Doctor in your Pocket

Dozens of Healthcare Information Startups today are scrambling to design the Doctor In your Pocket (DIP) application which tests in real time your cholesterol, monitors your circadian rhythms and insulin levels, recommends restaurants around that offer the healthiest options given your disease predisposition, anticipates your blood sugar spikes and potential cardiac arrest. Exciting possibilities all!

Imagine just like Alexa, if you could speak to the DIP app about your back aches and acquainted with your medical history, genetic profile and environmental exposures it could produce a very targeted list of possible causes. Sometimes the cause might compel you to immediately visit the emergency ward and on other occasions it grants you the peace of mind to follow through on the day’s agenda. Wouldn’t that be something?

Mired within this fantasy are some thorny technical and business challenges. I felt compelled to write a post to enumerate them

The DIP by definition will be a consumer app but I can envision version of it offered through employer’s payors and government agencies who incentivize their employers and customers to use it with lower premiums on their insurance plans. How does one convince these purse holders of the Healthcare industry of the utility of DIP apps

Then there is the question of Data DIP apps will be housing. Ensuring the safety of such data is mandated by such a patchwork of global regulations that will make the heads of the most veteran biopharmaceutical executives spin. Companies like DNA Nexus and platforms like AWS are focused solely on OMIC data security leaving the analytics to their consumers for a reason. Because it takes enormous resources to get it right.

And even after complying with the laundry list of rules that apply to Biomedical data, it still might not suffice customers to trust DIP developers with the same. Therefore DIP companies will need to deploy the most cutting edge understanding of Consumer Behavior to win their trust and acquire their information

Now consider the effort it will take to compute an Individuals biometrics, query their genomics data , store their lifestyle data and interpret it in a responsible and legally satisfactory way. Well it would entail a mammoth collaboration between Bioinformaticians, Stack Developers, Microbiologists and Data Scientists. The investments in R&D in to build DIP will be gargantuan and both the enterprise and its investors must have long term investment horizons. These stakeholders must not self prevaricate assuming they are building a run of the mill app

Eric Lander famously remarked “Genome, bought the book hard to read” which pithily encapsulates the central challenge for bioinformatics DIP apps will need to run. Sure, we can list the order of the three billion base pairs in the genome but what is the significance of any variations it has with a reference genome? How do genomes act in concert to produce the symptoms of Polygenic disease? Do we understand the role non-coding parts of the DNA play in preserving human health? The relationship between Human Genome and Bioinformaticians today is very similar to the relationship between Hieroglyphic Literature and Linguists in the 17th century. The cipher for the ancient script had not yet been discovered and therefore the social, political and economic life of ancient Egyptians remained a mystery to us. Similarly, Data Scientists and Geneticists today are trying to crack the cipher for the Human Genome to help us understand the microscopic life of cells, genes, proteins, antibodies etc.

As a second challenge of adding features to DIP apps take the example of liquid biopsies which have proven themselves to be curmudgeonly scientific problems. Companies like Grail and Guardant Health had devoted themselves singularly to the diagnostic test. Notwithstanding, A study by John Hopkins university found that two Liquid Biopsy tests drew different conclusions from analyzing the same sample of blood which gives an inkling of how far we are from using these tests in clinical settings. If we cannot run liquid biopsies consistently in a lab, imagine what it will take to design a pin prick contraption that tags on to your modular phone and uses to deep sequencing to measure the nucleic acids shed by tumors in your blood sample it gathers. I think its safe to say such a device still belongs in the realm of science fiction.

There will be trials and tribulations galore on the path to building DIP apps but the dent it will allow us to make on the human condition will make it all seem worthwhile.

Let me know your thoughts.