iPhones as Infrastructure, Part I

The Generation of Healthcare Big Data, and Why Big Data is a Part of Personalized Healthcare

There are often two sides to every technology story: hardware, and software. Since the lifeblood of personalized healthcare delivery is centered on data, we will first focus on the hardware that generates said healthcare data in Part I, and then move onto the software platforms and applications in Part II.

While the main topic of this blog is not about Big Data (it will not explore analytics deeply, focus on pharma/life sciences, nor talk about hospital usage of Big Data for ops efficiency gains), it is important to understand that Big Data is a crucial component for delivering more targeted healthcare to patients, so Part I will touch on this topic in the greater scheme of ICT’s impacts on the field. In this blog, the focus will always be the tools that provide better benefits to the individual patient, enabled by ICT.

Macro Trends: Consolidation as Infrastructure

In my last post, introduced the idea of tech-based personalized medicine, and left with a teaser: how do the seeming ubiquity of iPhones play a role in such services? We begin with a macro perspective: comScore reports that with a 44.2% market share as of Sept 2015, Apple keeps converting Android users (for platforms: Android down -1.3%, with Apple up 1.1% for percent share of smartphone subscribers). The ascent of Apple’s share of the mobile market continues to increase:

Apple ascendant.

Here is the important point: one company controls ~44% percent of the smartphone market, and this trend is expected to continue.

As such, the consolidation of hardware by a singular entity, like Apple, is a key driver for any Big Data play.

This consolidation of the smartphone market is important for ICT to impact personalized healthcare delivery at scale, especially from a data perspective. Apple, as a vertically integrated company, very tightly controls its user experience and ensures data linkages between their mobile device categories (compared to Google’s Android OS, which has multiple software forks and iterations across multiple physical devices). Speaking from experience in the software development space, working across cross-compatibility between Android platforms is a huge pain — the relative ease of technical implementation due to Apple’s tightly integrated ecosystem is of significant advantage. The consolidation of hardware therefore drives consolidation of the data source, which in personalized health, is the person themselves.

The Merriam-Webster definition of infrastructure is:

the underlying foundation or basic framework (as of a system or organization)

I would argue that the iPhone, as well as the Apple Watch, provide this underlying foundation for healthcare data and personalized healthcare:

  • Access to scale (number of people with iPhones is high; and iOS users are more active online)
  • Communication between different units (iPhone to Apple Watch data-transfer, tying other sensors to those mobile units as presented at Apple’s WWDC 2014)
  • Standardization (all mobile Apple products run the same version of iOS)

This is in stark distinction from traditional data collection methods in clinical research or the means of healthcare delivery, with hospitals limited in their geographic scope, electronic health records beguiled with interoperability problems, and a lack of standardization.

Mobile First: Why Personalized Healthcare Necessarily Must be Mobile

I also alluded to the idea of Mobile First in the prior post. What then, does this mean? As Ben Thompson from the tech blog Stratechery summarizes:

This distinction is critical: what is essential to understand about Mobile First is that everything flows from, well, mobility. The relative importance of implementation details and even the underlying OS fade relative to the significance of nearly every person on earth having an Internet communicator with them at every single moment. Every single product and service must start with this fundamental assumption.

Therefore, we can see that mobility reigns king — this is even more true in personalized health. With every single person, on their person, connected to the Internet, this opens up significant opportunities for time-series and continuous data sets that could not possibly be achieved otherwise. It is important to distinguish this from traditional data sets collected by “traditional” clinical trials, which often collect data at discrete timepoints. Over the last decade, the shift to a Mobile First strategy has therefore allowed the following Big Data qualities to prosper. Apple, with its huge network of iPhones, is looking to develop this infrastructure even further:

  • Volume — users of iPhones as end-points
  • Variety — iPhones/Apple Watch does not only track biometric or clinical data through apps, but “digital biomarkers” can even be geographic (GPS-based) or even non-clincial interactions with other apps might be used as predictors
  • Velocity — the speed that data can be tracked, as well as continual data monitoring, this allows for the change in any data point to be recorded for an individual person

As mentioned in class, managers — in healthcare, these are clinical researchers or physicians — often track stock information. This can lead to erroneous or non-robust information. With the iPhone infrastructure, personalized health companies (and maybe some providers) that are entering this space will more easily be able to track the flow information for a particular person, or a population in order to deliver more precise healthcare services.

Therefore, based on the individual habits tracked on iPhones and associated health apps, the aggregate data can help provide predictive power and personalized recommendations to that specific person. With an n=1; this was previously impossible — now companies and clinicians can better do inference for an individual.

For example, given a health profile for a person, one could actually infer their individual risk of heart attack. An app, connecting a person to a physician, could therefore recommend a check-up based on these risks or motivate behavioral lifestyle changes. That is the value of this infrastructure and the value of ICT on healthcare delivery.

Part II, coming soon:

As mentioned previously, there are two sides to every technology story. We will transition into interesting instances of software leveraging iPhone infrastructure next, especially with respect to those software companies providing digital preventative/predictive services.

Like what you read? Give Gavyn Ooi a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.