Testing errors, misdiagnoses and a sluggish public health response left patients out of official tallies.

When newscasters announce the latest tally of coronavirus cases, Arthur Hall chuckles and turns to his wife.
“Whatever number they say, I’d add one,” he said.
Hall is an uncounted survivor of coronavirus.
The 51-year-old school administrator spent five days in the hospital with what doctors told him was severe respiratory distress caused by COVID-19. Although he had tested negative for coronavirus, his Delaware care team said the tests available in early April were unreliable.
Experts estimate tens of millions of Americans contracted coronavirus but are not included in official tallies because of testing errors, misdiagnoses, a sluggish public health response, and ignorance about the disease during its early days in the United States.
The Centers for Disease Control and Prevention (CDC) reported Thursday that there have been about 3.5 million confirmed cases of COVID-19 nationwide and at least 136,000 deaths. The actual number of infections is likely 10 times the number of reported cases, CDC Director Robert Redfield said in a news conference on June 25.
The nation’s understanding of the pandemic is hampered because so many victims remain uncounted. Americans who paid a physical and emotional toll for COVID-19 describe how being excluded from official statistics exacerbates their isolation, uncertainty, distrust and fear. And without an accurate count, researchers might draw flawed conclusions about the virus’ behavior, local leaders might make poor re-opening decisions, and citizens might underestimate their own risks.
USA TODAY interviewed dozens of people who were diagnosed by a doctor after a presumed false negative test or who were unable to get tested for coronavirus at all, as well as the friends and family of some who might have died from undiagnosed COVID-19 early this year. 
Whatever the reason official tallies missed them, most uncounted survivors said they just want their experience to matter.
“Who are we?” asked Emily Talkington, an uncounted survivor in California. “Did we just battle this out for nothing?”
On her worst nights of illness, Talkington said she awoke soaked in sweat, unable to sleep because of difficulty breathing, joint pain, severe itchiness and burning eyes.
When the 45-year-old woke up gasping for air, Talkington thought she might die. She wondered how long it would take people to notice since she lived alone. Sometimes, she hallucinated that her dog lying in bed next to her was dying and that she, a vet, could do nothing to keep him alive.
“I travel worldwide quite a bit with my work. I’ve had Zika, Dengue, Malaria,” she said. “This felt very different.”
She visited a drive-up testing site and received a negative result that left her wondering why she was ill if it was not COVID-19. After four weeks of bed-bound sickness and isolating at home, Talkington coughed up blood and nearly blacked out. She took herself to a Santa Cruz urgent-care clinic on a day she had the strength to go.
The physician assistant who treated her in early April didn’t believe she was sick at all.
“He thought I was making it up,” Talkington recalled. “He said, ‘Well you just have menopause … and you’re coming up with things in your head.’”
Days later, she said her regular doctor told her that the physician assistant had emailed him to warn she might be on the verge of a mental break.
Her doctor also reassured her she did not have perimenopause — which does not cause women to cough up blood — and agreed that she likely had been suffering from COVID-19. He told her the urgent care clinic staff probably hadn’t read the latest CDC guidance, which recognized a broader array of symptoms for COVID-19 than just a dry cough and fever.
Talkington still suffers from heart and blood pressure symptoms, which she says are documented daily as part of two medical research projects on coronavirus. Information about older cases like hers is critical to accurately understanding how the virus behaves to develop better treatments, vaccines and public health interventions.
But Talkington doesn’t tell most people what she’s been through. Some relatives and friends have told her they are skeptical she really was so sick for so long. When she runs into people she knows, they often comment on the 20 pounds she lost while ill.
Talkington doesn’t tell them the truth. Instead, she says, “I’ve been eating healthy and walking my dog.”
Christine Orrick, 47, was sick for nearly a month in Denver, Colorado.
For the first two weeks, she thought she was just anxious and exaggerating a cold as details of severe coronavirus cases flooded the news in late March and early April. After all, she already had a history of chronic illness, including asthma, pain and hypertension.
In the third week, Orrick became much sicker: fever, chest pain, difficulty breathing, body aches, coughing, swollen lymph nodes and a dysfunctional gastrointestinal system. Her doctor said she almost certainly had COVID-19, but told her to stay away from the hospital unless she could not breathe.
Christine Orrick, was sick for nearly a month in Denver, Colorado.
It was harrowing. As the chest pressure and pain ramped up, it was difficult for me to know, is this a worthy time to call my doctor? Will I die in my sleep? How bad does it have to get before I go in.
After more than a month of escalating symptoms, her doctor fudged the rules that limited who could be tested. She approved one for Orrick, a former mental health therapist, on the grounds she was an essential health care worker. The result was negative, but her doctor told her it was likely inaccurate. The tests available to her had a known 20% false negative rate and few people still have active virus cells in their system so many weeks after being infected.
Orrick is grateful for her doctor, who helped her navigate a “terrifying” few weeks when no one really understood how the virus worked. Like other survivors, Orrick said it is important their experiences are recognized even if they are not studied or added to official counts.
That validation is important, Orrick said, in part because survivors’ trauma continues long after physical recovery. Doctors still do not know the long-term health consequences of COVID-19. Friends dismiss their stories as attention-seeking tall tales. Survivors watch neighbors resume normal activities as if the coronavirus isn’t still spreading. 
As Orrick continued to recover, she recognized the symptoms of Post-Traumatic Stress Disorder. Without a positive test, she faced additional scrutiny and personal uncertainty about whether her illness was real.
“We were put in a double bind — told by the experts that we should not seek help until it was too late, and then due to poor testing or testing too far out, we felt additionally invalidated in our illness (when we received a false negative),” wrote Orrick. “This left many of us in a powerless place.”
Orrick, Talkington and others found solace in one place: Facebook.
Coronavirus survivor groups have proliferated on social media sites. For some, the posts spark new anxiety about what they might face next. Most, however, said they were comforted by the informal networks, reading personal stories that often reflect their own.
“Thank God for that group,” Talkington said. “I’m not the only one feeling disheartened and completely defeated.”
In Cheyenne, Wyoming, Chase Brumley received an email in March informing him that a person with a confirmed case of coronavirus had been in his office at the Department of Transportation.
Brumley, 28, was already ill. He had asthma but was deemed a low-risk case because of his age and was denied coronavirus testing. Public health officials told him to stay home.
Brumley became so weak he couldn’t walk across the apartment without collapsing from fatigue. His lungs felt like they were so full there was no room for air. One night, his lips turned blue, and his wife begged him to go to the emergency room. Brumley refused, remembering a nurse practitioner had told him during a video appointment that they wanted to preserve hospital beds for high-priority patients.
He described his experience as a paradox: “You simultaneously had it and not had it. It was Schrodinger’s coronavirus.”
Accurate and available testing is critical, not only to better treatment for patients like Brumley, but also for tracking, tracing and stopping coronavirus’ spread. If people with COVID-19 can be identified and isolated before they infect someone new, communities can avoid widespread shutdowns, hospitalizations and deaths. Often the virus is spread before people develop symptoms as bad as Brumley’s or without showing any signs of illness at all.
“Most cases are not going to lead to outbreaks or super-spreading events, but some are. And you don’t really know which ones those are going to be,” said Jaline Gerardin, an assistant professor of preventive medicine at Northwestern University’s Feinberg School of Medicine. “They can be devastating. They can restart your outbreak.”
A complete count also helps people determine when it’s safe to return to normal activities.
“It’s important, both for directing resources at the time, but also to plan for future needs,” said Dr. Donna Hansel, who leads the pathology department at Oregon Health Sciences University’s School of Medicine. “We can’t forget what’s happened and think it won’t happen again.”
Two months after falling ill, Brumley described lingering symptoms and chest pain. On bad days, he can’t go downstairs without getting winded. On good days, he can go to the grocery store but returns exhausted.
“I’m just one person,” Brumley said. “How many people did this happen to?”
Early research suggests entire communities might have been dramatically undercounted. Experts say gaps in testing can be widest in low-income and marginalized neighborhoods unless health officials and local elected leaders make concerted efforts to reduce barriers.
One April study highlighted how income, work and cultural barriers to testing contributed to undercounting cases of coronavirus early in the pandemic. 
Researchers from the University of California, San Francisco collaborated with community organizers and the local government to test nearly 3,000 people in San Francisco’s Mission district, a historical immigrant neighborhood where one-third of residents are Hispanic.
“The results suggest that as many as 1 in 50 people living and working in the Mission could be actively infected with the virus, and that many are likely to be asymptomatic,” the group wrote in its initial report.
Latinos accounted for 44% of residents in the tested area, but they made up 95% of the positive tests.
It was the first time neighborhood leaders had data to back up what they had heard anecdotally about coronavirus’ spread, said Susana Rojas, who is executive director of the Calle 24 Latino Cultural District in the Mission.
“It gave us a direction and a focus,” she said. “Before, we didn’t know for sure. We didn’t have the data to confirm it.”
The challenges that make it easy for Latino workers and families to contract the illness, as well as those that make it difficult for Mission residents to access testing, are the same structural problems that lead to poor health outcomes for many other diseases: low-income, no health insurance or paid leave, high-exposure jobs, crowded homes, and complicated rules about safety-net services for people without permanent residency or citizenship. Many of those factors also were confirmed and quantified by the UCSF study.
Among those who tested positive, 90% could not work from home and 89% earned less than $50,000 a year in one of the nation’s most expensive cities. Almost 60% lived in homes with three to five residents and another 29% reported even larger households.
 “The communities that have been historically underrepresented and underfunded because of race are going to be the ones that bear the brunt of the pain any pandemic will bring,” said Jon Jacobo, a community organizer with the Latino Task Force on COVID-19 and a lifelong resident of The Mission. “If we all understand that framework, we are able to prepare and hopefully push the resources to always help those with the least first.”
Arthur Hall isn’t included in official COVID-19 tallies because of a false negative test result.
Hall, a 51-year-old school administrator in Wilmington, Delaware with no pre-existing health conditions, had taken several trips in the early spring to Washington D.C. for work and to the West Coast to visit friends. His family had planned a spring break trip in mid-March, but their flights were canceled amid coronavirus concerns.
Three days later, Hall started to feel back pain. Then, he became fatigued, and his breathing was labored.
“I was out of breath taking a shower, just standing there,” he said.
He said his doctor told him not to go to the hospital, which would send him home. She said to call 911 if he couldn’t breathe. By that night, he couldn’t. And he started vomiting. Paramedics loaded him into the ambulance so quickly that Hall’s wife didn’t have a chance to ask which hospital they would take him to.
“That experience was pretty traumatic for all of us,” Paula Hall said. She developed mild symptoms but was never tested because supplies were limited to serious cases at the time. 
Arthur Hall’s symptoms worsened at the hospital: fever, chills, hallucinations and low oxygen levels. He was hospitalized for five days and tested negative for coronavirus twice. Doctors said not to be bothered by the tests, which they described as unreliable.
“There’s no way this was just pneumonia,” Hall said. “I don’t necessarily begrudge the fact that I came up negative. I’m just happy to be alive and with my family.”
The day after he returned home, one of his daughters turned 12 years old. He was supposed to isolate himself in a room away from family until he had a second negative test. He broke the rule briefly to join the celebrations from six feet away, wearing gloves and a mask.
“It took my family a while to get used to me being home and even longer to make sure I wasn’t contagious anymore,” he said.
His wife agreed that it was nerve wracking to not know when exactly her own husband was safe to be around her and their children.
“I just did not want to be in the same position knowing we had to take care of the girls,” she said. “And with these new cases in kids in New York City, I guess as a parent that’s my next worry.”
Melissa Hennings, an IT worker at a hospital in Berkeley, California, suspects she contracted COVID-19 in December.
When her fever shot up to 103 and she developed a bad dry cough, Hennings feared pneumonia. In a tele-appointment with a doctor, she had trouble describing her symptoms because talking triggered coughing fits that made her nauseated. After a few days of medication, Hennings still did not feel better, so she drove to the ER.
“They diagnosed me with pneumonia in both lungs and multiple nodes,” the 48-year-old said. “They hooked me up to four bags of IV fluids. I had two different nebulizer treatments in nine hours and my oxygen didn’t come up by much.”
Doctors conducted several tests but found nothing. Against their recommendation, she went home to ride it out. Hennings said it took months to fully recover.
“That part is really crazy, having actual doctors say to you, ‘You are really, really sick, and we have no idea how you got this sick this fast,’” she said.
Like Hennings, many interviewed by USA TODAY said they were seriously ill before the World Health Organization declared a global pandemic and as states struggled to gather testing supplies. 
Because their coronavirus antibodies might dissipate before antibody tests are developed that doctors broadly trust, they might not receive formal confirmation of their COVID-19 experience.
Melissa Hennings
I dont need to be counted in the cases necessarily, Im just wondering, what if we could be helping somehow?
Gerardin, from the Feinberg School of Medicine, said identifying early cases like Hennings is helpful in understanding how the virus spread before public health leaders intervened with shut-down orders.
“It’s important to understand what the baseline looks like,” she said. “If we gave up and went back to 100% normal life, that’s what spread would look like.”
Months before federal leaders identified and acted on community spread in the U.S., some people were dying from coronavirus. Many will never be identified, experts said.
Understanding how many more people died from COVID-19 may not change public health policies, said Nicholas Jewell, a professor of biostatistics at The University of California – Berkeley. But it matters deeply to communities that have experienced loss.
“It’s recognizing death and the passage of loved ones. The impact we had on families is incredibly important as humans,” he said. “We have to get beyond just hearing a number.”
Angie Summers’ friends and family believe she might be one of those early victims.
Friends of the 57-year-old former Fort Worth newspaper columnist said she had not traveled recently nor did she have medical conditions that would have put her at high risk for COVID-19.
But on Feb. 18, Summers texted friends and family, saying she was sick but a flu test was negative. Doctors told her that some of her lab results had been “troubling” and “puzzling.”
Three days later, Summers died. An ambulance brought her to the hospital to be treated for pneumonia complicated by sepsis. Within hours, kidney failure triggered a heart attack that killed her.
That same day, Dr. Nancy Messonnier, director of the CDC’s National Center for Immunization and Respiratory Diseases, told reporters that they had not yet seen community spread of coronavirus in the United States but warned that a pandemic was “likely.”
It would be weeks more before officials confirmed that the virus had, in February, killed several Americans who had no international travel history.
For months, Summers’ friends and family checked the Tarrant County Medical Examiner’s Office’s website over and over to see if officials had identified her cause of death.
“For me, the fact that we still don’t have answers has provoked a range of painful emotions,” said her friend Rhonda Aghamalian. “I think that it’s human nature to seek an answer to the question of what took her from us. And I also feel quite strongly that if Angie succumbed to COVID-19, she deserves to be counted and recognized as a victim of the pandemic.”
The Tarrant County Medical Examiner’s Office cited privacy laws as it declined to discuss Summers’ case.
Four months after Summers’ death, family members learned officials had finally finished their investigation.
Summers’ official cause of death: “Undetermined.”
Investigative reporter Jessica Priest contributed to this story.
Illustrations by Veronica Bravo, Jennifer Borresen and Javier Zarracina/ USA TODAY