Dirk Stanley, MD, CMIO, UCoon Health

Hi fellow CMIOs, CNIOs, and other Applied Clinical Informatics and #HealthIT friends,

Today I thought I’d share the first ten (of 20) strong suggestions I put together into slides for other Applied Clinical Informaticists, or those considering a career in the field. I’m hoping this helps shed light on the importance and value of this role in modern healthcare, and how it helps to evaluate, implement, and maintain clinical technology and content.

First, my number one piece of advice to newcomers: Always map the current-state and future-state workflows. While some might argue this is an unnecessary step, this exercise will benefit you in some very important ways:

• It will help you understand and relate to your end-users.
• It will help you determine just how much work it takes to get from your current state (Point A) to your future state (Point B), which is necessary to help plan and allocate resources.
• It will help you develop blueprints, develop downtime forms, identify stakeholders, and scope/prioritize your projects.

For my second recommendation, I’d like to share how to write a good task, and then a good procedure. Learning to write a good task and procedure is so instrumental in building or untangling workflows, that it can even be used as a quick substitute for swim-lane diagrams (e.g., when trying to quickly document a workflow during a video chat with clinical end-users).

Recommendation number 3 involves something that sounds dull, until you learn about how it impacts your infrastructure and operations: document management. Learning how to create, edit, and archive documents can actually be a very powerful way of shaping or augmenting workflows in your electronic medical record. My mantra for newcomers: “Learn to control your documents, before they control you.”

My next recommendation (number four) is to learn the basic structure of healthcare operations by understanding the relationship between Administrative, Academic, Research, and Clinical Enterprises (Note: smaller community hospitals typically only have Academic and Clinical enterprises.)

In short, Administration supports the needs of the Academic, Research, and Clinical Enterprises. Learning how to navigate the people in these areas will help you break down silos, untangle workflows, and improve collaboration.

Coming in at number five is my recommendation to care about hard work, details, and precision. “In healthcare, there are no shortcuts.” While timelines are short, and there is often pressure to move ahead, try to resist the temptation to serve workflows that are not complete. (They may get you across the project finish line, but you risk having to do the whole project again — especially if end-users are not satisfied with the results.)

Recommendation number six might come as a surprise to some: When working in a team, file-naming conventions really matter. Group files should be both easy-to-find and easy-to-identify. My own personal favorite:

DRAFT/FINAL – ARCHETYPE – Descriptor – Created/Updated/Approved mm-dd-yyyy.ext

To offer some clarification:

• DRAFT/FINAL = Use ‘draft’ for documents in development, and ‘final’ when approved
• ARCHETYPE = Describes the file type (e.g. Education, Budget, Order Set, Catalog, Index, Contract, Policy, Protocol, Guideline, Schedule, Bylaws, Notes, Slide, Screenshot, etc.)
• Descriptor = Describes a unique identifier for the file (e.g., “ICU DKA Treatment Discussion”, “Meeting with Dr. Smith”, “Malaria Workup”, etc.)
• Created/Updated/Approved = Use ‘created’ when first creating a file, ‘updated’ when updating a file, and ‘approved’ when creating a final version
• mm-dd-yyyy = Describes when the file was created, updated, or approved
• ext = File extension (e.g., “.docx” or “.PDF”, etc.)

My next recommendation (number 7) is to learn the 24 basic tools that shape all clinical workflows: 12 are typically outside of the EMR, and the other 12 are found inside of it. Understanding the basic functions and design of each of these tools will help you to better plan projects, identify deliverables, identify stakeholders, and create smooth, complete clinical workflows.

Coming in at number eight is my general recommendation to all Applied Clinical Informaticists to care about the entire ‘Informatics tree’, including both the ‘Data In’ and ‘Data Out’ branches. While most people will gravitate toward one area, understanding the whole tree will broaden your perspectives and skill sets, and overall help you plan workflows.

My ninth recommendation for Applied Clinical Informaticists seeking to design smooth workflows comes from this 2015 blog post, where I talked about the importance of learning the relationship between concepts, terminology, templates, documents, and workflows.

Organizational Support (#8) is necessary to:

• Identify the concepts and ontologies (#7) that help you…
• Develop the definitions, terminology, and standards (#6) that you need to…
• Develop the templates and archetypes (#5) that will help you…
• Create the documents and tools (#4) that, combined, will help to…
• Create and support the workflows and processes (#3) that, if designed properly, will…
• Align with your goals and regulations (#2) which should…
• Align with your Mission and Vision (#1).

Typically, after first understanding #2, Applied Clinical Informaticists will concern themselves with aligning levels #7 to #3 of this pyramid. (Learning how pyramid levels #8-5 impact the documents in #4 can help you troubleshoot even the most complicated workflows in #3.)

Finally, my tenth recommendation for Applied Clinical Informaticists seeking to design smooth workflow is to care deeply about change management. While Kotter’s 8-step change management model is an excellent foundation, I recommend beginning with a standard, linear waterfall project model and then expanding it slightly for healthcare purposes, to include the following:

1. Conception, Determination, and Documentation of Need for Change
2. Evaluation, Analysis, Scoping, Presentation, Prioritization, and Approval for Change
3. Project Planning
4. Drafting of Change
5. Building of Change
6. Testing of Change
7. Final Approval of Change (go/no-go discussion)
8. Communication and Education of Change
9. Implication/Publication (‘Go-Live’) of Change
10. Monitoring and Support of Change

Once these ten steps are laid out, you can begin looking at the tasks beneath each step and developing your own ‘waterfall-meets-healthcare’-type change management strategy.

I hope this set of slides is helpful. Feel free to leave comments below with any thoughts or feedback. In my next post, we will look at another 10 of my strong recommendations for Applied Clinical Informaticists seeking to design smooth workflows.

This piece was written Dirk Stanley, MD, a board-certified hospitalist, informaticist, workflow designer, and CMIO, on his blog, CMIO Perspective.

John Halamka, MD, President, Mayo Clinic Platform

We have written often about the social determinants of health (SDOH), a problem that continues to challenge healthcare policy makers and bedside clinicians alike. In a recent New York Times article, Nicholas Kristof described the profession’s inability to adequately address the issue as “the scandal that is American Health Care.”

His statistics certainly justify this indictment:

• The average person in Mississippi will not make it past age 72 (71.9), compared to Bangladesh, where life expectancy is 72.4.
• Life expectancy in Alabama, Arkansas, Kentucky, and Louisiana are less than 74 while those living in Japan, Australia, and South Korea typically live to age 84 or beyond.
• Neonates in India, Rwanda, and Venezuela have a longer life expectancy than Native American newborns.
• About 150,000 toe, foot, and leg amputations are performed in the United States each year, making us “a world leader.” They’re mostly the result of poorly managed diabetes.

These troubling statistics are inconsistent with the amount of money the U.S. invests in healthcare. According to the CDC, the nation spends about $4.3 trillion per year, about twice as much as other wealthy countries. Although there are numerous contributing causes for this American scandal, SDOH play a major role.

The CDC describes these determinants as “conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies, racism, climate change, and political systems.” Addressing these issues will require systemic changes in national policies and changes in the attitudes of decision makers nationwide. But there are also measures that clinicians and technologists can implement that will have an impact.

Addressing “underlying problems”

In recent years, healthcare providers have become much more aware of the impact of social factors on their patients’ health and have taken steps to address these problems. Many have come to realize that improving clinical outcomes will never happen without fixing these underlying problems. Expecting patients to lower their HbA1c levels when they can’t afford the medication needed to lower their blood glucose is unreasonable; so is asking them to arrive for a clinic appointment when they have no viable means of transportation. Equally unreasonable: scolding them because they won’t replace their empty calorie snacks with fresh fruits and vegetables when they live in a food desert.

Many clinicians are tackling these issues by referring patients to a variety of community-based non-profit organizations. While these referrals can impact SDOH, some clinicians are moving beyond this screen-and-refer approach and taking a more direct interventional approach. Rahul Vanjani, MD, with Brown University, and associates suggest it may be time for physicians to serve as “street-level bureaucrats.”

Tapping social resources

Vanjani et al point out that there are many social resources can be tapped with the help of a physician’s direct input and signature. They cite the case of a diabetic patient living with his mother who was about to die, at which point he would no longer be eligible for the same subsidized housing his mother had been eligible for. The physician in care of the case spent three minutes printing out and signing a federal form — the US Department of Housing and Urban Development (HUD) Verification of Disability — which enabled the patient to remain in his mother’s apartment.

The authors point out that many clinicians don’t know anything about the “often untapped power of medical documentation… A lack of adequate paperwork is often used to justify denial of social services for which people are otherwise eligible, and clinicians can help patients get the necessary paperwork to document medical eligibility for services such as disability housing, prevention of utilities shut-off, disability income, improvement of housing conditions, prevention of incarceration, waiving of court fines and fees, and more.”

Of course, fixing social problems is only part of the solution. Our healthcare ecosystem devotes far too many resources to the treatment of late-stage disease and far too little on prevention and early detection. One study found that early detection of cancer would save $26 billion per year in the U.S. Several digital solutions are emerging to make early detection more readily available to clinicians and patients. For instance, a deep learning algorithm called Sybil, which is designed to predict the risk of lung cancer, has been validated on three independent data sets and has the potential to identify those at risk based on a single low dose chest CT scan.

At Mayo Clinic Platform, we have spent the last several years developing similar digital tools to improve early detection of disease. In partnership with Lucem Health, we have created an agnostic AI platform that can help detect prediabetes and diabetes and improve the early diagnosis of colorectal cancer. And in partnership with Anumana, we have created deep learning-based algorithms for the early detection of atrial fibrillation and other cardiovascular disorders.

The healthcare ecosystem faces major challenges, but with the right combination of compassion, teamwork, and machine learning, we see a bright future ahead.

This piece was written by John Halamka, MD, President, and Paul Cerrato, senior research analyst and communications specialist, Mayo Clinic Platform. To view their blog, click here.

Dirk Stanley, MD, CMIO, UCoon Health

Hi fellow CMIOs, CNIOs, and other Applied Clinical Informatics and #HealthIT friends,

Today I thought I’d share the first ten (of 20) strong suggestions I put together into slides for other Applied Clinical Informaticists, or those considering a career in the field. I’m hoping this helps shed light on the importance and value of this role in modern healthcare, and how it helps to evaluate, implement, and maintain clinical technology and content.

First, my number one piece of advice to newcomers: Always map the current-state and future-state workflows. While some might argue this is an unnecessary step, this exercise will benefit you in some very important ways:

• It will help you understand and relate to your end-users.
• It will help you determine just how much work it takes to get from your current state (Point A) to your future state (Point B), which is necessary to help plan and allocate resources.
• It will help you develop blueprints, develop downtime forms, identify stakeholders, and scope/prioritize your projects.

For my second recommendation, I’d like to share how to write a good task, and then a good procedure. Learning to write a good task and procedure is so instrumental in building or untangling workflows, that it can even be used as a quick substitute for swim-lane diagrams (e.g., when trying to quickly document a workflow during a video chat with clinical end-users).

Recommendation number 3 involves something that sounds dull, until you learn about how it impacts your infrastructure and operations: document management. Learning how to create, edit, and archive documents can actually be a very powerful way of shaping or augmenting workflows in your electronic medical record. My mantra for newcomers: “Learn to control your documents, before they control you.”

My next recommendation (number four) is to learn the basic structure of healthcare operations by understanding the relationship between Administrative, Academic, Research, and Clinical Enterprises (Note: smaller community hospitals typically only have Academic and Clinical enterprises.)

In short, Administration supports the needs of the Academic, Research, and Clinical Enterprises. Learning how to navigate the people in these areas will help you break down silos, untangle workflows, and improve collaboration.

Coming in at number five is my recommendation to care about hard work, details, and precision. “In healthcare, there are no shortcuts.” While timelines are short, and there is often pressure to move ahead, try to resist the temptation to serve workflows that are not complete. (They may get you across the project finish line, but you risk having to do the whole project again — especially if end-users are not satisfied with the results.)

Recommendation number six might come as a surprise to some: When working in a team, file-naming conventions really matter. Group files should be both easy-to-find and easy-to-identify. My own personal favorite:

DRAFT/FINAL – ARCHETYPE – Descriptor – Created/Updated/Approved mm-dd-yyyy.ext

To offer some clarification:

• DRAFT/FINAL = Use ‘draft’ for documents in development, and ‘final’ when approved
• ARCHETYPE = Describes the file type (e.g. Education, Budget, Order Set, Catalog, Index, Contract, Policy, Protocol, Guideline, Schedule, Bylaws, Notes, Slide, Screenshot, etc.)
• Descriptor = Describes a unique identifier for the file (e.g., “ICU DKA Treatment Discussion”, “Meeting with Dr. Smith”, “Malaria Workup”, etc.)
• Created/Updated/Approved = Use ‘created’ when first creating a file, ‘updated’ when updating a file, and ‘approved’ when creating a final version
• mm-dd-yyyy = Describes when the file was created, updated, or approved
• ext = File extension (e.g., “.docx” or “.PDF”, etc.)

My next recommendation (number 7) is to learn the 24 basic tools that shape all clinical workflows: 12 are typically outside of the EMR, and the other 12 are found inside of it. Understanding the basic functions and design of each of these tools will help you to better plan projects, identify deliverables, identify stakeholders, and create smooth, complete clinical workflows.

Coming in at number eight is my general recommendation to all Applied Clinical Informaticists to care about the entire ‘Informatics tree’, including both the ‘Data In’ and ‘Data Out’ branches. While most people will gravitate toward one area, understanding the whole tree will broaden your perspectives and skill sets, and overall help you plan workflows.

My ninth recommendation for Applied Clinical Informaticists seeking to design smooth workflows comes from this 2015 blog post, where I talked about the importance of learning the relationship between concepts, terminology, templates, documents, and workflows.

Organizational Support (#8) is necessary to:

• Identify the concepts and ontologies (#7) that help you…
• Develop the definitions, terminology, and standards (#6) that you need to…
• Develop the templates and archetypes (#5) that will help you…
• Create the documents and tools (#4) that, combined, will help to…
• Create and support the workflows and processes (#3) that, if designed properly, will…
• Align with your goals and regulations (#2) which should…
• Align with your Mission and Vision (#1).

Typically, after first understanding #2, Applied Clinical Informaticists will concern themselves with aligning levels #7 to #3 of this pyramid. (Learning how pyramid levels #8-5 impact the documents in #4 can help you troubleshoot even the most complicated workflows in #3.)

Finally, my tenth recommendation for Applied Clinical Informaticists seeking to design smooth workflow is to care deeply about change management. While Kotter’s 8-step change management model is an excellent foundation, I recommend beginning with a standard, linear waterfall project model and then expanding it slightly for healthcare purposes, to include the following:

1. Conception, Determination, and Documentation of Need for Change
2. Evaluation, Analysis, Scoping, Presentation, Prioritization, and Approval for Change
3. Project Planning
4. Drafting of Change
5. Building of Change
6. Testing of Change
7. Final Approval of Change (go/no-go discussion)
8. Communication and Education of Change
9. Implication/Publication (‘Go-Live’) of Change
10. Monitoring and Support of Change

Once these ten steps are laid out, you can begin looking at the tasks beneath each step and developing your own ‘waterfall-meets-healthcare’-type change management strategy.

I hope this set of slides is helpful. Feel free to leave comments below with any thoughts or feedback. In my next post, we will look at another 10 of my strong recommendations for Applied Clinical Informaticists seeking to design smooth workflows.

This piece was written Dirk Stanley, MD, a board-certified hospitalist, informaticist, workflow designer, and CMIO, on his blog, CMIO Perspective.

John Halamka, MD, President, Mayo Clinic Platform

We have written often about the social determinants of health (SDOH), a problem that continues to challenge healthcare policy makers and bedside clinicians alike. In a recent New York Times article, Nicholas Kristof described the profession’s inability to adequately address the issue as “the scandal that is American Health Care.”

His statistics certainly justify this indictment:

• The average person in Mississippi will not make it past age 72 (71.9), compared to Bangladesh, where life expectancy is 72.4.
• Life expectancy in Alabama, Arkansas, Kentucky, and Louisiana are less than 74 while those living in Japan, Australia, and South Korea typically live to age 84 or beyond.
• Neonates in India, Rwanda, and Venezuela have a longer life expectancy than Native American newborns.
• About 150,000 toe, foot, and leg amputations are performed in the United States each year, making us “a world leader.” They’re mostly the result of poorly managed diabetes.

These troubling statistics are inconsistent with the amount of money the U.S. invests in healthcare. According to the CDC, the nation spends about $4.3 trillion per year, about twice as much as other wealthy countries. Although there are numerous contributing causes for this American scandal, SDOH play a major role.

The CDC describes these determinants as “conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies, racism, climate change, and political systems.” Addressing these issues will require systemic changes in national policies and changes in the attitudes of decision makers nationwide. But there are also measures that clinicians and technologists can implement that will have an impact.

Addressing “underlying problems”

In recent years, healthcare providers have become much more aware of the impact of social factors on their patients’ health and have taken steps to address these problems. Many have come to realize that improving clinical outcomes will never happen without fixing these underlying problems. Expecting patients to lower their HbA1c levels when they can’t afford the medication needed to lower their blood glucose is unreasonable; so is asking them to arrive for a clinic appointment when they have no viable means of transportation. Equally unreasonable: scolding them because they won’t replace their empty calorie snacks with fresh fruits and vegetables when they live in a food desert.

Many clinicians are tackling these issues by referring patients to a variety of community-based non-profit organizations. While these referrals can impact SDOH, some clinicians are moving beyond this screen-and-refer approach and taking a more direct interventional approach. Rahul Vanjani, MD, with Brown University, and associates suggest it may be time for physicians to serve as “street-level bureaucrats.”

Tapping social resources

Vanjani et al point out that there are many social resources can be tapped with the help of a physician’s direct input and signature. They cite the case of a diabetic patient living with his mother who was about to die, at which point he would no longer be eligible for the same subsidized housing his mother had been eligible for. The physician in care of the case spent three minutes printing out and signing a federal form — the US Department of Housing and Urban Development (HUD) Verification of Disability — which enabled the patient to remain in his mother’s apartment.

The authors point out that many clinicians don’t know anything about the “often untapped power of medical documentation… A lack of adequate paperwork is often used to justify denial of social services for which people are otherwise eligible, and clinicians can help patients get the necessary paperwork to document medical eligibility for services such as disability housing, prevention of utilities shut-off, disability income, improvement of housing conditions, prevention of incarceration, waiving of court fines and fees, and more.”

Of course, fixing social problems is only part of the solution. Our healthcare ecosystem devotes far too many resources to the treatment of late-stage disease and far too little on prevention and early detection. One study found that early detection of cancer would save $26 billion per year in the U.S. Several digital solutions are emerging to make early detection more readily available to clinicians and patients. For instance, a deep learning algorithm called Sybil, which is designed to predict the risk of lung cancer, has been validated on three independent data sets and has the potential to identify those at risk based on a single low dose chest CT scan.

At Mayo Clinic Platform, we have spent the last several years developing similar digital tools to improve early detection of disease. In partnership with Lucem Health, we have created an agnostic AI platform that can help detect prediabetes and diabetes and improve the early diagnosis of colorectal cancer. And in partnership with Anumana, we have created deep learning-based algorithms for the early detection of atrial fibrillation and other cardiovascular disorders.

The healthcare ecosystem faces major challenges, but with the right combination of compassion, teamwork, and machine learning, we see a bright future ahead.

This piece was written by John Halamka, MD, President, and Paul Cerrato, senior research analyst and communications specialist, Mayo Clinic Platform. To view their blog, click here.