Chester County to overhaul purchasing after spending millions on faulty antibody testing program

Chester County will adopt sweeping changes to the way it purchases products and services in the wake of a failed COVID-19 antibody testing program that is now the subject of an $11 million county lawsuit against Advaite, the unproven local company that developed the product.

”While it was a crazy, crazy time, we want to make sure it doesn’t happen again,” Marian Moskowitz, chair of the county Board of Commissioners, said during a Thursday morning meeting held on Zoom. “It sounds like there will be major changes.”

Raymond McGarry, a lawyer hired by the county to review the aborted testing program, and county controller Margaret Reif, presented a raft of recommendations for improving procurement, including centralizing the process, and adding new layers of oversight — even in emergency situations.

But they stressed that they found no fraud or unethical behavior behind the no-bid deal with Advaite. Governments are permitted to skip the time-consuming bidding process during a public health emergency.

”Nobody acted inappropriately. There was nothing criminal or unethical,” McGarry said. “County staff and administrators were dealing with an unprecedented emergency. It was a fast-paced, emergency situation.”

In May, Chester County became the first in the state to offer first responders and health-care workers finger-prick blood tests to look for disease-fighting antibodies, a sign of a past infection that confers at least some level of immunity. COVID-19 antibody tests were just beginning to be developed around the world, and Advaite had no track record in diagnostic testing.

The county shelved the program weeks later after a spike in positive test results — far more than was plausible, given the low prevalence of the coronavirus in the population in the spring.

After The Inquirer reported in September on the false positives and the program’s abrupt but quiet end, county officials said they would notify the residents who had received “questionable test results.”

» READ MORE: Chester County spent $13 million on coronavirus antibody tests. Then it quietly shelved the program.

In March, federal regulators granted Malvern-based Advaite permission to commercialize its test, pending an evaluation of the product’s performance. Last month, Advaite received U.S. Food and Drug Administration authorization for emergency use of the antibody test. The original product that Chester County bought has been changed to measure one antibody instead of two. “There is no identified advantage” to either approach, according to the CDC.

The county’s breach-of-contract lawsuit, which was not mentioned during Thursday’s review, contends it received only about 39,000 tests by the end of May, when the agreement specified delivery of one million tests by then. About 2,000 tests were returned because of “defects and component issues,” the complaint says.

The county originally contracted to pay $20 million in federal pandemic aid, and actually gave $13 million to Advaite. Chester wants $11 million returned. Ajay Raju, a lawyer for Advaite and its founder, Karthik Musunuri, said on Thursday that the county never picked up available tests, and the company stands by the quality of its product.

”They purchased a product. Why would we give a refund?” Raju said. “If you order food and don’t eat it, you don’t get a refund.”

Earlier this week, Raju issued a statement that “the lawsuit is a thinly veiled attempt on the part of Chester County to divert attention from its own internecine squabbling and incompetence.”

During Thursday’s meeting, controller Reif recommended that the county create “risk management teams” so key departments, including hers, could assess proposed contracts and purchases. Reif also said that the Health Department should “take the lead” in health emergencies, with emergency services following that lead.

It was emergency services director Michael Murphy and county administrator Robert Kagel who pushed to make Chester County Advaite’s first customer, according to internal emails and interviews with former county health physician Renee Cassidy. She quit in protest after officials decided not to alert test-takers to potentially false results.

“How fast can you get me some samples for us to test?” Murphy emailed Musunuri on March 23. “Once tested, we would like at least 10,000 kits to start, with the option to order much more.”

Advaite was founded in 2017 by Musunuri, 26, while he was majoring in pharmaceutical marketing at St. Joseph’s University. He got advice from his father, Shankar, a biotech veteran who set up his own firm in 2013. The young entrepreneur also got the endorsement of a family friend, former Pennsylvania Sen. Andy Dinniman.

Dinniman, a champion of the state’s biotech industry, knew that Chester County — like everyplace in the nation — was hamstrung by the shortage of molecular diagnostic tests in the spring. He suggested Advaite’s test — which gives fast results on the spot, like a pregnancy test — might be an alternative.

Cassidy, the physician, and county health director Jeanne Casner wrote a report on the pros and cons of COVID-19 antibody testing in general. Around the world, such testing was raising hopes for a way to identify people with some protective immunity, and track the spread of the life-threatening disease, which has so far killed more than 400,000 people in the United States.

Antibody tests — unlike the molecular diagnostic tests that search for viral genetic material in a nasal swab sample — can be thrown off by antibodies to other coronaviruses that merely cause colds, or because the level of COVID-19 antibodies is too low to detect, among other problems.

Advaite’s own performance evaluation data, included in the FDA-required package insert for the “RapCov” test, is telling:

Assuming 5% of the population has been infected with the coronavirus (the CDC estimates it’s actually 3.2% in the Philadelphia region), then RapCov would accurately identify people with no antibodies 99.5% of the time. However, it would accurately identify people with a past infection only 49.7% of the time — about as good as flipping a coin.

RapCov performed better in an independent evaluation at a federal laboratory, which estimated it would accurately flag 80% of blood samples containing antibodies.

But since those estimates are based on lab tests using known samples, the FDA warns that “real-world performance of the device” may be different.