Can Speech Disrupt Healthcare Decision Making?

Nitin Dubey, Co-Founder, PrayagadElectronic Medical Records (EMR) have been a failure in India, until now! Over the years, data collection in health care has evolved globally from scanning handwritten records, to typing details, to transcribing dictations for doctors to having Tele ICU & wearables giving a continuous feed of a patient’s vitals. Then why is healthcare decision making still not driven by clinical data in India?

Dr. Krish (name changed) owns and runs a multi-speciality hospital in Bengaluru. For several decades, his hospital was a prominent landmark and was thriving, but now, even its front façade looks grim. With many new competitors in the locality, his profits have dried-up and he is having a tough time retaining his best doctors & nurses. He believes he needs to invest in new technologies and increase his marketing spends, but he is not certain about the departments or machines to invest in and doesn’t know where to focus his marketing efforts. So, Dr. Krish reaches-out to his friends who guide him to a Management Consultancy firm. He bites the bullet and signs a fat cheque to engage them. Several sessions and 184 man hours later, the much-awaited report arrives. Dr. Krish is shocked to see that the fancy report has prescribed him back his own assumptions on marketing & investments and has provided a rolodex of possible technology partners. Total disappointment! And the report says – further analysis & recommendations could not be made due to insufficient data. Dr. Krish is back to square one and wonders how to get DATA for his decision making?
Sadly, Dr. Krish isn’t alone. Even some of the largest hospital chains in India struggle to get specific details about their doctor-patient interactions, patient life cycles or the missed opportunities. Thanks to HIS, LIS and POS systems, Indian healthcare industry does have some data around financial & operational metrics, but structured clinical data exists only in areas of diagnostic reports & imaging, which doesn’t help the likes of Dr. Krish. It does not require a Sherlock Holmes to arrive at the root cause – i.e., ‘stakeholder dissonance’. In India, doctor hours are inadequate & unevenly distributed, insurance doesn’t even cover 30 percent of the healthcare expenses, per capita government spending on healthcare is among the lowest on the planet and healthcare infrastructure lacks in quality & accessibility. While most of the other stakeholders want every last detail of clinical data, the key stakeholders, i.e., Doctors & Patients are ill equipped to capture and provide this data. They lack time, infrastructure and incentive alignment. What does a doctor get by spending 10 percent extra time on typing detailed clinical records? Nothing tangible! On the contrary, in the developed world, the hospital doesn’t even get paid if detailed records are not presented.

However, there is a ray of hope! i.e., Speech-based technologies. Giants in U.S. like Nuance, M*Modal and now even Google and startups like Prayagad in India are cracking this exact piece of the puzzle. A doctor can now just speak, instead of typing her notes, and she can record much more data in a matter of seconds. In turn, the product provides drug contraindications, alerts the doctor on allergies and contextually relevant medical history of the patient, shows statistics based on the doctor’s speech queries, and all of this in real time, in spite of the diversity in Indian accents. The secret lies behind the Natural Language Understanding algorithms, which understand speech, covert doctor’s inputs to structured data and provide insights, and these models adapt quickly to an individual’s accent & style of practice over a very short period of time. What I just mentioned is not fiction, it is already happening and happening in India as well.

Healthcare data models built using such systems provide critical insights for decision making, whether these decisions are clinical or business oriented. Dr. Krish now has access to details about his patients, the potential life-time value from them, his best performing specialization, the lost opportunities on pharmacy, diagnostics and procedures, future opportunities with possible ailments, and is able to provide better care to his patients. These solutions are slowly minimizing stakeholder dissonance by simplifying the strenuous lives of physicians and improving healthcare outcomes for patients, while delivering value for every rupee spent. For India, implementing an ambitious government sponsored health plan like Ayushman Bharat for half a billion people, requires very high-data integrity, tracking and monitoring. Execution on such a humongous scale is impossible without data capturing done at grass-roots by Medical Officers & healthcare workers. Since smart phone penetration, data availability, UID coverage, biometric technologies and an awakened end-user are all coming together, time is ripe for adoption of speech-based smart systems which will take the power of technology to the masses.