Tuesday, October 18, 2011

An Update on Massachusetts Health Information Exchange

As I've described previously, Meaningful Use Stage 1 was focused on the electronic capture of data into EHRs.  The standards we specified included content and vocabulary but not transport.

Stage 2 will be more focused on Health Information Exchange.  Transport standards will likely be included in the Notice of Proposed Rulemaking.

In order for Health Information Exchange to work, I believe we need
a.  A transport standard which can be implemented consistently for multiple senders and receivers (such as Direct)
b.  A certificate distribution infrastructure to secure the endpoints (such as via DNS)
c.  A directory for routing information between organizations (such as a RESTful API to a SQL database)
d.  Connections to the last mile -  sending/receiving directly into EHRs (such as via XDR/SOAP) or into a standalone web-portal as a short term solution while vendors build transport features into EHRs.
e.  Governance to guide the technology and policies that support the above

On October 17, the Massachusetts HIT/HIE Advisory Committee (think of it as the state equivalent of a Federal Advisory Committee) presented its recommendations to the HIT Council (the governance for Massachusetts HIT activities).   The Advisory Committee asked for the input of 88 stakeholders divided into 5 groups - legal and policy, technology and implementation, finance ad sustainability, consumer and public engagement, provider engagement and adoption.

The state HIE implementation plan we presented:

*Aligns with national interoperability standards and emerging MU stage 2 requirements
*Maximizes State Medicaid Health Plan/Medicaid Management Information System Federal Financial Participation funding,  a 9 to 1 match
*Focuses Medicaid funding on building infrastructure for statewide services, and the ONC Cooperative Agreement HIE funding for last mile implementation

What is last mile implementation?

1. Our regional extension center (REC) will do a pareto analysis of EHR adoption in Massachusetts to identify the most commonly used EHRs.  I believe that 20 EHRs are common but 5 EHRs (eClinicalWorks, AthenaHealth, NextGen, Allscripts, GE Centricity) cover 90% of the providers.   We'll negotiate with those vendors to create the software which is necessary to connect their EHRs to the statewide Direct backbone in 2012.  Yes, we could wait until 2013 when Meaningful Use Stage 2 requires them to implement transport standards, but coordinated procurement now will accelerate HIE integration.

2.  We'll provide System Integrator services to support on boarding of small practices and subnetworks of clinicians to the state HIE backbone.

3.  We'll provide education and training to foster adoption and use of HIE services.

4.  We'll devise grants and subsidies that serve as incentives for HIE adoption

5.  We'll facilitate the addition of value added services to the backbone such as public health reporting, clinical registries, and quality measurement services

As I've spoken with vendors, many have noted that State HIEs tend to build central infrastructure but assume endpoints will connect to it on their own.   My experience is that a centralized project management office and a single coordinated plan is needed to connect the endpoints.  Once every provider, payer and patient is connected, the value proposition of the HIE will increase significantly per Metcalfe's law.

We agreed that by November the Advisory Committee will complete:

*Initial State Medicaid Health Plan Implementation Advanced Planning Documents (IAPD) for review by Advisory Committee workgroups and the Secretary of EOHHS before submission to CMS
*An outline of Statewide HIE Policy Guidance that complements the technology to be built via IAPDs
*A high-level plan for entire project, so that all stakeholders understand what we're creating

I'm extremely optimistic - we have a plan with alignment of support, appropriate funding, and mature technologies that are low risk for failure.   The HIT Council, MeHI,  and Federal government stakeholders are reviewing our recommendations over the next few weeks and we'll seek their approval to move forward in November.

The momentum for Health Information Exchange in Massachusetts is building among payers, providers, patients, government, and industry.   2012 will mark the tipping point that enables us to stop talking about barriers to Health Information Exchange and instead focus on the accelerators.

Monday, October 17, 2011

Exploiting Privacy Breaches

I've described information security as a Cold War, requiring constant investment and vigilance to innovate faster than the hackers and criminals who are stealing data to commit identity theft.

I'm spending an increasing percent of my resources on regulatory compliance and data protection.

Over the past year, Federal and State governments have

1.   Specified standards to protect healthcare data during transport
2.   Required encryption of data at rest.
3.   Required breach notification to patients and prominent media 
4.   Created policy to define meaningful consent and other important patient privacy rights
5.   Launched a new initiative on data segmentation in an effort to support more granular healthcare privacy preferences

CIOs and Chief Information Security Officers are working as hard as they can, hackers are intensifying their attacks, and the world is accelerating its adoption of mobile technologies that make perfect control of data more challenging.  Despite all our efforts, breaches will occur.   Even the most sophisticated security companies have been breached by increasingly sophisticated malware.

There's a dark side to all of this that is the subject of today's blog post - using the new privacy breach reporting laws for personal gain.

There are many good attorneys.   My parents are attorneys (patent and business law).    Some of my favorite colleagues are attorneys working hard in the public interest (Deven McGraw at CDT, Jodi Daniel  at ONC).

As with any profession there are those attorneys who use the law for personal gain.    Here's a list of privacy breach class action suits, comparing payments to attorneys versus their clients.

There are many good  investors.    Accelerating new technology by providing funding to those who can build high value businesses is a good thing.     As with any profession, there are investors who put profits ahead of societal benefits.

I've heard discussion about an alarming new business model.   Investors paying attorneys to file class action suits related to privacy breaches in return for a portion of the profits.

Prviacy Breach reporting is now public.   Identifying a class is easy.

However, if the risk of harm from the privacy breach is low, attorneys may not want to bear the expense and burden of filing a suit, given that recoveries might be minimal.   If investors underwrite the risk, realizing that most healthcare organizations will want to settle rather than spend time and resources on litigation, we'll likely see a lawsuit following every reported privacy breach.

To me, there are different kinds of privacy breaches - those which are caused by true carelessness and those which occur because of sophisticated attacks that the Pentagon could not even repel.   We should hold organizations accountable for implementing best security practices to protect privacy.   We should report breaches to patients and prominent media, since breach reporting regulations provide a great incentive to invest in appropriate security.   However, we should do this in an effort to enhance the society we live in, not generate profits.

As we all work together on electronic health records and healthcare information exchange, let's try to create regulations that do that right thing

1.  Protect the data
2.  Respect patient privacy preferences
3.  Recognize the difference between hard to prevent breaches and those that occur because basic protections were not in place

Investing in class action suits that asymmetrically benefit the finance and legal professions is not something that benefits society.

As the eternal optimist, I'm convinced we can all work together for the common good and make every day better than the last.   If you hear about someone using privacy breach reporting for their own personal gain, shout out that it's the wrong thing to do.

Friday, October 14, 2011

Cool Technology of the Week

Big data has arrived.   At BIDMC, I oversee 1.5 petabytes of clinical and administrative data.   At HMS, I oversee nearly 3 petabytes of research data.

As Blackberry's recent outage illustrates depending on single monolithic infrastructure has its risks and impact of failure can be enormous.

How can we leverage commodity hardware infrastructure, reduce risk, and meet user demands for mining big data?   Apache Hadoop is a cool technology worth knowing about.

Hadoop is an open source framework that allows for the distributed processing of large data sets across clusters of computers, designed to scale from a single server to thousands of machines.  Rather than rely on hardware to deliver high-availability, Hadoop detects failures and automatically finds redundant copies of data.  The Hadoop library includes

*The Hadoop Distributed File System (HDFS), which splits user data across servers in a cluster.

*MapReduce, a parallel distributed processing system that takes advantage of the distribution and replication of data in HDFS to spread execution of any job across many nodes in a cluster.

Microsoft has just introduced support for Hadoop into SQL Server 12  as part of their  end-to-end Big Data roadmap.

A fault tolerant distributed file system using commodity hardware for big data that is even integrated into mainstream data mining tools like SQL Server.  That's cool!

Thursday, October 13, 2011

Columbus Day in New England

Every Columbus Day, my family visits New Hampshire for hiking, apple picking, and autumn color at peak.

Here's my ideal Columbus Day Weekend:

1.   Climb Mount Monadnock via the Marlboro trail - it's the road less traveled that includes easy rock scrambling, great views, and beautiful foliage.  I've climbed Mt. Monadnock over 100 times.

2.  Pick Apples and buy fresh chestnuts at Old Cider Press Farm in Westmoreland, NH.   My favorites are Gala, Empire, and Black Gilliflower.

3.  Spend the night at the Inn at East Hill Farm in Troy, NH - we've rented the upper floor of Sugarhouse on Columbus Day weekend for the past 15 years.

4.   Have a vegetarian lunch at the Zeppelin and Kaliedoscope in Marlborogh, NH.    My favorite is the Falafel platter.

5.  Meet artists throughout Southern New Hampshire as part of the Friends of Dublin Art Colony Columbus Day weekend open studios.

It's quintessential New England.   Robert Frost would be proud.

Wednesday, October 12, 2011

The Impact of Consumer IT on the CIO

I've written about the implications of staff bringing their own devices to the office instead of using corporate desktops and the challenges of keeping mobile devices secure.

I own a Blackberry Bold 9700, first released on December 18, 2009.  Since I do not read email when I drive, I hand it to my wife who instinctively tries to navigate my email by gesturing on the screen (it does not work).  I receive one email every 30 seconds (over 1000 per day), many of which contain web sites and attachments.   Reading them on a Blackberry Bold is not a satisfying experience.   It's a messaging device, not a comprehensive mobile platform.

In 2007, owning a Blackberry was cool and RIM interviewed me about the my mobile lifestyle.

The iPhone 3GS was released on June 19, 2009 and since then Blackberry has become less cool and RIM has lost $30 billion in market value.

Market changing consumer products are introduced every 6 months.  In the same timeframe, I've been working on hundreds of important enterprise initiatives.  The average hospital IT project, given limited resources and large scope, takes 18 months.

The consumer IT marketplace moves at such a fast pace, applying thousands of people to create a single device, that the average employee now expects every hospital IT project to proceed in the same manner, even though only one person may be working on a niche project that serves a small number of users.

I sometimes describe the job of the CIO as a quest to minimize negative reinforcement - "I only received 100 challenging emails, it's been a great day".  The accelerated pace of consumer IT multiplies impatience, intolerance, and emotion.

It would be great if mobile devices and the Cloud solved all our problems, but unfortunately, the enterprise world has compliance requirements, security constraints, complex business processes, controls, and workflows that are not addressed by consumer technologies.

What's a CIO to do?

1.  Embrace mobile devices and the cloud when they make sense.   We do support iPhones and Android devices for email.   We do support the cloud for image exchange and a private cloud for hosting community physician electronic health records.  BIDMC Information Systems is considered "consumer device friendly", which helps my reputation.

2.  The expectation of new infrastructure and applications every 6 months tempered by the reality of 18 month hospital project plans requires intense communication.   This week, I sent my staff a plan to create  "IT concierges" for our key stakeholders, ensuring that monthly project updates keep users informed and better align expectations with reality.

3.  Meetings with disgruntled stakeholders are really important to maintain credibility.  "Presence" of the CIO can really make a difference when customers perceive the pace of enterprise IT innovation to be slower than the consumer marketplace.

4.  Maintaining agility and flexibility without being dogmatic ensures the CIO is not the rate limiting step to innovation.  I've always said that if emerging companies can provide superior service at lower cost to any product we have currently, we should openly evaluate it.   Customers will appreciate that IT has a culture of innovation even if product life cycles are longer than the consumer marketplace.

5.  CIOs need to accept that 10% of users will dislike you on a given day because enterprise technologies are unlikely to keep pace with consumer technologies.   Rather than get frustrated, realize that by focusing on continuous measured progress, you'll create a trajectory that prudently moves forward, balancing security, innovation, and functionality.   As my daughter would say, Ganbatte!

Complexity, unrealistic customer expectations, and resource limitations make the job of CIO increasingly difficult.   By focusing on the possible, communicating your plans broadly, and accepting consumer technologies for the use cases that make sense, the CIO can continue to thrive.

Tuesday, October 11, 2011

The BIDMC FY12 Operating Plan

This is my 1000th blog post.  Readers have visited the geekdoctor blog over a million times.   The discipline of writing every day has helped me become a better communicator in my work and personal lives.    Thanks for being a part of it.

Today, I'm publishing the BIDMC FY12 Information Systems Operating plan which includes all the high priority projects I identified in my earlier post about preparing for a new CEO.    

In the Clinical Information Systems area there are numerous important goals, but five that deserve a more focused discussion:

ICD-10 - A challenging and expensive compliance project that must be done because 70% of the revenue cycle is at risk.  The scope of the project includes BIDMC, it's community hospital affiliate BIDH-Needham, the owned physician group APG, the academic affiliated group HMFP and the private physicians in BIDPO.    The likely budget - $5-10 million.   The likely safety, quality, efficiency improvements - none at all.   Many in the industry recognize ICD-10 as a colossal waste of time and effort.  I've tried very hard to communicate this broadly and will continue that effort in articles, lectures and personal meetings in Washington.

Clinical Documentation - although all the ambulatory areas and ICU areas of BIDMC have online documentation, progress notes in the wards are still written on paper.   In order to improve care coordination, decision support, and provide the granularity necessary for ICD-10 coding, we will work hard to create multi-disciplinary documentation in our ward areas.

Electronic Medication Administration Records - We're redesigning  medication management from the supply chain to the bedside.   We'll adopt innovative tools to automate the new idealized processes.

Healthcare Information Exchange - Accountable Care Organizations will require increased health information exchange analytics.   All the stakeholders in Massachusetts have worked together on an HIE plan that connects every provider, payer and patient.   Now we need to build it.

Analytics - Turning data into information, knowledge and wisdom will require a new generation of analytics.   We're building those now.

The Fiscal Systems area includes several important compliance goals such as 5010 and enhancements to our human resources management systems.

Knowledge services will continue curation of our electronic library resources.

Media Services will enhance our streaming and conference room infrastructure.

In the Infrastructure area, there are important 15 important themes:

Data Protection - Develop and implement a strategy for protecting BIDMC authentication credentials and sensitive data accessed from personally owned computers including mobile devices.

Patch Management - Improve patch management sufficiently to reduce vulnerabilities by 50 percent or more on I.S. managed systems.

Voice Mail System - Replace the current voice mail system that supports over 6,000 accounts with a more up-to-date, IP-based system without incurring extended disruption of service or loss of old content.

Clinical Systems Service Level - Implement high availability features using Cache ECP and Red Hat Enterprise Linux clustering technologies.  Increase service levels from 99.9 percent to 99.99 percent due to Cache or Linux related problems.

DR Data Network - Virtualize the distribution layer of the data network in the disaster recovery center to permit more flexible repositioning of assets without regard to VLAN assignment.

Life Cycle Management - Replace one-third of the approximately 120 data network access switches in the east and west campuses with more up-to-date and efficient devices.

Security Dashboard – Develop metrics that in combination provide a way to quantify the risk, by subnet, of devices attached to our data network.   Using this information, apply security measures that are tailored to the relative risk each subnet presents.

Vulnerability Assessments – Deploy a commercial Vulnerability scanner along with updated processes and procedures to enhance our ability to identify and remediate vulnerable systems network wide.

Windows 7 – Upgrade all kiosk/public managed workstations to Windows 7 to enhance the user experience, improve security, and take advantage of better management features.   Upgrade at least 50 percent of business builds to Windows 7 with the balance targeted for FY13.

SQLServer Enterprise Edition – Migrate two-thirds or more of the 32 data bases still using Standard Edition to Enterprise Edition.   EE provides higher availability, larger hardware, improved auditing, and other features.

Rights Management – Improve the utility of the Varonis rights management system to include surveillance of unusual access patterns that may indicate a security issue.

Exchange 2010 – Plan the upgrade and deployment from Exchange 2007 to Exchange 2010.

Internet Access – Develop and implement a strategy that reduces the exposure of BIDMC IT technical resources and sensitive data to Internet-based malware that strikes a workable balance between the security threat and business need for Internet access.

Service Level Monitoring – Extend the use of Nagios software monitoring to 75 percent or more of the data center assets to improve the reliability, granularity, and accuracy of system alerts.

Storage – Migrate all new image acquisition PACs to Atmos and begin conversion of all Centera assets to Atmos.

We'll do all of this within the FTE, operating budget (less than 2% of the BIDMC Operating budget) and capital budget assigned.

I look forward to a great year ahead and share with all of you the lessons learned along the way.

Friday, October 7, 2011

Cool Technology of the Week

As the discussion of the functionality to be included in a Nationwide Health Information Network (NwHIN) continues, there are 3 different secure transports being evaluated:

Exchange:  “NHIN Messaging Platform Specification”, which uses SOAP for transport and WS-I Basic Security Profile for security (TLS + XML signature + WSDL + AES + X.509 + SAML)

Direct:  “Applicability Statement for Secure Health Transport”, which uses SMTP for transport and S/MIME for security (AES + X.509)

Secured REST:  specification to be done, but will use HTTP for transport; candidates for security include TLS, X.509, and OAuth.

Each has different characteristics and different strengths.   The barrier to RESTful implementation is lack of a consistent implementation guide.

The folks at MITRE have implemented project hData  noting that "Current electronic health data standards are complex, hard to implement, and difficult to manage”.

hData separates transport and packaging from content – something the HIT Standards Committee has supported.   This FAQ provides more details.  Clearly hData is still in development and not yet adopted, but I do think they are pursing an appropriately simple approach to transport.

The hData content format has been balloted by HL7 and a Draft Standard for Trial Use (DTSU) is expected this month.  The hData transport format (RESTBinding) is in the Open Management Group comment resolution phase.

A RESTful implementation guide for healthcare that separates content and transport, providing easy to implement,  secure transport.    That's cool.