Why Kentucky Is Changing Some COVID-19 Data: The Positivity Rate Explained
Kentucky’s Public Health Commissioner Steven Stack on Tuesday announced some changes in how the state will gauge the spread of COVID-19 after the state came off a week that saw a surge in cases largely driven by infections in rural counties.
While the commissioner’s plan does take cues from the models used by many states across the country, he didn’t say whether any new data will be published that could help draw an exact picture of the spread.
The notable changes include: using incidence rate to spot active cases: removing antibody tests from positivity rate calculations; and excluding old tests for calculating average rate of positive tests.
According to the Centers for Disease Control and Prevention, the positivity rate is defined as the percent of total reported tests that come back positive (where the total reported tests equals positive plus negative tests).
The nation’s top public health agency lists three variations of the positivity rate as a standard method that offers a great deal of insights into the community spread of the deadly virus.
A high positivity rate means many test results are coming back positive, a sign of community spread. It could also occur if there is a localized testing drive in a community that is at the greatest risk for the infection. Moreover, it can also determine whether there are delays in reporting processes.
Experts suggest that a 7-day moving average or a 14-day moving average positivity rate should be read in conjunction with other metrics such as death and hospitalization rates. The changes Kentucky officials are making should help to make the positivity rate a more useful tool for public health decisions.
Under the new guidelines, the state will exclude tests outside a 7-day window because a trove of backlog reports creates irregular swings in the positivity rate and makes it hard to make sense of numbers.
Like some states, Kentucky has issues with numbers getting skewed as the delay in manual reporting creates a backlog partly because a lot of records are compiled manually. The new automated system will filter old records and exclude the ones that don’t fall in a 7-day period.
“When we have a lab that comes on board and starts reporting electronically and dumps a large number of historical lab reports, those will be auto filtered, so only their most recent seven days of lab reports are included,” Stack said. The positivity rate is supposed to be a seven-day trailing average, he said. “So, we don't want data from two or three months in there.”
The 7-day moving average rate of positivity normalizes the irregular daily swings to give a true sense of the direction of the transmission.
Viral vs. Antibody Tests
For months, Kentucky combined two different types of coronavirus tests when calculating the positivity rate: viral tests — which show an active infection — and antibody tests. Antibody tests detect antibodies the immune system produced to fight the virus, but don’t necessarily indicate an active infection.The commissioner said that the state won’t use antibody and rapid diagnostic tests for calculating positivity any more.
“More than 90% of all the tests we do for COVID-19 in the state of Kentucky, are polymerase chain reaction or PCR tests. Those are the gold standard. Those are the most reliable for finding active disease in currently infected people,” Stack said.
Beginning Monday, Stack said, the state would use electronic lab reports to calculate the positivity rate in hope of creating a faster and stable datastream. (The commissioner said that the average turnaround time for PCR tests is 2 days.)
Excluding the antibody tests is a good move, according to experts who track the data.
There are multiple issues with the previous reporting method and calculations based on antibody and rapid testing numbers can create incorrect readings.
In Kentucky, exclusion of antibody tests will help the health department reach a more accurate rate of positivity for multiple reasons.
For one, people who recover from COVID-19 have antibodies so they aren’t shedding active virus load.
Also, many people who take diagnostic tests end up checking themselves for antibodies as well. To include antibody tests would be a double count and would falsely decrease the positivity rate if the number of these types of tests is substantial.
About 5% of total tests done by the department are antibody tests. Excluding these tests will increase the positivity rate slightly, but not by a lot.
The state had 1,346 new coronavirus cases as of Wednesday, bringing the state’s total to 83,013 while 1,276 have died. Kentucky has been reporting a positivity rate between 4% and 5%.
State health measures of positivity rates have frequently been out of step with those calculated by others tracking the disease. For example, data from Johns Hopkins show Kentucky’s 7-day moving average rate of positivity of up to 7% as of Oct 12.
The difference is so large because the testing denominator used by the Kentucky’s health department includes duplicate entries that inflate the testing numbers and thus decrease the rate of positivity.
In order to target community spreads and break the line of transmission, the state needs to evaluate the number of people tested for the coronavirus both on state- and county-level.
The public health commissioner’s decision to exclude rapid tests from their calculations is largely due to the delays in reporting and the inaccuracy of these tests.
The Food and Drug Administration doesn’t have full confidence in rapid or antigen tests because these tests can produce false negatives in as many as half of the instances.
A false negative test means that the report comes back negative even when a person is positive. This, in turn, could mean letting people mingle with their families and friends while they silently carry the virus and spread it to others while thinking that they aren’t transmitting the virus. It is a dangerous proposition that could create superspreader events.
Although the accuracy of all types of Covid-19 tests is uncertain, PCR tests are considered the best available to diagnose an active infection.
Incidence vs. Positivity Rate
The commissioner said that the incidence rate — defined as the number of positive tests per 100,000 people — is a metric that they are using to track active cases. He added the test positivity, a somewhat standard measure for tracking COVID-19 spread, should be used to make sure there are “enough resources to test people who need to be tested.”
Stack said test positivity alone does not tell us how active the disease is. “We cannot use this measurement to say there's a lot of disease,” he added.
Incidence rate and positivity rate used in conjunction can better assess the spread of the virus.