Throughout the COVID-19 response, ADHS has continually reevaluated how we are collecting, analyzing, and presenting COVID-19 data for cases, hospitalizations, and laboratory tests. As the COVID-19 testing landscape changes, with new tests, new laboratories, and new guidance on who should be tested, and as data come to ADHS in new ways, we have examined whether any changes need to be made in order to share information in the best way possible.
Location, whether at the county or ZIP code level, has always been important information for COVID-19 cases and testing. A full patient address is not always available when a case or a lab result is first reported. While we do our best to make sure the case is counted within the county of residence, we sometimes need to rely on the address of the provider facility or reporting entity in the absence of a complete patient address. More than 90% of cases have been shown in the ZIP code of their residence throughout the response, but we have recently re-examined how we attribute cases to ZIP codes and labs to counties. Using the experience of the large quantity of reports now available to us, we decided we can provide better, more useful information by reshaping how we use the address information submitted. The impact of these changes is described below.
For the case data shown on our data dashboards, there is no effect on the state counts or on the county-level numbers. However, the ZIP code dashboard may reflect some changes as we shift some cases from a provider ZIP code to the ZIP code of a patient’s address. By implementing the change, we increase the proportion of cases shown using patient address from 94% to 97%. Some areas are more affected than others, with Coconino, Navajo, and Graham Counties having the biggest shifts towards patient address.
For the laboratory testing data, there is likewise no change in the statewide numbers of people tested or tests conducted, but the county numbers are affected. This change involves better use of the ZIP code data submitted with the lab report, with the overall effect of attributing many more laboratory results to a specific Arizona county, and moving a smaller number between counties. The total number of persons tested increases for all counties; Maricopa County gains 6,018 people tested while Pima County gains 3,455. The largest percent increase is for Santa Cruz County (2.4%). Again, there is no impact on the statewide data as these tests have been counted all along in the Arizona numbers; instead, these increases are the result of attributing records for more than 12,000 people tested into county data.
While we do not anticipate additional changes in these methods in the near future, ADHS will continue to examine the data we have as the COVID-19 response continues, to ensure we can provide the most useful information to everyone involved.