BACKGROUND: South Dakota is one of many jurisdictions (e.g. CT, MA, MN, NYC, NC) using the Maven disease surveillance and case management system. Although the core functionality of Maven is the same for all jurisdictions, in South Dakota we have been able to utilize the South Dakota Electronic Disease Surveillance System (SD EDSS) for enhancing foodborne disease investigations. South Dakota built and implemented a NHGQ survey for case investigations of Salmonella, STEC and Listeria (SSL). The design supports high-quality data capture in an efficient, low cost manner requiring minimal staff time, and enables xml extracts to be shared with CDC often before PFGE and other methods of characterizing the isolate are available.
METHODS: South Dakota initiates investigations for all cases of SSL within 3 days of receipt of positive laboratory test results. During the patient interview, email and mobile phone numbers are requested and entered in SD EDSS. SD EDSS generates a web link specific to the case and sends it via State email client server for case/self-administration. Epidemiologists use automated workflows for email/mobile phone entry-notification, and sending and reviewing completed NHGQ surveys. Epidemiologists extract completed questionnaires in the NHGQ compliant xml format. Using a document editor, confidential person and local business-identifying information in comment sections are redacted. Finally, xml extracts are sent to CDC for inclusion in NHGQ-SAS datasets.
RESULTS: A response rate of 29.3% (133/454) was observed for case-submitted NHGQ surveys between 07/20/2013 and 10/27/2016. Of these, 72.2% (96/133) were completed within 3 days of email delivery. Nearly sixty percent, 59.4% (79/133) answered at least 90% (352/391) of questions with a valid value. Stratified by disease, the response rate for Salmonella was 28.4% (86/303), and for STEC 31.1% (47/151). Stratified by season, the response rates were: Spring (Mar-May) 22.9% (19/83); Summer (Jun-Aug) 28.0% (53/189), Fall (Sep-Nov) 32.3% (42/130) and Winter (Dec-Feb) 36.5% (19/52).
CONCLUSIONS: South Dakota developed the capability to obtain supplemental NHGQ histories from cases of SSL as an effective, timely and inexpensive means of data capture in SD EDSS. The technology enables cases to receive the questionnaires during optimal recall for maximizing data quality. Response rates are favorable toward continued use and development. Caveats include selection bias among cases (i.e. those having experience with, and ready-access to, email and web technologies). Disease type and season may influence response rates.