Babesiosis National Surveillance: Data Uniformity, Completeness, and Timeliness, 2011–2013

BACKGROUND: Babesiosis became a nationally notifiable condition in 2011; states in which babesiosis is a reportable condition submit surveillance data to CDC using the Nationally Notifiable Diseases Surveillance System (NNDSS). To evaluate the babesiosis surveillance system, we reviewed the first 3 years of national data.

METHODS: Core variable data (e.g., demographics and county of residence) were submitted electronically via NNDSS. In addition, CDC requested data for supplemental variables (e.g., symptoms and route of exposure), which were submitted directly from health departments to CDC’s Parasitic Diseases Branch (PDB) as either paper case report forms or exported line lists from state surveillance systems. Currently, no mechanism exists in NNDSS for electronic messaging of supplemental data.

RESULTS: Since 2011, CDC has been notified of 3,797 confirmed and probable babesiosis cases. Core variable data on sex and county of residence were available for 97% (3,675/3,797) and 99% (3,770/3,797) of cases, respectively. Supplemental data were received for 84% of cases (3,176/3,797). For most of these cases (96%; 3,059/3,176), supplemental data were provided in a line list, which may not have included all of the requested variables. The median time from illness onset or diagnosis to the initial report of core variables to CDC was 112 days (range: 4–718) and from illness onset or diagnosis to submission of supplemental data was 353 days (range: 38–1,016). Data necessary to classify cases (e.g., laboratory results and clinical signs/symptoms) were more complete than data for other variables; for example, symptom data were provided for 93% of cases with supplemental data (2,953/3,176). Data regarding potential routes of exposure (e.g., tick bites or transfusions) were available for only 25% (785/3,176). Data indicating the likelihood that cases were acquired via transfusion were available for only 18% (571/3,176); some line lists included data on recent transfusions but without the requested assessment of the likelihood of transfusion transmission.

CONCLUSIONS: Core variable data had higher levels of completeness compared with data for the supplemental variables. The fact that some exported line lists did not include all of the supplemental variables may indicate that those variables have not been incorporated into state surveillance systems. Electronic messaging for supplemental variables might improve data uniformity, completeness, and timeliness. CDC’s PDB plans to evaluate variables with high proportions of missing data to provide insights into the variables’ utility at state and national levels.