Finding a Way to Better Decisions: Lupus, Pregnancy, and the Challenges of Ensuring that Therapeutic Advances Actually Help Patients

August 19, 2019

“TB” (not her real initials) is a 26-year-old Asian-American woman with systemic lupus erythematosus (lupus), a chronic autoimmune disease that can cause inflammation in multiple organs. She recently visited my clinic when she was about midway through a challenging pregnancy. This patient lives about 5 hours from my office, and I’d spoken with her local rheumatologists about 2 months earlier to help her make decisions about how to manage lupus during pregnancy.

Around that same time, TB was admitted to her local hospital with chest pain, which turned out to be caused by pericarditis, an inflammation around the heart. Her condition improved over the next week with prednisone therapy, and her rheumatologist and I planned that she would continue TB’s hydroxychloroquine, azathioprine, and colchicine through pregnancy to prevent recurrences of the pericarditis. This plan was working only partially when she came to see me – she’d been admitted twice more with episodes of chest pain and her pain hadn’t fully resolved.  She also had a lupus-related rash on her face that seemed to be worsening.

We were in a quandary. Her lupus was only partially responding to the best therapy we have that is compatible with pregnancy. There are several other medications that we can use to manage lupus outside of pregnancy, but these medications cause birth defects and pregnancy loss, and so are not prescribed during pregnancy. Prednisone can be effective therapy for lupus and it can be used in pregnancy, but in my experience, using high doses for long periods puts the pregnancy at risk for early delivery.

I resolved to check her blood levels of azathioprine and hydroxychloroquine to see if we could bump up her already high doses. To my surprise, her blood levels came back showing no measurable drug.

In other words, TB wasn’t actually taking the medication that had been prescribed.

I’ll be honest: I felt some unflattering emotions. My first reaction was anger. It felt like an affront to my authority as a physician – her rheumatologist and I had talked through the plan, yet she appeared to be flouting these recommendations. I was also dismayed – how could this woman put her unborn fetus at risk for short- and long-term complications? Then, I felt despair. I care a lot about helping women with lupus have healthy pregnancies, and this seemed like a complete failure.

But then I took a step back, as I’ve generally found that giving way to such emotions don’t solve problems. I know that many women make the choice to avoid therapies that will improve their chances of having a healthy offspring is surprisingly common. There are many possible reasons. Some stem from inadequate medical care, some from cultural norms that run counter to data, and some from personal circumstances.

Addressing Fears About Medication Use During Pregnancy

Unfortunately, the medical system itself – from clinicians to insurers – keeps some woman from accessing the care, information, and therapies that would improve the health of their offspring.  National guidelines for state-of-the-art care, covering approaches to pregnancy monitoring, medications, and delivery, are published widely each year. But frustratingly, these guidelines are often overlooked.  For example, The US Preventive Health Task Force declared in 2014 that all pregnant women with lupus should be treated with aspirin in the 2nd half of pregnancy. However, a recent analysis of lupus pregnancies managed at leading universities across Western countries found that only 25% of these pregnancies were treated with aspirin.


The prevailing cultural norm that pregnancy should be “natural” and that medications are inherently harmful to a fetus may drive a woman away from therapies that would benefit her offspring. The thalidomide disaster in the 1950s led to widespread public concern about medications in pregnancy, and while few medications have been shown to cause birth defects to a similar degree, the sense that medications are dangerous has continued. I hear this desire for a "drug-free pregnancy" expressed by many women with rheumatic disease, but for women with lupus, a "natural" pregnancy is fraught with danger. I’ve personally seen only a few women with lupus successfully carry a full-term pregnancy without the assistance of medications. For patients with lupus in particular, continuing medications has transformed pregnancy from a very high-risk proposition, resulting in numerous pregnancy losses and maternal deaths, into a relatively routine event.

Suspicions about the risks of medications are often fueled by well-meaning clinicians, pharmacists, family, and friends. While the rheumatologist or maternal-fetal medicine specialist may be best trained to prescribe medications for lupus during pregnancy, less knowledgeable clinicians and bystanders may cast doubt on their recommendations. Given a pregnant woman’s proclivity to be worried about medication safety in pregnancy, any authority figure amplifying these concerns can easily undo logical discussions and recommendations from well-informed professionals. Without a full array of reliable sources, support, and cheerleaders, a woman may be easily scared away from taking essential and safe medications during pregnancy.

I’ve found that clinicians and patients both can make skewed decisions about the risks and benefits of medications in pregnancy. For example, we have decades of evidence demonstrating that having very active lupus in the first half of pregnancy at least doubles the risk for pregnancy loss (to as high as 40%) and doubles the risk for preterm birth (to up to 66%). While active lupus doesn’t cause birth defects, it can cause life-long consequences for offspring through the many potential complications of prematurity, including vision loss, chronic lung disease, cerebral palsy, autism, and learning deficits. 

On the other hand, neither hydroxychloroquine nor azathioprine have been shown to increase pregnancy loss or birth defects.  Azathioprine is primarily prescribed to pregnant women with significant illness – lupus, inflammatory bowel disease, prior organ transplants – and these pregnancies deliver preterm more often than pregnancies in women who are not taking azathioprine; it isn’t clear whether azathioprine may increase preterm birth modestly or if it is the underlying maternal condition that increases this risk. Nevertheless, the degree of risk for preterm birth with azathioprine is not close to the 66% expected with very active lupus.

Taken together, it’s clear that untreated active lupus (40% pregnancy loss, 66% preterm birth) is far riskier to mother and offspring than a medicated pregnancy with quiet lupus (20% pregnancy loss, 33% preterm birth).  I find that some patients and clinicians may have an unanswered concern that “something bad” might happen from these medications. Nevertheless, all of the data we have strongly suggests that any possible unknown adverse effect is far rarer than the clearly documented risks of not taking these medications. 

Encouraging and Empowering Better Health Decisions

Taking medications every day can be really challenging, whether inside or outside of pregnancy. I think it’s worth thinking about medication use as a “health behavior” just like maintaining a healthy diet, avoiding junk food, and exercising. We can all see that following these health behaviors every day is a major challenge for many people; therefore, it should be no surprise that taking medications every day can also be tough. The potential barriers to daily medication use are numerous, and include factors both intrinsic and external to the patient. Intrinsic factors prevent a woman who has the medication from actually taking it and are often caused by limited motivation, side effects, and forgetfulness. In pregnancy, ongoing nausea and vomiting can make taking medications even more complicated. External factors, which prevent a patient from obtaining the medication in the first place, may include lack of access to the medication due to expense, limited funds, or difficulties in getting to the pharmacy.

When I spoke to TB, she admitted that she was taking her medications perhaps 4 times a week. Based on her blood levels, I suspected that this was an overestimate, but I didn’t press the issue. She told me that she often forgot to take her medications, rushing out of the house in the morning without her pills. I suggested that she keep a pill box with some extra pills in her purse, her car, or at work, so that she could take them whenever she remembers. I also suggested using a reminder app on her phone, or enlisting friends and family to help her take them, as additional alternatives. I suspect that the primary reason she wasn’t taking her pills was that she didn’t fully grasp the safety of these medications, or their value to the long-term health of her baby. However, she was eager to get new prescriptions, which left me hopeful that she would take them more reliably.

How can we solve the dilemma of women with lupus forgoing medications during pregnancy that will improve the health of their offspring? While there are many different options that might help, I’ve decided to focus on rheumatologists—the doctors who are most informed and involved in lupus care.


My first stab at this can be found at, a website designed to enable honest and accurate conversations about pregnancy planning and management. After talking with rheumatologists and women living with lupus, we identified the absence of honest and accurate conversations as a starting point for skewed decision-making. Rheumatologists weren’t giving women the opportunity to share their hopes and plans for a family and women weren’t insisting that their rheumatologists listen to them. In response to this, we’ve introduced the One Key Question to rheumatology— “Would you like to become pregnant in the next year?”— as a data-driven approach to opening these conversations.

We also found that rheumatologists lacked the kinds of information about medication safety in pregnancy that would enable them to provide accurate and trust-worthy recommendations. We therefore developed a clear handout and accompanying website to inject reliable, trustworthy information into these conversations. I’m now traveling the country, training providers to have these conversations and instilling the skills they’ll need to help patients make decisions about lupus treatment in pregnancy that will give their offspring the best chance to live a healthy life. 

While improving the patient-provider interaction might help, it won’t vaccinate women against the onslaught of anti-medication sentiment they will encounter outside of the clinic. For this, we’ve built a patient-oriented website and recorded patient videos. These are designed not only to provide information, but also to share the personal stories of women who have lived through similar challenging decisions themselves.

Women living with lupus need the support of other women to make the healthiest decisions. Some find this online in an array of Facebook groups and other social media spaces. While I haven’t seen them all, the groups that I have encountered appear to be supportive, positive, and medically accurate. For example, a woman may log on and ask about taking azathioprine in pregnancy, and within several hours a dozen women with lupus from across the globe will write about their positive experiences taking the medication during pregnancy, often accompanied by pictures of smiling babies. I imagine that this feedback may be even more valuable than a physician’s assurances about the safety of the drug.

Advances in lupus management in pregnancy have made possible great gains – fewer pregnancy losses, fewer preterm deliveries, and fewer women living with permanent scars from the physical and emotional trauma of these challenging pregnancies. But if we are to realize the full benefit of these advances, it’s essential that they be put into practice.

See more blog posts for Dr. Clowse


Portrait of Duke Forge Co-Director Megan E.B. Clowse, MD. MPH