“Er, um. A week or two ago? I think…” you answer.
Patients often have an estimate, but rarely can they pinpoint an exact date. “For a lot of patient care, that’s fine. But for dating a pregnancy, it actually matters a lot, because that is one of the guidelines we can use—either the last period or the first ultrasound. More and more recently, when I ask that question, women pull out their phone, scroll back, and can show me the exact date,” explains Dr. Nathaniel DeNicola (@NDeNicolaMD), an ob/gyn doc from Penn.
In addition to being a practicing ob/gyn, Dr. DeNicola is also a fellow of the Penn Social Media & Health Innovation Lab (@PennSoMeLab), part of the Penn Center for Innovation (@PennPCI), a group that aims to conduct research at the intersection of healthcare, mobile technology, and social media as well as disseminate information to the healthcare community and general public.
Because social media and digital health are emerging fields, the Social Media Lab is charting new territory in this space. Dr. DeNicola explains that while it is important to do no harm as a physician, it is also necessary to have, “…the flexibility to try new things. In medicine, if we are always relying on the precautionary principle, as we do with patient care—which is a good thing—we never advance forward at the speed we need to, to really stay ahead.”
The potential of patient-generated data sets, from iPhone apps, Fitbit, and other smart watches is currently a largely untapped collection of clinical datasets. What would clinical practice look like if your primary care physician could see changes in weight over the course of the year? Spikes in blood pressure? Alterations of blood glucose levels? How much you walk or exercise? Or, even more powerful, what if researchers could use these large datasets to find relationships between simple health stats and long-term health outcomes?
However, understanding how these trackers can best help patients is still an area that requires more research—it’s a “coming attraction”. It is likely that the burden of all this data would be too much for a single clinician to manage for each patient. “There is an explosion of data that could be available and will, I think, be available [in the near future]. We want more data. But you need someone who could manage it and answer questions that come up, like, what do you do with an alarming value? Who’s responsible for that? How do you verify that these data are real? How accurate are they? There are a lot of things to figure out.”
We want more data. But you need someone who could manage it and answer questions that come up, like, what do you do with an alarming value?"
Although all patients stand to benefit from the expanded use of patient-generated big data in clinical practice, young women may be one of the first groups benefiting from these advances. “Young, reproductive women specifically, are using apps more than other demographic groups. Some version of app use: tracking pregnancy, tracking your periods, becoming pregnant—it’s becoming standard.”
Although step counters generate by far the most data, some studies put fertility and pregnancy apps above fitness trackers when looking at total number of users. With increasing numbers of women utilizing these apps, Dr. DeNicola says that he is excited to see where the field will go. “[I am interested in the] return of the rhythm method. (Also known as fertility awareness methods.) It’s the method of contraception that uses no birth control at all—it only uses cycle tracking, measuring basal body temperature, and cervical mucus thickness—to either avoid pregnancy or, on the flip side, try to optimize time for conception. This is really an emerging area because healthcare apps are exploding, including AppleKit coming out in the fall, which will include reproductive health.”
A small minority of women rely on the rhythm method exclusively. However, “…it has generated buzz, because it’s perfectly synched with using an app to track these things. It’s hard to imagine a condition more perfect to tracking on an app. I can see why manufacturers targeted this. However, they often overstate the scientific claims.” The best methods for birth control are surgical and implants, such as Mirena and Nexplanon. “Fertility awareness is nowhere near that—we know it has about a 25% failure rate because anything, a fever or infection, can throw that off.”
Despite the inaccuracies associated with using fertility awareness as a form of birth control, for women trying to become pregnant, these apps can provide additional support and information during this process. And careful tracking of periods might allow for increased accuracy when dating pregnancies. “There is a good reason to study this. Is tracking periods with an app more accurate? Can we have higher accuracy in dating pregnancies?”
The importance of accurately dating pregnancies really comes into play when managing a high-risk pregnancy. “As we get further along, those days really matter [and influence] whether we recommend delivery at a pre-term gestational age. A lot changes between 28 and 32 weeks. A lot changes again between 32 weeks and 36 weeks. So if your dating is off by up to two weeks—which we know that it can be—you can be looking at a pretty big difference there. That’s maybe a baby who needs antenatal steroids for lung maturity or to protect the brain versus who doesn’t. That’s maybe a baby who needs to get certain medications, such as magnesium sulfate, to protect the brain from cerebral palsy or who doesn’t. So, it’s not a small thing. It’s important to be able to nail down the dating criteria. And this may be an area where mobile devices and patient-generated data can help.”
Young, reproductive women specifically, are using apps more than other demographic groups. Some version of app use: tracking pregnancy, tracking your periods, becoming pregnant—it’s becoming standard.”
Dr. DeNicola explained that he is currently involved in a study where patients, “…give us [limited and defined] access to their Facebook posts and Twitter content—to read what material they have out there—and use the data. There are a number of ways this study will go, but one of the things we are looking at for the obstetrics population is if, amongst the Facebook and other social media data, there are signals for [post-partum] depression. Number one, it is a serious condition, we know it can lead to both suicide and infanticide. And we know that these catastrophic consequences are more common in younger women—who are also using social media platforms more. And, we know it is under-diagnosed.”
Although post-partum women are screened for depression in the hospital after delivery, many, many at-risk patients are missed because the full form of depression does not develop until weeks after delivery. Docs will also perform screening at the 6-week follow-up visit, but this appointment can be difficult for some patients to make. Medicaid patients lose coverage after six weeks, for example, so if there are even minor scheduling issues with this visit, the follow-up may be missed.
“Having some type of interface with this vulnerable population is very important,” explains Dr. DeNicola, and social media is one way these outcomes can be predicted.
At first, the Social Media Lab was unsure patients would be willing to give Penn researchers access to this personal information. “The response has been tremendous. It’s so exciting that we are actually getting a lot of enrollment.”
So—if all goes well, how could health care providers incorporate this type of social media screening into clinical practice? “We first have to show verifiable results and that for some patients this can provide a signal. That has been demonstrated in other studies, but for the post-partum group we are helping to define this. [These techniques] may not apply to all groups [of patients].”
However, these predictive techniques take time, effort, manpower, and math skills. “I could see a new clinical position developing, that is devoted to generating and monitoring social media content [with patient consent], and bundled within that, is patient-generated data. And we’ve seen [these new positions] develop in medicine in the past. There was a point when you didn’t need to have a billing department—it was fee for service. Now there are seminars on how to do coding for billing! There was a time when you didn’t have electronic medical records, didn’t have an IT person in every office. Now almost every clinic has an EMR or IT person. Even how we have redistributed who sees patients [has changed]. The transition from a doctor-only model to doctors supervising and working with physician’s assistants.”
Could statisticians and data management experts be the newest additions to the medical field? Maybe hospitals and clinics will be hiring Clinical Data Scientists in 2025? Who knows? However, it is clear that the field is moving in a data-driven direction, particularly for women’s health, hopefully resulting in improved quality of care for patients and active engagement in lifelong health.
To learn more, check out the American College of Obstetricians and Gynecologists (ACOG), one of the first professional medical societies to issue social media guidelines, as well as information about reproductive health updates to Apple Kit, due out in November.
My Favorite Health Apps
1. Pedometer Tried and true, this is as easy as it gets. No extra devices required except for your phone. This helps me get a gauge on my activity levels throughout the week.
2. Sleep Cycle This helps me figure out how much sleep I am actually getting and zero in on causes of poor sleep quality. Before I used the app I thought I was sleeping 8 or more hours a night. I’m actually getting more like 6.5-7 when you account for waking up and reading or TV before bed. It also wakes you up when you are in a lighter phase of the sleep cycle.
3. Heart Rate This app records your heart rate using the iPhone camera and will keep a record of your heart rate readings.
4. Calm Although this is not necessarily a health tracker in the conventional sense, this app provides peaceful nature noises and guides you through mindful meditation sequences.
5. Fitbit Though using this app also requires you to get a Fitbit, it is really great. I find that the readings are more accurate for sleeping and walking than using the iPhone only and it tracks heart rate throughout the day. So. Much. Data.