During the last Apple Event, Tim Cook proudly unveiled ResearchKit, an iOS interface bringing researchers to our iPhones. This line of tools and apps enables researchers to publish apps for the purpose of collecting data from iPhone and iPad users from around the world. In theory, this sounds fantastic, especially for large scale studies that normally requires tremendous amounts of administrative and financial support just to initiate a project. ResearchKit eliminates these expenditures through empowering individual iOS users to sign up and participate in research as a participant at their leisure. All the research team needs to do is obtain approval from the institutions human subjects committees or Institutional Review Boards (IRB).
So where’s the previously mentioned flaw?
When a study requires cross sectional data, or data from various demographics, anything that limits participation skews the final results. When examining a study that requires an iPhone 5S or newer, or an iPhone with an activity monitor, plus ones ability to measure other vital signs, collect biometrical data, or obtain lab results, the study criteria immediately eliminates those with barriers to these medical resources, like those of lower SES.
To validate my claim that the iOS ResearchKit will produce skewed results based on SES, look no further than the My Heart Counts by Stanford. This research study hopes to collect information about heart health from around the world. Data is collected over a 7 day period. This is broken down into daily events, like a daily questionnaire gathering activity information and sleep duration; twice over a week, personal perceptions about their health; and once over the week, like a walk test and lab results. Each of these items derive from known indicators for heart health, and are not surprising.
However, what if you are unable to afford a new iPhone 6 and keep it on you all day for step counting? What if you are unable to collect your blood pressure due to lack of access to a cuff or knowledge on how to even do this? What if you cannot make it to a doctor’s office and request your cholesterol levels and resting blood glucose levels? What if you are an older adult with heart disease, but do not use a smart phone? What if you are a budget Android user?
Well, your data stands a great chance of being discarded due to incomplete results or simply excluded due to access. Those familiar with public health advocacy or healthcare disparities would not be surprised by this statement, since it highlights the core principals on why healthcare disparities exist in the first place.
Within the medical and research community, the term, social determinants of health, groups together the five factors that accurately summarizes why health disparities exist. The Center for Disease control identifies social and socio environments, physical environment, genetic predisposition, behaviors, and healthcare access, as their determinants of health.
For example, an University of California in Davis study found that lower SES populations possess a roughly 50% chance of developing heart disease. This same population also would struggle to afford a shiny new iPhone 5S or iPhone 6, the necessary resources to obtain the required vital and lab results, and other aspects of the My Heart Counts study. This is without even bringing up the Stanford’s aims to conduct this study globally, where many populations definitely cannot afford these devices or ability to upload their data on a daily routine for 7 days.
This leaves me to question the overall beliefs behind the research studies being conducted with the Apple’s ResearchKit. If studies are designed to understand prevalence rates or healthy living trends through luxury consumer electronics that omit those most likely to experience the greatest health disparities, what does ResearchKit actually provide researchers? Will results actually advance the knowledge base for medicine?
On one hand, I applaud Apple for considering the implications and designing ResearchKit, since it can serve to help the research community obtain hard to gather cross sectional data and even Longitudinal data from a large sample population at a cheap price. On the other hand, validity remains questionable if we, as researchers and consumers of studies, need to know more about populations with the greatest health disparities.