Intelligent healthcare automation - 3 problems worth solving

Practicing, receiving & paying for healthcare these days is an ignoble, hugely expensive & low-reliable experience. While the data is slowly getting better (thank you FHIR), stories of efficient, effective, and satisfying healthcare experiences are exceedingly rare.

Case in point, you're healthcare provider spends on average 16 minutes of every 30-minute appointment documenting your care and charges. Another, only 25% of patients get a 6 month chest x-ray to follow up for a suspicious (read cancer) lung masses.

This whitepaper intends to address the mass inefficiencies and situational awareness gaps that stand in the way of quality care across provider organizations. This paper will also highlight how an open standards, data fabric approach can be used rapidly to drive efficiency and quality in high risk cohorts.

The Health IT Paradox

As modern consumers have witnessed information technology revolutionize so many other aspects of our lives. Given the connectedness in the rest of our lives, I am shocked that we tolerate these experiences in healthcare as patients, providers, and payers.

Can you imagine if our air traffic control IT worked as poorly together as our health IT systems?

Would you ever order an item from an online retailer again if the package never showed up and you didn't know how much it cost until getting the bill 3 months later?

Last week I refinanced my house using an app on my phone. The process was quick & somehow painless. I know the customer experience wasn't so seamless and easy. We purchased our first home in 2003 to bring our babies into the world As a young newly minted father and physician, I remember being overwhelmed by the complexity, data needs, and verifications.

As consumers in every other industry we have witnessed artificial intelligence & automation have made complex buying cycle better, faster, cheaper & more reliable.

As consumers in healthcare, we still operate in the dark ages of information technology. As patients we wander into the dark cave of healthcare with our record clutched to our chest. We stumble forward in the dark and bump into well meaning healthcare professionals along the way. If we are lucky, they can help us with a set of goods and services. As a consumer we then face a stark choice on the path forward. As consumers healthcare is the only industry where we buy something yet still have very little idea as to:

• Alternate options at other locations or with other healthcare teams.
• How much it (e.g. knee replacement) will cost.
• The quality of the care and services being offered,

When we buy or work in healthcare, why is everything still so ridiculously hard and expensive?

3 problems worth solving.

Healthcare has an abundance of problems. That said, not every problem has the same priority nor impact ratio when solved. As we boldly stumble into the age of artificial intelligence & automation solutions in healthcare, it's exquisitely important to consider which problems should be solved first.

As always with healthcare, I choose to look to the Theory of Constraints when trying to uncover the highest value targets. IMHO the problems below the biggest constraints to Flow in healthcare. If the theory holds up, then solving them should drive the greatest value for all stakeholders.

1) Information muling

I remember when computers first started showing up in healthcare. I applauded the ability to actually read previously illegible clinical notes and I loved the safety of direct order entry. As the years rolled by more and more applications began to clutter my workflow. As most applications aren't connected well, I find myself now hunting for patient data across many different applications. Instead of spending time talking to and counseling patients, I am now forced to tend to a myriad of health IT applications. As providers, much of our time is spent digging for data, normalizing the data, computing on it in our heads, and then shoving that data into an endless set of hungry, disconnected machines.

Have you seen a Doctor lately and gotten frustrated with how much time they look at the computer screen rather than at you?

Unfortunately, that's now because we work for & tend the needs of the computer applications rather than the other way around.

Fortunately; a properly implemented, #fhir data fabric quickly drive efficiencys and remove meaningless healthIT from healthcare provider workflows.

2) Situational awareness of emerging risks (and everything else).

As patients, we have this idea that our healthcare teams are monitoring our healthcare processes and risks. We believe that they must have some air traffic like control systems that identify when bad things are about to happen to patients and at-risk populations. The truth is unfortunately far, far from that notion. The modern healthcare team is awash in disconnected data. Emerging risks generally aren't "detected" until weeks or months after the event occurred. Generally speaking, healthcare teams are in the dark and simply do not know until it's too late that a patient needs rapid attention. At the same time, identifying and mitigating risks in a near real-time fashion is one of the keys to avoiding costs and patient suffering.

Can you imagine if we didn't know that planes in the sky were in trouble until 6 weeks after they crash?

Unfortunately; most healthcare provider teams aren't aware of critical, emerging patients risks until it's too late.

Fortunately; a properly implemented, #fhir data fabric along with CQL & DQMs knowledge services can be tuned extract the signal from the noise that our healthcare teams need to avoid the inevitable costs and suffering from unmitigated disease.

3) Low-reliability information technology

Writing orders for future interventions and studies is a crapshoot in healthcare. Sometimes the thing happens but all too frequently the important thing never occurs. I equate ordering in healthcare to writing the requested thing on a helium balloon in black marker. I let it go in the sky and hope that someone sees it before it disappears. So many critical things simply never occur because they get lost in the everyday noise of our broken system. The truth is that health IT systems are open looped and are not designed to easily track fulfillment. As patients we intuitively feel this and know that if we don't track, advocate and push our own healthcare needs forward then they will immediately get lost "in the system".

Can you imagine ordering something from an internet retailer and not being able to track your package to your front door?

Unfortunately, as clinicians most of the things we order never happen given. This is not only frustrating, it is dangerously unreliable way to use our scarce time.

Fortunately; a properly implemented, #fhir data fabric along with CQL & DQMs knowledge and BPM+ Health processes can more simply and reliably guide the healthcare buying process towards beneficial outcomes.

Is all hope lost?

No. These problems are solvable with intelligent automation approaches running off an open standards, healthcare data fabric. As a matter of fact it isn't even that hard. Apply the following beneficial pattern cyclically:

1) Aggregate and harmonize all of your patient data streams to FHIR.

2) Tune your data fabric using near real-time data and simple risk algorithms

3) Connect downstream tools and processes for critical processes using safe, scalable FHIR APIs for the high traffic conditions in your network.

4) Use near time risk information to bend the cost curve for high burden healthcare conditions.

5) Rinse, wash, repeat

The successes of Smile CDR and our open source HAPI version over the last few years offer proof that FHIR apis can deliver great value along the the purchased healthcare chain. If you are a solutionist, I'd strongly recommend you start by trying to fix these global and hugely negatively impactful problems and help healthcare flow towards better quality.

It is now actually possible to deploy air traffic like control systems for healthcare. At this point it's no longer a technology problem but a legacy business problem.

#betterglobalhealth isn't an aspiration. It's what we do everyday at Smile CDR.