Calculating Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Laboratory Test Percent Positivity: CDC Methods and Considerations for Comparisons and Interpretation
SARS-CoV-2, the virus that causes coronavirus disease 2019 (COVID-19), laboratory data from public health laboratories, U.S. hospital laboratories, private and commercial laboratories, some large chain drug stores, and other testing entities are reported to state and local health departments. Laboratory data are reported to state and local health departments in accordance with applicable state or local law and transmitted to CDC in accordance with the Coronavirus Aid, Relief, and Economic Security (CARES) Act (CARES Act Section 18115pdf iconexternal icon). De-identified COVID-19 laboratory data from 6 large commercial and public health laboratories are also reported directly to CDC. Hospital laboratories in states that are not yet sending line-level data to CDC report data directly to the Department of Health and Human Services. These entities voluntarily submit electronic files of COVID-19 de-identified laboratory data to CDC or HHS on a daily or periodic basis. At CDC and HHS, the data are processed and analyzed, with the results made available for federal response efforts (in HHS Protect). Laboratory data provided to CDC is made publicly available on CDC’s COVID Data Tracker website in a way that protects the privacy of individuals and can be downloaded for analysis from healthdata.govexternal icon.
Using Percent Positivity Calculation for Public Health Surveillance
The formula for calculating percent positivity used by CDC is the number of positive tests (numerator) divided by the total number of resulted reported tests (denominator): (positive/total) x 100% where the total equals positive plus negative test results and excludes indeterminate results.
In general, for federal COVID-19 response reporting purposes, laboratory test percent positivity has represented the percentage of all RT-PCR tests conducted that are positive. Although the methods used by different organizations to calculate percent positivity can differ (see below), percent positivity has provided insights into transmission of infectious diseases, including COVID-19 (SARS-CoV-2), in a geographical area (e.g., national, regional, state, county). The interpretation of percent positivity depends on the volume of COVID-19 diagnostic laboratory testing reported to state and local health departments and the criteria used for determining what populations are tested (routine screening of asymptomatic persons vs diagnostic testing of symptomatic persons or case contacts).
A high COVID-19 RT-PCR percent positivity occurs when many of the test results among those being tested and reported in a community are positive. This can mean that
- there are widespread infections in the community tested; or
- only a subset of the community at greatest risk for SARS-CoV-2 infection is being tested; or
- there are reporting processes or delays that skew the results (e.g., prioritizing reporting of positive test results over negative results).
The laboratory test percent positive goes down when more people tested are negative. This happens when the number of infections goes down (the numerator gets smaller), or testing is expanded to more people who are not infected (the denominator is larger without uncovering a lot of new infections). In general, percent positivity will go down as more persons are being screened in non-outbreak settings (e.g., routine screening in schools, long-term care facilities, workplaces) and the results are reported. Expanded testing does not always reduce the percent positive when there is widespread transmission; it may reveal more people who are infected.
Different Methods Used to Calculate Percent Positivity
CDC, state, and jurisdictional health departments may calculate percent positivity differently, which may include:
- Differences in the numerators or denominators used (e.g., tests/tests, people/tests, people/people). State and jurisdictional health departments have access to personal identifiers in their datasets and can identify and de-duplicate persons with multiple positive tests whereas CDC is unable to perform this function.
(See Figure 1 below.)
- Differences in the timeframe in which data are included (i.e., a seven-day verses a 14-day rolling average), as well as what dates (e.g. specimen collection date, test date, result date) are used to assign tests to specific timeframes.
- Differences in the inclusion or exclusion of antigen test results. Antigen tests may be used for screening or diagnostic purposes. Positive test results for COVID-19 antigen are considered a probable case as outlined in the CSTE position statement, and might not be confirmed using FDA authorized RT-PCR tests for COVID-19. Antigen test results might not be consistently reported to public health by clinics or sites where routine screening is conducted (e.g., long-term care facilities, schools or workplaces).
- Differences in inclusion of screening tests results. With increased screening using both antigen and RT-PCR tests, the ability to confidently interpret the meaning of percent positivity results will be impacted by the unknown criteria for testing (routine screening versus diagnostic testing of symptomatic persons).
- Differences in how test results are assigned to jurisdictions, including by the person’s place of residence, the provider’s clinic location, the location the test specimen was collected, or the location of the laboratory.
Read Text Equivalent (for Section 508 access).
CDC Methods for Calculating Percent Positivity
Currently, for federal reporting purposes, CDC defines percent positivity as the number of COVID-19 positive RT-PCR tests as the numerator (includes diagnostic tests and screening tests, when analyzed at CLIA-certified laboratories), divided by the total number of RT-PCR tests with positive and negative results as the denominator: (positive tests/total tests) x 100% (Method 1 in the figure above). This excludes antigen tests, antibody tests, and RT-PCR tests conducted by non-CLIA laboratories for surveillance purposes. Data received at the federal level are de-identified and, therefore, are not able to be linked at the person level for de-duplication. This prevents CDC use of methods 2 (people over test) and 3 (people over people) in Figure 1 above.
Resulted tests in data provided to CDC by states and local health departments are assigned to a 7-day timeframe based on the top test-related date available in the data according to the following hierarchy:
- Test date (day the laboratory had a test result)
- Result date (day the laboratory sent the result to the requestor)
- Specimen received date (day accessioned by lab)
- Specimen collection date (day specimen collected from patient)
Test results are assigned to a geography based on a hierarchy of test-related locations:
- Patient location
- Provider facility location
- Ordering facility location
- Performing organization location
Data resources where percent positivity for COVID-19 are currently reported
CDC currently provides data at the national level on RT-PCR laboratory test percent positivity on the CDC COVID Data Tracker website.
The Center for Medicare and Medicaid Services publishes percent positivityexternal icon by county to guide the frequency of COVID-19 screeningpdf iconexternal icon of long-term care facility residents and staff.
The White House Coronavirus Task Force reports percent positivity at the national, state and county levels in the Governor’s Report, a weekly report that sent by the Vice President to each Governor.
- Different methodologic choices for calculating percent positivity when consistently applied are useful for monitoring trends and magnitude for surveillance purposes and decision making.
- State and local officials may consider adopting some or all of the methods outlined above if these methods suit their surveillance needs.