A Tool That Turns Infectious-Disease Statistics into On-the-Ground Alerts

A lookup table became material for commercial services

The Korea Disease Control and Prevention Agency (질병관리청), South Korea’s national disease-control authority, operates a public API, or software connection point, that lets programs automatically retrieve notified infectious-disease data from its integrated disease-control information system. It offers eight query functions: period, patient classification, domestic or overseas infection area, region, age, sex, cases by infectious disease, and deaths by infectious disease. The data are aggregated counts by condition, not individual patient records.

On October 15, 2026, the terms of use changed from Korea Open Government License Type IV (공공누리 제4유형) to Type I (공공누리 제1유형). Users can now modify the data or include it in commercial services if they attribute the Korea Disease Control and Prevention Agency. Online services should also show a link to the original source where possible. The same notice replaced the connection URLs, and the old URLs are scheduled to end after 90 days.

There is no fee, but the data are not immediately usable. A user must register with the Public Data Portal (공공데이터포털), apply to use the API, and send an issued authentication key with every request; applications are approved automatically. The portal shows 857 use applications, but that does not mean 857 services are operating.

The basic development limit is 1,000 calls per day. Pulling all eight queries for one region every 15 minutes already uses 768 calls a day. A service that frequently updates multiple regions and diseases will need to request a higher limit or store retrieved results for shared use. The detail page labels the update cycle as “real time,” while the agency’s statistical guidance says routine statistics are updated daily. The actual reflection time should therefore be checked with test calls.

The meaning of the numbers also needs care. They are provisional statistics aggregated from reports by medical institutions and others. Regions generally reflect a patient’s address, not the place of infection, and some diseases may include suspected cases or pathogen carriers. Category 4 sample-surveillance diseases, such as influenza and hand, foot and mouth disease, are not all available in the same way because they are not subject to full-case reporting. A service for childcare centres would need to add separate data or narrow its disease scope.

Where the work breaks between statistics and parent notices

Imagine the director of a childcare centre with 80 children and 12 staff members. When several children are absent with similar symptoms, the director opens websites from the Korea Disease Control and Prevention Agency or local health authorities to search for the disease name and region. Without dedicated health staff, this happens between preparing arrivals and assigning teachers.

The director copies the numbers into a note or work document. They then find official prevention guidance on another screen and rewrite it separately for teacher guidance and parent notices. To explain whether the number is based on report date or onset date, and whether it is final, they must check the original source again.

The work most often breaks when moving from a statistical screen to action guidance. Infectious-disease counts do not include operational instructions such as classroom disinfection, symptom observation, or contacting guardians. The director moves among a statistics site, prevention-guidance documents, notice tools, and absence records, repeatedly copying the same disease name and date.

Using the new connection point, a tool could automatically retrieve the latest counts each morning for selected regions and diseases, and show them alongside the previous week or several earlier weeks. Next to each number, it could permanently show the data reference date, an explanation that it is a provisional report-based figure, and a source link. The director would review only items that changed rather than searching for every number.

When the director selects an item with a large change, a teacher checklist and a parent-notice draft could open on the same screen. The statistics would come from Korea Disease Control and Prevention Agency data, while action guidance would come from separate official prevention guidance, with the source and applicable date shown for each. The director would set only the class name and notice time, then send it through the existing notice tool.

Some work must still stay with people. Whether a particular child can attend, whether to recommend a medical visit, and whether to close the centre or cancel an event cannot be decided from public statistics alone. Public data contain no individual records, and a rise in reports does not mean infections are occurring inside the childcare centre.

So the product worth building is not a tool that predicts infections or makes decisions for people. It connects four steps in one place: finding local statistics, confirming what they mean, attaching official guidance, and turning it into operational language. To earn trust in the field, it must not hide which diseases are missing or when the data were last updated.

Elsewhere, products have sold the connection point

Kinsa FLUency in the United States provided connected thermometers and an app to households, then aggregated consented temperature and symptom data at school level so school health staff could see trends and guidance messages. Kinsa offers home thermometers and an app, FLUency for schools, and regional disease analysis services for businesses.

Kinsa says it has operated in more than 5,000 schools in the United States and Canada. It also reports a 27% reduction in illness-related absences during flu season, but this is the company’s own claim and needs separate verification. The important point is that it did not stop at showing measurements; it connected them to school notices.

Brazil’s InfoDengue combines notification data with climate and environmental data to show dengue, Zika, and chikungunya risk by municipality in colour. Municipal officials and citizens can access it for free, and it provides both a public API for other services to retrieve results automatically and programme source code.

InfoDengue has operated for about 10 years and now provides analysis for all Brazilian states and municipalities. Users can select a disease, municipality, and epidemiological week to receive an alert level and estimated cases. Rather than simply republishing public notification data, it combines multiple datasets into a single unit for judgement.

Canada’s BlueDot provides infectious-disease teams in governments, hospitals, and businesses with outbreak signals and travel risk from around the world. It automatically collects and organizes news and official announcements in multiple languages, then has experts review them and sells institutional dashboards and connected data alongside air-travel information.

BlueDot says it alerted customers to unexplained pneumonia in Wuhan on December 31, 2019. It also says customers’ recurring surveillance work fell by 88% per week, but this figure comes from customer responses collected by the company and needs separate verification. This case also charges not for data collection itself, but for reducing the time needed to turn information into reports and response decisions.

Four small places to start

A childcare-centre infectious-disease notice helper. This service would help directors of childcare centres with 50 to 100 children and no dedicated health staff check local figures and official prevention guidance, then create draft notices for teachers and parents. Now that commercial use is allowed, it could first reduce the repeated searching and copying work.

Its first screen would include the region, target diseases, latest count and reference date, change from the previous week, an indicator for unavailable diseases, and a “Create notice draft” button.

A morning inventory check for neighbourhood pharmacies. This service would help a pharmacy owner near medical clinics, who also handles ordering alone, use local reporting trends to create a checklist for non-prescription items such as masks, thermometers, and hand-hygiene products. The key is to use public statistics as a reference signal for one more check before ordering, not to package them as a sales forecast.

Its first screen would include infectious diseases with the largest recent changes, the data reference date, the pharmacy’s own current inventory, and a list of “Items to check today.”

An infectious-disease preparation checklist for local events. This would help a first-time operator of a sports event or local festival with participants arriving from several regions prepare hygiene supplies, notices, and contact routes for suspected symptoms based on the event location and date. Because regional and period-based data can be commercially processed, it could reduce the work of moving information into event documents.

Its first screen would include fields for the event date, host region, and participant age range, plus relevant disease trends, data limitations, and a “Create operations checklist” button.

A response board for community welfare centres. This service would help staff at local welfare centres that manage several senior centres answer repeated questions from older residents and guardians consistently, while separating questions to send to public health centres. Managing sources separately for numbers and official action guidance can reduce the reuse of outdated wording.

Its first screen would include today’s local changes, frequently received questions, approved response wording, the last review date, and a “Send to public health centre inquiry” button.

What to check today

Ask one childcare-centre director to show you a recent parent notice about an infectious disease. Ask them to describe, in order, every screen they used from finding the number to sending the notice. If they moved across three or more places, copied the same disease name or date three or more times, and took more than 20 minutes, that is a reasonable signal to try a small prototype that connects statistics and notices.

Why this matters where you are

This is a Korean case shaped by Korean public-health reporting, licensing, API access, and the availability of official prevention guidance. In your market, check whether local health data can be commercially reused, what the data actually represent, and whether official guidance can be linked to specific operational actions. The transferable opportunity is not the Korean data source itself, but the manual handoff between a trusted number and a useful local notice.