Smart Contact Tracing
Smart algorithms for reducing the time spent on data entry, data matching, and information retrieval
Airport Border Screening Case Study
Fuzzy name matching enables lookup of new COVID-19 cases against recent travel manifests to find potential close contacts in-flight.
A seamless UI for border agents to match health data with travel records
NLP accelerates the largely manual contact tracing process hampered by disconnected information systems and the inability to spot when two contact tracers have the same person on their lists.
As flu season approaches, and with it the anticipated “second wave of COVID-19” coupled with the reopening of schools and universities, NLP delivers the boost that contact tracers need:
- Automatically matching names of new cases to plane, bus, and train passenger information or school rosters
- Enabling search across disparate systems, such as a state department of public health system, hospital records system, and a Salesforce database used by contact tracers, while adhering to system access permissions
- Detecting the same person on different contact lists through fuzzy name matching. In any outbreak, there is overlap among close contact cases, and multiple tracers may waste efforts trying to call the same person. A hastily spelled name and a phone number may be the only information contact tracers have.
“The availability of passenger locator data, particularly for airline passengers, is extremely important for the success and effectiveness of contact tracing operations for communicable diseases.”
Lookup COVID-19 patient names in rosters
A newly diagnosed case may have recently traveled or attended a college party or wedding. Fuzzy name matching by Rosette® helps automate the lookup of the case name against air/bus/train passenger lists or school rosters to swiftly provide contact tracers with the contact information of potentially close contacts.
Search across unconnected information systems
State and local health boards often keep records in a different system (e.g., Maven, NBS, Sunquest, Epi Trax) than what contact tracers are using (e.g., Salesforce). Instead of integrating or importing/exporting data, KonaSearch by Basis Technology finds all the search results from Salesforce and other systems while adhering to system permissions, so that contact tracers only see the data they are allowed to see, so that privacy and confidentiality is maintained.
Spot duplicate names on different contact lists
The European CDC estimates contact tracers may spend up to seven hours tracing each case. Overlapping networks of contacts at one hotspot means multiple tracers may be attempting to reach the same person. Transcription errors and duplicate entries make it hard to spot the same names on different lists. Name matching by Rosette overcomes entry errors to find matches and avoid redundant phone calls.
“[You] don’t check into a hospital with your passport number. When you have no common identifiers between health and travel records besides name and date of birth, you have to go with the best name matching you can get, and for SITA that is Rosette from Basis Technology.”