When they called to tell me my COVID-19 test was positive, I thought there must have been a mistake. I felt perfectly fine, and in the long months of the pandemic my husband, Jeff, and I had been behaving the way much of the United States had: hyper-vigilant about where we went and who we saw, and careful to follow the recommended public health precautions.

Besides, Jeff had taken the same test at the same time, and his was negative. Since we did almost everything together, how could you explain that?

"It's a PCR test — they're very accurate," said the woman on the phone from the Delaware Board of Public Health when I expressed some disbelief. We had taken the test not because we suspected anything but because, having driven from our home in New York City to Delaware for a weekend break, we wanted to be good citizens and follow New York's rules regarding travel out of state when we returned.

The woman on the phone went down a list of the things I needed to do: Stay home. Don't leave the apartment for any reason. And isolate from Jeff.

That last bit surprised me — my husband and I certainly hadn't been isolating from one another between taking the test on Nov. 7 and getting the results late on Nov. 11, so what was the point in doing so now? If I was going to expose him, surely the most dangerous period had been those four crucial days when I was supposedly shedding coronavirus in the bed we shared, the sips of bourbon we exchanged, and the three-plus hours in the car together, windows closed, during our drive home to New York on Nov. 9 after our weekend away.

As a science journalist who's written about emerging viruses for 30 years, I knew enough to ask the Delaware public health rep whether the lab report included a Ct number. Experts have been saying that the Ct (cycle threshold) number is crucial to understanding how the coronavirus spreads and how it causes disease. It can offer helpful information not only for epidemiologists but for individuals, too — those of us desperate to predict, as I suddenly was, the likely trajectory of our illness after a positive test result.

The Ct number tells you how many amplification cycles through the PCR machine are required before the lab can get a detectable level of viral RNA. If the lab had to go through the cycle just 20 times or so, that means you started out with a relatively high viral load, and it would make sense to expect that you'd get pretty sick. (Though even that is conjecture; no one has kept good track of the relationship between viral load and severity of symptoms, so what we're left with here is mostly an educated guess based on correlations seen in some observational studies.)

If, instead, it took more rounds of amplification — say, 35 or 40 cycles — to be able to detect virus in your sample, maybe that means you started off with very little virus at all.

According to an analysis by The New York Times last summer, the labs that keep track of Ct numbers tend to report them at 37 to 40 — meaning they run a sample through 37 to 40 amplification cycles, if needed, before deciding whether to call it positive or negative. That could mean at least some of the people deemed positive for COVID-19 started off with minuscule amounts of virus. And while there's no direct evidence that people with those high cycle numbers would be less sick than those with low numbers, there's good evidence that they would at least be less contagious.

The Times analysis found, among other things, that the New York state laboratory uses a Ct of 40 to make a positive designation. If it used a Ct of 30 as a cutoff instead, the Times calculated, 63% of the people identified as positive would instead be told that no virus could be detected — in other words, they'd be told they did not have COVID-19.

I was curious to know on which side of that boundary my own test fell.

The woman on the phone had never heard of the Ct number. She said I would get a follow-up call in two or three days from a Delaware contact tracer, and I could ask them.

And with that, I entered into a real-world demonstration of how little we really know about the coronavirus.

To begin with, no one from Delaware ever called me for a list of my contacts — an indication of how, even before Thanksgiving, COVID-19 was already outpacing public health officials' ability to corral it. I did have a few contacts — Jeff and I had been cocooning, but even cocoons are permeable, and though we'd dutifully worn masks in public, we had been in range of about half a dozen people in the previous week. I was left to alert those contacts on my own.

In addition, every time I tried to understand more about my own situation, I ran into a stone wall of incomplete information. Which is a terrifying prospect at this moment in the pandemic — as transmission, illness, and death rates from COVID-19 continue to ramp up exponentially.

After the call from Delaware, I put on my mask and made up the bed in the spare room, preparing for my 10 days of isolation inside our home, as the CDC recommends. Even though the fact that I felt perfectly fine made it all seem a bit surreal, Jeff and I decided to assume the test was right and to operate as if I really was infected, which meant staying at different ends of the apartment from each other. We're both lucky enough to be able to work remotely — I'm a freelance, he's a college professor — so quarantining didn't have any financial repercussions for us.

"No, it's not wrong; PCRs are very accurate," my doctor said on a video call when I asked her whether the test might have been a false-positive. She said symptoms would probably show up within the next week. I told her that was not reassuring. "Anything I can do to try to stave them off?" I asked her.

Maybe I should take some Vitamin D, I thought. Or get a humidifier? How about my oxygen level; should I keep track of things with the pulse oximeter we bought months ago in our first round of pandemic panic?

"Nope," she answered, a bit too cheerfully, I thought, though it was hard to tell for sure with her mask on. My doctor seemed to think symptoms would be showing up soon, which was a scary prospect, and she told me to schedule another tele-visit if (when?) I started to feel really sick.

I went to bed the night of Nov. 11 completely symptom-free, yet anxiously waiting for the other shoe to drop.

All the reading and reporting I had been doing since March suddenly became intensely personal. I was no longer just playing the odds by following public health advice about "mitigation," trying to limit spread in the highly unlikely event that I was spewing virus from my mouth and nose. This was the real thing now. For my husband's sake, I was trying to spread less of the virus I knew I was carrying.

I wore a mask, all the time. I opened all the windows in the apartment even though it was cold outside. I washed my hands so maniacally that the little diamond ring I inherited from my grandmother shone as if it were brand new. I closed the cover on the toilet whenever I flushed.

On Nov. 12 I still felt fine. I contacted the people I had interacted with just before my test: two couples in Delaware, in each instance outside and at a distance; and our housekeeper in New York, who had been in the apartment for a few hours on Nov. 4 and had worn a mask, as had Jeff and I. I told them I was sorry to have endangered them unwittingly. They all went for COVID-19 tests.

My doctor thought I was probably infected on Nov. 3 or 4, but wasn't that just a guess? Maybe it was Nov. 1, when I'd opted for in-person early voting. Maybe it was Oct. 31, when Jeff and I'd spent a few hours with our daughter, son-in-law and two young granddaughters, ages 2 and 5, in their Brooklyn neighborhood. We'd been mostly outside and mostly masked during the visit, but when it started to rain we had moved our bagel brunch into their garage-cum-party room, leaving all the doors and windows open. After they heard I'd tested positive, my daughter and son-in-law got tested, too, along with their 5-year-old.

But maybe it was none of these times. Maybe I got infected months ago, and this was just the long-simmering tail of a completely hidden infection. As far as the CDC knows, tests for COVID-19 can still be positive up to three months after the symptoms resolve. So can you count three months backward, too, for someone whose symptoms never appeared? This is when it would have been really handy to know the Ct number for my test.

On Nov. 13 I still felt fine, though I kept my mask on during a Zoom call with three old college friends, just to reduce the amount of virus I might be emitting into the apartment. I wore my mask during a call with both our daughters on Nov. 14, too, when we Zoomed in to do the Sunday Times crossword puzzle as a group.

"You're not going to infect us through the screen," my older daughter teased, but I kept the mask on anyway, I guess for the sake of Jeff, who was eating some of his meals in the room I was calling from. I really, really didn't want him to get infected; research finds men tend to get sicker with this coronavirus infection than women do.

All my contacts tested negative for COVID-19. That was reassuring. But not completely so, since most of them used a rapid antigen test that can have disconcertingly high rates of both false-negatives and false-positives. Once again, the details of trying to chart one individual's experience with coronavirus revealed the patchiness of our ability to track the virus down, to test reliably for its presence, and to stay ahead of its devastation.

My lack of symptoms was beginning to feel, to be honest, too good to be true. Why should I be this lucky? Harboring a virus that has flattened America, causing disease, death and endless despair for hundreds of thousands of us. It made no sense that I should be spared the worst of its effects, when so many others with this diagnosis have been so grievously harmed.

I wasn't even an especially good candidate for getting off easy. I'm 67 — part of the cohort of people at higher risk of death from COVID-19. And while I'm basically healthy and don't have medical conditions that can be associated with a bad COVID-19 outcome, my aging immune system is without doubt more sluggish than it used to be.

I researched further, trying to explain my inexplicable good fortune. Epidemiologists don't know much about what differentiates people with symptoms from those without, partly because asymptomatic cases — who the CDC estimates make up some 40% of people with COVID-19 nationwide — tend to be invisible.

Did I remain symptom-free because I always wear a mask? Studies suggest that the proportion of asymptomatic cases is higher in regions where a greater proportion of people are wearing masks in public. Maybe masks reduce the amount of virus you take in when you do get exposed, and maybe less virus means fewer symptoms.

Or maybe my luck could be traced to other habits. I encountered one study, for instance, that indicated that getting certain vaccines might offer some protection from COVID-19 as an unexpected benefit. Two weeks before my positive test I had gotten a pneumonia vaccine, and back in September I had gotten a flu shot. Could one of those immunizations have been relevant?

I'm just spitballing here. But, in a way, that's what even the experts have to do in many cases. Though they're learning all the time, they still know very little about exactly how this virus new to humans behaves in the context of any one person's particular mix of genes, physiology, environmental exposure or any of a number of other factors that could help explain why some are hit so much harder than others.

Jeff went for a second PCR test on Nov. 14 and, because of the overload of testing facilities in New York, he was still waiting for results on Nov. 17 when I emerged from my isolation period. I'd never developed any symptoms.

When Jeff's results came back the next day — negative for COVID-19 — we were greatly relieved, but also, perhaps perversely, a little disappointed. Now what lessons were we supposed to draw?

Maybe the caution we'd used during my week of isolation had protected him. Maybe my lack of symptoms showed I had a very low viral load and was never really going to pass it on. Or maybe my own test was a false-positive — perhaps from contamination or a mix-up in the Delaware lab? — and I never had COVID-19 at all.

Three weeks after my positive COVID-19 test, I walked over to a testing site in Manhattan to get my blood drawn for an antibody test, just to see my story through. It was negative — another disappointment, and one more bit of data I wasn't quite sure how to interpret.

Antibody tests have a high rate of both false-negatives and false-positives. Combined with my positive PCR test for the virus, I would have considered a positive antibody test to be confirmation that I really had at one time been infected with the coronavirus, no matter how healthy I felt all along. But a negative antibody test? That presented more of a puzzle.

It could mean I was never infected with the coronavirus. Or it could mean I was just slow in building up a supply of antibodies, but they would show up eventually. Or it could mean, as I find myself musing now at 3 a.m., that I did in fact have an asymptomatic case, and I'm silently harboring a nice healthy store of T-cells — another protective form of immune cells that most commercial labs don't look for yet. This last bit, I realize, is no doubt just middle-of-the-night wishful thinking.

But there's one thing I do know, in light of this experience: Despite a global scientific effort that has led to a spectacular burst of new information this past year about a previously unknown pathogen, we can say very little with confidence about how the coronavirus will behave inside any one of us.

We're still unable to tell a fully fleshed-out story about a particular individual's encounter. Not the beginning, about how and when the virus was transmitted; nor the middle, about what symptoms it will cause; nor, especially and most distressingly, anything about how it will end.

Copyright 2020 NPR. To see more, visit https://www.npr.org.