Interactive Seminar

Day 1: Tuesday, 13 September


DiD IT?: A difference-in-differences investigation tool for public health

Session convener: Roger Morbey, Public Health England

A key element of health security is being able to detect and respond appropriately to acute public health events. When a potential threat is identified, the first requirement is to identify the scale of any impact as quickly as possible. Real-time syndromic surveillance systems have been developed to provide early warning and situational awareness for a wide range of potential multi-hazard threats, including infectious disease and environmental hazards. Syndromic systems are designed to be as sensitive as possible to detect all potential threats, however this can make it difficult to quantify the specific impact of incidents. For example, an increasing trend during an incident may be partly due to other seasonal factors. Also, differences between an area affected by an incident and its neighbours may partly be due to structural differences like deprivation or local health care provision. Therefore, UKHSA's real-time syndromic surveillance team have developed a novel investigation tool which uses a difference-in-differences (DiD) method to quantify the impact of local incidents, called "DiD IT?". The (DiD) method uses two sets of controls, both the time immediately before the incident occurred and areas unaffected by the incident. Thus, we can estimate, spatial and temporal differences and calculate the direct effect of the incident on our syndromic indicators. Importantly, the tool can also provide reassurance when no impact is detected. Estimates are always given with uncertainty intervals so that incident directors can assess what confidence to give to the syndromic data.