An app to prevent unwanted pregnancies by tracking a woman’s body temperature has scored a first-of-its-kind marketing approval from the Food and Drug Administration, the agency announced.
The US stamp of approval—which clears the way for similar apps to get the green light—lands as the app’s Swedish maker faces investigations by European authorities into its advertising claims, plus criticism from health experts and reports of dozens of unwanted pregnancies.
The sleek mobile app, called Natural Cycles, boasts 900,000 users worldwide as well as approval from the EU to act as a form of contraceptive. Yet it’s essentially riff on an old-school “natural family planning” method dressed up for the digital age. An $80 annual subscription for the app comes with an oral thermometer and relies on a user’s basal body temperature (BBT) to estimate the time of ovulation (when an egg is released from an ovary and wanders down the fallopian tube for a potential sperm-rendezvous, which happens at approximately day 14 of a textbook, 28-day cycle).
The aging BBT method hinges on the curious quirk that women’s body temperatures tend to rise slightly during and after ovulation. In this instance, “slightly” means between 0.5 to 1°F or about 0.3 to 0.6°C. Those faintly balmy conditions are linked to a surge in progesterone in the latter half of the menstrual cycle, called the luteal phase. A woman’s temperature will stay elevated if she becomes pregnant or, if not, until the end of that menstrual cycle, when her period starts and the unfertilized egg gets flushed.
With the app, a user ideally takes her oral temperature every single day, first thing in the morning—before getting out of bed, moving around, eating, etc., all of which could raise her temperature slightly and skew her basal reading. The user then combines that temperature data along with a log of her periods in the app. Over several menstrual cycles, the app uses a proprietary algorithm to gradually refine an estimate for the timing of the next ovulation and the window of fertility. The app then warns users which days she’s fertile or not—in other words, when there’s a chance of sperm and egg meeting or not—noting those days in red and green, respectively.
The diligent, data-collecting user is then expected to either abstain from sex on the red fertile days or turn to another contraceptive method, such as a condom. On green, non-fertile days, the app gives the all clear for unprotected sex.
Out of rhythm method
Of course, these are the ideal conditions; the reality is much messier and, all in all, BBT is not a reliable proxy for ovulation. Though the sultry cycle is indeed well documented across women, it’s easily thrown off, and the erratic temperature swings can be difficult to interpret. Everything from thermometer calibration to stress, poor sleep, fevers, and jet lag can alter temperature readings.
In a small clinical study conducted in Italy in 1990s, researchers compared BBT and other natural methods (such as tracking ferning and luteinizing hormone in urine) to pelvic ultrasonography data—a surefire way to pinpoint ovulation. BBT accurately identified the day of ovulation only 30 percent of the time.
Another glaring problem is that picking out ovulation is not exactly key to blocking or achieving pregnancy. Sperm can linger in a woman’s reproductive tract for up to five days, and many fertility experts will suggest that those to conceive aim to have swimmers treading water at the gate rather than aiming for a tightly timed tryst. In other words, sex in the days just before ovulation are also critical. Of course, the app’s span of red days tries to account for this and predict the days prior to ovulation in which a woman is fertile. But if that timeframe is based on a faulty ovulation estimate, it can leave users vulnerable to several days of unexpected fertility.
Last, even if temperature readings and/or the app manage to nail ovulation in one particular month, many women have irregular cycles, making it difficult to use that information to predict future cycles.
The American College of Obstetricians and Gynecologists summed things up simply: “BBT by itself is not a good way to prevent or promote pregnancy.”
Many of the apps early adopters would probably agree. Numerous media reports have highlighted horror stories of unwanted pregnancies after users were seduced by the idea of a hormone and latex-free contraceptive. Swedish media reported earlier this year that a hospital in Stockholm filed a complaint against Natural Cycles with the Swedish Medical Products Agency. The hospital identified 37 women who sought abortions after relying on the app for contraception. Those women were among 668 seeking abortions at the hospital between September 2017 and the end of the year.
At the same time, the UK’s Advertising Standards Authority (ASA) is also investigating the app and some of its advertising that appeared on Facebook. The ASA told that it had received three complaints about the Facebook ads, which described Natural Cycles as highly accurate.
“We would require robust substantiation from any company to support such a claim,” an ASA spokesperson told the newspaper.
Last month, the newspaper also ran a feature written by a woman who had an abortion after being duped into trusting the app. She highlighted the stories of several other women in similar situations. She also noted that she didn’t report her pregnancy to the app maker, mostly out of embarrassment. That means the pregnancy (and potentially many others) wouldn’t be included in any data on the app’s efficacy calculations.
This is an important point because the company behind the app, also called Natural Cycles, argues that the rate of unintended pregnancies in media reports squares with the app’s efficacy rates, which are on par with other birth-control methods. In an initial response to the pregnancies earlier this year, Natural Cycles released a statement saying:
At first sight, the numbers mentioned in the media are not surprising given the popularity of the app and in line with our efficacy rates. As our user base increases, so will the amount of unintended pregnancies coming from Natural Cycles app users, which is an inevitable reality.
Natural Cycles claims that the app has a 93-percent typical-use success rate (or a seven-percent failure rate). That means that, of 100 average users, seven will experience unwanted pregnancies. In comparison, condoms have an 18-percent failure rate, the pill has a nine-percent failure rate, and various intrauterine devices (IUDs) have failure rates of less than one percent, according to the Centers for Disease Control and Prevention(PDF).
Natural Cycles goes further, saying that for “perfect” users, the app has a success rate of 99 percent. Perfect users are those that faithfully log their data and follow the apps recommendations, as opposed to average users who occasionally slip.
The trouble is, unlike the standard efficacy ratings for other birth control methods, the efficacy rates claimed by Natural Cycles based on rigorous clinical trials. Instead, the company based its rates on a prospective study of data harvested from women already using the app. As such, the study is missing typical data, controls, and standards that would be in a clinical trial. The company published the results in December 2017 in the journal .
Briefly, it includes a scattershot of data from 22,785 women who registered for the study. Natural Cycles ran the study from August 2014 to August 2016, but 54 percent of those women dropped out within the first year, and the company doesn’t have any follow-up data to help understand why. The app maker was able to identify 1,273 pregnancies from the overall group of users during the study, but it didn’t have enough data from 402 other women to tell whether they were pregnant or not by the end. Only 2,684 women logged data from more than 18 months.
The company also didn’t have nearly enough data to know when or if users had sex during menstrual cycles, which is rather useful for determining a contraceptive’s efficacy. Users recorded having sex during only 32 percent of all menstrual cycles collectively logged into the app. Yet, among the women who got pregnant, nearly half did not report having sex. More specifically, in 47 percent of the menstrual cycles in which a pregnancy occurred, users did not record if or when they had sex.
Following the study’s publication, obstetrics and gynecology experts were quick to point out its failings. In a May 2018 letter also in , experts noted that Natural Cycles researchers did not sort out women with irregular cycles or by age—which is a standard practice in clinical trials for birth control because these are well-known to influence fertility. For instance, a woman’s ability to get pregnant gradually declines with age, regardless of contraceptive method used. So, by clumping women older than 35 into an analysis, researchers could artificially skew the results, making the app seem more effective than it is at preventing pregnancy. The researchers also dinged the company for the lack of sexy time data.
“[W]e believe these missing data limit an accurate calculation of typical-use efficacy as well,” they concluded.
In a published response, Natural Cycles developers stood by their analysis. They argued that requiring users to input intercourse data and other health data would be intrusive and may leave them dissatisfied with the app. Interestingly, they also noted that “it is very hard to conduct real-life large-size studies such as the one conducted in this study and at the same time expect to have full information about the users (which is what can instead be expected in a clinical trial).”
Most tellingly, perhaps, is a nugget revealed in a interview with one of Natural Cycles’ researchers, Kristina Gemzell Danielsson. In it, she bluntly told the paper that the app is not a good option for women who absolutely want to avoid a pregnancy.
None of this seemed to phase the FDA. In its announcement, Dr. Terri Cornelison, assistant director for the health of women in the FDA’s Center for Devices and Radiological Health, was quoted as saying:
Consumers are increasingly using digital health technologies to inform their everyday health decisions, and this new app can provide an effective method of contraception if it’s used carefully and correctly. But women should know that no form of contraception works perfectly, so an unplanned pregnancy could still result from correct usage of this device.
The FDA also highlighted the fact that the approval “creates a new regulatory classification,” meaning that other birth control apps can achieve FDA approval simply by demonstrating that they work similarly to Natural Cycles.