Evaluation of Legionellosis Surveillance in Virginia, 2005-2013

BACKGROUND:  Virginia state regulations require physicians, medical directors, and laboratory directors to report all suspected or confirmed legionellosis cases to the Virginia Department of Health (VDH) within three days of identification.  VDH then investigates to confirm the case definition is met and to ascertain demographic, risk factor, and exposure information.  Since 2005, legionellosis incidence in Virginia has more than doubled, from 55 cases in 2005 to 118 cases in 2013.  Most of these cases are not associated with a known outbreak, and questions have arisen from VDH personnel about the usefulness of the exposure information that is collected for each case.  A study was conducted to evaluate the legionellosis surveillance system in Virginia with a focus on risk factor and exposure information.  

METHODS:  Qualitative and quantitative data sources were examined in accordance with current Centers for Disease Control and Prevention public health surveillance system evaluation guidelines.  Data collected in the Virginia Electronic Disease Surveillance System (VEDSS) were analyzed to evaluate several system attributes including simplicity, flexibility, acceptability, and data quality.  Qualitative data were gathered from state and local health department stakeholders to describe system performance, advantages, and challenges.  Additionally, case report forms from 2012 were examined for completeness and usefulness in identifying possible exposure sources.  Data were analyzed using Epi Info 7.1, Microsoft Excel 7, and SAS 9.3 software. 

RESULTS:  Legionellosis surveillance and investigation is relatively simple and well integrated with the rest of notifiable disease surveillance in Virginia.  The biggest challenges stakeholders identified were that VEDSS does not record exposure information and VEDSS lacks the flexibility to create disease-specific modules.  Data quality issues, especially among laboratory data, were identified.  Among the 39 VEDSS variables that were classified as important for legionellosis surveillance, on average the data were found to be missing or unknown 16% of the time.  In 2012, the state health department received 65 (86%) of the 76 confirmed case report forms, of which 48% had “no” or “unknown” selected for every exposure question.

CONCLUSIONS:  While legionellosis surveillance is considered simple and acceptable among stakeholders, issues with data quality and limited exposure identification lead to questions about its usefulness in assessing trends and identifying possible outbreaks.  Recommendations to improve legionellosis surveillance in Virginia include: 1) replace the current case report form with the VDH Legionellosis Cluster Questionnaire; 2) provide disease-specific VEDSS data entry guidance to the field; and 3) further explore the capacity for adding more risk factor and exposure variable fields in VEDSS.