In Maryland, drive-through coronavirus testing sites are now open to all residents, whether or not they show signs of illness.

In Oregon, by contrast, officials have said that generally only people with symptoms of covid-19, the illness associated with the coronavirus, should be tested even in the case of front-line health-care workers.

In Rhode Island, officials have proactively tested all of the state’s 7,500 nursing home residents, including those with no symptoms, and are developing plans to test more people in high-risk workplaces, such as restaurants and grocery stores.

The wide range of approaches across the country comes as the federal government has offered little guidance on the best way to test a broad swath of the population, leaving state public health officials to wrestle on their own with difficult questions about how to measure the spread of the virus and make decisions about reopening their economies.

Faced with conflicting advice from experts in the field, states are using different tests that vary in reliability and have adopted a variety of policies about who else should get tested and when — particularly when it comes to asymptomatic people who are considered low-risk for the illness.

“The states are on their own,” said Kelly Wroblewski, director of infectious diseases at the Association of Public Health Laboratories, noting that the kind of guidance the federal government routinely gives in screening for flu and other outbreaks “has been absent” in the covid-19 pandemic. “There has been no coordination.”

That means that while tests are available to anyone who wants them in states such as Kentucky and Georgia and some large cities such as Detroit and Los Angeles, state officials in Idaho and Louisiana continue to recommend that only sick people get tested.

The lack of a unified national strategy has left Americans uncertain about whether and how to be tested and is hampering reopening plans, experts warn.

Many officials now worry that protests in more than 100 U.S. cities in recent days after the death of George Floyd in police custody, which have drawn thousands of people packed closely together, could spark new infections.

So far, about 460,000 Americans are being tested a day — 0.15 percent of the population, and still shy of the 900,000 to 30 million that experts say need to be tested daily to capture the extent of the virus’s spread.

“The case numbers we’re seeing are probably massively undercounted,” said Divya Siddarth, a researcher who helped devise a testing strategy for Harvard University’s Safra Center that emphasizes finding and suppressing the disease in areas with fewer cases. “These [lower prevalence] regions are likely to reopen, and they’ve barely done any tests.”

The lack of clear information is forcing businesses large and small, schools, universities and professional sports organizations to make their own decisions about how much testing they need to be safe.

Some institutions have announced their own plans for universal testing. The National Hockey League, for example, has said it plans to test all players daily as part of a plan to resume play in June. The University of Arizona has developed its own antibody test that’s available to all students and local health-care workers.

Under a law passed earlier this year, the Trump administration is required to develop a national testing strategy. But an 81-page document submitted to Congress by the Department of Health and Human Services late last month was not released publicly and offered few detailed recommendations.

The Washington Post obtained a copy of the plan, which set a goal for states of testing at least 2 percent of their residents in May and June. But how to meet that benchmark and whether to go further was left up to state leaders who were required to submit plans this month to HHS for review.

The Centers for Disease Control and Prevention has recommended universal testing for residents of nursing homes, which have been especially hit hard by the coronavirus. But the HHS document said the CDC was still working on guidelines for other large populations of mostly asymptomatic people — including at universities, prisons and “critical infrastructure worksites” — as well as those for integrating testing into reopening work places.

Mia Palmieri Heck, a spokeswoman for HHS, said the federal government “has provided prescriptive criteria about testing asymptomatic individuals when they affect highly vulnerable populations such as individuals who live in nursing homes, working in or visiting health-care clinics or communal dining spaces.” She added that federal experts have also been advising states on developing plans to more broadly test people without symptoms to determine community spread.

The question of asymptomatic testing is particularly tricky given that the CDC late last month said that its researchers now believe as many as 35 percent of people infected with the coronavirus never show symptoms of disease.

Typifying the kind of conflicting information facing states, a World Health Organization official sparked global confusion on Monday when she said it is “very rare” for people with no symptoms to transmit the disease. After significant pushback from researchers, the official said Tuesday that scientists continue to believe that people without symptoms do in fact spread the virus — but more research is needed to understand by how much.

She noted that some modeling shows as much as 41 percent of transmission may be due to asymptomatic people.

“In some ways, this may be the Achilles’ heel of the entire testing challenge for this virus,” said Ashish Jha, director of the Harvard Global Health Institute, who has advocated for increasing the number of people getting tested.

Local and state health officials worry that the lack of coherent strategy could result in tests becoming widely available for the affluent, while remaining limited for those with fewer resources, including minority communities that have already been disproportionately affected by the virus.

At the University of Arizona, officials plan to reserve molecular swab tests, which determine if a person is currently infected, for symptomatic students and their contacts. Each test is about $50 to $75 dollars; there are 60,000 students, staff and faculty and each would have to be tested repeatedly.

“Maybe the NFL can afford that; we can’t, and I don’t know any university that can,” said Robert C. Robbins, the university’s president.

'Box the virus in'

When coronavirus cases began to mount in March, a severe shortage of test kits and supplies meant tests were sharply rationed. Even after it was clear that the virus was spreading in the United States, the CDC at first recommend only testing people who had visited China or been in contact with someone who had.

Later, federal officials suggested that younger, healthy people did not necessarily need testing even if they were experiencing coronavirus symptoms, reasoning that the tests should be reserved for hospitalized patients for whom a positive result might make a difference in treatment plan.

As tests have become more available, officials have begun to recommend that anyone who is experiencing signs of illness, even a mild cough or sore throat, get one.

The goal is to identify and quarantine people with the disease, and then use contact tracers to track down people who have interacted with that person and quarantine them as well.

“Testing is just part of a comprehensive strategy,” former CDC director Tom Frieden said. “As you emerge from that sheltering situation, you box the virus in.”

But when it comes to testing people without symptoms, state recommendations vary.

About at least half of states aim to test people identified as contacts of known positive cases, according to a Post tally, as was recommended in new guidance from the CDC this week. But many others tell those people to self-isolate for 14 days.

“Every state is figuring this out on its own, little bit by little bit,” said Philip Chan, medical director for the Rhode Island Department of Health.

Nearly all states have set aside thousands of tests for people in congregate settings — residential settings where large numbers of people live in proximity, especially nursing homes and prisons.

But only a handful of states have so far satisfied the CDC goal to test everyone living in a nursing home, where the age and underlying medical conditions of residents make them especially vulnerable to covid-19 outbreaks.

Some states have also prioritized testing front-line health-care workers and other people working elbow-to-elbow in manufacturing facilities, particularly meatpacking plants, which have been hit hard by the virus.

Even states that have conducted widespread testing in such facilities face difficult questions about whether a single round of testing is sufficient, given that people could easily contract the virus at any time, including after testing negative.

“There’s not a lot of communication between the states and there’s not a lot of specifics, so everybody’s kind of going on their own,” Wroblewski said.

A tricky disease

A number of states and large cities, such as Detroit and Los Angeles, have opened drive-through testing sites like those offered in Maryland, a mode of mass testing used effectively overseas in South Korea and elsewhere.

Experts have warned that drive-through sites often fail to collect enough information from those tested to follow up effectively. They also prioritize people who choose to show up, tending to mean tests go to better educated and informed residents and not necessarily those most likely to have been exposed to the virus.

In Macon, Ga., the Moonhanger Group set up drive-through testing for employees returning to work at their four restaurants. But they did not wait for the results, or for all employees to get tested, before reopening on May 26.

“We were confident, based on the low number of positive results reported in Bibb county, that none of our employees would test positive and we hoped to share that news with the public,” owner Wes Griffith wrote on Facebook. “Unfortunately and surprisingly, we have employees who have tested positive. All of them were a-symptomatic.” Griffith did not respond to a request for comment.

Three of the four restaurants had to quickly close again, pending further testing.

In Georgia, public officials are advertising on radio and social media to encourage anyone to get tested at drive-through sites.

Those tested have included political leaders, who got tested largely to encourage others to do so too, only to find themselves “shocked” when their results came back positive, said Phillip Coule, chief medical officer of the Augusta University Health System, which is partnering with the state on testing.

“It’s a great demonstration of how tricky this disease is,” he said.

Other states have downplayed asymptomatic testing as unreliable or a poor use of resources.

Coule noted that the message, “If you want a test, you can get a test,” puts the onus for deciding who should get tested on individuals, rather than prioritizing the highest-risk or the most vulnerable. One of his patients, he noted, sought a test because he wanted to honeymoon in St. Lucia and needed a negative result to enter the country.

Oregon only opened testing to front-line workers and long-term care residents without symptoms in April and continues not to recommend asymptomatic testing, saying on the state website that it is “not useful” because the false negative rate is high. Viral tests have been estimated to have up to a 20 percent false negative rate.

At a recent news conference, Oregon Health Authority Chief Medical Officer Dana Hargunani said people without symptoms are “unlikely or certainly less likely to cause transmission of the virus.”

'It's like a war'

For states looking to figure out who to test and when, advice from national experts has been abundant — but not always consistent.

Proposals from academics and other experts vary widely in their recommendations of the numbers of tests that should be performed each day, and many do not offer guidance about who should be tested.

Some researchers have recommended focusing on parts of the country that have few cases in hopes of stamping out the disease.

“We should quickly get resources to places where the disease can be suppressed, then backfill tests in the places currently overwhelmed,” said Glen Weyl, an economist at Microsoft, who worked on the Harvard University proposal. “It’s like a war — you have to more troops than the enemy in order to win a battle.”

Other researchers have proposed blanketing the country with tests, with a focus on places experiencing clear outbreaks.

Paul Romer, an economist at New York University, said there should be mass testing in hot spots that is quickly expanded to near-universal, constant testing for everyone — 23 million tests a day, noting that the cost of tests have dropped.

“It would be feasible if we just invested and made it happen,” he said.

Other countries have used aggressive and organized testing to help stop the spread of the virus. South Korea — where the first case of the coronavirus was diagnosed on the same day as in the United States — quickly started mass testing at drive-through sites to spot and isolate cases.

The government has also instituted a sophisticated and aggressive effort to trace contacts of any known case, to squelch outbreaks. After several people who visited nightclubs in Seoul tested positive in early May, the government within two weeks tracked down 46,000 people who might have been exposed and tested them all.

In Wuhan, China, the site of the world’s first major coronavirus outbreak, government officials said they tested nearly 10 million of the city’s 11 million residents since mid-May, part of an effort to test universally and ensure the city doesn’t experience a new wave of infections.

Still, many experts agree that completely random asymptomatic testing is not an effective strategy.

A report issued late last month by the Center for Infectious Disease Research and Policy at the University of Minnesota called for ramping up testing nationwide, including in some congregate settings and as part of public health research. But the report found that widespread testing of people without symptoms was not advisable in most workplaces, in schools or in the broader community.

Researchers at the center found such testing could waste precious resources and could cause problems for communities, given that the tests are not fully reliable.

“There’s been far too much of this group think around, ‘test, test, test,’ without understanding what it’s accomplishing,” said Michael Osterholm, the director of the center. “You need the right test, at the right time, for the right reasons.”

The report’s central recommendation: that HHS form a blue-ribbon commission with national experts to formulate advice for states.

Amy Goldstein and William Wan contributed to this report.