I waited for my COVID-19 test in July, coughing and mildly feverish, in a parking lot in a small Colorado town called — no joke — Delta, and I thought to myself:

Seriously?

Like so many others, I have spent the last year and a half being as careful as possible, and I got vaccinated as soon as I was eligible. Yet now, after being exposed to the coronavirus by an unvaccinated nephew at a family reunion, it looked as if I’d been nailed by the delta variant.

But the answers I got were not 100% clear. The result from the first test was negative, while a second test, back at home in Pennsylvania, came back as “inconclusive.”

Lucky me, I get paid to ask questions, and I had a bunch:

Notwithstanding my unclear results, was it nevertheless likely that I had COVID-19? If so, did that mean my protection from the vaccine was waning, perhaps because my antibodies were not a perfect match for the delta variant? Did the vaccine nevertheless protect me from more severe illness? And is delta harder to detect with the standard nasal-swab tests?

I spoke to Brian DeHaven, a La Salle University associate professor of biology who studies viruses and the immune system; Matt Tugwell, director of genomics at Mako Medical Laboratories, the Raleigh, N.C., company that analyzed the sample in my second test; and Alfredo Penzo-Mendez, scientific quality manager at ECRI, a Plymouth Meeting-based nonprofit that evaluates the safety and quality of health care.

My main takeaway: It’s clear that the vaccines remain a powerful tool, even against delta, the best option we have for returning to normal. But with just half of the United States fully vaccinated, and many areas well below that threshold as other variants emerge, the pandemic is sure to drag on for months.

» READ MORE: Some people’s minds are changing about the coronavirus vaccine. Here’s how doctors persuade them.

I recovered just fine from my apparent case of COVID-19, though I spent two days in bed. But my mother-in-law — a lymphoma survivor who turned 82 this month, and thus was at higher risk despite being vaccinated — had a harder time. She tested positive and had to go to the hospital, though now is doing better at home with supplemental oxygen.

The nephew, who is in his early 20s, has recovered, and of course he feels bad that his grandmother got sick. The reunion included 15 people in a cabin in the San Juan Mountains, and we learned too late that he, his two siblings, and his mother all were unvaccinated. He was under the impression that his older relatives, having been vaccinated, were adequately protected, and that his choice to skip it affected only himself. Nope.

What my test results meant

My symptoms started three days after we all arrived. Two days later, at the parking lot in Delta, in western Colorado, a technician stuck one of those familiar swabs up my nose. The sample underwent a process called polymerase chain reaction (PCR), meaning a device copied genetic material in my sample that matched any of three target regions in the coronavirus.

Such devices are run for dozens of cycles, roughly doubling the amount of viral genetic material each time. At a predefined point — say, after 35 cycles, though some labs use a different cutoff — the test is considered positive if the target sequences can be detected. If not, it is judged to be negative.

In other words, I had no detectable signs of virus in the first test. That might have meant that I did not have COVID-19. Or maybe I did, but I got tested too soon in the course of infection. It can take a few days for a patient’s “viral load” to reach detectable levels.

But nearly two weeks later, I still was coughing a bit, and I figured a retest was in order. This time I went to a Montgomery County test site, at Willow Grove Park mall. The sample was processed at Mako’s 65,000-square-foot lab in North Carolina, which has processed more than eight million COVID samples to date, according to company chief operating officer Josh Arant.

When I learned on July 30 that my test was inconclusive, the result was accompanied by this message: “Amplification of viral targets is present, but the amplification is not above the analytical cutoff value required to interpret the sample as positive.” That sounded to me as if I’d been infected.

Very likely, agreed Tugwell, the genomics director, when I called him the next week. Inconclusive results, which occur in fewer than 1% of cases, can mean the person was tested too early during infection, or perhaps on the tail end — which would be consistent with my timeline, he said. Or maybe there was contamination (Tugwell says that is rare), or not enough gunk on my swab. But it sure looked plenty gunky to me.

» READ MORE: As delta continues to spread, some Philadelphians embrace city’s new mask mandate

Is my vaccine wearing off?

More cases are occurring in people who have been vaccinated, prompting concern that the protection is wearing thin.

At our family reunion, for example, three people developed symptoms of COVID-19, in each case several days after being exposed to the nephew, who had been suffering from a sore throat, dizziness, and fatigue. (I’m including myself in the three new cases, despite the inconclusive test results. The two others were my vaccinated mother-in-law and the nephew’s unvaccinated mother, both of whom tested positive.)

That meant two of the three new symptomatic cases occurred in vaccinated people.

But to frame it that way is misleading, said DeHaven, the La Salle biologist. To gauge the level of protection afforded by the vaccines, the 14 people exposed by the nephew should be separated into two denominators: the 11 who were vaccinated and the three who were not.

In other words, one out of three unvaccinated people got sick, and two of the 11 vaccinated people got sick.

That suggests the vaccines may have prevented disease in as many as nine of the 11 — assuming that we all had ample opportunity to catch it from the nephew, cooped up in the same cabin for days. And my illness was mild. Had I not been vaccinated, it might well have been worse, DeHaven said.

“My guess might be, yeah, you had a little bit of a breakthrough case,” he said, “but the vaccine did its job.”

Then again, two of the three unvaccinated people did not get sick, either. Some luck is involved.

A similar pattern holds true at the national level. More and more breakthrough cases — such as my mother-in-law’s and maybe mine — have been occurring, simply because so many millions are vaccinated. Even though the vaccines continue to prevent most infections, the share of vaccinated people who get sick — a small percentage of a large number — likely will total in the thousands.

But what about delta?

My nephew, who lives in California, got sick as delta was overtaking that state, so it’s a fair bet that he (and later I) was infected with that strain. But how would he know?

The PCR tests can detect delta just as well as they do other variants, Mako’s Tugwell said. But they are not designed to distinguish among strains. For that, the entire genome of the virus must be sequenced.

» READ MORE: Philly-area Holocaust survivors say vaccination is a simple choice, not a political fight: ‘I want to live’

Mako does that type of analysis, too, under a contract with the U.S. Centers for Disease Control and Prevention. But for now, the information is shared only with public health officials. The company is seeking clearance to share the name of the specific variant with patients, as well, he said. What one would do with that information is less clear.

Evidence suggests that delta spreads more easily, but someone with COVID-19 should stay away from others no matter what the strain. So that advice now seems even more prudent, even if the others are vaccinated.

But do not worry too much about recent studies that suggest the vaccines may offer dramatically less protection against delta, said Penzo-Mendez, the data whiz at ECRI. Those studies, from the Mayo Clinic, Israel, and elsewhere, are real-world, observational studies, meaning the results can be skewed by factors beyond researchers’ control. For example, cases may have risen in areas with high vaccination rates in part because local officials relaxed other precautions, he said.

The best evidence, from ongoing controlled, clinical trials, suggests that there is a “modest decline” in vaccine effectiveness after six months, but that the injections continue to prevent most cases of severe disease, he said.

So sign up for that shot. And watch out for unvaccinated nephews.