The US doesn't know exactly how widespread the Delta coronavirus variant is because of inadequate virus sequencing.
Sequencing tells you what strain of the virus is causing a COVID-19 infection.
Data from the Global Initiative on Sharing All Influenza Data (GISAID), a highly-regarded central genetic sequencing database, showed on Tuesday that in the past six months the US had sequenced and shared results for 4.12% of all positive tests.
Over the same time period, the UK had sequenced and shared 18.6% of positive tests, Singapore 46.5%, Australia 58.7%, and New Zealand 41.5%, per GISAID.
The Delta variant was detected in the US on February 5 and became the primary cause of new US infections on July 7, according to the Centers for Disease Control and Prevention (CDC). The Delta variant now accounts for more than 80% of new cases in the US, according to Scripps Research's Outbreak.info, which is backed by the CDC and uses GISAID data.
Outbreak.info said on its website that the sequencing data it provides is a "best estimate" because the tests sequenced and shared to GISAID don't always represent an average American. For example, they often come from an outbreak, a breakthrough infection from someone who's vaccinated, a patient with unusually severe disease, or those with a travel history.
This could lead to overestimates of the number of infections caused by variants such as Delta, Outbreak.info said.
The CDC said in January that it had partnered with commercial labs and academic institutions to try to sequence up to 20,000 samples a week, which would have amounted to 18.4% of the US's weekly cases on Tuesday. The GISAID data suggests it is far short of this goal.
The same GISAID data showed that it took the US an average of 22 days to upload sequencing data onto its database, which could mean estimates for the number of Delta infections are outdated. The UK took an average of 11 days.
The US was not lagging behind every rich nation: it's 4.12% rate of positive tests sequenced and shared was higher than Israel's, at 1.26%.
A modeling study posted in January estimated that a country needed to sequence 5% of positive tests to detect an emerging strain, assuming 0.1% to 1% of the population were infected with it.
The CDC told the Financial Times that it was "confident" that with the current surveillance strategy "new and emerging variants would be detected well before they become prominent."
Will Lee, vice-president of science at the genomics company Helix, which has a sequencing contract with the CDC, said that piecing data together was a challenge. "With so many different places and means of getting tested, it can be hard to get the sequences," Lee told the FT.
Bronwyn MacInnis, director of pathogen genomic surveillance at MIT and Harvard's Broad Institute, said that the biggest barrier to "closing the gap" with other countries was getting positive tests to sites to be sequenced, per the FT.
Thomas Friedrich, professor of pathobiological sciences at the University of Wisconsin-Madison, told the FT that federal regulations designed to protect people's privacy can get in the way of the "rapid sharing of information we need." States interpret these regulations in different ways, he said.
Oklahoma has sequenced 736 positive tests, whereas Texas has sequenced more than 59,000, according to Scripp's Outbreak.info.
