The disruption triggered by the coronavirus (COVID-19) has induced unplanned growth across the healthcare industry. Despite these challenges, leaders in healthcare see tremendous potential in AI and analytics to deliver on the promise of higher quality care at a lower cost by empowering their executives, business leaders, clinicians, and nurses by harnessing the power of predictive and prescriptive analytics.
Many healthcare organizations are seeking to harness the vast potential of Artificial Intelligence (AI) and its four components — machine learning (ML), natural language processing (NLP), deep learning, and robotics — to transform their clinical and business processes. They seek to apply these advanced technologies to make sense of an ever-increasing “tsunami” of structured and unstructured data, and to automate iterative operations that previously required manual processing.
I have analyzed and calibrated these technologies leveraging a seminal strategy framework from John Gourville, Harvard Business School professor, predicated on the resistance to patient adoption, as well the degree of change behavior needed from physicians, clinicians, nurses, providers, payers, policy makers and the government, which will likely assure a high probability of success, in my humble opinion and will inform post-pandemic strategy blueprints and scenario/policy planning from these entities.
Strategy blueprint: The evolution of “digital healthcare” in the post-pandemic era.
This strategy matrix articulates five promising digital health capabilities and technologies which will astronomically grow over 2020-21 and find increased adoption even after the pandemic has subsided.
I have modified and adapted this strategy blueprint from “Eager Buyers and Stony Sellers – Understanding the Psychology of New Product Adoption,” by John T. Gourville, to craft this 2×2 matrix.
The x-axis maps “patient resistance to digital health adoption” high to low while the y-axis maps the “change in behavior” needed from the other key stakeholders in healthcare such as physicians, providers, payers, policy makers, and the government (given that it is by far the largest payer for healthcare services today) from high to low.
The predictive power of this matrix can be leveraged as follows:
- Upper Right Quadrant: Low patient resistance to adoption with low change in behavior required from other stakeholders like physicians and providers can be prognosis for “Smash Hits” — technologies like telehealth, telemedicine, remote patient monitoring, and associated medical devices and instrumentation and analytics.
- Upper Left Quadrant: Relatively high patient resistance (demands some effort) to adoption but demanding low change behavior from other stakeholders is classified as “Easy Sells” — technologies like wearables and medical devices and apps to capture vital sign data from patients with manual/automated entry into patient portals.
- Lower Right Quadrant: Low patient resistance to adoption demanding high changes in behavior from other healthcare stakeholders like providers, physicians, and payers signify “Long Hauls” — investments in next generation analytics and artificial intelligence (AI) including machine learning, deep learning, natural language processing (NLP) and robotics that will enable predictive and prescriptive analytics capabilities across the healthcare enterprise.
- Lower Left Quadrant: High patient resistance to adoption with high changes in behavior required from other key stakeholders is a non-starter or recipe for “Sure Failures” and little to no potential traction. Key message — proceed here at your own peril and risk!
The key premise is that the greater the level of change in customer (patients, physicians, nurses, providers, and policy makers) behavior needed, the greater the barrier to customer adoption, despite the promise of value delivered by the new product or technology. John Gourville makes the point that producers of innovation often overestimate the customer adoption by a factor of 3x while consumers allocate significant value to their current product or service and demand a value proposition that is practically 9x that offered (perceived) by their current product or service.
Let’s leverage this strategy framework to map the five innovative and disruptive technologies which are not only growing astronomically as we speak but are here to stay for the long term even after the COVID-19 pandemic is a distant memory. Matter of fact, these five technologies will be segments of astronomical growth that will contribute to the overall growth of the healthcare and life sciences industry, unlike every other industry that will be demand constrained over the pandemic and beyond.
- “SMASH HIT” — Telehealth for screening, triage, and follow up appointments.
Telehealth is growing astronomically as we speak given the advent of COVID-19 and is here to stay beyond the pandemic. Seniors over 65, as well as patients with chronic conditions like cancer, chronic obstructive pulmonary disease (COPD), cardiovascular disease, etc., are most vulnerable to death if afflicted by the COVID-19 virus and are also locomotion challenged. Leveraging telehealth to diagnose and treat them remotely for conditions that are not acute, will lower risks of infection for both these vulnerable patients as well as caregivers.
But are patients open and receptive to embracing telehealth as a medium to engage with their PCPs and specialists, as well as nurses and care coordinators, which is critical to long term adoption and sustenance?
According to a recent survey by Amwell (formerly American Well) consumers remain interested in telehealth, with 66% reporting they are willing to use it. Unsurprisingly, the younger demographics are most open to telehealth, with 74% of 18‒34-year-olds and 72% of 35‒44-year-olds saying they are willing to use it. While the senior population has the lowest interest overall, 52% of seniors are still open to using telehealth. That means that of the 47.8 million Americans over the age of 65, 24.85 million are willing to use telehealth, which makes this a “Smash Hit” already.
- “SMASH HIT” — Analytics and Remote Patient Monitoring for Care Teams and Clinicians.
Like telehealth, Remote Patient Monitoring (RPM) adoption is exploding as we speak and will be here to stay in the post-pandemic era. For high risk seniors, patients with chronic diseases like cancer, COPD, cardiovascular diseases, diabetes with complications and kidney disease, having care coordinators leverage RPM to reduce the number of office visits by patients, while still receiving substantial compensation from CMS presents a sustainable opportunity for healthcare providers to reduce 30 and 90 day re-admission risk, while assuring better patient outcomes.
Innovative and disruptive RPM systems help care teams monitor, manage, and engage patients in the comfort of their homes which contributes to reducing cost, mitigating risk while improving outcomes and increasing reimbursements.
These systems continuously stratify patient risk via artificial intelligence (AI) driven algorithms, alerting and empowering care teams with optimal windows of opportunities to intervene when needed. Automated visual/audio reminders and phone calls enable higher patient engagement, medication adherence, integration with telehealth, and virtual video conferencing and visits which enables rapid patient health assessment and optimized for various chronic diseases.
According to a 2019 Spyglass Consulting report referenced in a recent article in Healthcare IT News, over 88% of hospitals in the U.S. have plans to invest in RPM in 2020 which have been accelerated in the wake of the pandemic, validating my assessment that RPM, like telehealth, is a “Smash Hit” that will grow and see adoption post-pandemic.
- “SMASH HIT” — Pandemic/Disease Outbreak Readiness Monitoring and Response Center.
A key challenge experienced at the national, state and local level was/is the lack of readiness to the pandemic which culminated in the crisis on a state-by-state basis with hospitals in hard hit areas being completely overwhelmed.
Going forward, thought leaders in healthcare will institute a Pandemic and Disease Outbreak Readiness Monitoring and Response Center as part of their strategic, scenario, and policy planning initiatives, even at a health system level. Equipped with data streams from the CMS, HHS, etc. and leveraging self-service analytics platforms and visual analytics tools, these leaders will stand up a war room(s) with dashboards that will not only monitor disease and epidemic outbreaks in other parts of the world, but also provide actionable insights via visual analytics dashboards which will proactively, monitor, analyze, and measure both the demand as well as the supply side implications.
- “EASY SELLS” — Cloud and patient portals compatible AI and mHealth apps offering data capture and recording of vital signs.
One of the biggest barriers triggered by the COVID-19, was/is the availability of test kits to detect the coronavirus infection in patients. Further exacerbating this serious challenge was that it took 10-20 days to capture samples, process and send them to the diagnostics vendors (who were overwhelmed by the volumes) and then waiting for the results to return and then communicate those to patients. Many patients who were in fact positive deteriorated and often had to be admitted to ICUs and ERs further overloading these hospitals, given their constrained capacity of ICU/ER beds with ventilators as well as testing kits, drugs, devices and therapies to treat these patients.
Happily, leaders as well as disruptive start-ups in the healthcare industry have responses to this dire crisis with a plethora and portfolio of cloud-based AI-powered apps deployable in mobile devices and tablets (mHealth), many of which can capture and record vital signs onto patient portals for instant access by care teams and clinicians.
- “LONG HAUL” — Artificial intelligence (AI) and analytics investments to enable predictive and prescriptive analytics insights across the healthcare enterprise.
In my previous blog, Covid-19 and its Impact on Healthcare and Public Health, I predicted the convergence of AI (comprising machine learning, natural language processing (NLP), deep learning, and robotics) along with analytics to drive measurable improvements in business, clinical, and patient outcomes. This has also been validated in a recent research survey, “Data Analytics in Healthcare,” by HIMSS Media.
There are significant challenges to adoption identified by the research, including working with unstructured data and collecting data from diverse, disparate sources, investing enough resources in AI, and optimizing/automating analytical models. At the same time, the benefits and outcomes (for both patients and healthcare providers) far outweigh the barriers. These include:
- Higher levels of safety for patients (especially high-risk patients) and physicians, nurses and technicians treating them while assuring better patient outcomes.
- Lower cost of healthcare delivery while assuring quality and safety, inviting competitive reimbursement from both payers and the government.
- Mitigating risk of 30- and 90-day re-admissions, denials and loss of revenue while ensuring the best possible patient outcomes from the treatment.
- Optimal utilization and return on investment (ROI) on constrained, high value resources like ICU and ED beds, and medical equipment like ventilators and complementary medical devices and diagnostic equipment.
- Higher levels of disease/pandemic outbreak readiness as well as responsiveness to ensure that the unforeseen and shocking vignettes articulated in the previous section can minimized or precluded.
20 innovative use cases
Healthcare leaders will be deploying AI (machine learning, NLP, deep learning and robotics) + Analytics over 2020-21 to empower their executives, line of business (LOB) leaders, physicians, clinicians, nurses, analytics and data scientists with predictive and prescriptive analytics powering proactive and actionable insights to enable superior business, clinical, patient, and population health outcomes. Here’s a look at some opportunities:
- Proactively identifying patients at risk for adverse health events like heart failure and heart attacks, as well those most at risk of being impacted by infectious and contagious diseases and epidemics.
-
Predicting patient utilization patterns (e.g. missed appointments and optimizing inpatient/clinical throughput.)
-
Population health stratification, including risk scoring for chronic diseases like cardiovascular disease, COPD, diabetes, and cancer.
-
Projecting (what-if and scenario analysis) the number of total hospital beds as well as the constrained capacity of ICU and ED beds that will be needed to deal with a potential outbreak, as well as the ability to deploy additional buffer capacity, if needed.
-
Monitoring the number of usable ventilators and vital signs monitoring medical devices and equipment on hand, potential shortfall, as well as sources like state and national stockpiles from where these can be resourced in an emergency.
-
Predicting demand for telehealth and Remote Patient Monitoring services and then mapping and matching with physicians, clinicians, nurses, care-coordinators, and technicians to meet the demand and ensure superior outcomes.
-
Blending data from multiple EMRs, standardize curated data sets to secure a “single version of the truth” re: the patient.
-
Enabling a 360-degree view of the patient for clinicians and nurses at the patient bedside with predictive analysis that willalert them about deteriorating conditions.
-
Advancing precision medicine and personalized treatment of patients predicated on an analysis of genomic profiles and
mapping, comparing these to the mainstream population of that country or continent, with gap identification triggering proactive treatment. -
Improving supply chain management (SCM) efficiencies, including predicting stock outs for drugs, devices, life-saving equipment like ventilators and supplies, as well as expired drugs, while saving money.
-
Proactively identifying and mitigating cybersecurity risks, including medical fraud.
-
Identifying and classifying anomalies through imaging and incidental findings from automated scanning of X-ray, MRI, and CAT scans to proactively alert radiologists on patients at risk of a cardiovascular event like heart attack or stroke, for instance.
-
Proactively detecting fraud, waste, and abuse pertaining to employee overtime and agency costs.
-
Minimizing issues with reimbursement and payments including fraud, waste, abuse.
-
Ensuring ICD-10 coding and billing accuracy given new protocols, testing kits, instruments, devices, and supplies.
-
Predicting denials, identifying root causes, and managing these to revenue realization.
-
Forecasting demand for all levels of staff needed to meet growth projections, and matching demand with supply of doctors, nurses, and technicians from known sources, as well as tapping into pools of retired doctors and nurses, as well as soon-to-graduate MDs and nurses to meet unforeseen demand when disease/pandemic outbreaks happen.
-
Predictive analytics on cellphone data and sentiment analytics to identify disease outbreaks and hotspots for pandemics, which was done in China recently.
-
Medical robots for capturing vital signs like temperature (touchless sensing), pressure and glucose levels, as well as those for serving food, medicine, and applying bandages to patients, ensuring minimal contact in the event of contagious and infectious diseases.
-
Automating executive and performance dashboards delivering key metrics and performance indicators instituted by senior executives, and then cascading these across the enterprise with complete transparency on performance againstcompany and department targets, like the “Balanced Scorecard” system with its origins from Harvard Business School (HBS) that is now ubiquitous acrossother industries.
STAY PUT.
LEARN
Dive deeper into how COVID-19 is impacting healthcare and public health.
READ
Check out 11 ways data science in combating the coronavirus.
GET
The
Essential Guide to Analytics Process Automation (APA).
LEARN
The Convergence of Data Analytics and Artificial Intelligence (AI) in Healthcare.