This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Urological mobile medical (mHealth) apps are gaining popularity with both clinicians and patients. mHealth is a rapidly evolving and heterogeneous field, with some urology apps being downloaded over 10,000 times and others not at all. The factors that contribute to medical app downloads have yet to be identified, including the hypothetical influence of expert involvement in app development.
The objective of our study was to identify predictors of the number of urology app downloads.
We reviewed urology apps available in the Google Play Store and collected publicly available data. Multivariate ordinal logistic regression evaluated the effect of publicly available app variables on the number of apps being downloaded.
Of 129 urology apps eligible for study, only 2 (1.6%) had >10,000 downloads, with half having ≤100 downloads and 4 (3.1%) having none at all. Apps developed with expert urologist involvement (
The explicit participation of urologists in app development is likely to enhance its chances to have a higher number of downloads. This finding should help in the design of better apps and further promote urologist involvement in mHealth. Official certification processes are required to ensure app quality and user safety.
Medicine is constantly evolving, and medical research and development are greatly influenced by available and new technology. Mobile health (mHealth), defined as “the delivery of healthcare services via mobile communication devices” [
mHealth has had an impact in several medical specialties, including anesthesia [
The economic literature indicates several factors that affect app downloads, with price being one of the significant predictors [
The exchange of opinions and experiences online, that is, online word-of-mouth, influences ecommerce sales [
Previous studies have shown that app demand decreases with the app file size. As apps become more complex they increase in size, meaning that they take longer to download and for users to try them. Moreover, they occupy additional space in the device memory [
The developer’s textual and visual description of an app can undoubtedly contribute to the willingness of users to download an app. Prior studies have shown that textual information and visual images affect consumer purchase decisions [
Other factors that may positively influence the number of downloads are the app’s age (ie, how long the app has been available) and availability of updates (ie, whether the app has been updated since launch) because these are surrogates of the app’s evolution [
The delivery of mHealth in urology will, as in all medical fields, largely depend on app availability, benefits, and user friendliness. Although economic studies have identified some of the factors that influence app downloads [
We conducted a commercial review of all urology apps for the Android mobile operating system in the Google Play Store (Google Inc, Mountain View, CA, USA) up to August 31, 2015: we examined all apps containing the term “urology” in their metadata (ie, the title, description, keywords, or version history). We included only urology-specific apps in this study; hence, we excluded apps containing content related to other medical specialties (ie, generic apps targeting multiple subjects; eg, an anatomy atlas), product advertisements (ie, apps only promoting pharmaceuticals or clinical equipment), and apps solely allowing the user to schedule private appointments.
We selected only Android apps for study because, in contrast to Google, Apple does not report the number of individual app downloads. Furthermore, Apple only lists the top 200 medical apps ranked by a nondisclosed proprietary algorithm. However, no urology apps were present in the top 200 medical apps listed in either app store.
For each app, 2 reviewers (NP-A and MR) recorded all available information according to 12 predetermined variables: (1) number of downloads, the dependent variable, (2) number of written user reviews, (3) price in euros, (4) average user rating (number of stars from 1 to 5), (5) app size (in megabytes), (6) number of screenshots (ie, an actual app image that showcased its features and functionality), (7) length of app description (number of characters in the app description, not including spaces), (8) app availability in the Apple App Store (ie, whether the app was available for iOS mobile phones or tablets), (9) new versions available (ie, whether the app had been updated since launch), (10) app age (number of days available in the Google Play Store), (11) absence of age restriction (ie, defined by the developer as having content appropriate for all ages), and (12) availability of in-app purchases (ie, the opportunity to buy extra content).
Variables included in the model to predict the number of downloads of urology apps.
Variables | Description |
Level of downloadsa | Level 0: no downloads |
Level 1: 1–5 downloads | |
Level 2: 6–10 downloads | |
Level 3: 11–50 downloads | |
Level 4: 51–100 downloads | |
Level 5: 101–500 downloads | |
Level 6: 501–1000 downloads | |
Level 7: 1001–5000 downloads | |
Level 8: 5001–10,000 downloads | |
Level 9: 10,001–50,000 downloads | |
No HCPb participation | 0: Other |
1: No HCPs mentioned | |
Other HCP participation | 0: Other |
1: Other HCPs, pharmacists, and nurses | |
Urologist participation | 0: Other |
1: Urologist or urological association participation | |
Number of reviews | Number of reviews in the Google Play Store |
Actual price | Actual price of the app in euros |
Average user rating | User evaluation on a scale from 1 to 5 stars |
App size | App file size in megabytes |
No age restriction | 0: Age restriction |
1: No age restriction (ie, appropriate for all ages) | |
Number of screenshots | Number of screenshots in the Google Play Store |
Length of description | Number of characters (without spaces) in the textual app description in the Google Play Store |
Availability in the Apple App Storec | 0: Not available |
1: Available | |
Version | 0: One version |
1: New version exists | |
App age | Number of days available on the market |
In-app purchases | 0: No in-app purchase |
1: In-app purchase available |
aThe exact number of downloads is not available from the Google Play Store. We categorized it according to the system used by Google in the Play Store.
bHCP: health care professional.
cAvailable for iOS mobile phones or tablets.
To test the hypothesis that urologist involvement influences app downloads, we added a further variable to our model: HCP participation. We identified HCP participation by examining the app’s description and considered it to be present only when explicitly mentioned. We classified the participating HCP as urologist (ie, urologist or urological association), other HCPs (ie, other medical doctors, pharmacists, or nurses), or no HCP (ie, no explicit mention of an HCP). The 2 reviewers gathered download data based on the classification system of level of downloads used by Google in the Play Store (
Analyses were performed using IBM SPSS Statistics v20 (IBM Corp). We considered
A total of 250 Google Play apps contained the term urology in their metadata. We excluded 121 apps: 109 were generic apps (ie, not designed specifically for urology, eg, ArchieMD 3D Health: PREVIEW), 11 were for making appointments (eg, Dr Fateh Singh Appointments), and 1 app was designed solely for product advertisement (Actient Pharmaceuticals).
Of the 129 included apps (
Summary descriptive statistics for continuous variables for apps containing the term urology.
Mean | SD | Range | Median | ||
Number of reviews | 0.84 | 2.08 | 0–12 | 0 | |
All apps | 2.55 | 9.89 | 0–83.15 | 0 | |
Paid apps | 8.45 | 16.68 | 0.68–83.15 | 2.69 | |
Average user rating (no. of stars) | 2.65 | 2.13 | 0–5 | 3.5 | |
App size (MB) | 7.37 | 10.36 | 0.01–48 | 3.2 | |
Number of screenshots | 5.4 | 3.86 | 1–25 | 4 | |
Length of description (nonspace characters) | 896.08 | 872.24 | 3–3348 | 531 | |
App age (days) | 721.18 | 425.76 | 1–1733 | 699 |
Even though 2 (1.6%) apps had >10,000 downloads (level 9), half of all urology apps had ≤100 downloads (level 4 or less). At the time of this review, 4 apps (3.1%) had not been downloaded (
Frequencies for the categorical and binary variables.
Frequency | Percentage | Cumulative percentage | ||
0: no downloads | 4 | 3.1 | 3.1 | |
1: 1–5 downloads | 6 | 4.7 | 7.8 | |
2: 6–10 downloads | 3 | 2.3 | 10.1 | |
3: 11–50 downloads | 35 | 27.1 | 37.2 | |
4: 51–100 downloads | 16 | 12.4 | 49.6 | |
5: 101–500 downloads | 33 | 25.6 | 75.2 | |
6: 501–1000 downloads | 10 | 7.8 | 82.9 | |
7: 1001–5000 downloads | 18 | 14.0 | 96.9 | |
8: 5001–10,000 downloads | 2 | 1.6 | 98.4 | |
9: 10,001–50,000 downloads | 2 | 1.6 | 100 | |
Other | 104 | 80.6 | 80.6 | |
No HCPs mentioned | 25 | 19.4 | 100 | |
Other | 111 | 86.0 | 86.0 | |
Other HCPs, pharmacists, and nurses | 18 | 14.0 | 100 | |
Other | 43 | 33.3 | 33.3 | |
Urologist or urological association participation | 86 | 66.7 | 100 | |
Age restriction | 72 | 55.8 | 55.8 | |
No age restriction | 57 | 44.2 | 100 | |
Not available | 36 | 27.9 | 27.9 | |
Available | 93 | 72.1 | 100 | |
One version | 66 | 51.2 | 51.2 | |
New version exists | 63 | 48.8 | 100 | |
No in-app purchase | 118 | 91.5 | 91.5 | |
In-app purchase available | 11 | 8.5 | 100 |
aHCP: health care professional.
Although most apps, that is, 86 of 129 (66.7%), were developed with specialist urological input and other HCPs were involved in a further 18 apps (14.0%), 25 apps (19.4%) had no documented HCP involvement. A total of 57 apps (44.2%) had no age restriction. Only 11 apps (8.5%) had in-app purchases available.
Multivariate logistic regression revealed the factors contributing to urology app downloads (
Multivariate ordinal logistic regression of factors contributing to number of urology app downloads.a,b
Variables | Estimatesc | SE | 95% CI | |
App age | 0.001 | .0004 | .24 | –0.0003 to 0.001 |
Other HCP participation | 0.469 | .602 | .44 | –0.71 to 1.65 |
Urologist participation | 1.43 | .479 | .003 | 0.49 to 2.37 |
Number of reviews | 0.440 | .102 | <.001 | 0.24 to 0.64 |
Actual price in euros | –0.071 | .020 | <.001 | –0.11 to –0.03 |
Average user rating | 0.337 | .089 | <.001 | 0.16 to 0.51 |
App size | 0.018 | .017 | .30 | –0.02 to 0.05 |
No age restriction | 0.498 | .346 | .15 | –0.18 to 1.18 |
Number of screenshots | 0.047 | .048 | .33 | –0.05 to 0.14 |
Length of description | –0.0004 | .0002 | .09 | –0.001 to 0.00006349 |
Availability in the Apple App Store | –0.641 | .441 | .15 | –1.5 to 0.22 |
Version | –0.372 | .340 | .27 | –1.04 to 0.29 |
In-app purchases | 1.67 | .682 | .01 | 0.33 to 3.0 |
Nagelkerke |
.48 |
aThe dependent variable is the level of downloads.
bThe reference level for health care professional (HCP) participation is “No HCP participation.”
cEstimates are the ordered log-odds regression coefficients and they show the relative magnitude (ie, relative impact of the factor) and direction (ie, positive or negative) of the impact of the listed variables on the level of downloads.
Number of urology apps per level of downloads and health care professional participation.
The lack of studies on the predictors of the number of downloads for medical apps in the PubMed database suggests that this is the first study of its kind in mHealth. However, economic studies determined the predictors of downloads for generic apps, which we tested in this study. We showed that inexpensive apps developed with expert urological input and with optional in-app purchases were more likely to be installed. Furthermore, apps with higher user ratings and with a larger number of written user reviews were more likely to have a greater level of downloads. These results confirmed, for the first time, that urologist participation in app development positively influences urology app downloads.
Although the availability of various medical apps has been thoroughly documented, the factors that predict their downloads have, until now, not been studied. Given that mHealth is a rapidly evolving and novel field, these data are useful for practitioners and app developers interested in developing urology apps. Furthermore, the data are important for mHealth policy makers and regulators because no best practice guidelines exist with respect to medical app development.
mHealth is still a relatively new concept, and its full potential has yet to be fully explored. The number of downloads of mHealth apps will depend not only on available technologies, but also on the apps and their safety, effectiveness, and usability. However, concerns have been raised about medical apps, namely their scientific accuracy and user security [
Apps have the potential to be hazardous to uninformed users, either by error, such as miscalculation when using an opioid dose calculator [
Our results confirm our initial hypothesis that the explicit participation of an expert in urology in the app development process increases its chances to be downloaded. Given the lack of external certification of mHealth apps, one possible explanation for this result is that users are reassured to know that a health care specialist collaborated in the app design. Expert involvement could be equivalent to a “quality mark,” guaranteeing that the app is safe and scientifically valid. However, users must be aware that, because there is no official way to authenticate the veracity or the extent of the expert participation, unscrupulous developers could potentially misuse this approach via deceptive advertising or false endorsement. Interestingly, however, our findings also indicated that there is still a deficit of HCP participation in urology app development, with only two-thirds of apps having expert participation. This is consistent with previous reports on expert involvement in app development in other disciplines, perhaps signifying a wider trend across mHealth that needs to be addressed [
Cheaper apps with optional in-app purchases were associated with a greater level of downloads. As with mobile game users, mHealth users seem to prefer to pay less initially but to have the opportunity to buy additional benefits, features, or functionalities via in-app purchases, rather than paying a higher upfront price [
A systematic review has shown that eHealth adoption by HCPs is dependent on multiple factors, namely the involvement of users in the development and implementation phases, ease of use, demonstrated advantages of the system, and adequate training and support [
Even though, in the generic mobile market, factors such as app size, number of screenshots, length of description, app age, availability in other mobile stores, availability of new versions, and absence of age restriction have a significant impact on the number of downloads, we found that it was not the case in urology apps. Further studies are needed to determine whether this trend is specific to urology apps or also happens in other medical fields.
Future research may consider the number of positive or negative reviews as a potential factor to predict app downloads. It should also focus on what types of urological apps and what segments of this specific market (ie, patients, HCPs, or both) have higher downloads. Furthermore, subsequent investigations should compare the number of downloads of urological apps with those in other medical fields in order to gain insights into the state of mHealth.
This study has some limitations. We limited our commercial review of urology apps to the search term “urology.” We included only urology-related apps and collected app data solely from information available in the Google Play Store and developers’ websites. Nevertheless, the Google Play Store and developers’ websites are the main sources of information available to potential new users before downloading the app; therefore, our study mimics the real-life information available to the user before purchase.
Android leads the mobile phone market with over 80% of market share, and there are more apps available in the Google Play Store. This is in part explained by the 2 platforms’ different approval processes: iOS apps have to undergo a thorough review process developed by Apple, but Android apps are immediately published online [
Displayed information about the level of downloads in the Google Play Store can in itself influence downloads: if a user has to choose between 2 similar apps, most of the time they will download the most popular app first. For the sake of clarity, we studied only explicit expert participation, and it is possible that some app developers consulted medical experts during app design but did not mention it; there is, therefore, a risk of misclassification for this variable. However, when medical involvement was reported, there was no objective way to determine the extent of participation. The lack of a standardized format for the disclosure of expert participation and the absence of readily available tools to quantify it requires further study and future recommendations.
To our knowledge, this is the first study to determine predictors of urology app downloads. The explicit participation of urologists in app development is likely to enhance its chances of having a greater number of downloads. Furthermore, in-app purchases, cheaper apps, and those with higher user ratings and number of written reviews are more likely to have more downloads. Until a regulated approval process is implemented by government health authorities, analogous to the one that exists for medical devices, two pragmatic changes to urology mHealth app publishing could promote user safety and assure content quality: first, medical apps should include a full disclosure, similar to that provided in scientific papers; and second, urological societies could be involved with certifying the scientific integrity of mHealth apps by issuing a professional, peer reviewed app quality mark or standard. The efforts of the Health on the Net Foundation to guide users toward trustworthy health information online were justified by the results of 10th HON survey [
List of included apps.
Included apps ordered by the number of installs.
health care professional
None declared.