Thursday, September 8, 2011

Being a Good Steward of the Land

As a member of Wellesley's Community Garden, I do my best to serve the other gardeners by being a contributor to the entire property as well as a good steward of my plot.

There are basic rules and regulations covering membership in the community garden, but there has not been a advisory group of gardeners to recommend ongoing policy development or resolve disputes.   Reflecting on the work of the HIT Standards Committee and other Federal Advisory Committees which help inform regulatory decision making, I believe a group of gardening stakeholders can make a real difference by crafting policy that balances the needs of the many with the needs of the few.

Here are my early thoughts about being a good steward of the land:

1.  At the beginning of the season, gardeners should clean their spaces, removing weeds and debris that may have accumulated during the winter and early spring.   The previous year's plantings should be removed, the soil raked/turned, and fences mended.

2.  The town provides new compost that can be used to top off raised beds and planters.   Gardeners should add compost as needed to keep their spaces productive.

3.  The town provides wood chips that can be used to cover paths, controlling weed growth and covering muddy walkways.    Gardeners should weed and clear the paths around their space so that all gardeners can easily traverse the paths and so that the spread of weeds is reduced.

4.  Although the garden is primarily intended for annual fruits and vegetables, perennials can be planted, realizing that spaces are not owned, but are lent to each gardener for a finite period of time.    Brambles such as raspberries/blackberries should not occupy more than 25% of the space, since planting a bramble patch does not constitute gardening.

5.  Gardening is an ongoing labor of love, not a plant once and forget type of activity.   Plantings should be tended throughout the season with trimming, thinning, and harvesting as needed to keep the garden productive.

6.  The garden will always have a waiting list and in the interest of accommodating as many people as possible, new plots should given to 2 families.   Existing plots should be voluntarily subdivided.  Experienced gardeners who would like physical assistance with the garden should consider subdividing and serving as mentor to a gardening partner.

7.  If the Community Garden Advisory Committee (CGAC)  notes that a gardener is not being a good steward of the land, the person should be contacted and advice offered.  If no improvement occurs over 1 month, the space should be considered abandoned and given to a new tenant

8.  Gardening can be an expensive proposition and community garden improvements may benefit from grants or donations.    The Community Garden Advisory Committee should raise funds that can be used to improve fences, the water supply, and conceivably fund tools/seeds/plants that could benefit the entire community.

9.   At the end of the season, all tools and non-natural materials should be removed from each plot to ensure that the garden area is attractive during the winter to neighbors and visitors.

10.  In general, use of the garden should be viewed as a privilege, not a right.   We are being given 500-1000 square feet of valuable Wellesley land for $40/year.   We should use it as good stewards, balancing our own needs, the garden's needs, and the town of Wellesley's needs.

Our Community Garden Advisory Committee will begin meetings this Fall and then formally report out to the Wellesley Natural Resources Council.   After all my time in Washington, I look forward to playing a role in governance and policymaking at the local level.   If any of my readers have experience with community gardens, I'd welcome your input!

Wednesday, September 7, 2011

Vermont Information Technology Leaders

Today I'm in Vermont, meeting with the stakeholders of Vermont Information Technology Leaders (VITL), the federally designated Regional Extension Center and Health Information Exchange for Vermont.

I feel a close affinity to VITL as one of our New England collaborators (Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, and Connecticut) but more importantly because of my longstanding relationship with VITL's CEO, Dr. David Cochran, MD who was senior vice president for strategic development at Harvard Pilgrim Health Care and an influential driving force in the rollout of electronic health records in Massachusetts.

On a personal level, Vermont is one of my favorite places with remarkable countryside, a strong willed people, and a can do attitude.    Hurricane Irene has deeply affected the state but everyone is pitching in to accelerate the recovery.

My keynote will reflect on the journey from EHR implementation to true quality improvement using decision support, advanced analytics, and novel care management tools.  

For the first time in my career, I am seeing a cultural transformation such that the majority of clinicians believe an EHR is a necessary part of their practice.  Emerging accountable care organizations are stressing the need for health information exchange and financial/clinical analytics as a foundation for the healthcare reform work ahead.  

Meaningful Use Stage 1 sets the stage for quality measurement by moving clinician offices from paper to structured data entry.  Stage 2 will require more data exchanges and increasing use of controlled vocabularies that will make quality measurement easier.   Stage 3 will include new levels of decision support and data mining to prospectively and retrospectively help clinicians manage population health.

BIDMC's strategy to prepare for future stages of meaningful use and healthcare reform includes

*Embracing health information exchange by serving as a pilot site for government, academic, industry, payer, and patient engagement experiments
*Capturing the value of unstructured data by testing novel Natural Language Processing software
*Empowering users with new analytics using new business intelligence platforms
*Embracing novel sources of information including data from home care devices and patients themselves
*Exploring the implications of gathering and using genomic data for clinical care and clinical trials

I'll share examples of each of these with my colleagues in Vermont and in upcoming blog posts.

One of the most exciting developments is that we now are sending HL7 CDA document summaries from every patient visit to a community-wide registry which generates our meaningful use quality measures, submits our PQRI measures and supports ad hoc clinical queries.   By requiring every EHR vendor to send a CDA document over the New England Healthcare Exchange Network (NEHEN) to the Massachusetts eHealth Collaborative Quality Data Center, we have created the technical foundation for our emerging ACO that is in production now.   I'll post a more complete technical overview next week.

I look forward to exchanging lessons learned with Vermont.

Tuesday, September 6, 2011

Decision Fatigue


We're all suffering from information overload.  More projects with fewer staff on shorter timeframes mean more email, texts, blogs, online meetings, and phone calls.  

We make more decisions and have more accountability than ever before.  Regulatory complexity and the need for risk management has increased.  We're pressured to make decisions faster and there is less tolerance for mistakes.   Making all those decisions in a high stakes environment like healthcare leads to decision fatigue,  that numbness you feel at the end of an overloaded day when you decided what to spend, who to hire, and what to do, hundreds of times.

I believe decision fatigue is an escalating threat to our ability to manage the events of each day and keep balance in our lives.

When I think back on my early career as a leader, in the 1980's, there was no email, no overnight shipping, and limited numbers of fax machines.

Issues were escalated by writing and mailing a letter.    The time it took to compose, type, mail, and deliver a letter meant that many problems solved themselves.  Since the effort to escalate was significant, many problems were never escalated.

Today, everyone can escalate everything to anyone.  The barrier to communicating is nearly zero and communication is real time.   There is no mail room or team of middle management filters between you and the CEO.

This creates an interesting conundrum for leaders.  Should everything be answered in a very timely way with Solomon-like decisions about every issue?   Should everything be ignored unless truly emergent with the hope that someone else will solve the problem?  Should everything be deflected to those in middle management who would have read paper-based mail?

My goal is to never be the rate limiting step.  That means that I make hundreds of decisions every day.   Some are right and some are wrong, but they are the best answers given the information that I have.   In the IT industry,  timely action that is good enough is often more important than a delayed perfect action.

Thus at the end of every day my brain is whirring with thousands of inputs, and hundreds of decisions made.   I'm not physically tired after any workday, but I can feel mentally tired from decision fatigue.  

The problem with decision fatigue is the that quality of decisions can diminish as the quantity of issues increases.  

There are two ways to address decision fatigue

1.  Reduce the scope of your authority and hence the decisions you need to make and the risks you need to manage.  I'll post a blog next week about span of authority and risk management.

2.  Spread decisions over a wider group of people, reducing the volume of decisions that fall to any one person.

#2 depends upon having a great boss, who is supportive, responsive, and willing to share decision making risk with you.    #2 also requires great staff whom you can empower to make decisions on their own.

Thus, I make the decisions that I am uniquely qualified to make, while pushing others up and down the organizational hierarchy so that risk is mitigated (seeking approval up the org chart) and trusted staff are given the resources and authority to solve problems on their own (delegating down the org chart).

Here's an example of how I managed decision fatigue today.   Between 3pm and 4pm, I was asked to make several decisions:

1.  The regional poison control center sought my input on a mushroom ingestion case.   A 1 year old  had taken a large bite from a mushroom growing in a backyard.   Since I uniquely have mushroom toxicology knowledge, this was my decision.   The mushroom was a harmless Lactarius Fragilis  and I decided that the child would be fine.

2.  A leak in the Longwood Medical Area chilled water supply caused a 5 degree rise in our disaster recovery data center.   What should we do?   I ensured that all appropriate facilities and IT people were organized to address the problem, and asked to be informed if the temperature exceeded 90F.     The incident management decision making was delegated to others.

3.  A researcher in one of the Harvard buildings suspected that the network had been hacked because www.ups.com was unavailable.  Should I page security and networking staff to urgently investigate this on a holiday weekend?   I used my Blackberry to replicate the problem and escalated it to IT security, who found the problem was unrelated to our network/DNS servers.    The incident management decision making was delegated to others.

In the next few months, I'll be finishing the FY12 Operating Plan for BIDMC IS, so there will be plenty of decision making to spread among governance committees and executive management.

One other cure for decision fatigue that I recommend is a "time out".   On my way home in the evening, I stop at our community garden space to sit on the small bench we've placed there, eat a few cherry tomatoes wrapped in basil, and watch the birds peck at our sunflowers.   I leave my Blackberry in the car.   By the time I get home, the decision fatigue of the day has passed, so when my wife and I discuss dinner choices, I'm ready to act boldy.

Friday, September 2, 2011

Cool Technology of the Week


I recently received the press release below, which illustrates a cool trend in the healthcare IT industry.

eClinicalWorks and other EHR vendors have been piloting standard transport interfaces that are compatible with Nationwide Health Information Network Exchange transactions (pull/push) and Direct transactions (push).

Intersystems and other integration engine vendors have been building HIE appliances, cloud offerings, and software tools to connect the EHRs which support these transport standards.

The end result is a significant reduction in implementation burden and cost.

Once connectivity is enabled, novel transactions of many types can flow.

EHRs with standards-based transport interfaces and integration engines that can connect them without significant development time.   That's cool!

First Implementation of One-Click EHR Rolled Out on Brooklyn Health Information Exchange
August 31, 2011

InterSystems Corporation, a global leader in software for connected care, today announced that the Brooklyn Health Information Exchange (BHIX) is now providing breakthrough electronic health record (EHR) connectivity that is expected to dramatically speedup care delivery for BHIX clinicians. "The advanced functionality enables physicians to securely access patient records via the InterSystems HealthShare-based BHIX network directly from an eClinicalWorks EHR with just one click," said BHIX Executive Director Irene Koch. "This seamless interconnectivity makes it possible to share patient data and improve patient care coordination across the entire community served by BHIX."

Thursday, September 1, 2011

The College Drop Off


I have a very hard time giving up roles and responsibilities.    Rather than change jobs, I add jobs.

In 1996, I oversaw the CareGroup Center for Quality and Value, the data warehousing and analytic operations of a Boston-based integrated delivery system comprised of Beth Israel Deaconess and 4 other hospitals.   When I became the CIO of CareGroup in 1998, it took me a year to separate myself from the operational responsibilities of the CQV.

In 2000, I oversaw the Harvard Medical School learning management system as Associate Dean of Educational Technology.   When I become the CIO of HMS in 2001, it took me a year to delegate my educational technology role.   Early in my HMS tenure, I was asked to serve as temporary CIO of Harvard Clinical Research Institute (HCRI).  That temporary job lasted a year.

Today, my wife and I spent the day at Tufts, helping our daughter move into her dorm and begin her journey as an independent adult.   In many ways, my job as parent, that began 18 years ago, fundamentally changed today.   It's very hard to let go.

I'll want to hear about my daughter's experiences each day, the decisions she's making, the challenges she's facing, and the successes she's achieving.  I'll want to offer advice, assist when I can, and give her the benefits of my 50 years of experience.

However, all of these activities are the wrong thing to do.   She needs to fly on her own, knowing that we're here when she needs us.

The Deans at Tufts emphasized 3 goals for Tufts undergraduates - develop internal curiosity for learning, be responsible for your own actions, and become an advocate for yourself.  

The only way my daughter will become a mature, experienced, and assertive young woman is to do her best, explore a college world that is much more diverse than her high school experience, and be responsible for her own decisions.

Today, my wife and I became a safety net rather than a guiding force.

Lara has fledged and we have an empty nest.

We shed tears of loss when saying goodbye, followed by tears of joy for her new possibilities.

I may have had a hard time with the CQV, HCRI, and Harvard Medical School, but for Lara, I can morph my parent job so that she that she can thrive in our increasingly complex and confusing world.   My job needs to change, so that she can change.

Lara, we only cried part of the way home.   We're ok.

Now do great things.  The world is your oyster.

Wednesday, August 31, 2011

The Standards Summer Camp Deliverables

On September 28 2011, the HIT Standards Committee (HITSC) will officially deliver to ONC its 6 months of hard work from Standards Summer Camp.  HITSC subcommittees and workgroups have met every other day since April to prepare the standards recommendations needed to support Meaningful Use Stage 2 rule making.

The S&I Framework teams have been working in parallel on important issues - Certificates, Provider Directories, Lab Result Reporting, and Transfer of Care Summaries.

Here's how it all fits together.

Certificate Recommendations - HITSC recommended specific implementation guidance for X.509 certificates.  The S&I Framework teams developed a strategy  for certificate authorities to issue trusted credentials that will eventually be cross-certified with the Federal Bridge Certificate Authority (FBCA), enabling exchange with Federal agencies.

Metadata recommendations -  HITSC recommended CDA R2 headers for patient, provenance, and security metadata.  These were included in the Advanced Notice of Proposed Rulemaking.   HIEs should use these standards as metadata envelopes for content payloads that are sent between different organizations.

Provider Directory recommendations - HITSC considered LDAP but noted that federated LDAP directories and internet-based LDAP queries between organizations have not yet been widely deployed.   HITSC also considered microdata and web search engine retrieval of structured directory data.   The S&I Framework teams concluded that pilots of federated LDAP queries and microdata are a reasonable next step, because no provider directory standard is mature.  Additionally the S&I Framework teams recommended DNS for certificate distribution with the addition of LDAP if an organization's implementation of DNS does not support certificate discovery.

Vocabulary recommendations - HITSC recommended a parsimonious set of standards for vocabularies supporting quality measures including SNOMED-CT for problems, LOINC for labs, and RxNorm for medications.   The September HITSC meeting will include a transition plan for those vocabulary standards required for Stage 1 that are being retired/replaced in Stage 2.

Patient Matching recommendations - HITSC recommended a set of best practices that will guide implementors who want to match patients using demographic data elements with appropriate specificity and sensitivity.

ePrescribing of Discharged Medications recommendations - HITSC recommended NCPDP and HL7 standards that are widely implemented and compliant with Medicare Part D requirements.

Public Health recommendations - HITSC recommended HL7 2.51 implementation guides for syndromic surveillance, reportable lab, and immunizations - one highly constrained implementation guide for each transaction.

NwHIN recommendations - At September's meeting, HITSC will recommend one set of building blocks to support Nationwide Health Information Network Exchange transactions (pull/push) and Direct transactions (push).

Lab Results recommendations - The S&I Framework teams recommended an HL7 2.51 transaction that is very similar to the public health implementation guides already approved by HITSC.  It also includes vocabularies and code sets that constrain the optionality of the transaction.  The Implementation Guide  is broadly supported by ELINCS developers, commercial labs, and numerous informatics experts.   It will be balloted by HL7 in the next few weeks and then piloted before any regulations are written.

Transitions of Care recommendations - The S&I Framework teams recommended a transfer of care summary that is a natural stepwise evolution of the work we've done for the past 10 years  - CDA --> CCD --> C32 --> transfer of care CDA templates.   These CDA templates are easier to implement than C32 and more flexible,   Given that CCR is a declining standard (little new work is being done on it), CDA templates are a reasonable next step.   The HITSC will be asked to comment on the trajectory of this work and will evaluate the results of pilot testing.

At the September meeting of the HITSC, we'll review all the work we've done as well as the S&I Framework efforts on Certificates, Provider Directories, Reportable Lab and Transfers of Care.

What evaluation criteria should we use?   In the words of Doug Fridsma, who oversees the ONC Office of Standards and Interoperability

"While it might not be perfect, does it represent the best we have so far?
Does it point us in the right direction?
Is it the next step in an incremental approach to refining the standards and implementation guides?
Does it support our policy objectives?
Can we update it as needed through the SDO community?

All standards, even those that have wide-spread uptake, require constant updating and refinement. Vocabularies, terminologies, and other existing standards will require piloting to make sure that we maintain relevance.

We can't let the perfect be the enemy of good. Standards will require continued support and community refinement. If we can generally answer 'yes' to the questions above, then we need to continue to push forward toward the goals of interoperable health exchange."

I look forward to the September meeting and the delivery of all the great work done by the HITSC and S&I framework teams, bringing closure to this phase of Stage 2 preparation activities.

Tuesday, August 30, 2011

Lessons Learned from Steve Jobs

I recently spoke with several reporters about Steve Jobs' impact on healthcare , thanking him for the past 15 years of innovation.   In preparing for those interviews, I reviewed Steve's career milestones,

In 1997, Apple Computer was in trouble.  Its sales had declined from 11 billion in 1995 to 7 billion in 1997.  Its energies were focused on battling Microsoft.   It had lost its way.

Steve Jobs made these remarks at MacWorld 1997, a few months before becoming Apple's CEO.  He outlined a simple go forward plan:

1.  Board of Directors
2.  Focus on Relevance
3.  Invest in Core Assets
4.  Meaningful Partnerships
5.  New Product Paradigm

How can we apply these 5 ideas to the work we're doing in HIT?

It's clear that Health Information Exchanges across the country are in trouble - CareSpark closed its doors,  the CEO of Cal eConnect resigned, and Minnesota Health Information Exchange ceased operations.

Let's consider the application of Steve's principles to Healthcare Information Exchange in Massachusetts.

1.  Board of Directors - Governance in general is very important to healthcare information exchange.   HIEs need a multi-stakeholder governance body to set priorities, monitor progress, and ensure all stakeholders are engaged.   In the past few months, state government and the private sector experts have worked together to define roles and responsibilities.   The State's HIE coordinator, Rick Shoup, and I presented this consensus plan to the state's HIT Council, the decision making body established by state regulation Chapter 305.   Governance will be done by the HIT Council plus an HIT HIE Advisory Group consisting of payers, providers, employers, patients, academics, and government.    This "Board of Directors" of the Massachusetts HIE activities is top notch.

2.  Focus on Relevance -  HIEs can do many things.  They can push data among payers, providers, patients, and public health.   They can create master patient indexes, record locator services, and registries.    However, what will the market pay for today?   At the moment, simple secure transport that connects every stakeholder with easy to use web applications and native EHR interfaces seems to be the answer.   Rather than do everything simultaneously, we need to tightly focus on just secure routing, making 2012  the year of the state "information highway".

3.  Invest in Core Assets -  Massachusetts already has production HIEs that serve the business needs of several customers.   We have NEHEN, CHAPS, SafeHealth, North Adams HIE, and the MAeHC Quality Data Center.   Rather than reinvent these, we need to focus on the gaps, creating a state backbone that will connect every stakeholder, establishing a network of networks that leverages existing investments.

4.  Meaningful Partnerships - The State Medicaid Health Plan includes 14 projects that cover over 90% of the providers in Massachustets.   Since Medicaid is eligible for 90/10 matching funds (90% Federal/10% State), it makes great sense to do as much as we can via Medicaid.   Multiplying our purchasing power by 10 is a meaningful partnership!

5.  New Products - Once connectivity from every stakeholder to every stakeholder is in place, we can create novel functionality such as clinical registries and the ability to query data to support the "unconscious in the emergency department" use case.  

Thus, if 2011 was the year of governance, 2012 will be the year of connectivity, 2013 the year of registries, and 2014 the year of queries.

Thanks Steve, for an approach that gives us focus and momentum.    I'm completely confident our Massachustets HIE activities will succeed by embracing your 5 principles.