Sunday, October 27, 2013

Designing the Intro to an Innovation Curriculum for Medical Residents

[ ] Indicates language that should be verbally communicated to the audience or an exercise that should be performed

Target Audience: medical trainees interested in innovation

Course to Proceed as Follows:
Step 1:
[Ask residents the following to understand their pain-points and level of understanding:
-What do you want to get out of this experience?
-What do you know about value-based care?
-What do you know about innovation?]

Assuming they are not opposed to learning about skills that will allow them to excel in their careers a clinician leaders (not necessarily start companies or work in the private sector), then proceed to Step 2.

Step 2:
Provide bridge from clinical outcomes-oriented thinking to value-oriented thinking on a level that medical trainees understand: inspiring leaders

Simon Sinek: How great leaders inspire action
[Show Video of Simon Simonek] 

[Summarize Simon’s talk and apply to the perspective of a trainee:

Medical trainees at top-tier Academic Medical Centers (AMCs) commonly aspire to be leaders in healthcare. Although Simon Simonek uses a lot of business jargon, trainees can likely emphasize with his central thesis: “Great leaders inspire action by why they do not by what they do.”

Simon Simonek elaborates on his thesis of inspiring leadership through his idea of the Golden Circle.]


[Exercise: Have residents complete the Golden Circle for their own careers]

One way a medical trainee could fill in the Golden Circle with their own experience is:

What: delivery excellent clinical care, publish lots of papers
How: learn the most clinical knowledge and practice at the most prestigious institutions; position for best ability to get research funding via K-award, R-award, etc
Why: Eliminate suffering, Eliminate social injustice

[Provide the following interpretation:

With the rapidly evolving health care system, the current "how" and “what” or career path may not help medical trainees achieve the full potential of their impact.

More specifically, the health care system is rapidly changing from volume-based care to value-based care. There is decreasing incentive for scientific discoveries that lead to incremental improvement in outcomes (3% increase in 5 year mortality rate) at an enormous cost (billions of R&D budget for a drug). There is a shift away from fee-for-service reimbursement and toward managed care. Readmission penalties are climbing to 2% in 2014 and 3% in 2015. Hospital revenue will be plummeting 5-15% over the next several years as care is redirected to the community. (References available upon request)

The bottom line is that our livelihood and ability to impact change are at risk with the current limited skills that we are acquiring in clinical training. In order to survive and thrive in the future health care delivery and reimbursement system, medical trainees need to be familiar if not expert in value-based thinking.]

Step 3: Define Value
Value = outcomes/cost [Porter]

*of note, blind cost containment does not offer real value to patients, just fudges the equation

Triple Aim [Berwick]

(From IHI, 2009)

Step 4: Building the Lexicon: Tools and Methods for Describing and Delivering Value
[Verbally summarize:

To preface this section, it’s important to note that the following content and exercises may help you in the long term in your career, but in the short term may also help residents identify projects for Academic Development in residency or Fellowship Projects.

In order to achieve the triple aim, residents should be familiar with a combination of improvement science and innovation methods.

Improvement Science
Improvement science methods, such as Quality Improvement, provide a skill set that enables the clinician to measure and impact outcomes on a real-time basis. QI offers the opportunity for a clinician to impact at least one of the triple aims, improving population health.]

Methods in Innovation
[Define and describe steps of Lean Startup Thinking and Design Thinking:

Other approaches to delivery innovation facilitate not just practice improvement, but rather the creation of entrepreneurial, potentially high-risk new products or processes driven by consumer adoption.

Lean Startup Thinking
Lean Startup Thinking is a management process that aims to create a minimal viable product (MVP), which is an intervention with the fewest features needed to get an end-user to pay with a scarce resource such as money, time, or attention.[Reis] Similar to the Plan-Do-Study-Act cycle of Quality Improvement [Langley], the Build-Measure-Learn cycle of LST is a continuous process of building a prototype, testing and measuring the impact of the intervention, and learning from the results of testing to inform the refinement of the prototype. [Figure 3] This cycle is rapidly repeated until there is sufficient validation of an MVP or until there is sufficient invalidation to change course or pivot toward another MVP.


(From Ostrovsky A & Barnett, 2013 in review)

Although traditional implementation science and LST share the same origins of Lean Management, Six Sigma, and the scientific method, the outcome being tested in LST is not a statistically significant difference between two interventions or an improvement in clinical outcomes. Rather, LST focuses on the confirmation that a patient, provider, or payer is willing to pay for a new product or service.[Reis] The benefit of using LST as the foundational methodology for incubation at AMCs is its emphasis on end-user validation, which creates consumer-driven rather than clinically driven incentives to prioritize patient-centeredness, minimize costs, and optimize clinical outcomes. The alignment of these incentives and the inherent for-profit orientation of LST works towards generating sustainability through generating revenue. In contrast, traditional implementation science approaches may produce or discover interventions that lead to improved clinical outcomes, but if end-users do not perceive value in those outcomes, then the interventions risk never being utilized or the cost of those interventions may go unchecked. The end goal of LST is the creation of a product or service that builds on the three pillars of quality improvement [Berwick] to achieve the “quadruple” aim: improve patient satisfaction, decrease costs, improve care quality, and be commercially self-sustainability.

Design Thinking
Unlike LST, which is rigorous iterative process of testing that converges on a single solution, DT is a divergent process that combines empathy and creativity to generate insights into the deep-seeded problems faced by end-users.[Kelly] The design process can be outlined in 5 steps: 1) empathizing with the end-user to understand the root of their problem, 2) defining that problem, 3) thinking outside the box for ways to solve the problem, 4) building a prototype of the solution, and 5) testing the ability of the solution to solve the original problem. [Plattner] [Figure 4] Design Thinking adds granularity and a deeper understanding of the end-user’s context, and it complements LST in discovering solutions that fit the unique needs of the end-user. With empathy as the central tenet underlying both clinical practice and DT, clinician-innovators are uniquely qualified to be Design Thinkers.]


(From Ostrovsky A & Barnett, 2013 in review)


Explore problems faced by medical trainees by having trainee present a problem in front of group or have participants document their problems on post-it notes. Have facilitators review the problems/pain-points as a group and dissect the problem into its component parts (empathize). Then go through the Design Thinking process to come up with potential prototype solution to the problem.]


Berwick D, Nolan TW, & Whittington J. The Triple Aim: Care, Health, and Cost. Health Affairs. May 2008: 27(3). 759-769.

Porter ME. What Is Value in Health Care? N Engl J Med 2010; 363:2477-2481.

Reis E. The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses. 2012

Langley et al. “The Improvement Guide: A practical approach to enhancing organizational performance.” Jossey-Bass Publishers. 2009.

Kelly T. The Art of Innovation: Lessons in Creativity from IDEO, America's Leading Design Firm. Random House. 2007.

Plattner H. The Stanford Approach to Design Thinking. 2013. Institute of Design at Stanford.

p.s. Great quotes from Simonek talk:

Why matters more than what you do:
“Apple- everything we do we believe in challenging the status quo and thinking differently; user friendly products; computers”

“Wright brothers- Change the course of the world by redefining boundaries between man and bird; new way of travel, commerce, communication; airplane”

“TIVO- if you’re the kind of person that likes to have total control over every aspect of your life, boy to do we have a product for you”

“MLK- not until all laws created by man were aligned with laws of authority”

“[MLK] gave the “I have a Dream” speech not the “I have a plan” speech”

Now apply the “why matters more than what” to a career as a clinician innovator. 

Sunday, October 13, 2013

Eliminating Health Disparities by Filling the Digital Health Investor Knowledge Gap

Digital health startups secured $1.4 billion in investment in 2012 with healthcare administration, personal-health, and consumer-engagement emerging as some of the leading investment categories.[1] Despite the growing interest in healthcare, early-stage investors’ reluctance to invest in unfamiliar industries[2] has limited funding for more nascent digital health opportunities like managed long-term supports and services (MLTSS).[3]

Managed long-term supports and services entail risk-bearing organizations, like health insurers, coordinating care for frail elderly that are vulnerable to experiencing health disparities.[4] The goal of MLTSS is to triage the patient to the lowest cost care and maintain sufficient functional supports to keep the patient in the community rather than in the institutional setting.

Between the evolving healthcare regulatory environment, clinically complex patients, and dispersed post-acute care market, the multifactorial nature of MLTSS makes for an insurmountably steep investor learning curve. The deficit in the average investor’s knowledge base about MLTSS not only precludes investment consideration, but more importantly stymies the funding channels needed to fuel innovation for vulnerable populations. Furthermore, the digital health technologies currently garnering the most investment do not emphasize the elimination of health disparities for vulnerable populations[3]; in fact some may be at risk of exacerbating disparities.[5]

Filling the gap in investor understanding about the market opportunity and emerging business models for digital health startups in MLTSS could help more investors realize the potential for compelling ROI in these companies and simultaneously achieve the Triple Aim[6] for vulnerable populations.


Source: IHI Innovation Series. Cambridge, MA. 2012.

Market Opportunity

The market opportunity for MLTSS is grounded in the substantial potential for payers to profit from managing complex patients. Typical margins for payers managing generally healthy patients reimbursed by Medicaid range between 2-5%. Higher-risk patients, particularly those that are Medicare-Medicaid Dually Eligible (Duals), can lead to margins as high as 10-15%, if managed effectively. Conversely, some Duals can cost $60,000 per year.[7] And when payers fail to risk stratify appropriately and only negotiate a capitated $2,500 per member per month (PMPM) premium ($25,000 per year), they can very quickly lose a lot of money.

The profit potential from managing Duals arises from their disproportionately high utilization of acute care hospitals, with almost twice the hospitalization rate for Duals than general Medicare non-Dual patients.[8] Duals are also more likely than non-dual Medicare patients to get readmitted, with 25% of Duals and 20% of Medicare non-Duals being readmitted within 30-days of discharge.[9]

In addition to being higher utilizers, Duals also account for a disproportionate share of Medicare and Medicaid spending: Duals comprise 20% of the population and 31% of spending for Medicare and 15% of the population and 40% of the spending for Medicaid.[10] 

Source: Jacobson G, Neuman T, & Damico A. Medicare’s Role for Dual Eligible Beneficiaries.  APRIL 2012. Kaiser Foundation.
While 19% of the average PMPM premium for Duals is spent on hospitalization costs, Duals are also high utilizers of post-acute care services with 23% of the PMPM premium being spent on Home Care and 33% on Skilled Nursing Facilities (SNFs). Even though acute care is one of the largest contributors to Duals spending, post-acute care is the fastest growing cost.[11] Within post-acute care, institutional long-term care, such as SNF care, is $1614 more expensive PMPM than home and community based services.[12] 

Not only does MLTSS decrease cost of care, a review of 9 randomized controlled trials showed that integrated and coordinated care improves outcomes and reduces health care utilization for frail elderly people.[13] This creates an enormous cost differential for payers to take advantage of, particularly when up to 30% of “low care” SNF patients could age in their homes instead of the SNF if they had the right resources or technology in place.[14] 

With the MLTSS market doubling since 2004 [15], digital health has enormous potential to reduce hospitalizations and unnecessary utilization of expensive post-acute care by streamlining, augmenting, and scaling MLTSS.

Emerging Business Models

In addition to outlining the market opportunity for digital health in MLTSS, preparing investors for more complex business models may help them become more comfortable investing in this space. A major challenge to creating scalable business models for MLTSS digital health companies is the limited purchasing power of the end beneficiary, the patient. These patients are typically financially compromised in addition to being clinically high-risk. The confluence of socioeconomic and medical risk factors limit the ability of Duals and other vulnerable populations to purchase digital health technology directly. Additionally, their technological illiteracy due to language or age barriers introduces more challenges to freemium or ad-based models.

A promising business model for MLTSS digital health startups is selling directly into risk-bearing organizations that have the most to gain from optimizing care coordination and care management for Duals. Several policy instruments have created incentives for such organizations to keep patients healthy and in their homes. These organizations, typically payers, have substantial purchasing power and can drive scale of a technology once it is adopted. A major limitation with a B2B sales strategy with payers is the long sales cycle. But the latter is a surmountable barrier, especially when investment dollars provide a startup in this space with the runway to test their product and business model a sufficient number of times to find the approach that is scalable.


By educating investors about the opportunities that lie in the MLTSS digital health space, we hope to facilitate more informed investment into a healthcare space that is pregnant with opportunity to create profit for investors and achieve the Triple Aim for vulnerable populations.

Other stakeholders in the startup ecosystem can help to eliminate risk for investors in MLTSS and other spaces that leverage digital health to eliminate health disparities. Incubators can vet startups for top entrepreneurs and provide a conduit for startups to difficult to reach resources such as development talent or legal guidance. Foundations can help to further educate investors and establish best practices for investment in digital health to serve vulnerable populations. The government can continue to liberate data and push for rational interoperability standards. This glimpse into socially-minded digital health innovation and investment is just one of many opportunities to leverage commercial growth to eliminate health disparities.


1) Tecco H. 2013 Midyear Digital Health Funding Update. Rock Health. 2013. < >
2) Feeney et al. Private investors' investment criteria: Insights from qualitative data. Venture Capital: An International Journal of Entrepreneurial Finance. 1999, Vol 1(2). 121-5.
3) Startup Health, IHI, & Ostrovsky A, et al. Characteristics of Top 100 Funded Startups. Startup Health Insights. 2013. [Unpublished]
4) Medicaid. Managed Long Term Supports and Services. < >
5) CPEHN. Equity in the Digital Age: How Health Information Technology Can Reduce Disparities. 2013. < >
6) Berwick D, et al. Health Aff May 2008 vol. 27 no. 3759-769.
7) Young K, et al. Medicaid and the Uninsured: Medicaid’s Role for Dual Eligible Beneficiaries. April 2012. Kaiser Foundation.
8) Segal M. CMS Policy Brief: Dual Eligible Beneficiaries and Potentially Avoidable Hospitalizations. 2011.
9) Jencks SF, et al. NEJM 2009; 360 (14):1418‐1428.
10) CMS. A Data Book: Health Care Spending and the Medicare Program. 2012.
11)Chandra et al. Large Increases In Spending On Postacute Care. Health Aff. 2013 May; 32(5): 864–872.
12) Milligan C. Innovations in Integrated Care. February 23, 2011. Hilltop Institute
13) Eklund K, Wilhelmson K. Outcomes of coordinated and integrated interventions targeting frail elderly people. Health Soc Care Community. 2009 Sep;17(5):447-58.
14) Grabowsk et al. Prospects For Transferring Nursing Home Residents To The Community. Health Affairs, 26, no.6 (2007):1762-1771
15) The Growth of Managed Long-Term Services and Supports (MLTSS) Programs: A 2012 Update. Paul Saucier, Jessica Kasten, Brian Burwell, Lisa Gold. July 2012

Thursday, October 3, 2013

Have a rash, consult Derm; Have a UX/UI problem, consult the Design Team

I have the privilege of training at one of the most advanced, cutting-edge pediatric medical centers in the world. We have the most #1-ranked sub-specialty services in the country according to US News and World Report. As if having so many top-ranked sub-specialties wasn’t enough to be a center of clinical excellence, Boston Children’s has a particular niche because of its abundance of sub-subspecialists. I have consulted pediatric metabolic neurologists, cardio-geneticists, and neonatal pulmonologists. We have specialists in syndromes with acronyms that are common vernacular within the hallowed walls of BCH, but are usually perceived as alphabet soup in the real world outside of Children’s. And yet, I am continuously surprised to see that we (and most other medical centers in the U.S) lack a basic specialty that could be transformative in achieving the Triple Aim and increasing the value of care we are delivering. I hypothesize that large academic medical centers could dramatically improve clinical outcomes, decrease costs, and improve patient satisfaction if they had a “Design” consult team.

The Design service would be similar to most other consultation services whereby experts in a sub-specialty would be paged on an as needed basis to help the primary inpatient team answer questions outside of their scope of practice. Just like dermatologists are consulted to inquire about a perplexing rashes or neurologists are consulted about unusual palsies, designers could be consulted when there is a user-experience glitch that is impeding optimal care delivery.

I had an experience today that highlighted the need for a Design consult. I was working in an outpatient sub-specialty clinic and observed one physician attempt to train another physician on compliance requirements for meaningful use through our EMR. Both the trainer and trainee were frustrated. Clinic workflow was interrupted. Patient care was delayed. Although no patient was directly harmed, the experience left all parties involved dissatisfied. If this level of dysfunction can occur at the #2 children’s hospital in the U.S., I am afraid of the challenges that may be faced by other care delivery systems when implementing meaningful use. Design consult, stat!

If there was a Design team available by page, we could have consulted them to at least assess the problem. Like many sub-specialty services, there are many tools at the specialist's disposal to assess and manage a consult, but I propose we start with the basics; the Design consultants could have come to the clinic and gone through a very quick design exercise including the following steps: 

The designers could have observed the frustrating workflow inefficiency of an overqualified trainer disrupting the busy trainee in their care delivery (empathize). They could then define a major problem by eliciting the frustration from the trainee with an imposed new workflow that they do not perceive as a benefit to the provider or the patient (define). The designers could then empower the involved parties to brainstorm what an ideal scenario for training may look like, such as a no-pressure, reimbursed training experience with ample opportunity to practice and education about the intended goal of the meaningful use intervention (ideate). This could be followed by implementing a very small sub-component of the ideal scenario, such as having the trainer and trainee both look up the intention behind meaningful use certification on EMRs (prototype). By quickly searching the internal or external web resources, the providers could quickly discover the published justification for meaningful use which includes evidence suggesting fewer medication errors and better patient engagement when certain elements of meaningful use are implemented. The impact and learning from the prototype intervention could then be used to inform future meaningful use training experiences to improve workflow and provider satisfaction (test).

The benefit of the aforementioned design consult is a potential improvement in patient or provider experience, or at least identification and codification of the core elements of frustration. Defining the problems allows for measurement of the problem and measurement enables improvement.

The challenge that the design consult creates is an additional time burden on the provider team in order to provide the requisite information for the designers. But this is not much different from the primary clinicians providing a background history and rationale for their consult to a dermatology or neurology consultant.

Similar to other consult services, the Design team could complete part of its assessment and/or intervention immediately or inpatient and then ”follow up outpatient” for more long-term or in-depth design work on the problem requiring consultation.

Unlike most other consult services, the Design team is a generalist service that could be called upon to help with provider workflow issues, patient satisfaction issues, management challenges, branding concerns, or any other non-clinical issue whereby the problems may not be perfectly clear and the solutions to those problems are even more nebulous.

The ideal composition of the Design consult service could mirror traditional teaching-hospital consult services: a full-time designer experienced in healthcare, a full-time clinician (does NOT have to be an MD) experienced in design, couple of fellows, and some med students. And there you have a swat team that will guide patients and providers through their user experience challenges, potentially propose some solutions, and even prototype in real time with follow-up “outpatient."

Having a dedicated design team within a care delivery system is not a novel idea; Kaiser Permanente and Mayo Clinic have fully fledged design armies dedicated to solving large and small challenges within and outside of their institutions’ walls. Both are top-ranking healthcare delivery systems, which begs the question when U.S. News and World Report will develop a measure for ranking sub-specialty programs in healthcare design services!

Wednesday, October 2, 2013

Digital Health Startup Entrepreneur’s Cheat Sheet: #CustomerDevelopment #LeanStartup #InnovationAccounting #DesignThinking #PirateMetrics

As a physician, I carry around cheat sheets to condense critical information into an easily accessible stack of pocket-sized, laminated tri-folds. They include important numbers to call, medications to use in a code-blue situation, and antibiograms to help select the right drug for the right bug. This is all information I know how to use but may not be able to regurgitate off the top of my head, especially in life or death situations.

As an entrepreneur, there are no such cheat sheets. There’s no pamphlet or app that condenses the essentials of Customer Development by Cooper and Vlaskovitz, Lean Startup by Reis, Innovation Accounting by Maurya, Design Thinking by McCarthy and Berkowitz, and Pirate Metrics by McClure. Although a startup may not be as tenuous as a critical airway, easily accessible information on startup techniques can help an entrepreneur focus on the process rather than on the technicalities of entrepreneurship.

I heavily borrowed from the above mentioned thought leaders in entrepreneurship to create the following cheat sheet.

The x-axis of this cheat sheet is oriented to the stages of customer development. On the left is the beginning of the value discovery process: problem/solution fit. On the right is scale, every entrepreneur's goal. Of note, I switched the wording order of product/market fit to market/product fit to show an evolution from problem to market and solution to product.

Just above the axis are a series of build-measure-learn cycles from Lean Startup. The build-measure-learn cycles are only valuable in the context of the type of risk they are designed to test and eliminate, which is shown above the build-measure-learn cycles in the risk-matrix.

Based on Maurya’s "Running Lean" and "Innovation Accounting" literature, the three major risks to the business model (product, market, customer) are outlined with their respective “critical questions.” The questions along each type of business model risk are organized based on a progressive sequence of steps including understanding the problem, defining the solution, and qualitatively and quantitatively validating the elements of the business model. Due to space limitations, I did not incorporate the lean canvas, which itself is a very powerful tool. These elements of innovation accounting have corollaries to elements of design thinking such as "empathize", "ideate", etc (italicized above).

As build-measure-learn cycles are completed, the questions about business model risk in the innovation accounting matrix increasingly become answered. The learning from these experiments helps to inform best practices for optimizing the conversion funnel shown at the top of the image.

An important note about this presentation of the conversion funnel is that in enterprise level sales in healthcare, I have found that the “retention” and “revenue” sections appear transposed. When selling into payers or providers, the customer is often retained first for a pilot then they pay some time later. And referrals only come later in the sales cycle.

As startup teams go through build-measure-learn loops to conduct experiments that eliminate risk to the business model and inform the conversion funnel, they will hopefully efficiently move toward the right side of the x-axis.

Moving beyond cheat sheets, startup entrepreneurs would greatly benefit from a set of validated, peer-reviewed best practices in digital health innovation to guide them to creating the next disruptive technology.