As a high-risk obstetrician, I’m used to making tough decisions. Then pregnant women showed up with COVID-19.

I’m a high-risk obstetrician, or maternal-fetal medicine specialist, working with pregnant women in some of the poorest ZIP codes in New York City. I usually take care of preterm labor, diabetes, kidney infections and preeclampsia, things that happen frequently to pregnant women. If a patient needs intensive care, I treat her in the ICU. My job has always been complex and interesting.

Then COVID-19 started. Sick people began coming in, including sick pregnant women. We had so little information in the beginning about what COVID-19 would be like in pregnancy; what we should look for and do. Back in February, all we had was a small study out of China of seven women hospitalized with COVID-19, which was somewhat reassuring. The women didn’t seem to get sicker than the general population, and there was no evidence their fetuses or newborns were affected.

Before long, COVID pneumonia patients filled our emergency department and our labor and delivery units. We struggled with questions that had not been addressed in the small studies available. First we had those questions once or twice — and then several times every day.

In the hospital, a woman at 29 weeks of pregnancy — very preterm — was admitted with severe COVID pneumonia. Ordinarily, I would give steroids to a critically ill woman at that stage of pregnancy, to help the fetal lungs and brain mature so if we had to deliver the baby early it would potentially have fewer complications. But some studies out of China showed that COVID-19 patients treated with steroids had worse outcomes, and maternal health is always paramount in these situations. Should we give steroids? Should we not?

The next day, a patient at 36 weeks came in, breathing hard. X-rays showed the signature COVID-19 patches all over both her lungs. Thirty-six weeks is fairly late in pregnancy. A baby born at this stage would do very well. So what should we do if the woman’s breathing deteriorated further? Should we have the ICU intubate her and hold the course, rather than start a delivery plan that would stress her body even more? Or do we start her delivery process to make more room for her lungs to work?

These were the impossible questions, with very little evidence to guide us, our patients or our partners in critical care and infectious disease. In the absence of data, the medical team talked with the patient and her family about options and risks and benefits, and about what we knew and couldn’t know. We talked to each other, constantly and worriedly. We talked to doctors in our specialty and others, in hospitals nearby, in other regions, and around the world.

In the end, the medical team watched these women closely. If we had to make a decision, we did  — give steroids, don’t deliver —  and then watched closely again and changed the plan if things weren’t getting better. I'm used to making difficult decisions based on limited information; a lot of obstetrics is like that. But I've never had to do it with a strange new disease wreaking wide havoc — one that is still, months later, poorly understood.

As COVID-19 patients poured in every day, we saw more strange things about this disease. We found COVID-positive patients in labor who had lab abnormalities that made it almost impossible to tell whether a woman had severe preeclampsia, a dangerous condition that affects women at the end of pregnancy with high blood pressure and signs of kidney damage — or whether her anomalies were all COVID-related. The only way to tell seemed to be to wait. If a patient got better after she gave birth, it was preeclampsia. If she didn't, it was probably due to COVID-19.

Around the world, doctors were observing, researching and writing, just like us. From them we learned that COVID-19 has a disproportionate impact on the very patients that we take care of — people of color in some of the poorest ZIP codes in the United States, women who already have high risks of complications during pregnancy and birth.

As COVID-19 patients poured in every day, we saw more strange things about this disease. We found COVID-positive patients in labor who had lab abnormalities that made it almost impossible to tell whether a woman had severe preeclampsia — or whether her anomalies were all COVID-related. 

Later we learned that many COVID-19 patients die from cardiac failure rather than lung problems. And that perhaps clots — in the lungs and elsewhere — are part of the reason why some young people don’t get better. We learned and made decisions based on that imperfect data, writing protocols for blood thinners in COVID-positive patients and then revising them again and again and again as new information came to light. In that rewriting is the attempt to do good and not harm, without being able to know which way that harm might lie.

I had a patient with a pregnancy at the earliest limit of when a newborn can survive outside the uterus. She came in, bleeding heavily. Was this just a terrible-bad-luck placental abruption, where the placenta separates prematurely from the uterine wall? Or was this COVID-19 affecting a placenta in ways that the research doesn’t completely understand?

We also saw many pregnant women come in with COVID-19, get care and recover. Many of them had healthy babies days or weeks or months later.  I wish I understood what made the difference. We are months or years away from truly knowing.

But we are getting there. And that is because I see my colleagues leave the hospital after an overnight shift, and I hear them on a conference call two hours later, trying to figure out how we can best study what we were seeing, report our experiences, learn from doctors on the front lines in other regions and share our knowledge with them.

Today, if that patient at 29 weeks of pregnancy came to our emergency room, I would have data to show we can give her steroids without worsening her COVID-related disease. But I would not know if she will get better or worse, or whether we will need to deliver her baby early. I would still have to tell her that it depends on how her body manages this illness, and that will change every day, sometimes every hour. In the end, we have learned so much, and we have learned not nearly enough.

For the past three months, I’ve had to stop reading the news and much social media because the larger world has been beyond my control, and it helps nobody when I am angry and scared. I’ve focused on my patients, my colleagues and my family. Our country’s response to COVID-19 has been imperfect and confusing. More than anything, what I feel is tired, deep-in-my-bones tired, with vacations and weekends cancelled, endless conference calls, new COVID-19 response protocols stacking up — and with no child care and the ever-present fear that I will infect my family.

Despite all that, I’m so proud of my colleagues around this country. We are working to keep our patients — and our colleagues — safe. We keep trying to learn and share so we can all do it better, and so someone else may not have to do it at all.

Chavi Eve Karkowsky, M.D., is a high-risk obstetrics specialist in New York and a writer. Her first book “High Risk: Stories of Pregnancy, Birth and the Unexpected” (Liveright/Norton) was published in March.