We conducted a systematic content analysis of articles contained in the highest circulated newspapers in the United Kingdom and United States that dealt with health care apps (2006-2015). A summary of the methodology is presented in Figure 1.

The initial search yielded a total of 714 UK and US newspaper articles reporting on health care apps between 2006 and 2015. After removing duplicate articles, 689 articles were retrieved, of which 220 articles met the inclusion criteria; 151 articles were published in UK newspapers and 69 articles were published in US newspapers. The distribution of the articles across the different newspapers is presented in Figure 2. In the United Kingdom, the Daily Mail and The Guardian were the newspapers that reported on health care apps most frequently; in the United States, health care apps were most frequently reported in the New York Times and the Washington Post.

In the United Kingdom, the number of published articles on health care apps increased notably over the study period, particularly during 2014, whereas in the US newspapers, the articles reached peak incidence in 2013 and 2015. The overall trends for both UK and US newspaper articles were upward across the study period as shown in Figure 3. Based on study title and abstract, a total of 944 research articles that involved health care apps were retrieved from PubMed over the study period. As indicated in Figure 3, the number of the scientific articles increased exponentially over the study period.

Of the 220 identified newspaper articles, 58 articles (26.4%) appeared in the features section and 42 articles (19.1%) appeared in the main news section. Less than 10% of the articles (19 articles) appeared in the health/life sections of the newspapers included in the evaluation. Although less frequently, health care apps also appeared in the business/financial section (13.2%, 29/220) and editorial section (1.8%, 4/220) of the newspapers.

The calculated interrater reliability (Cohen kappa) agreement between the codes from the two coders were all positive and ranged from .421 to .889. Cohen suggested the kappa result be interpreted as follows: values ≤0 as indicating no agreement, .01-.20 as none to slight, .21-.40 as fair, .41-.60 as moderate, .61-.80 as substantial, and .81-1.00 as almost perfect agreement [38]. Accordingly, the current interrater analysis revealed that seven of 19 items (37%) illustrated moderate agreement, seven items (37%) illustrated substantial agreement, and five items (26%) illustrated almost perfect agreement between the two coders, demonstrating objectivity in elemental judgment as illustrated in Table 1.

Within the 220 articles selected, the most frequent areas covered in the newspapers were general information about health care apps (101 articles; 45.9%), sport and fitness apps (63 articles; 28.6%), and disease-specific apps (20 articles; 9.1%). The remaining articles (36 articles; 16.4%) covered a range of topics, such as apps as sources of information, diet/healthy food apps, and apps used to communicate between health care professionals and between patients and health care professionals (Table 2). Most articles (Table 3) were written from a societal perspective (159 articles; 72.3%) followed by an industry point of view (30 articles; 13.6%); the other viewpoints included policy-related (16 articles; 7.3%), scientific (8 articles; 3.6%), legal (4 articles; 1.8%), and economic perspectives (2 articles; 0.9%). There was no significant difference between the key perspective of UK and US articles (χ²₆=6.5, P=.36).

Although most of the articles (124 articles; 56.4%) did not refer to a specific mobile operating system, 58 articles (26.4%) mentioned iOS (Apple) as the operating system; both Apple and Android operating systems were mentioned together in 42 articles (19.1%). There was no significant difference in the reporting of mobile operating systems between the United Kingdom and United States (Mann-Whitney U=12.0, P=.39).

Of the 220 articles, approximately 106 articles (48.2%) related to general health. including smoking cessation, alcohol use, weight loss, fitness. and exercise tracking, whereas 81 (36.8%) were linked to specific diseases. A smaller number of articles (33 articles; 15.0%) mentioned multiple diseases and/or medical uses (eg, first aid apps, thermometer apps, ear examination apps, diabetes apps).

Endocrine diseases (mainly diabetes) were frequently mentioned in relation to the use of health care apps; however, cardiovascular diseases (including blood pressure and heart rate monitoring) were the most commonly mentioned conditions in the relevant newspaper articles (Table 4). There was no difference between UK and US newspapers in terms of the frequency of mentioning specific disease indications (Mann-Whitney U=41.5; P=.08).

The vast majority of the newspaper articles (201 articles; 91.4%) mentioned at least one benefit related to the use of health care apps, whereas only 53 articles (24.1%) reported at least one risk. There was no significant difference in the frequency of benefits or risks reported between the UK and US articles (χ²₁=1.3, P=.44). The three main benefits associated with the use of health care apps related to personalized care were improvement in general health and fitness (in 124 articles; 56.3%), public/patient access to health information (in 32 articles; 14.5%), and improved health outcomes for the public/patients (in 32 articles; 14.5%). Conversely, a breach in confidentiality was the most commonly cited risk (in 25 articles; 11.3%). In addition, approximately 9% of the articles (20 articles) reported that health anxiety could be induced because of the use of health care apps (Table 5). For example, in one of The Guardian articles, it was debated whether fitness tracking apps are “untested and unscientific and they open a door of uncertainty,” and uncertainty may ignite anxiety in people [39].

Approximately half of the articles (101 articles; 45.9%) mentioned potential facilitators to the use of health care apps, whereas less than 10% of articles (16 articles) reported real or potential barriers to their use. The main recorded facilitator was improved technology (87 articles; 39.5%). No difference was found between UK and US newspapers in terms of the frequency of reporting of facilitators or barriers to the use of health care apps (P=.12) as shown in Table 6.

Most of the articles had a positive slant regarding the use of health care apps (146 articles; 66.4%), whereas 26.4% (58 articles) were classified as having a mixed slant. Relatively few articles reported negative views (11 articles; 5.0%) or a neutral view (5 articles; 2.2%). The majority of the articles (184 articles; 83.6%) were judged as having balanced judgments (ie, the authors did not exaggerate or understate the main message). The quality of the information presented in the newspaper articles was classified into three categories: poor, average, and excellent. Articles were classified as poor if judgment was not balanced and/or if only anecdotal evidence was presented; articles were classified as excellent if they presented information based on firm evidence, often also presenting quotations from experts in the field. Overall, 132 articles (60.0%) were scored as having poor information about health care apps, 75 articles (34.1%) were judged to have good quality information, and 13 articles (5.9%) were judged to provide excellent quality information. There was no significant difference between the quality of the reporting process in UK articles and US articles (χ²₂=1.8, P=.29).

Newspapers are read by a high proportion of the population in the United Kingdom and United States on a regular basis. There is also a strong correlation between newspaper publishing and other mass media platform reporting on the same topic or issue [40]. The methodology adopted for this research was based on earlier published research using content analysis of newspapers [41-44]. In this study, there was a substantial level of agreement between coders as the median kappa statistic score among the two researchers for all variables was .624 (range .421-.889). This indicated moderate to almost perfect agreement for all coding variables [35]. This finding is similar to the level of interrater agreement previously reported in published research on content analysis of newspaper articles [45,46]. The validity of this study was enhanced by precoding 10 articles followed by adjustment to the coding framework. There is no specific requirement for a particular sample size in kappa calculations. In this study, a random sample of 10% of the published articles was included.

A total of 220 articles were identified in the highest circulation of UK and US newspapers over the period of 10 years investigated in the study (2006-2015). For the same period, the number of scientific articles published in PubMed was 944, with the number increasing exponentially over time. In September 2013, the FDA issued guidance relating to medical apps that stated that the majority of health care apps pose a low risk for users or patients and indicated that they would only get involved in regulating those mobile apps that transform mobile devices into medical devices [47,48]. This may have caused the decline in the number of articles published about health care apps in 2014 in the United States. For the United Kingdom, establishing the NHS health app library in 2013 led to a more widespread consideration of health apps; this was reflected the relatively high reporting in UK newspapers in that year (Figure 2).

The main theme of articles on health care apps was found to be related to public health and well-being, including lifestyle management. Unsurprisingly, the related matters of sport and fitness were frequently reported themes. Many of the articles were reported from a societal perspective. This was expected because a high proportion of the newspaper articles focused on general public health topics, including fitness, well-being, and good diet. This finding concords with the finding of a recent report which stated that apps that remind, monitor, and track are intended for the general public, as is the case with social media apps [49]. These kinds of innovative apps can play an important role in general health and lifestyle and can help motivate people to make healthy lifestyle choices [3]. In 2014, a study was conducted to identify the number of health and fitness apps in the Apple Store and Google Play. The author found 23,490 and 17,756 health and fitness apps available in the Apple Store and Google Play, respectively. Moreover, the number of health and fitness apps had grown by 62% over the previous year (2014) in comparison with a 33% growth for apps in general [50]. This vast number of available health and fitness apps helps explain why such apps were reported more frequently in the media when compared with disease-specific apps.

Diabetes was the illness for which apps were most frequently reported by the newspapers; other diseases frequently mentioned were cardiovascular disease and those affecting the central nervous system. A total of 422 million people have been diagnosed with diabetes worldwide and the number is increasing in many countries due to an increasing age demographic together with an increased incidence of obesity and low physical activity [51]. Diabetes is a progressive disease that is associated with several complications including cardiovascular, kidney, and eye complications. Control of diabetes requires self-monitoring of blood glucose levels. Advanced mobile technology allows patients to record these blood levels electronically. Some available apps provide several advantages to manual recording including graphing daily blood sugar levels together with calorie intake and exercise undertaken [52]. A recent review of diabetes-related apps recorded a range of functions that these apps can perform, including documentation, data forwarding, information provision, analysis, recipe suggestions, reminder, and advisory functions [53]. Despite the high prevalence of asthma worldwide, reporting on apps for respiratory illness was infrequent. Available apps in this domain are used to record symptoms of asthma, peak flow readings, and number of attacks, and to provide information about asthma triggers.

As anticipated, in this study the benefits of health care apps were more frequently discussed within articles than were risks. Improved personalized health and fitness were the main benefits described and the issue of confidentiality was the most frequent risk discussed. The literature has documented that newspapers in general overreport the benefits and underestimate the risks when publishing information about health intervention research [47,54]. If the benefits of health care apps are overstated, this could raise public expectations about improved health, whereas overstating risk could generate anxiety and may negatively affect the uptake of what could be a helpful intervention.

In this study, the quality of reporting on health care apps was judged to be poor in most of the articles (60%). These articles in general were brief and did not add sufficient supporting evidence or refer to health agencies. A third of the articles were classified as having average/good quality of reporting, whereas only 5.9% were considered excellent. This was consistent with previous research articles that highlighted the general poor quality of information in the print media about health-related matters [55,56]. It has been reported that high quality newspaper articles often flow from a press release, generally from a scientific journal [19]. Therefore, researchers in the field should ensure that such press releases are written alongside their publications in order to help better information reach the public via the mass media.

There are several limitations to this study. This study examined reporting on health care apps in newspapers only and did not cover other mass media sources such as television and radio. However, as mentioned earlier, it has been reported that the content of newspaper articles correlates closely with other media sources [57]. In addition, some highly circulated US newspapers such as the Dallas Morning News, Chicago Tribune, and Chicago Sun-Times were not included in the study because they were not archived by Nexis. Moreover, limited articles were accessible in Nexis for the Los Angeles Times and Newsday as only articles from the previous 6-month period were available.

This study has revealed that health care apps are indeed reported on by UK and US newspapers. The number of articles reporting on health care-related apps increased over time in UK and US newspapers over the 10-year study period. Several benefits were reported relating to app use, especially in relation to promotion of a healthy lifestyle, whereas reporting on risk was less frequent. Improving personalized care was the most frequently mentioned benefit; confidentiality breaches were the most commonly reported risk. Diabetes was the disease most commonly linked with the use of health care apps. In general, the main claims about benefits had a positive slant, but articles generally were balanced in their judgment.