In Search of Accurate Fetal Heart Rate Monitoring Mobile Applications

Introduction

Medical mobile applications (apps) have continued to grow, with over 90,000 health apps in the U.S. Apple iTunes Store. ¹ Seven percent of these health apps relate to women's health and pregnancy, including those marketed to patients and healthcare providers alike. ¹

Regulation of medical apps has not been rigorous. The Federal Communications Commission, Office of the National Coordinator for Health Information Technology, and the Food and Drug Administration (FDA) have all issued statements regarding the need for regulations on mobile health apps. The FDA's statements in 2013 and 2015 explain which apps are considered devices and therefore subject to oversight by the organization. Apps that meet the definition of medical device “use a mobile platform's built-in features such as light, vibrations, camera, or other similar sources to perform medical device functions.” ² These regulations were placed to promote patient safety and provide a governing body for the legal implications posed by medical apps.

Although regulations exist, they do not impose strict quality control on the app software. ³ Moreover, these regulations have been newly issued, leaving many apps available to providers and patients that have escaped evaluation by the FDA. In addition, regardless of evaluation of the app itself, the FDA has issued statements warning against at-home fetal monitoring of any kind as they are “prescription devices” and should only be used by or with the assistance of a healthcare provider. ⁴ Despite these warnings, apps are available and in use promoting at-home fetal surveillance.

The need for patient safety in health apps is important as apps are consistently poor in quality. ⁵ There are inaccuracies of women's health and pregnancy apps, shedding light on the dangers for providers and patients using these apps. For example, only 50% of pregnancy wheel apps for providers and 20% of free menstrual tracking apps for patients were accurate. ^6,7

Women like to monitor their health more than their male counterparts and often use technology to gain more information regarding their pregnancies. ⁸ Studies detail the increased patient satisfaction with mobile monitoring of conditions like gestational hypertension and diabetes. ^9,10 Because of this patient interest in monitoring, we sought to identify and to determine accuracy of at-home fetal heart rate monitoring (FHM) apps to educate and inform patients and providers. The ramifications of inaccurate FHM may result in inappropriate clinical actions. ¹¹

Methods

This study was conducted as a case report, with all applications tested on a single pregnant patient and proper consent obtained. To perform this case report, first a complete list of FHM apps was generated from three sources: (1) using terms from an article detailing fetal heart rate tracing interpretations, ¹² (2) using terms from the 2005 and 2009 American College of Obstetricians and Gynecologists Practice Bulletins on Clinical Management Guidelines, ^13,14 and (3) adding terms that were considered to be missing by the research team. Terms that were considered too broad in scope or irrelevant to the research questions were excluded (Table 1).

Each term was entered into the search field of Apple iTunes Store between June 23 and June 29, 2016. We compiled a list of the resulting apps in the order of search. A master list of unique apps was then created. We considered apps as “unique” if they met the following criteria: (1) not found using a preceding search term, (2) not found in another iOS system (iPhone compared to iPad), and (3) not a different version of the same app (i.e., free compared with priced versions).

Each unique app was reviewed in the app store description for relevance to the research objective. Apps were considered relevant if they advertised the ability to use the app to listen to the fetal heartbeat. Apps with both free and paid versions were included, with the free version downloaded initially for testing. Apps were excluded if they were games, adult heart rate monitors, or other apps that were determined by the research team not to be focused on the research question, that is, pregnancy wheels, fetal kick counters, and pregnancy guides.

Characteristic information from the remaining apps was collected. This included the download date of August 8, 2016, developer or seller, version, device platform (iPhone, iPad, or both), price, and version modifiers (free, priced, pro). Product information was collected, and screen shots were taken of each unique app to ensure accurate analysis based on the information available at the time of data collection. Apps were then downloaded and analyzed. Devices used for analysis were an iPhone 5S and an iPhone 6S. Apple iPhone headphones and pharmacy purchased generic earbud headphones without a microphone were used when needed.

Data on the app contents were collected and included variables of necessary purchases of accessory items for optimal app operation and misinformation such as duplicate software, presence of disclaimer regarding medical advice, and accuracy of medical content. Apps were then tested on one of the authors (M.S.) at 34 weeks of gestation to assess if the app could accurately monitor the fetal heart rate. The author's fetal heart rate and vertex presentation were verified by pre-app testing at a routine prenatal visit to establish a baseline accurate fetal heart rate, as well as establish location of the fetal heart rate for optimal app testing. We excluded apps requiring a purchase because at-home dopplers ranged in price from $15 to 200. We wanted to allow for analysis of apps available to all users regardless of financial resources. In addition, these at-home dopplers may have transformed a mobile platform into a medical device. In that situation, the FDA would need to regulate the app.

Results

A search of 11 terms pertinent to FHM in the Apple iTunes^® Store resulted in 308 iPhone^® apps and 225 iPad^® apps. After excluding apps that did not meet study criteria, 30 apps remained for categorization.

Of the 30 FHM apps, one app required an in-app purchase to use and was removed from further analysis. Seven of the remaining apps required the purchase of an accessory device and were not analyzed. The accessory devices required for purchase were at-home dopplers which ranged from $15 to 200.

The 22 remaining apps were then fully evaluated. Of these apps, all (100%) were unable to detect the fetal heart rate of the author's pregnancy. The baseline fetal heart rate was 137 bpm during the author's routine prenatal care visit before the app testing. All of the 22 apps were unable to generate a fetal heart rate after the instructions for the app were followed and the headphones or iPhone itself was placed at the known location of the author's fetal heart. Displayed by the app was background static without generation of either any heart rate or a sustained heart rate reading for more than 1 s.

Characteristic information on the 22 FHM apps is shown in Table 2. Icon figures were included for recognition and to highlight duplications. To account for interobserver differences, the apps were individually downloaded and evaluated by the authors, with an error rate of 4/270 (1.5%).

In addition to the nonfunctionality of the apps, 5 were found to be duplicates, 13 lacked a disclaimer regarding medical use of the app, and 15 provided false information in the app content.

The false information provided in the apps included incorrect medical statements (i.e., normal heart rate range 120–160 instead of 110–160). Additional false information provided included statements about gender heart rate differences based on fetal sex and incorrect statements regarding foods and pregnancy (i.e., recommending cold foods and warning against hot foods and rinds of papayas causing miscarriage). Other statements were made regarding the dangers of cell phone devices with respect to the fetus. Cell phones emit radiofrequency energy, a form of electromagnetic radiation. ¹⁵ To date, there is no direct relationship between the use of cellular devices and adverse health outcomes. ¹⁵ Therefore, statements in the evaluated apps warning users against cell phone radiation and the harm to the fetus were considered false statements. Statements in the evaluated apps asking users to place the device on airplane mode, stating that cell phones emit electromagnetic radiation, were not considered false statements.

Discussion

This study demonstrates that despite a large number of FHM applications available on the market, none of the reviewed apps was able to detect a fetal heart rate using iPhone. In addition, the majority of the apps were found to carry risks to the patient as discussed further. Evaluation of the fetal heart rate is an important component of the obstetric evaluation, and any patient's self-assessment of this data point is important to obstetric healthcare providers.

The first risk of these apps is providing false fetal heart rate readings to the patient. This could prompt a patient to seek medical care, present to the office for monitoring, or present to triage for a visit on the labor floor. These visits would cause an increase in cost to the healthcare system, as patients would be presenting for extra visits without medical indication to do so. In addition, these false readings would cause increased anxiety on the part of the patient, further compounding the risks posed by these apps. Readings, alternatively, could be falsely reassuring to patients keeping them at home instead of seeking medical attention when necessary. There also exists the possibility that the patient may be prompted to download such an app in the case of a perceived problem (i.e., decreased fetal movement) and should the app yield a falsely reassuring heart rate, this app may prevent the patient from seeking care at an appropriate time.

The second risk of these apps is that of false advertising. The patient is told that this app could allow her the ability to listen to the fetal heart. This study has clearly shown the falsehoods in the advertising of these apps as they are unable to monitor the heart rate in any reliable capacity. One-third of apps do not even include a medical disclaimer.

This gives way to a third risk, that of expense to the patient. The patient may purchase an app for any of the price amounts found on review, costing the patient money that will not give her the monitoring she desired. This may pose further cost should the patient then seek another app to allow for a perceived better, yet possibly more expensive, product or even a duplicate of the same app.

A stratification of risks in apps has been proposed by Lewis and Wyatt. ¹¹ The risks involved in this study include poor quality patient data, poor lifestyle or clinical decision, and inappropriate but reversible clinical action. Based on their proposed risk stratification schema, these FHM apps would be included in Category C, as the apps pose significant risks to patients due to potential for harm if misused. Some apps would also fall into Category B, causing harm if used inappropriately. Lewis proposes that Category B apps should be regulated with a self-certification model or peer review, while Category C apps should undergo formal assessment for use. ¹¹

The results of this study must be interpreted in the context of the study limitations. Other FHM apps may have been missed because only Apple iTunes Store was searched for mobile apps. The present study would be more comprehensive with the inclusion of Android and or Google Play Store apps. However, we limited our study to apps from the Apple iTunes Store for several reasons. First, the store is superior in the number of medical apps compared to Google Play Store. ¹⁶ Second, there is integration between iOS devices; an app can be used on both the iPad and the iPhone. Third, there is a lack of fragmentation for iOS apps; that is, an app can function along a continuum of operating system iterations. ¹⁷ Fourth, the Google Play Store optimizes the search by yielding apps that have “off-metadata” elements like links, ratings, reviews, and social signals, and thus, the results are less specific. ¹⁸ Searching all platforms was beyond the scope of the study, but would have strengthened the current study.

Additional FHM apps may have been missed as technology is fast paced. Availability, content, and special features may have changed during the time period of our study as technologic advances move at a faster pace than research. ¹⁹ In addition, only 22 of the 30 unique apps were fully evaluated given that the study did not obtain additional purchases. It is possible that these apps provide accurate at-home monitoring; however, the study was not able to test these apps fully.

The dangers of at-home monitoring and surveillance are well known, prompting the warnings from the FDA. ⁴ However, these apps are in circulation and in use by patients seeking to increase their own surveillance of their pregnancies. These apps are not only dangerous by design but also their inaccuracies pose significant health risks to the pregnant patient. Healthcare providers should be aware of the inherent risks of these apps and should counsel patients regarding their usage.