Inflammatory bowel disease (IBD) is a chronic, relapsing, and remitting autoimmune disorder comprising Crohn disease (CD) and ulcerative colitis (UC). There are over 1.6 million Americans living with IBD and as many as 70,000 new cases are diagnosed annually [1]. IBD is frequently diagnosed in the second to fourth decades of life, with the highest incidence between 20 and 29 years of age, during peak female reproductive years [2].

Treatment of IBD patients during their reproductive years provides a unique clinical challenge for clinicians, who seek to maintain remission during conception and pregnancy while ensuring the health of the fetus. Adverse outcomes such as preterm delivery, low birth weight, and increased risk of miscarriage are all associated with the degree of disease activity at the time of conception [3,4]. Women who plan their pregnancies while in remission are less likely to experience disease flares during pregnancy and have similar fertility rates as the general population [5]. A growing body of literature now suggests that the majority of IBD medications are low risk for use during pregnancy and even during lactation [6]. Despite this, observational studies note that women—particularly those with CD—have higher rates of voluntary childlessness due to concerns surrounding medication side effects, passing the disease to their offspring, fear of infertility, and advice given by treating physicians [7,8]. Moreover, many women perceive medications to be unsafe during pregnancy [9-11]. Consequently, women are uncertain about taking medications during pregnancy and many believe that medication use should be highly restricted during pregnancy, even when discontinuing medication use might threaten the health of the mother [10]. In one study, one-third of women believed that all medication use should be halted during pregnancy, and 20% of patients surveyed said they would stop medications even if they were advised by a physician to continue medication use [12]. Of the women surveyed, 68% reported anxiety or worries related to the effects of drugs on their pregnancy [12].

Social media use is widespread in the United States. More than two thirds of Americans report using Facebook every day, and the typical American is on 3 social media platforms [13]. Additionally, millions of individuals around the world use the internet to describe and share their health care experiences while seeking a sense of community for information-seeking purposes. A study reported that 34% of Americans reported reading about someone else’s health care experience and 15% consulted online reviews of medical facilities [14]. Social listening, or examining social media and forum posts, can reveal trends in public attitudes and behaviors about various health care topics. For example, our group has used social listening to examine patient beliefs about the use of opioid medications and their gastrointestinal side effects, public opinions about the use of virtual reality in health care, and patient beliefs about biologic medications in IBD [15-17]. Researchers have used social listening methods to examine e-cigarette attitudes and posts in the United States, estimate trends in influenza epidemics, and monitor the use and misuse of antidepressants [18-22]. This approach offers unique advantages compared with traditional interviews or focus groups, including the ability to capture discussions among social media users without a moderator present. As a result, social listening may capture opinions from individuals on sensitive topics or from those who might not necessarily choose to participate in a research study [23]. Participants also create communities within social media and network forums that draw patients from many different countries; gathering such a diverse and large group of participants for a focus group or interview-based study would be both logistically challenging and prohibitively expensive.

The percentage of adults who use at least 1 social media site has increased steadily since 2006. Individuals of certain demographic groups are more likely to use and engage in social media compared with other groups. One exception is race—approximately the same proportion of black, white, and Hispanic Americans use at least one social media site [24]. However, income, gender, and education have been shown to be associated with different levels of social media use. A larger proportion of individuals making more than $75,000 per year (77%) reported using at least 1 social media site, compared with 63% of Americans making less than $30,000 per year [24]. Women are also more likely to use social media compared with men [24]. Education also plays an important role—with 79% of Americans with a college degree reporting use of at least 1 social media site compared with 60% of Americans with high school or less education [24]. Age is also an important factor. Although older Americans are increasingly using social media, a greater proportion of younger adults are on social media compared with older adults. A total of 88% of 18- to 29-year-olds in the United States report using some form of social media compared with 37% of Americans 65 years and older [13]. These numbers have some implications for social listening, with a potential bias toward capturing voices from younger individuals with higher incomes and more education, and potentially capturing more female voices.

In this analysis, we examined social media and health forum posts to explore how patients with IBD understand, discuss, and act on the perceived risks and benefits associated with taking IBD medications during conception, pregnancy, and breastfeeding outside of the physicians’ office. The objective of this study was to gain a more thorough understanding of how individuals taking IBD medications during key reproductive periods make decisions about their medications to inform clinical practitioners. Having a better understanding about why patients stop or refuse to take certain medications during key reproductive periods may allow clinicians to address specific beliefs and attitudes during office visits.

We collaborated with researchers from Treato, a social media data mining service, to find relevant patient experience data. Treato uses Natural Language Processing computer algorithms to collect and index patient, family, and caregiver content from over 3000 websites and social media sites such as Facebook, Healingwell, and TheBump. Posts are indexed using a lexicon of over 100,000 medical terms based on the Unified Medical Language System.

To find IBD-related posts, we extracted relevant data from an e-forum and social media database, using a set of relevant keywords developed through a literature review (Textbox 1). We sought to identify posts about medication use among individuals with IBD and developed 2 category lists—the Conditions category, which included names and abbreviations for IBD, UC, and CD; and the Medications category, which included brand and generic names and abbreviations for medications indicated to treat IBD, including biologic therapies. Posts were selected for analysis if they contained a keyword from the Conditions category as well as a keyword from the Medications category. We further culled posts by using pregnancy and fertility keywords to identify discussions relevant to this study. We analyzed English language posts published online between June 27, 2012, and June 27, 2015. Textbox 1 provides the list of the keywords used.

After identifying relevant posts, we coded the resulting text corpus using ATLAS.ti (Berlin, Scientific Software Development), a qualitative coding software program. We used qualitative content analysis methods to code the data [25,26]. The coding group was composed of a multidisciplinary team, including MSK, a social science researcher; SM, a board-certified internist; ERC, a board-certified internist and gastroenterology fellow; and JLK, a research intern. The multidisciplinary composition of the team allowed for rich and varied discussions about the codes and the organization of the codes into larger thematic categories. The first level of analysis, open coding, identified individual concepts present in the social media posts. The unit of analysis was the entire post. We coded individual user’s actions and focused on what users were doing (eg, sharing personal experiences with medications, expressing fear about medications, not taking medications due to beliefs about harm). We subsequently created a codebook from significant or frequently occurring codes in the data. After creating the codebook, we reviewed the data and coded using our focused codes. Posts were assigned a code if they mentioned experiences regarding pregnancy, breastfeeding, infertility, and experiences with IBD, IBD medications, or IBD-related complications. Each major theme was constructed by aggregating several focused codes identified through the open coding process. For example, focused codes for the major theme “Fear, Anxiety, and Uncertainty” included the following: fear about transferring IBD through genetics, fear of birth defects, fear of delivery issues with IBD, fear of fertility issues with IBD, fear of risks from medications during breastfeeding, fear of risks from medications during pregnancy, fear that pregnancy would worsen IBD symptoms, and general fears and anxieties.

The data collected contained no personal identifiers, and posts included only the website where the comments were posted. Although we were not able to ascertain the gender of the user, given that most social media posts referred to personal experiences with pregnancy, we assumed that most posts were written by women. However, we did find a few exceptions where partners posted on behalf of their wives or where men posted about male fertility issues related to IBD. The social media data were considered part of the public domain. The study was reviewed and approved by the Cedars-Sinai Medical Center IRB. No individual subjects were contacted.

We identified 1818 unique posts discussing the use of medication in IBD related to fertility, pregnancy, birth outcomes, or breastfeeding. We identified 7 major themes through the open coding process, presented in Table 1. Individuals used social media to (1) seek and share information and advice about medication use during pregnancy, infertility, and breastfeeding, 13.92% (252/1818); (2) express beliefs about the safety of therapies during pregnancy and breastfeeding, 7.43% (135/1818); (3) discuss decisions around personal experiences with medication use, including changes in medications or decisions to stop using medications during pregnancy and breastfeeding, 16.72% (304/1818); (4) articulate fears and anxieties about the safety and use of therapies during pregnancy and breastfeeding, 11.55% (210/1818); (5) discuss physician-patient relationships and related issues of trust and distrust, 3.14% (57/1818) and recommendations from providers about medication use during pregnancy and breastfeeding 3.74% (68/1818); (6) share issues around infertility, including concerns about male infertility and IBD medications, 16.97% (309/1818); and (7) discuss issues around pregnancy and breastfeeding health, including IBD symptoms and medication side effects during pregnancy and post-pregnancy, 11.33% (206/1818).

Our analysis revealed that many factors influence the development of beliefs about IBD medications, including anxieties and fears about taking medications during pregnancy and breastfeeding, the individuals’ trust in their health care providers, beliefs about taking medications associated with infertility, and information gathered from online sources and online communities. Beliefs about IBD medications, in turn, are associated with decisions to start, stop, or change IBD medications during pregnancy, breastfeeding, and infertility treatments.

We identified 252 posts (13.92%) where individuals sought and shared advice with others on taking various types of medications, treating flare-ups, and addressing labor complications during pregnancy and breastfeeding. Many individuals expressed distress about recommendations from physicians or about information they had gathered online, particularly with regards to medication changes. For example, a woman 20 weeks into her pregnancy wrote about continuing Remicade infusions (a biologic treatment for IBD) until the 32-week mark and then discontinuing infusions and starting prednisone. She wrote:

The woman in the post above expressed uncertainty in both the medical and the scientific community and the clinicians’ recommendations. Hesitation in trusting clinical findings and the medical community as a whole recurred throughout the posts in our analysis.

Individuals also expressed anguish about the lack of definitive safety information available for newer medications, particularly biologics, and sought advice on these medications. For instance, 1 individual seeking advice about his/her partner’s treatment wrote:

The individual noted that his or her partner had received conflicting information from various providers on medications and delivering via Caesarian section and shared the frustrations with the online community.

Individuals also expressed their implicit and explicit beliefs about taking IBD medications during pregnancy, breastfeeding, and infertility treatments. Despite many concerns shared about the safety of IBD mediations, there were twice as many posts expressing beliefs that IBD medications are safe during pregnancy and breastfeeding (n=68, 3.74%) as compared with beliefs that IBD medications are not safe (n=35, 1.93%). When sharing their experiences, many individuals explicitly noted that their medications were “very safe” and mentioned whether their providers agreed with their beliefs. One individual emphasized that she had consulted with 3 different clinicians when considering the safety of the IBD medications she was taking while pregnant, sharing both her personal opinion and viewpoints of clinicians in her life:

Other social media users expressed doubts or firm beliefs that IBD medications were unsafe during pregnancy, echoing other qualitative work that has found that many women are uncertain about taking any medications during pregnancy.

Some individuals expressed beliefs that conflicted with recommendations they had received from their providers. Many of these patients expressed distrust in the evidence supporting the safety of IBD medications. For example, 1 patient posted a comment about a bad experience she had heard of regarding an individual using a biologic medication during pregnancy:

We also identified 32 (1.76%) posts where individuals weighed the benefits and risks of taking IBD medication and concluded that the benefits outweighed the risks of a severe flare-up during pregnancy. Avoiding flare-ups by taking IBD medications was a common narrative. One 29-year-old patient with CD discussed taking azathioprine for 8 years and staying on it throughout her pregnancy. She wrote:

She notes the various sources of information she consulted throughout her pregnancy, including information from social networks and formal sources (ie, professionals), highlighting the fact that many individuals seek multiple sources of information about medical decisions and must weigh the validity of these sources when making medication-related decisions.

Many individuals reported changing or stopping medications during pregnancy and breastfeeding, either as a result of the advice of a clinician or based on personal beliefs about the safety of taking IBD medications. There were 198 posts (10.89%) where individuals specifically cited taking or planning to take IBD medications during pregnancy or breastfeeding, 93 (5.12%) posts where individuals said they stopped or planned to stop medications during pregnancy or breastfeeding, and 13 (0.72%) posts where individuals said they planned on changing or changed medications during these reproductive periods.

Individuals cited reasons such as beliefs about the safety of medications, advice from providers, or advice from fellow sufferers for their decisions. As an example, one social media user discussed her decision to avoid breastfeeding while taking Remicade, a biologic therapy for IBD:

This individual expressed that although she understood that other people had made the decision to breastfeed while taking these medications, she preferred to take the personal risk of going off the medication to avoid any perceived risk to her baby. This example highlights a tension that many individuals expressed between caring for their own bodies and avoiding perceived harm for the baby. The desire to avoid medications during pregnancy extended to individuals who had already given birth and taken medications while pregnant but wanted to avoid all medications for the next pregnancy. In one post, a woman expressed how she took 2 different IBD medications—Azulfidine and Asacol—during her first pregnancy, which she mentions was unplanned. However, for future planned pregnancies, she describes that she would:

The individual posting here emphasizes her desire to use nonmedication alternatives (supplements), perceived as safer, to stay “healthy” and thus potentially be able to avoid taking medications she perceives as harmful.

Among those who cited either a personal reason or provider advice to stop taking a medication, nearly 3 times as many patients who stopped taking a medication did so as a result of personal reasons (N=25, 1.38%) versus those who stated they stopped based on the advice of a provider (N=8, 0.44%). One post exemplifies this decision-making process:

Patients frequently expressed fear, anxiety, and uncertainty around the use of IBD medications during periods of infertility, pregnancy, and breastfeeding. A major concern was the fear of birth defects and higher risk of miscarriage due to taking any medications during pregnancy. Although many patients had received medical advice, they frequently looked to others online to validate their concerns and gather more information. For instance, an individual with UC who was 7 months pregnant wrote:

Patients were conflicted between their fears of potential harms to the baby and the benefits of taking medications during pregnancy and breastfeeding, despite reassurances from clinicians. In many cases, providers appeared to express uncertainty with regards to the safety of the medication, leading the patient to also feel a heightened degree of uncertainty. One individual wrote:

Patients discussed beliefs that their physicians expressed about the safety and appropriate use of IBD medications during pregnancy and breastfeeding. There were 35 reports (1.93%) of providers stating that IBD medications were safe during pregnancy/breastfeeding, 23 (1.27%) where patients said the provider indicated that IBD medications were unsafe during these periods, and 10 (0.55%) where patients said their provider was unsure about the safety of IBD medications during pregnancy and breastfeeding.

In this example, an individual discusses her experiences with her maternal-fetal medicine specialist, the beliefs her provider has expressed, and the relationship she has with this provider:

In the post above, the individual expressed how the maternal-fetal medicine doctor was the one who was able to bridge the information divide between the 2 other specialists.

This experience echoes that of many individuals who posted on social media, many of whom expressed frustration that their obstetricians were not well informed on how to counsel women with IBD and how their gastroenterologists did not provide sufficient counseling during pregnancy. Many patients wrote that they were frustrated with their providers’ unwillingness to discuss pregnancy and fertility issues. One patient expressed anguish regarding her conversations with physicians about IBD medications and pregnancy:

We found that even when patients discussed provider recommendations, there were elements of distrust or uncertainty regarding the physicians’ advice. One individual noted:

We identified a group of posts focused on the effects of IBD and IBD medications on infertility (n=316, 17.38%). In these posts, social media users also discussed the effect of IBD or IBD medications on infertility in general or whether these affected the effectiveness of in vitro fertilization. Although most posts were related to female infertility and IBD, we identified 19 posts (1.05%) discussing male infertility or risks of birth defects due to medications taken by male IBD patients. Several patients reported confusion or surprise that medications used to manage IBD could affect male fertility. One patient with CD taking Cimzia and prednisone with a persistent flare-up sought advice on taking methotrexate while he and his wife attempted to conceive:

As with other posts, the individual decided to turn to social media to seek information rather than turn to a medical professional.

Patients used social media to discuss IBD symptoms and potential causes during pregnancy. We found 202 posts (11.11%) that discussed flare-ups during and after pregnancy/during breastfeeding. Patients discussed how changes during pregnancy and breastfeeding affected their IBD symptoms; 65 posts (3.57%) discussed an IBD flare-up during pregnancy, 65 posts (3.57%) discussed complete remission during pregnancy, and 64 posts reported postpartum flare-ups. We also found 37 posts (2.04%) where patients reported that the onset of IBD coincided with their pregnancy, underscoring the importance of patient education on IBD and reproductive issues during the reproductive years.

We found that social media users often expressed significant tension between taking IBD medications for their own health and fears about potential birth defects. Our findings are in line with another research that finds that individuals often overestimate the teratogenic effects of medications during pregnancy [10]. A study using traditional survey methods also found that one quarter of participants thought it was more important to tolerate flares than take medications during pregnancy and 36% expressed beliefs that IBD medications are harmful to the fetus [27]. Although we were not able to measure education in our study, other studies have found that education level and health literacy are inversely associated with beliefs about medication safety during pregnancy, with less educated women/women with lower health literacy stating beliefs that medications are harmful during pregnancy [10,28].

Additionally, other studies have also found that adherence to medications for chronic conditions is low during pregnancy and is strongly related to women’s beliefs about medication safety [29]. Our findings demonstrate that social media users indeed express that they stop taking medications for IBD during pregnancy when they are feeling better to reduce perceived risks.

Discussions with gastroenterologists about reproductive health in IBD are vital to bridge information gaps. Women with IBD are more likely to seek information about contraception and reproductive health from their gastroenterologist than from a family care physician, although most of this counseling is initiated by patients [29]. Yet, most patients seeing a gastroenterologist have no documented reproductive counseling, and 41% of general practitioners report that they never raise the issue of family planning with their IBD patients [12]. A timely discussion with a provider can alleviate some of these anxieties and help women make better decisions about managing their IBD during pregnancy. Discussion of family planning with a physician is associated with higher IBD-related pregnancy knowledge scores and lower odds of voluntary childlessness [8]. Gastroenterologists can play an important role in transmitting information about reproductive health and IBD during office visits to increase adherence to medications and reduce flare-ups.

One potential issue raised by social media users in this study may be the dearth of evidence-based information about medication safety during pregnancy, particularly for newer medications such as biologics. A 2010 study of 305 obstetricians and gynecologists found that 42% selected lack of sufficient information on medications as a significant barrier to discussing medication safety with pregnant women [30]. Increasing pregnancy registry participation could improve the quality and quantity of information available to clinicians and patients.

Strengths of this research include the ability to analyze the varied perspectives of hundreds of patients from all over the globe, which would be logistically unfeasible with traditional focus groups or interviews. However, there are also several limitations to this analysis. First, we cannot verify the identities of the patients posting on social media, particularly because many post using screen names. As a result, we cannot definitely confirm whether all patients have clinically verified diagnoses or are self-diagnosed. Another limitation is that not all patients are equally likely to write about their medical experiences online. As a result, we may only have the opportunity to view and analyze the experiences of a certain segment of the population.

As other researchers have found, social media use is associated with female gender, higher income, higher level of education, and younger age [13,24]. Moreover, these same factors are associated with characteristics of individuals who seek health information online [31]. Health status may be associated with online information-seeking, but the findings are conflicting—one study identified that healthier individuals are more likely to seek information online, whereas another study found that health status was not associated with seeking online information [31,32]. Thus, it is possible that the social media posts included in this analysis reflect a younger, more educated, and higher-income cohort, and thus, the results may be more applicable to individuals with similar backgrounds. As with traditional focus groups, another limitation is that certain individuals may be more likely to engage and participate in the conversations [33]. However, we examined more than 1800 posts from a variety of different websites, which should allow for discussions and posts from a variety of different communities and individuals, which strengthens the findings from our analyses.

In this study, we used social listening methods to survey how individuals with IBD think about and make decisions about medications during key reproductive periods. We aimed to understand the types of beliefs that individuals have about IBD medications and birth defects, fertility, or outcomes in children. One important finding of our study is that individuals’ needs may not be adequately addressed in clinical practice and that many patients are frustrated with their providers for failing to discuss the effects of their pharmacological regimens on fertility, pregnancy, and breastfeeding. We found that both men and women have concerns about the safety and effects of IBD therapies during reproductive periods, and that there are significant fears and anxieties about the use of these therapies that are largely not addressed during office visits. As result, individuals may turn to online groups for information and emotional support. We also found that individuals often questioned the recommendations of their clinicians and sought second opinions via online forums. However, as these online forums may not provide evidence-based or updated medical information, individuals might receive erroneous or dangerous information. Future research should examine why clinicians might be hesitant to bring up reproductive concerns during office visits for individuals with IBD. Future research might also elucidate successful strategies that individuals can use to raise these topics that foster effective shared decision making in the IBD and reproductive health settings. Improving patient-provider communication around this issue may help improve the health of individuals with IBD during important reproductive periods.