Tag Archives: context analytics

Healthcare in India – still Miles from the Internet of Things Vision

28 Dec

My 81 year old dad has been having a rocky couple months.
It started with weakness and giddiness; then a feeling of numbness in his limbs; then incidences of blurred vision and slurred speech – classic stroke symptoms.
My brother and I started the usual merry go round that all caregivers of senior citizens go through – emergency-neurologist-cardiologist-nephorologist-endocrinologist-back to emergency-repeat cycle. Threw in sessions with a GP, and then a couple operations of the eye at the ophthalmologist.
Went to two hospitals, and 4 branches of one of them.

Was completely appalled at the isolated nature of patient information access and diagnostists available, even at the best hospitals, in the Silicon Valley of India. And this, when we are debating the use of Wearables, Internet of Things and Big data all the time!

Lets look at the possibilities for the Healthcare IoT first:

– At the individual/ patient level –

The ultimate beneficiaries, patients (both current and potential/ future), will basically create and consume larger amounts of data than ever before – due to wearables (e.g. Tempo from CarePredict) / implants/ other gadgets and devices owned.

This will enable better monitoring of behaviour, and, when coupled with context – facilitate hugely improved diagnostics. Diagnostics that are derived from integrated data – multi disciplinary, as well as omnichannel/ source. (So, no more – “your sodium is low, eat more salt – says nephrologist — uhhh, but BP is high, reduce salt says cardio”! )

These diagnostics will then not only treat any emergencies or negative incidences quicker/ more competently, but also act proactively – thus alerting users and discouraging them from harmful behaviours (too much smoking/ not enough walking/ too many carbs/ not enough sleep….)

This personalised patient care, is after all, the raison-d-etre of the wearables market, which, according to a recent IDC report, is set to reach 45.7 million units worldwide in 2015, and 126.1 million units in 2019.

The IoT approach can then be taken a step further – and use gamification techniques to reward compliant medical behaviour. (My constantly sudoku playing dada and candy crushing mom I’m sure will be very vulnerable to these approaches!)

– At the macro/ aggregated level –

Obviously, when taken as a large body of data, analytics organisations can develop and fine tune personas and segments, making them richer and more detailed. The way to do this would be to combine biomed research using life sciences databases with aggregated patient behaviour and symptoms. This should ideally ultimately result in predictive models, thus leading to better treatment. Basically, a higher emphasis on real time monitoring and intervention combined with traditional lab work becomes the new clinical research methodology! In long term vision, companies and nations get a jumpstart on reducing burden of disease in the overall population.

– At the enterprise level –

Many studies are saying that providers may be able to save up to a quarter of their business costs with the Internet of Things.  This would be done by integrating new technologies into their operational strategies. Think about it – a vast, dynamic mass of wearables, biometric sensors, smartphone add-ons, and other medical devices and implantables – all growing, shifting, and ripe for an IoT approach

wearables

– Miscellaneous –

The serious side of Healthcare IoT is only one side of the picture actually – the wearables landscape is fairly glamorous even from the early adopter tech consumer’s Point of View! Think Ping, a social networking garment developed by designer Jen Darmour; Smart socks that tell you when to change, and where the pair is; Electrozyme or Motorola digital tattoos (that always remind me of the Bourne Identity series); tweeting bras (to remind one of breast exam monthly); dream inducing headbands; Peekiboo, a cap that lets you see the world through your baby’s eyes; smart pajamas that help you with bedtime entertainment when your kids are ready for bed; i-Grow income hair growth helmet; Drumpants which is Multiple touch pads sewn into a pair of classy jeans give drummers a six-piece kit; and, the coolest of all, Necomimi, a set of feline-like ears that strap on to your head, supposedly measure brainwaves and then move and flick based on how you’re feeling at a given moment!

The Flip Side

But the real truth, atleast in India (as experienced by my dad and me) is a far cry from the vision. At the moment, there is NO healthcare Internet of Things – maybe just a random collection of medical devices and consumer monitoring technologies that don’t quite know what to do with each other! 

Treatment is isolated – diagnosis is gut and experience based (which actually is not necessarily a bad thing – since, given India’s huge population, doctors do get to practice a lot and so see a wide variety of patients); there is very little digitisation even in the best hospitals. (In fact, in the hospital where we treated our dad, even within the 3 buildings in the same health city campus, there are different degrees of digitisation – the newest, the Eyecare hospital introduced EMR maybe 3-4 years ago, and so it’s all modern; the next oldest Multi Specialty Hospital now scans the written prescriptions of the doctor after you come out of the treatment room, (so, atleast keeping some records – though I’m reasonably confident nothing is being done with those scans yet); and the oldest, the Cardio facility I think still does nothing digital)

To answer the question – “why” – I think it really has to do with a basic question of population – in our country, we are too busy solving the Maslow’s Hierarchy lowest rung questions of food, clothing and shelter to spend too much money and time on other stuff. Coupled with legacy systems and old habits dieing hard, it’s a tough combo to beat!

And then there’s the question of reliability – many medical practitioners suspect non manual devices and gadgets – I have met with scepticism on the results of digital BP monitors from all doctors and nurses! (and my 10-times-a-day-BP-measuring dad’s experience shows significance differences between same time readings of manual/ analog BP monitors and the digital ones. Despite repeated calibrations!). So, reliability is suspect.

And, as yet, I’m not even stepping into the whole compliance/ medico-legal/ security aspect of the data

I think this is a bit like the “obvious” steps in customer centricity across functions and kind of business that are visible to everyone now with Big Data, but the fact is that no one is exploiting them – due to lack of vision, corporate lethargy, or even mal-intent in the form of vested interests! (watch out for the next blog on that topic in a few days). The good news – In a PwC study, 95% of healthcare CEOs said they were exploring better ways to harness and manage big data. The bad news – this is not confined to India alone – Read this article on how this state of affairs sadly seems true globally :(.

Maybe it’s just a matter of time – the question is how much time! Meanwhile, for my dad and me, it’s onwards and upwards to the next “o-logist” – this time the ENT specialist for new hearing aids! Yawwwnnnn….

Advertisements

The Problem With Context

15 Sep

Jack and Jill

Jack and Jill


Credit

“I do not fear truth. I welcome it. But I wish all of my facts to be in their proper context”: Gordon B Hinckley

Growing Up, my marketing 101 lesson taught me “Customer is King”. The Father of India, and probably its best known global personality, Mahatma Gandhi, said – “A customer is the most important visitor on our premises. He is not dependent on us. We are dependent on him. He is not an interruption in our work. He is the purpose of it. He is not an outsider in our business. He is part of it. We are not doing him a favor by serving him. He is doing us a favor by giving us an opportunity to do so.”

However, It could be argued in today’s world that the all important C word is being replaced with another C word – Context. So, now, Context is king – mainly, because Context is what allows an organization to understand its (first C word) Customer better.

We as prospects, shoppers, switchers; movie goers, diners, Home owners, Business people, car drivers, facebook users, are today cumulatively generating Yottabytes (10^24) of data, which is going to increase very soon to Brontobytes (10^27)! Intelligent organisations are (or atleast are thinking of, and if not, then should be thinking of) investing in technology, infrastructure and analytic decision processes to use this data for higher revenue generation as well as cost optimization. Context, the accumulated historical data generated by people, places, and things, is almost a mandatory component of these analytic processes.

In simple, laymen terms – context makes up the circumstances in which an event or an idea is set, and that therefore is what makes something clearer to you —

– Does listening to a special song make you think of a special person or a special situation? That’s context! (As “The way you look tonight” was Julia Roberts’ and Dermot Mulroney’s “special song” in My Best Friend’s Wedding)

– Think of nursery rhymes – when you delve deep into them or take them out of context of being repeatable pieces of music for kids, they are fairly disturbing! (Jack broke his crown? Humpty couldn’t be put back again? Whoaaa!!!)

For the traditional marketing folks, context analytics is a bit similar to behavioural + psychographic segmentation in the good ole fashioned days, and not just demographic – so, you add parameters of date/ time/ purchase event/ mood/ place/ company/ actions/ attitudes/ usage etc., and layer them in – to create better profiles of people/ events/ data….

When you marry Context with Data, you trigger unique, new relationships between hitherto unrelated data points – this helps you derive trends and patterns – and generates new business opportunities. Context makes data become richer, more meaningful. Someone very famously once said “Context is worth 80 IQ points”

Conversely, without Context, business conclusions might be flawed. It’s the old analogy of “knowledge is knowing that a tomato is a fruit, but wisdom is knowing not to put it in a fruit salad”

So, is data and context a marriage made in heaven?

I would say, it is certainly high on the “will not make the divorce courts within the next 6 months – ok, 2 years” side of the coin, but, like all relationships, it needs work! One of the thing that contextual analytics ends up doing, counter intuitively, is still not giving the whole picture (A bit like the blind men and elephant story – where, depending on CONTEXT, each blind man ascribed a different name to the elephant – thus describing a piece part of it, but no one realized it was an elephant)

blind men elephant relativity

So, what then is the secret sauce required to turn this marriage into one made in heaven?

a) Disruptive Discovery: An inherent flaw in most predictive analytics algorithms – whether based on big data or not, is that their results are incremental – because what you know/ what has happened/ inbuilt assumptions – i.e., CONTEXT are what predict the future, they will to a certain extent pre-determine the outcome of the prediction.

As a friend said, facebook always recommends Metallica to me when because like Iron Maiden – it doesn’t realize that I also like Jazz! Similarly, Amazon always shows me more Nora Roberts – as I bought some romance once, and leaves a huge opportunity gap because it doesn’t realize that I have bought say a Khaled Hosseini elsewhere!

Not sure how many of you read this hilarious piece that was circulating recently on how “liking” everything on facebook for 2 days turned the news feed into a strange animal! Much recommended for an empathetic laugh

A good case study in how pre ordained behaviors, or context, or for that matter assumptions, to a certain extent skew actual results can be read here (a pithy commentary on Windows 8’s failure)

b) Response Timeliness: This one again is a bit counter intuitive. Experts and proponents of context analytics would say – but PART of context is the real time nature of the tracking! One is continuously refreshing data with newer and newer inputs, that turns the context richer and richer that then gives better and better outputs — and on and on and on.

The tragedy, however, is that while most organisations are (or atleast are thinking of, or if not, then should be thinking of) collecting and processing data on a real time basis, the RESPONSE to a lot of this data is NOT real time – why? Mainly because it CANNOT be! They don’t have the capabilities – this is the classical bottleneck/ the Blackwell’s limiting factor/ the critical path in the Gantt chart!

As a parent, I realized very early on that the concept of “quality time” that was created to essentially rationalize lack of adequate time spent with kids by working parents was a lot of bs. It fell into shambles the minute my tearful 1 ½ year old asked me – “when I fell down and was bleeding so much, where were you”!!

As business people, we also know the value of being at the right place at the right time. This time criticality determines a lot of what we do – how we behave as consumers – what time we eat, when we exercise, when we buy… – but also how we plan say marketing promotions, or budget calendars, or for that matter the pitch to the HR guy for a raise!

As one of my fav musicians George Harrison said:
“It’s being here now that’s important. There’s no past and there’s no future. Time is a very misleading thing. All there is ever, is the now. We can gain experience from the past, but we can’t relive it; and we can hope for the future, but we don’t know if there is one.”

What this then means in the big data world is, that when a customer sends out data – you translate it into a signal/ information/ wisdom – and then you jolly well figure out a way to respond to it at the time that the signal is asking for a response! May be immediately, may be a few days later – but, figure it out! Use the context, to not only tell you parts of the elephant, but the whole mammoth! So, its not only to find out what you can sell to the customer and where – but when/ and how quickly to lend a sympathetic ear/ when to fix a broken washing machine or a crashed site!!

Unless you can gear up your response mechanism to be time dynamic and responsive, all that context married to your data – is going to —- to use the marriage analogy – not produce a baby!!!

c) The Human Touch – AI, the Internet of Things and all automation notwithstanding, atleast for now, we humans are just wired to need human intervention

a. This is because technology is still flawed – Watson notwithstanding, most humans actually instinctively and judgementally can scan a wide breadth of data and more or less give reasonably accurate judgements – after all, that is what experience and seniority lend to a man.

b.Man is a social animal after all – most of us want, nay need, human contact – a good case is customer service. Automated responses just don’t cut it for many – they want to talk to a real person, the more serious the issue, the higher the need for human contact (maybe shortly that “real person” will also be a drone, but that’s besides the point).

After all, how else but due to a human would you see these 2 really delightful examples of ‘customer service” interactions – one by a Netflix representative, and this one by an amazon guy.

Make no mistakes, human beings could just as easily make bad mistakes – like seen in this http://www.theverge.com/2014/7/15/5901057/comcast-call-cancel-service-ryan-block!

But, by and large, if one could only afford it, human beings would be the best response mechanism.

All in all, its great to see the progress made due to big data and the Internet of Things etc, but the organization who can take the best of the context plus data marriage, and add the disruption, the time criticality and the human interface element would be the one on which I would put my money!