality and a major cause of mortality and morbidity throughout childhood (Mathews 2007, Hoyert 2006). Data on birth defect prevalences generally come from birth defects registries maintained by specific states. National prevalence estimates have been done in the past using hospital discharge data (Edmonds 1990). The National Birth Defects Prevention Network (NBDPN) is an organization of state birth defects surveillance programs which collects data from 34 state population-based surveillance systems. Recently the Centers for Disease Control and Prevention (CDC) and the NBDPN developed national prevalence estimates for 21 selected major birth defects (Canfield 2006). The defects were chosen because they are recognizable at or shortly after birth and they are likely to be ascertained similarly across states. After examining the data, the NBDPN chose to present numbers using only data from the 11 state registries which used 'active' ascertainment. 'Active' ascertainment systems use field staff to visit hospitals to ascertain and abstract information on cases. These registries were thought to provide more 'reliable and valid' data for the estimates. The national estimates are useful and can be used as a standard for comparisons among registries, for health care planning and to evaluate interventions such as folic acid.
The CMR relies on hospital reporting and thus is termed a 'passive' registry; therefore data from the CMR were not used in creating the national prevalence estimates. We will be using these estimates as a benchmark for our registry (see Section 5). CMR staff recognizes that completeness, accuracy and timeliness are the hallmarks of a good surveillance system. However, these attributes exist in tension, "conflicting principles" (Kallen 1988). Steps taken to improve completeness and accuracy may actually reduce timeliness. From the very beginning, the CMR has built in procedures to improve the quality of the data in the CMR. These systems have changed over time (Sekhobo, Druschel 2001) and the CMR now has three major approaches to improving data quality: 1) matching to hospital discharge data, the Statewide Planning and Research Cooperative System (SPARCS) for completeness; 2) the web-based reporting system, the Health Provider Network (HPN) for timeliness and completeness; 3) on-site hospital audits for completeness and accuracy. In addition, we also periodically request medical records and compare them to the hospital's report for an additional review of accuracy.
CMR staff have developed on-line SAS/IntrNet applications which empower the users to search and retrieve hospital submitted cases, generate real-time reports and perform simple statistical analysis using the CMR's database. For instance, CMR staff can select a reporting hospital and discharge years of interest and then, generate a real-time report table which lists the number of cases by discharge year and month. By reviewing this report, CMR staff is able to see if the hospital has been submitting an appropriate number of cases routinely or if the hospital stopped or skipped reporting for certain months or years. A study that evaluated the completeness of submitted case information and timeliness of reporting to the CMR and the effectiveness of the HPN communication and query system when compared to the previous manual, paper-based system found that the implementation of the HPN system has resulted in more timely submission of cases and promoted effective communication between the CMR and reporting hospitals. There was a nearly 50% reduction in median days used for reporting. (Wang 2007b).References
