Advice to Healthcare Startups

Like many in my role, I am constantly pounded by young entrepreneurs with the "next great innovative idea for healthcare".  I appreciate their energy and enthusiasm, and in some cases they really do have something cool and special.  However, I do find myself repeating many of the same thoughts and "rules" - so I thought that I would put some down on paper to prepare them ahead of time.

This is in part inspired by an GREAT blog by Todd Dunn (Director of Innovation, Intermountain Healthcare Transformation Lab):  The Seven Deadly Healthcare Startup Sins (and his follow up advice).  The summary:
Sin 1: Healthcare startups assume hospitals will let them host patient data in “their portal.”
Sin 2: Startups assume that clinicians will be willing to access yet another portal for their data.
Sin 3: That one doctor or hospital lends enough credibility for other organizations to simply accept a startup’s solution.
Sin 4: Believing that ONE key leader inside a hospital is the decision-maker, influencer, etc. all in one role….
Sin 5: Thinking that conducting a “proof of concept” and/or pilot is a simple endeavor.
Sin 6: There isn’t anyone else out there solving the problem.
Sin 7: Believing that startups need to have more answers than questions.

His Advice:

• Use the Lean Startup tools! Regardless of where you start, it comes down to your value proposition as a starter or non-starter. 
• This often tries the patience of entrepreneurs. I cannot overemphasize the need to use the learning loop in every single part of the Value Proposition and Business Model canvases. The only way to do that is to GET OUT OF THE OFFICE!
• Be curious about workflow  - Be empathetic to your user.
• Study the industry more deeply. While you may have a great value proposition for one or two hospitals, how does your solution fit into the regulatory landscape, workflow, etc. of multiple hospitals?
• Listen! Assume you don’t have enough evidence to scale your business yet. Act like you don’t know enough. While an entepreneur’s “go get ’em” attitude is appreciated, it isn’t appreciated when the entrepreneur isn’t open to feedback, seems to have all the answers, and has a condescending attitude toward the way “jobs” get done today. Test your assumptions! Come loaded with questions that are related to your assumptions.
• Last but not least, structure a learning plan. Embrace the Lean Startup tools and methods. Following this structure will cause you to write a learning plan. A foundational question to guide your learning plan in every part of your business model is “What do we need to learn before we invest more time and money?”

Some thoughts and Rules I would add to enhance the above

• There are basically no new ideas... a successful startup understands it is about execution.  So please don't tell me that you have a brand new idea and want an NDA because the idea is so priceless and if anyone else finds out about it they will copy it.  If it's that easy to replicate, then you really don't have a business.  I remember years ago when I was being mentored by the great informaticist Dr. Bob Greenes.  He took me into his office and showed my his PhD thesis from around 1966.  This was the dawn of the age of computers, and in his thesis was basically every idea we are now hearing from "startup" companies daily - computer guided interviews and diagnosis, telemedicine, artificial intelligence to read notes, etc...   The key is rarely the idea, but how you combine the right people with the right technology and the right timing to make it all work.  Bill Gross had a nice Ted Talk on the topic of "The Single Biggest Reason Startups Succeed".  So convince me you really understand a problem and solution well and that you can be THE company that executes on it better, faster, cheaper than anyone else!
• Truly understand and be able to explain your "Value Proposition" - specifically, clarify (1) Who Pays for your tool, (2) Who Uses your tool, and (3) Who Benefits from your tool.  In healthcare, the incentives are often not aligned - and the smart startup will fully understand and have a business plan that makes sense.  Nothing turns me off quicker than a company that expects a doctor to pay for and use a tool, when all the financial benefit then accrues to another party.  
• Bring me a solution, not just a tool.  A lot of startups are talking about how they use "big data" to identify problems and opportunities for improvement.  That is nice, but the truth is we have a lot of low hanging fruit in healthcare- I don't need to find more problems as much as I need solutions.  So if all you are selling is a way to find more problems, that will not resonate as well as a packaged solution that also "fixes" them.   For example, the analytics vendor  HealthCatalyst is soaring because they realized that they need to use analytics to identify both the problem and the potential solutions to be successful.  Another interesting company, Transfuse Solutions, combines analytics and process improvement techniques to focus on the specific issue of identifying when a hospital is doing too many transfusions and then offering solutions on how to improve on those metrics. 
• Be committed... healthcare is not for the faint of heart.  This is a big business, with long sales times, difficult implementations and hard change management.  When something works and can improve efficiency, quality and financials at the same time - and can scale well... then you will have a winner, but nothing happens overnight like in so many other industries.  So don't tell me how you have a part-time CEO, and you are out-sourcing all your IT work so some guys who have other jobs.  That is not going to build a company which has the DNA needed to succeed in this industry - show me executives and staff that wake up every day obsessed with fixing a specific problem, and an IT team that understands the nuances of healthcare and can react quickly to solve issues. 
• Make it easy to do the right thing, especially if this is doctor facing.  I often say that the best healthcare IT can make life easier for doctors and better for patients at the same time.  Do not try and tell me how "this system only asks the doctor to spend one more minute for each patient" - we don't have one extra minute!   We want you to tell us how you save us time from mundane tasks so we can have more face to face time with patients - that is what will win our hearts and minds!   This post from last year explains this thinking further: http://drlyle.blogspot.com/2014/11/the-three-keys-to-solving-our.html

So yes, please keep innovating and trying to make things better.  Our current healthcare system is clearly not sustainable as it stands, making for a "target rich environment".  But when pitching to busy providers and healthcare organizations, remember that their plate if often very full - so have your value high, your proponents lined up, your story straight, and your team ready to truly make a difference in the lives of both providers and patients.


Addendum: List of Other Relevant Blogs and Advice for Startup Entrepreneurs

• How Famous Start-ups Were Started (ScoopWhoop.com, 6/15)
• Six Reasons Why Digital Healthcare Startups Fail (1/17)

Perspectives on the Future of Healthcare and IT.. a Video Interview

I was recently "Video-interviewed" about my thoughts on the future of healthcare and IT.  These types of interviews are usually quick - two questions, five minutes... hopefully some value!  Here are my two questions and a summary of my answers:

Where is the healthcare industry headed?
I believe healthcare is currently a runaway train with an unsustainable model.  But there is hope if we can adapt reimbursement models to incentivize value over volume, and use HIT to simplif, automate and delegate all the care that needs to be done.  With respect to HIT, since over 80% of physicians have an EMR in place, we now an infrastructure or platform on which to build "EMR Extender Tools" which allow for better EMR functionality, efficiency, and effectiveness.  Furthermore, we need to focus HIT efforts on Population Health, Virtual care, and Workflow Efficiency to meet the increasing demands for care that are upon us.  With respect to population health; ACOs and other types of volume-based to value-based reimbursement changes will make it easier and financially viable to really manage the health population well - but we need the right HIT tools to risk stratify the population and then manage them more easily.  Meanwhile, we should see a rapid expansion of virtual care as technologies and demand sync up. Lastly, as physicians (and staff) are burning out quickly, using HIT to create workflow efficiency by simplifying, automating delegating care, is vital to the performance of doctors, as well as the health of patients (which is why I helped found healthfinch to build software solutions that allow medical groups to redesign care more efficiently and effectively). 

What is an HIT Innovation you would like to see happen soon?
I think we are getting closer and closer to “ubiquitous monitoring.” Wearable devices are available, but right now these are often just used by the “healthy and wealthy.” Although this is a good starting point, there is a need to develop patient monitoring tools that are fully ubiquitous - so that collecting biometric data becomes a simple byproduct of everyday life.  These may start as being embedded in smart phones, and now we are seeing them woven into in clothes, but soon we will have watches, patches and even injected nanotechnologies. As these evolve, doctors will be able to receive regular, real-time monitoring of their patients. From there, one can feed data into a rules engine to notify doctors (or even patients themselves) if something is medically wrong. This portends a fantastic future for remote monitoring so that doctors do not have to rely on patients to manually input data all the time and wait for them to come into the office to explain there is a problem.

The Three Keys to Solving our Healthcare System are Getting SAD, FAT and Innovative!

I was interviewed by a new magazine called Healthcare Innovation News for their September 2014 issue and they asked a lot of interesting questions - so below (a slightly editedversion) to learn about the following: How did my career path wander from engineer to doctor to entrepreneur; Why the future of the EMR rests on innovators; Why healthfinch is called The Doctor Happiness company & why we started with automating the medication refill workflow; and finally - What are the three key components to solving our healthcare system (hint - getting SAD and FAT are two of them!).

Q:   Did you plan to become involved in healthcare technology when you entered medical school? What was the impetus and what does it take for a physician to become a health information technology (HIT) entrepreneur?

A: I liked both medicine and computers growing up, so I studied Biomedical engineering at the University of Pennsylvania, where I wound up programming and working with a variety of PhDs and MDs. When I went to the University of Illinois College of Medicine, I was fortunate to have a mentor in Arthur Elstein, PhD, who had founded the Society for Medical Decision Making and created our med school's initial Informatics Department.  I became his research assistant, worked on a variety of informatics projects and realized that I wanted this to be part of my career. 

Over the years, in addition to being a PCP, I accumulated a diverse set of technology and business experiences, from serving as the Medical Director of IT for a large primary group to starting a consulting company to serving as the Chief Medical Officer for two publicly traded companies in the IT space. Then in 2008 I was able to merge technology and innovation when I received philanthropic funding to start the SzollosiHealthcare Innovation Program at Northwestern, and have been able to learn a whole new skill set of design thinking and methodologies.

Over the years, I also advised a number of startups, but eventually realized that to make the biggest impact (and have the most fun), I should be starting companies myself.  I knew I could bring an interesting variety of real world clinical, IT, Innovation and business skills to certain healthcare problems, but quickly learned that the key to success is always finding great people that can execute on a vision… and I've been very fortunate in finding the right people with which to partner.

For physicians interested in being entrepreneurs, my advice is to try and get a good variety of experiences, and then become part of a team so that you don't have to give up your "day job" of seeing patients.  You will find that keeping your day job will help financially during the early times, will allow you to add in more real world flavor to your company, and will satisfy a part of your soul that only taking care of patients can do.

Q:  Do you still believe that the electronic medical record is dead and that instead, innovation in HIT rests on apps?

A: I do believe that "EMR version 1.0" is dead, in the sense that this early vision of the EMR was mainly focused on putting data into electronic media for legibility and easy access, with an overall bias towards billing and compliance rather than making clinical care easier and better.  And while those were important building blocks, the result has been difficult to use EMR systems, especially in the primary are arena. 

So now we are evolving to EMR 2.0, where the classic EMRs become the platforms upon which innovators can build an amazing variety of apps that fulfill every doctor (and patient's) dreams!  Imagine what might happen if thousands of programmers were able to easily build "EMR Extender Tools" on top of (or within) all the EMRs which are deployed… how many great apps might we see that would never come from the traditional vendors because there are simply too many ideas and not enough time. 

Already we have seen a surge in EMR Extender Tools which provide content, decision support, and analytics programs.  And we are now starting to see apps that focus on truly improving the workflow of care to create both major efficiency and quality improvements. Ideally, future EMR Extender Tools should have enough flexibility to adjust for multiple styles and types of care, whether supporting a Cardiologist in California, a Rheumatologist in Rhode Island, or a Neurosurgeon in Nebraska.. 

We are also seeing that some EMR vendors have started to readily embrace this concept of being an open platform, while others have been slower.  But I suspect all will come along because in the history of technology, a well known truism is that "closed wins early, but open wins late". 

Q:  Why did you decide to focus on medication refills when creating your company healthfinch? How does streamlining the process affect population health?

A: We wanted to start with a workflow which affected the majority of primary care doctors, but one they would happily "give up" because losing it was non-threatening both clinically and financially. The medication refill process was the perfect scenario for us - it is a constant chore for any PCP, with the average doctor getting around 15-20 requests a day, which takes up to 30 minutes of their time to properly manage, and they never get paid for it!

We thus created RefillWizard as a workflow automation tool that intercepts any incoming refill, reviews it in a cloud-based rules engine, routes it to the appropriate person (e.g. RN, Pharm tech, MA) and instructs them on how to handle it.  With this tool supporting care redesign, the doctor then only needs to be involved in the 10-20% of refills where their judgment is truly required.  This made for easy adoption since it actually decreased the amount of work for physicians.  Additionally, we found that that the quality of care could actually improve due both to the speed of answering refills and the consistent use of evidence based rules. 

Our philosophy is actually a very different take on population health.  Rather than focus on computerizing the 20% of care which is most complex, we instead focus on building tools to automate and delegate the 80% of care that is relatively routine. This has two implications for population health. First, since physician time is a HUGE commodity for population health programs, freeing up their time from routine, repeatable tasks allows doctors to take care of a higher volume of patients, focus more on high risk patients, and/or simply catch their breath and keep their sanity. Second, by using the med refill process as a model for how to centralize and standardize certain workflows, we help create the type of team-based infrastructure and culture that will be needed in the future to efficiently manage large populations. Refills first, then other workflows later!

Q:   Why is healthfinch called the “Doctor Happiness” company?

A: In other industries, it has become clear that making the front line workers happy invariably results in better products and service to the customers (think Southwest Airlines, Google, or read Shawn Achor's "The HappinessAdvantage").  Healthcare should learn from these lessons, and we strongly believe that by focusing on creating tools which make doctors happy, the patient will invariable benefit as well.  In fact, a recent article (From Triple to Quadruple Aim: Care of the Patient Requires Care of the Provider) strongly suggests that we should expand the Triple Aim to include physician & staff satisfaction as soon as possible.

We believe in this deeply, and so one of healthfinch's driving mantras is to Save our Primary Care Physicians, and we do this by waking up every day thinking about how to use HIT to make life easier for physicians and better for patients.  In other words, this is not your normal healthcare software company.  Our first goal is to literally improve doctor happiness (without hurting quality of course). And at the same time, we believe we can "sneak in" a variety of quality improvements by helping to ensure consistent use of evidence based guidelines in routine workflows (and ps, doctors are happier when quality goes up as well).

I think that EMRs have gotten a bad rap as always creating more work for physicians.  Of course, that should be no surprise as the traditional EMR vendors are understandably focusing on functionality that is mandated by various regulations (i.e. documentation and billing and MU) over holistic usability.  And that is a problem since no matter how good a function might be, there is no quality without use, and there is no use without usability.  However, I believe that the Golden Era of using HIT to lessen the workload for physicians is just around the corner!  . 

Q:  What are the three key HIT strategies physicians should adopt to improve healthcare delivery?

A: First, we need to optimize our use of HIT to make doctors much more efficient and happier.  I call this Get SAD to Make Doctors Happy:

• Simplify their interaction with EMRs, such as by minimizing their direct hands on experience (e.g. scribes, voice recognition, Google Glass) or using data visualization to help make the cornucopia of data easier to view and understand
• Automate as much of the workflow as possible
• Delegate the things that can be delegated to the appropriate person on their team

Second, we need to better manage large populations of patients. I call this Get FAT to make the Population Healthy:

• Financial incentives need to be aligned around volume rather than value
• Analytics will be used to risk stratify and understand our populations better
• Team based care which will spread the workload appropriately across all the members of the team.

Third we need to devote time and resources to being more innovative. I call this The Three EEEs of  Innovation:

• Explore new technologies and thinking to get some insight and ideas on what you might do and how you want to prioritize
• Experiment by building prototypes and pilots to test new ideas, care models and technologies.  Iterate often until you have both a good clinical use case and business case.
• Expand the innovation to the rest of the organization

If we can do these three things effectively, we will find that we don't really havea shortage of physicians, just a shortage of using them efficiently. And the future will be one where the typical physician sees less patients face to face, but is able to take care of more patients every day using an IT empowered, team-based approach that utilizes everyone to the top of their license and improves the quality and experience of care for both patients and providers.

Social Media and Healthcare

I have not written much on social media in healthcare, which might range from a practice with a facebook site for marketing, to a surgeon tweeting that the gallbladder is out so the family can relax a bit sooner, to a Groupon for reduced botox, or to a system which texts patients to motivate them to eat better or take their meds on time.  But instead of commenting myself, this blog entry will mainly be a list of relevant links, including a list of great bloggers and interesting news stories. 

Healthcare Social Media Bloggers
* 33 charts focuses on social media in health and is written by Bryan Vartabedian, MD.
* http://www.tedeytan.com/ is written by Ted Eytan, MD - an extraordinary thinker and blogger who often writes about the impact of social media and web 2.0 in healthcare.

Stories of Interest
* Could Facebook be your Platform for Care Coordination?  (e-Care Mgt Blog, May, 2011)
* Social media could 'accelerate clinical discovery' (Article about PatientsLikeMe.com, April, 2011)

* Five social media tips for docs worried about HIPAA (April, 2011): Great advice about how to understand that HIPAA actually promotes email and other electronic forms of communication - and is often misunderstood due to paranoid legal beagles!
* What do Physicians Really Think about Social Media?:  A series of blogs by Dr. Ted Eytan based on interviews with doctors from Sermo as well as some Academic sites (Spring, 2011).
* Social media tools may reduce attrition in online health programs... and prove an effective way to boost participation in online health programs, according to researchers at the University of Michigan Medical School (Dec, 2010)
* AMA Guide to Social Media (Nov, 2010): The American Medical Association has adopted a new policy that gives guidance to physicians using social media.
* Tips on mitigating risk of social networking in healthcare organizations (Nov, 2010): like it sounds!
* 6 reasons to manage and archive your social media (May, 2011)
* Facebook friends with your doctor: good medicine or ethically 'icky'? (Aug, 2012) - Advice is to educate in general (eg professional FB or twitter page for education), not create a personal social interaction.

Our Healthcare System: Update

A variety of websites and stories which I found to be important or at least thought-provoking:

Key Web sites
• http://healthcareforamericanow.org/
-- The best site I have found to simply explain, "What does the new health reform law mean for YOU?"

• http://www.healthcare.gov/
-- The federal government's site that includes specific advice on how to find health insurance and how that is impacted by the new health reform law.


Healthcare IT stories
• Use of HIT Improves the Quality of Care
-- A Kaiser Permanente Study Finds Quality of Care Scores Increase as Patients and Physicians Communicate via Secure E-mail.

• Improving Usability of Health IT for Physicians
-- A great article in Healthcare Informatics which starts by pointing out that most health professionals do not use available health IT systems because they actually increase their work effort and can too frequently hurt quality, whereas the objective should be to help physicians decrease their work effort while increasing their quality consistently. They offer thoughts on a "physician-specific point-of-care system that continuously adapts to practice patterns that could result in dramatic improvements to the quality and efficiency of healthcare delivery".


General Healthcare Stories
• Process improvement to improve compliance with specialty visits
-- Turns out that when a PCP refers a patient to a specialist, they only make the appointment 70% of the time, and of those - only 70% show up - thus less than 50% of people go to the specialists when they are referred! This article talks about how a process improvement improved those metrics. Our medical group (www.NMPG.com) does something similar to help with this process and we believe it provides a higher quality and more efficient process for sure!

• Better ways to manage the flood of test results
-- New recommendations target how physicians and hospitals can best communicate test results and prevent harm to patients.

• Aftercare Tips for Patients Checking Out of the Hospital
-- NY Times article on how good discharge planning can keep patients from needing to be re-admitted after leaving a hospital, and could save Medicare billions.

• How the Performance of the U.S. Health Care System Compares Internationally (2010 Update)
-- Yet another report, placing the US healthcare system last among industrialized nations. US spends $7,300 per person per year on healthcare and gets the worst results. UK spends $3,000, New Zealand $2,500; Canada $3,900; Australia $3,400.

• Health overhaul may mean longer ER waits, crowding
-- Due to a shortage of primary care physicians (PCPs), Emergency Rooms may grow even more crowded with longer wait times under the nation's new health law since there will be many more patients with insurance, but no increase in PCPs.

• The Variability of Patient Care - by John Glaser
-- One of the smartest guys in healthcare explains the theory from one of my favorite books (Designing Care by Richard Bohmer),which I talked about in a previous post about Checklists and process improvement. The key point being that there are two classes of care in a hospital and in a physician's practice, and the importance of understanding that these two very diverse scenarios need to be recognized when designing process/workflows for care (especially including use of EHRs). Glaser explains further;
---- Sequential care is a form of production: It involves performing well-understood tasks in a well-understood sequence (e.g. routine heart surgery). Sequential care's mental image is that of a production line. With sequential care it is possible to engineer a preferred sequence of steps and have the EHR guide the care team in performing these steps. And it should be quite possible to measure the outcomes of these steps. (This is similar to Clay Christensen's Value Added Process)
---- Iterative care is a form of discovery: It addresses complex diagnoses and conditions for which the diagnosis and treatment are a repeating series of hypothesis-test/treat-revise hypothesis steps. Iterative care is different. The mental image should not be the factory floor but a group of scientists in the laboratory. In this scenario we must encourage collaboration, enable an unpredictable set of actions to be taken, and provide easy access to information and other experts that might help the team form and test hypotheses. Measuring the outcome of discovery is very difficult. (This is similar to Clay Christensen's Solution Shops)

The Dawning of the EMR as a Platform...allowing us to "get the health care that we build"

Joe Flower continues to be one of my favorite healthcare writers. In a recent article in HHN Online, he talks about the heroes in healthcare who are constantly trying to improve the system. I love the quote on which he ends his article:

As Aristotle famously shaped it, "We are what we repeatedly do. Excellence, then, is not an act, but a habit." We will not get the health care that we want. We will not get the health care that we deserve. We will get the health care that we settle for. We will get the health care that we build, where we are, with the tools that we have, with the courage and compassion and collaboration and hard insistence on excellence that lies within us.

At the recent HIMSS conference, I think we began to really see the first signs of an important paradigm shift in the EMR world which will help make this ability to build a better healthcare system more feasible. Specifically, we saw the rise of the "Ecosystem" or "Platform" - terms which will become the buzz word of the coming year as vendors are starting to "open" up their systems (e.g. via APIs, or other technical and business transparency).

Stepping back, the historical scenario for an EMR vendor is to sell you all three tiers (database level, application/functionality level, and user interface level) as a tightly integrated unit. The upside is they should all work well together, the downside is minimal ability to customize one layer without having to get involved with the other layer because they were so tightly linked. For example, if you wanted to display vital signs in a different way in your user interface - you would also have to change the underlying data model and application abilities. We can refer to this as the "Tyranny of the Three-Tier Architecture".

Unfortunately, what we have seen are quite bad user interfaces from the EMR vendors and minimal ability for real life users to improve upon them. The result has been poor adoption of EMR systems, as well as multiple instances of "unintended consequences" from poorly defined user interfaces. Fortunately, the EMR vendors must have realized this was the case (or they are getting spooked by the new crop of HIE vendors and system integrators who are trying to take their data and allow for more customized user interfaces).

So at HIMSS, I found that many EMR vendors are now allowing at least some ability for users or third parties to create new widgets and user interfaces to "put on top of" their EMRs. We are still pretty early in this phase, but eventually- the hope is that this will become analogous to Apple creating the "iPhone Platform": The EMR vendors will ideally compete to create the best platform which will then allow for some true innovation at both the application and presentation layers, or alternatively brand new vendors will come along to create platforms which take what is needed from legacy systems while allowing for others to build on top of them in a unified environment (e.g. GE's new Qualibria). Either way... the ideal result will be an Ecosystem where we can indeed Build the Healthcare System we need and deserve.