Monitoring blood glucose with your iPhone

Since the iTunes app store opened almost 3 years ago, over 9.8 billion apps have been downloaded. Given the time needed to conduct research and publish results in academic journals, there is little information available yet about the efficacy of mobile applications designed to assist in behaviour change.  A study published in July 2010 by Rao et al. is perhaps the first to examine the potential role of iPhone apps in providing feedback.(2)  The study assessed apps developed to assist diabetic patients by enabling them to record blood glucose levels and monitor entries over time.

Unfortunately, I believe that this study has not done much to clarify the role new mobile technology can play in improving health.  Nor has this study provided assistance to iPhone users who wish more information before selecting which app to purchase to help them monitor their own blood glucose levels.

Previous studies have shown that self-monitoring of blood glucose levels helps diabetic patients improve blood glucose control compared to patients who do not self-monitor. In turn, this leads to reduced financial and health costs associated with diabetes.(1)  The mobile platform may be able to provide advantages over other forms of self-monitoring through easier user interfaces, long-term tracking that can be viewed graphically, and the ability to share results easily with a health care practioner through email or other wireless means.

Rao et al., in their paper “Evolution of data management tools for managing self-monitoring of blood glucose results: a survey of iPhone applications“, identified the top 12 user-rated apps from the iTunes App Store that met the criteria of monitoring blood glucose for diabetes.(2)  A preliminary assessment of the apps ranked these 12 based on a checklist of items developed by the authors.  The checklist emphasized features that the authors believed should be included in the app and focussed primarily on app functions with minimal reference to the apps ease of use.

App functions included in the checklist (in addition to the ability to monitor blood glucose) were:

• carbohydrate tracking
• insulin/medicine tracking
• physical activity monitoring
• weight tracking
• blood pressure tracking
• meal time tagging
• notes/journal/comments
• calorie counting (food database)

Unfortunately, there was no discussion in the paper regarding why the features that the authors selected were chosen.  Are all of these features important for diabetic patients to keep track of?  If not, which ones are most important?  Can an app be effective even if not all of the functions are present? At what point does the number of functions overwhelm the user and the app becomes too complicated to use efficiently and effectively?

From the checklist, the top three ranked apps were selected for further review.  In this second step, 23 subjects (diabetic patients) were asked to complete several tasks (e.g. recording an entry, adding comments, emailing results, etc…) using each of the three identified apps.   After, each subject completed a written survey to help evaluate which features of the apps were most important and which app was the easiest to use.

What did the study find?  The app that subjects ranked as easiest to use, requested the least help using, and were able to complete the tasks most quickly, was the app developed by the company that funded the study.  Also of note, the primary author of the paper is employed by the company that funded the study.  Is it just wishful thinking that I was expecting an independent unbiased review?

What else do we learn from this paper?  For me, there are many questions left unanswered.  The present study did not test assess the potential benefits of using the apps over time (though perhaps this is planned as a future study).  Do mobile apps help produce the desired outcomes (in this case of improved glucose regulation and reduce costs)?   An intervention-based study would also help identify which functions of the apps (e.g. carbohydrate tracking or physical activity monitoring, etc…) should be included in the app and which may not be needed.  Did the app developers follow usability guidelines when developing an app?  A preliminary assessment of  which apps followed usability guidelines may also be helpful in identifying which apps will be successful at producing the desired outcome.  Finally the study also did not offer any discussion or new insights into how the apps might serve as feedback within a complex system.

My next blog posts will continue with the theme of mobile applications to help change health related behaviours, please stay tuned for more!

References:

1. Allemann S, Houriet C, Diem P, Stettler C. Self-monitoring of blood glucose in non-insulin treated patients with type 2 diabetes: a systematic review and meta-analysis. Curr Med Res Opin. 2009 Dec;25(12):2903-13.
2. Rao A, Hou P, Golnik T, Flaherty J, Vu S. Evolution of data management tools for managing self-monitoring of blood glucose results: a survey of iPhone applications. J Diabetes Sci Technol. 2010 Jul 1;4(4):949-57.