Detecting Gamification in Breast Cancer Apps: An Automated Screening Tool

We developed a predictive model to automatically detect gamification elements in breast cancer mobile apps, addressing the lack of scalable methods for analyzing large datasets. Using titles and descriptions from app stores, our model achieved high accuracy (85% ROC) in identifying gamified apps. This tool enables researchers to efficiently screen apps for gamification, supporting further exploration of its role in health outcomes.

What was the context of our study?

Gamification, the application of game design elements in non-game contexts, is increasingly used in mobile health (mHealth) apps to improve engagement and behavior change. However, its prevalence and impact in breast cancer apps remain underexplored due to the manual effort required for analysis. This study aimed to create an automated methodology for detecting gamification in breast cancer apps to facilitate large-scale research.

What did we do?

We collected data on 1,476 breast cancer-related apps from the Google Play and iTunes App Stores using systematic search methods. Gamification specialists created a list of keywords commonly associated with gamification, such as “points,” “rewards,” and “leaderboards.” These keywords were used to train a logistic regression model, with the presence of gamification verified by human reviewers as a gold standard.

What did we find?

The final dataset included 1,176 English-language apps, of which 20.5% contained gamification elements according to specialist review. Our predictive model successfully identified gamification presence with an area under the ROC curve of 0.89 during training and 0.85 during validation. Key variables contributing to the model’s accuracy included terms related to activity tracking, engagement, and entertainment.

Why is this important?

This study introduces the first automated tool for detecting gamification in breast cancer apps, significantly reducing the time and resources needed for analysis. By enabling large-scale research, this tool can help evaluate the effectiveness of gamification in improving health outcomes and inform the design of future health apps. Broader adoption of this methodology could extend its application to other health conditions.

Reference

Giunti G, Giunta DH, Hors-Fraile S, et al. Detecting Gamification in Breast Cancer Apps: An Automatic Methodology for Screening Purposes. 30th IEEE International Symposium on Computer-Based Medical Systems.* 2017; Thessaloniki, Greece. https://ieeexplore.ieee.org/document/8104273.

Share this post