Table 3 Concepts for quantifying uncertainty in residential geolocation of cancer incidence data

Leveraging discriminant power of residential address to quantify confidence in residential geolocation

We propose to use of CEL and sRGDP

Enabling citizen understanding of quantified confidence in residential geolocation in cancer incidence data

We propose that uncertainty metrics be no more complex than necessary. At least some evidence (e.g., ED address points) should be available to citizens for independent verification

Using the principle of maximizing information entropy during geocoding record linkage to ensure comparability of confidence in residential geolocation across the subset of cases

Geocoding against ED address generalizations helps to ensure that a geographic area of uncertainty corresponds to the extent of accuracy and precision of address components

Differentiating between the domains of emergency dispatch address point and residential address

We clarify that the ED address domain is not necessarily as extensive or discriminant as the residential address domain at any given time

Clarifying the role of attribute associations and data stream stewardship in earning citizen trust

We quantified uncertainty due to ED reference data vs. patient/medical facility errors

Adding metadata to geographic reference data to facilitate quantification of confidence in residential geolocation

We add value to ED address point data by storing the CEL count of each duplicate address, so that it is easily accessible. This reduces the number of cases for which CEL has to be interactively identified

Delimit responsibility for quantification of confidence in residential geolocation by CCR as a downstream steward

A proposed convention delimiting AA that enables and/or modifies spatiotemporal relationships in CCR data [15]