Dispatch from London

I'm in London for 48 hours, working with a group of international experts to define telehealth, care management, and big data opportunities for the UK, Europe, Australia, and US.

During the afternoon break I had a remarkable experience.

Unity Farm apples (40+ heirloom varietals) produce a crisp, well-balanced fermented cider that includes sweet, tart, aromatic, and astringent components.

Kathy, my wife, searched the web for the best cider pub in London and asked me to stop by so I could compare Unity Farm cider making with traditional farmhouse ciders from the UK.

The Cider Tap, a remarkable place, is, by good karma, 100 yards from my hotel in Euston Square.

Professor Justin Beilby, Executive Dean of the Faculty of Health Sciences at the University of Adelaide, and I  tasted six still and four sparking ciders, fresh from the cask.

Still:
1.  Severn Perry (made from crushed pears) - 6.3% alcohol, aromatic, dry, with a subtle pear flavor.

2.  Sandford Bumbleberry - 4% alcohol, sweet with a mixture of herbs and berries.    It reminded me of a traditional spiced mead - a metheglin.

3.  Bleangawney - 6% alcohol, dry, with an almost lime-like flavor.

4.   Upper House Farm Oak Barrel aged - 6.5% alcohol, medium/sweet with a clean, crisp flavor.   This was my favorite still cider and a style that I've tried to replicate at Unity Farm.

5.   Severn Farmhouse - 6.2% alcohol, medium, a classic full bodied english farmhouse cider

6.  Burrow Hill Alf n Alf - 6.0% alcohol  with a very complex taste that leads me to think it is half dry cider/half medium cider.


Sparkling:
1. Aspalls Harry Sparrow  - 4.6% alcohol, dry, crisp and clean.  This was my favorite sparking cider, a true scrumpy

2. Lilley's Stargazer -  5% alcohol, medium/sweet, with a great apple flavor

3. Sheppy's Oakwood - 4.8% alcohol, medium/dry without any overtones of oak, despite the name

4. Somerset draught - 5.5% alcohol, medium, well balanced and complex

The proprietress of the Cider Tap explained that the still ciders change weekly, with new fresh products produced in small quantities from local farms.

Tasting living, complex, handcrafted ciders made with centuries of experience was the highlight of my trip.    I only wish we had the local cider tradition close to Boston.    I guess it is up to Unity Farm to bring handcrafted ciders to the Metrowest!

Lessons Learned from the Health Insurance Exchange Launch

CIOs face many pressures - increase scope, reduce timelines, trim budgets.     After nearly 20 years as a CIO, I've learned a great deal about project success factors.

When faced with go live pressures, I tell my staff the following:

"If you go live months late when you're ready, no one will ever remember.

If you go live on time, when you're not ready, no one will ever forget."

I have hundreds of live clinical applications.    Does anyone remember their go live date?  Nope.

Were there delays in go live dates?  Many.

With even the best people, best planning, and appropriate budgets, large, complex projects encounter issues imposed by external factors (new regulations, competing unplanned events, requirements changes) that cannot be predicated during initial project scheduling.

It helps no one - the users, the business owners, or the IT department to slavishly adhere to a deadline when the project is not ready to go live.

I work on federal advisory committees in the Obama administration and truly believe in the goals of many administration programs - Meaningful Use, HIPAA Omnibus rule, and Affordable Care Act.

However, we've seen that in the interest of accelerating change, deadlines have been imposed that do not allow for sufficient testing, piloting and cultural change.  The result is that haste makes waste.

As I've written in my blog many times, ICD-10 will become a crisis for the Obama administration.   Payers and providers will not be ready by October 1, 2014.   Documentation systems and clinician billing process changes will not be mature enough to support a successful go live.   More time is needed.    My experience with IT crises is that you can survive one at a time, but a succession of problems creates a pattern that users and oversight bodies will no longer tolerate.   I hope the premature go live of the Health Insurance Exchange results in a review of ICD-10 go live dates.

Meaningful Use Stage 2 attestation criteria are good.   The certification scripts need very significant revision.   How did this happen?   They were created in a rush to adhere to an artificial deadline, not reviewed by the federal advisory committees, and not piloted tested/revised.     New regulation is needed fix them and that will take time.   Again, the lessons of the Health Insurance Exchange should cause us to extend Meaningful Use Stage 2 deadlines by a year, deferring future stages of Meaningful Use until we have consolidated our gains and understood our successes/failures with current stages.

The Office of Civil Rights is an important watchdog of patient privacy.   We all believe respecting patient privacy is one of our most sacred responsibilities.   However, at times government auditors have enforced policy for which the technology and infrastructure of the country was not ready.    Yesterday I received an email from Harvard Medical School noting that the laptop encryption software installed a few years ago was deemed too error prone and too hard to support so it would be retired.   Luckily in 2013, encryption is natively supported in current releases of Mac OSX, iOS, Android, and Windows.   The industry is ready to support robust device encryption now.    However, enforcement/breach penalties related to encryption on mobile devices started years ago when products were as stable as the Health Insurance Exchange.   We should have aligned enforcement with product maturity in the marketplace.   Similarly the HIPAA Omnibus Rule contains provisions like the self-pay redaction requirement that no hospital has figured out how to support.  However, enforcement is starting now.

Do we a see a pattern here?   Policies are good.   Policymakers are well meaning.  Timelines are set in such a way that none of these activities - Health Insurance Exchange, ICD-10, Meaningful Use Stage 2, or HIPAA Omnibus Rule have enough time for testing, piloting, and cultural change.

As I've written about previous in my post the Toad and the Snake, I'm not yet at that time in my life when I resist change or innovation.   I'm simply an IT leader and physician in the trenches who knows that 9 women cannot create a baby in a month.   There is a minimum gestation period for IT projects and our policymakers should learn from the lessons of the Health Insurance Exchange and re-calibrate the timelines shown in the graphic above so that everyone is successful.

Building Unity Farm - Fall Hoop House Planting

It's Fall in New England and the weather is turning cold.  Nights are in the 30's and days are in the 50's and 60's.   All the ferns in the forest are brown and most of the insects are gone.   On the farm, the apples are harvested, cider made, mushrooms dried, paddocks/pastures readied for winter, and the pace of harvest-related food preservation projects is slowing down.

We now turn our attention to Fall and Winter plantings.  Our 48x21 foot hoop house heats to 80 degrees F during the day by trapping solar energy under a 6 millimeter roof of UV resistant plastic sheeting.  We use barn fans to circulate air and hand cranked rollers to open the sides and prevent overheating.    The roof has a "gothic" cathedral shape which sheds snow and resists wind.

Over the summer, we built fifteen 4x8x1 foot raised beds and a gardening bench.   We brought water and electricity to the hoop house via a 200 foot trench from the house.   We placed a foot deep foundation of alpaca manure under each raised bed and now we're filling the beds with compost, moss, and perlite http://en.wikipedia.org/wiki/Perlite

In our first 2 beds we planted 200 bulbs of garlic, which will overwinter and produce new bulbs in the spring.   We've grown garlic for many years and enjoy bulbs of oven roasted garlic brushed with olive oil.

In our second 2 beds we planted romaine and oak leaf lettuces.   The heat of the hoop house should enable us to pick fresh greens every day during the winter.

In our remaining beds, we'll plant additional lettuces, kale, spinach, chard and other cold tolerant plants, keeping our kitchen and our barnyard stocked with fresh greens.   The ducks and the alpaca really enjoy a fresh head of romaine.   For the ducks, we chop the lettuce and mix it with water, creating a soup which they can easily slurp.

As the days shorten, we'll have less light to work in the hoop house, so I'll add 2 pendant galvanized barn lights http://www.barnlightelectric.com/pendant-lighting/barn-pendant-lights/the-original-stem-mount-pendant.html which will enable us to pick fresh salad greens in the evening after the work day.

It's our goal to become increasingly self-reliant over the next year as well as sell many of our products - honey, mushrooms, apples, blueberries, and vegetables at local farmers markets.   Although we've been farmers for a year, this will be our first winter with a hoop house, so I'm sure there will be many lessons learned growing vegetables as the snow begins to fall.

Reflections on My Trip to Asia

For 12 days, I served as guest professor in China, Japan and Taiwan, giving lectures, running meetings, and joining my colleagues for ceremonial meals.    In many ways, all three countries are trying to solve the same basic technology and policy challenges, but there are subtle differences.

China - China has a single payer system with universal, nationally funded healthcare in state operated hospitals and clinics.  Privately funded, entrepreneurial ventures  including high-end hospitals and clinics are emerging for those who want to purchase concierge care.   This article nicely summarizes the economic issues.

During this visit I had the opportunity to meet with Vice Minister of Health for China, as well as several hospital leaders, and informatics professionals.   Here's what I learned, although I will qualify my impressions with the fact that China is a large and diverse country and my visit was limited to Beijing, Hangzhou, and Shanghai.

There are over 500 vendors of EHR products and no market leader.   Many applications are home built or created by small, local companies which address the specific workflow needs of a single healthcare facility.    There is not a specific national healthcare IT policy, but there is recognition of the need for national interoperability standards and an incentive to use them.    Something like a certification program and meaningful use program may evolve.    There are many proprietary approaches to interoperability currently in use, and HL7 CDA is seen as a possible candidate for summary exchange.   Vocabularies such as SNOMED-CT, ICD9 or 10, and LOINC are not yet deployed but there is an understanding of the need for terminology services and an eagerness to work with US companies proving such services.  There are  pilots of healthcare information exchange, generally using a central repository model.   I asked an audience of 1000 people if they had ever used a PHR, and not a single hand was raised.   There is a desire to engage patients and families but no products in the marketplace yet.

The challenges faced by China include a population of 1.35 billion people, environmental concerns (air/water/land), an aging population, a significant migration of citizens from rural to city life,  and an increase in cancer/birth defects/respiratory diseases as a side of effect of rapid industrialization.

I look forward to next steps, which I hope include interoperability policy planning at the national level.  I've forwarded several US policy documents to the Chinese government and remain eager to work on Meaningful Use for 1.3 billion people and 20,000 hospitals.

Japan - Japan has broad healthcare coverage with a mixture of universal insurance and self pay.  This article nicely summarizes the economic issues.

There are a few large companies providing EHRs in China - Fujitsu, Mitsubishi, and NEC.  Hospital IT is frequently outsourced to such companies.   EHRs are typically client server architectures.   The web has not been widely embraced for EHRs, PHRs, or HIEs because of privacy concerns, but a few cloud computing pilots have been successful including an EHR for first responders in Fukushima.    There have been pilots of data sharing such as the Dolphin project in Kyoto and planning for PHR implementation.  

The challenges faced by Japan are an aging society, declining birth rate (1.4), lack of coordination of care because patients have access to any hospital/urgent care/clinician office on demand, creating a fractured record, and privacy policy that makes interoperability difficult.   Standards are emerging and there is an understanding of the need to use summary formats like CDA and controlled vocabularies.     A national policy requiring interoperability standards and encouraging data sharing with patient consent would significantly enhance quality, safety, and efficiency in Japan.

The government of Japan has recently changed and I'm told Mr. Abe will embrace innovation in healthcare IT as part of his economic recovery package.   Just as with China, I'm eager to help with a certification and meaningful use program, serving the 126 million citizens of Japan.

Taiwan - Taiwan has universal healthcare coverage with a mixture of universal insurance and self pay.  This article nicely summarizes the economic issues.

As with China, there are many small companies selling EHRs to local hospitals and clinician offices.  There's a long tradition of self-built systems as well.

Also, like China and Japan, most EHRs are client server, with little use of the web.

Privacy concerns restrict remote access, but I found the Taiwanese much more willing to embrace the internet and offer support for mobile devices, restricting information flows only to international locations because of privacy laws.

The hospitals I visited in Taiwan were high volume - 10,000 outpatients a day.  They implemented industrial process automation to help manage patient flow, laboratory specimen labeling, and medication dispensing.   I was impressed by their degree of EHR and ancillary system support.    Health Information Exchange and Personal Health Records are not widely used but there is a recognition of their importance for care coordination,

The challenges faced by Taiwan are an aging society, declining birth rate (.9), and ‘doctor shopping’ – patients visiting numerous practitioners simultaneously because of easy access to care at any facility without a primary care gatekeeper, creating a fractured medical record.

I look forward to further collaboration with my colleagues in Taiwan, helping create interoperability policy and technology to serve 23 million people.

In 12 days, I took 15 flights and worked long days to spend nearly 200 hours with government, academic and industry leaders.   I learned more than I taught.   Each country has its own unique history, culture and people.   However, the challenges of healthcare IT are very similar all over the world and the evolution that is taking place in our lifetime will ensure that Asia moves closer to the goal of  electronic health records for every citizen.

Losing the Popularity Battle, but Winning the Career War

CIOs are typically not very popular and are not known for their charismatic leadership i.e.

How do you distinguish an introverted CIO from an extroverted CIO?

An introverted CIO stares at his shoes.   An extroverted CIO stares at your shoes.

In the past, I've been able to achieve reasonable levels of popularity through rapid innovation and responsive agile application development, often delivering discretionary projects to individual departments.

As I navigate FY14, creating project plans that allocate resources and time, it's clear that I cannot deliver any discretionary projects.   If the Main Thing about leadership is To Keep The Main Thing The Main Thing (Stephen Covey) then I have no choice to but to keep IS resources focused on the Federal regulatory agenda that has been prescribed for FY14 and nothing more.

Now that I'm back from China/Japan/Taiwan, I will accelerate my efforts to communicate with stakeholders at all levels that the Federal government has set the strategy for healthcare IT departments in FY14 and as unpleasant as it sounds, IT management and healthcare stakeholders really have no flexibility to prioritize departmental projects.    Here's what I mean:

ICD10 - this federal requirement is pass/fail and involves the entire inpatient/outpatient revenue cycle.   Every piece of software, workflow, and process needs to modified.    It will cost the country billions, have limited benefits, and should be considered high risk, given the coordination needed among payer/provider organizations.   It's bigger than Y2K for healthcare and has a firm October 1, 2014 deadline that no one in government is willing to change.    Assume ICD10 will consume a majority of your IT resources for the next year.

Meaningful Use Stage 2 - this federal requirement is focused on stimulus in the short term, but penalty avoidance in the long term.   Hospital margins throughout the country are slipping, so it's very hard to turn down millions in Medicare/Medicaid stimulus.    ICD10 trumps Meaningful Use work, but hospital management really expects IT departments to deliver Stage 2 Certified software for everyone to use.   Eligible professionals who have already attested to stage 1 are looking to IT departments to provided updated  software so they can attest and claim the remainder of their $44,000 stimulus.    Meaningful Use Stage 2 has a 2014 deadline for hospitals and clinicians who attested to Stage 1 in 2011.

HIPAA Omnibus Rule/Compliance and Audits -  not a day goes by without a new audit by someone - OCR, OIG, DPH, internal, CMS.   I've heard that some IT departments are hiring full time staff just to respond to audits.  As with ICD10, these audits have limited benefit and consume resources that would have been applied to innovation in the past.   However, the work must be done.

ACA - The Affordable Care Act has required many new IT applications - health insurance exchanges, health information exchanges, quality registries, care management systems, and business intelligence infrastructure.  The Affordable Care Act required work is likely to improve efficiency and value in American healthcare.   However, the work displaces departmental priorities by consuming resources that might have been applied to local workflow enhancement projects.

Business imperatives with deadlines that cannot be missed - Healthcare reform has spawned a flurry of mergers and acquisitions that include fixed IT deadlines such as opening a building, extending networks, installing phone systems, merging clinical data, and expanding email coverage.  Although these are beneficial, the effort to support mergers and acquisitions takes resources away from optimizing local workflow and infrastructure.

What is the implication for CIOs?   Spreading a message that ICD10, MU2, HIPAA/Audits, ACA, and mergers have consumed all available IT resources for the next year is not going to be popular.   I truly expect many stakeholders to acknowledge that these priorities are reasonable as long as their departmental needs are also met.   The needs of the many are good as long as they don't outweigh the needs of the few.   Unfortunately, the answer for the next year needs to be "not now" if institutional survival is the main thing.

So over the next few months, I expect my waning popularity to wane even further.   I will lose the popularity battle.   However, when the regulatory mandates are done and the institution's longevity is assured, my career will be intact.   Losing the popularity battle but winning the career war for the benefit of the institution sounds like right long term strategy but certainly will require strength of will, a thick skin, and constant communication.

Building Unity Farm - The Poultry Grows Up

Last week I described our first year as farmers.   Just as we've grown, our animals have grown.

We started with 13 chicken chicks.   One was eaten by a hawk (Silver) and one died of egg impaction (Sunny).   The remaining 11 are happy and healthy, finishing their molt and preparing for winter

We started with 22 guinea fowl chicks.  11 have been eaten by predators because they have stayed out overnight sitting on nests or explored places they should not (fox dens, fisher cat habitat etc.).   However, we hatched 100 guinea fowl in August and kept 20 for ourselves.   They're now 8 weeks old and just about ready to leave the coop.   One alpha male (Mojo) was clearly very busy because 17 of the new guineas look just like him.

We received 10 ducks just 3 weeks ago and they've grown remarkably.  They enjoy eating a water soaked romaine lettuce and often stand in the bowl.   They live in a duck pen and spend the night in the duck house where it is warm.   Ducks don't care about wetness, so rain is not a motivator to go inside.    In another few weeks we'll let the ducks out to run around the property.  

Raising poultry is a remarkable experience.   All the chickens have distinct personalities and mannerisms.   There is definitely a pecking order.

Guineas are truly amazing to watch as they run around our 15 acres in search of ticks, worms, and other insects to eat.   They're excellent fliers and serve as a great alarm system, squawking whenever predators come close.  Their only downside is that are not skilled at survival in New England.   I'm sure that staying outside at night works in West Africa, but in the forests of New England, coyotes, foxes, raccoons, fisher cats, and weasels want nothing more than to eat Guinea fowl.

Ducks are remarkable.   They grow fast, live for water, and have distinct temperaments.  Some of ducks are calm, some of our ducks are nervous.  When we let the Indian Runner ducks out to circulate the property, it will be fascinating to see how they react to our wetlands.

At present we have 52 birds heading into the winter.  Hopefully all will survive until Spring.

Attestation verses Certification for HISPs

Of all the Meaningful Use Stage 2 questions I'm asked by vendors, HISPs, and providers, many involve confusion between certification and attestation.

As I've written about several times, the certification criteria are so extensive, often in unnecessary and confusing ways, that few vendors have been able to get through them.   Certification criteria exceed attestation criteria in many scripts.

I was recently asked about transition of care data exchange using Direct and the need for message delivery notification (MDN).   Micky Tripathi wrote the following excellent analysis, the bottom-line of which is that MDN is a narrow certification criteria, not an attestation requirement.   In the future, I think certification must be simplified to include the bare minimum necessary to support attestation.     Many people on the Standards Committee feel the same way and we'll support whatever polishing strategy ONC deems appropriate.

Micky wrote:

"1)  Organizations with self-developed systems may depend on a HISP as part of their certification, but organizations with vendor-based EHRs will not
a.   Although organizations with self-developed systems may chose to use the HISP as part of their alternative certification, most organizations will rely on their EHR vendor's complete certification
b.   For example, for Beth Israel Deaconess Medical Center certification, the HISP, acting as modular certified technology, needs to generate MDNs in response to incoming messages.
c.   For everyone else, the HISP does NOT need to generate any type of MDNs.  Sending providers only need to have reasonable assurance that messages sent via the HISP have been delivered to the intended recipients.

2)      There are three requirements that are relevant here:  the transitions of care attestation requirement, the technology certification for receiving messages, and the technology certification for creating/transmitting messages:
a.       The Meaningful Use Stage 2 transitions of care attestation requirement is that “the summary of care record must be received by the provider to whom the sending provider is referring or transferring the patient” (see page 4 of the measure)
b.      2014 Edition certification requires that an EHR be able to receive a Direct-compliant message and send an MDN for successfully received message (see page 5 of the NIST test script)
c.       2014 Edition certification requires that an EHR be able to transmit a Direct-compliant message to a Direct address recipient.  There is no MDN requirement on the transmission certification.  (NIST test script).

3)      The MDN requirement is SOLELY a certification requirement (it is NOT an attestation requirement) and it applies only to the requirements regarding receiving messages.  There is no certification requirement for MDN in transmission transactions.  There is no attestation requirement for MDNs (or any other technical means) to demonstrate assurance of receipt of transmitted transitions of care.
a.     While attestation does require that the intended recipient actually receive the message, there are no requirements on what type of assurance the sending provider must have in order to meet the Meaningful Use transitions of care measure.  Indeed, the ONC commissioned white paper on the topic of assurance states that:  “It is up to the Certified EHR Technology vendor to determine how to assist its customers and provide them with assurance that transmissions have reached their intended recipients.  This assurance could include a presumption of success on the provider’s part of subsequent transmissions if they have reasonable certainty that initial transmissions were successful.”  (see page 2 of the white paper)

4)      The MDN issue thus applies only to organizations that are using alternative certification and using the HISP as relied upon software because they need to be able to meet the “receive” and “create/transmit” criteria.  It does NOT apply to users who are using off-the-shelf Certified EHR Technology to transmit Direct messages over the HISP.
a.      Any other organization with off-the-shelf Certified EHR Technology which wants to use the HISP for transitions of care transmission does NOT need the HISP to be certified
b.      Their own Certified EHR Technology will generate a Direct-compliant message and pass it to the HISP for delivery
c.       The sender will have met their transitions of care requirement at this point as long as they have reasonable assurance that the HISP delivered the message to its intended recipient
d.      This assurance could be provided by MDNs delivered back from the receiving EHRs, but it does not have to be, and indeed, since recipients are not required to be Meaningful Use compliant, many recipients won’t be able to generate an MDN anyway
e.      In any case, it is NOT a HISP responsibility to generate and transmit MDNs back to the sending EHRs (except in the case where the HISP is acting as relied upon software for alternative certification)

5)      Some organizations may want the HISP to be certified for their attestation purposes.  For the purposes of attestation, the HISP will be certified ONLY for “generate/transmit” and NOT for “receive”, and thus it has no obligation to create MDNs
a.       In order for the HISP to receive modular certification for “receive”, it would have to be able to display CCDA documents, accurately match patients, and consume structured information for medication, problems, and medication allergies.  That would require that the HISP to include EHR features beyond the scope of HISP operations.
b.      Any organization that would like to attest with the HISP could thus only use the HISP for the “generate/transmit” requirement.

6)      Regardless of whether the HISP is used as just a conduit (#4) or as a certified module for transmission (#5), the HISP does need to provide some type of assurance of delivery back to the senders
a.       The current plan for the Massachusetts State HISP is to send back MDNs to provide assurance of delivery, which is ideal – however, it is NOT required
b.      The HISP could assure delivery by contract such as a service level agreement
c.      And/or the HISP could make available transaction audit records back to senders periodically or on-demand"