top of page
  • Writer's pictureAaron Neinstein

Thoughts on a Career in Digital Health and Informatics… Would I Do It Again?

Updated: Feb 12, 2021

I’m frequently approached by medical students, residents, and fellows interested in careers in informatics and digital health.  I love these conversations.  It is so fun and rewarding to be part of someone’s journey to discovering their career passion.  During my most recent conversation, I decided to try to capture it to create a post that might help many many more future students and trainees in the future.  So, this will read like part blog-post, part-interview, because much of it is an edited conversation transcript (which means some run-on sentences too!).

What was my journey to a career in informatics? Bob Wachter has always said that he is what happens when a political science major goes to medical school.  Well… I was an American Studies major and also a bit of a tech nerd growing up.  I lived on the computer growing up in the 1980s, was the first person I knew to have a modem and have email (Los Angeles Freenet!), played a whole lot of Civilization, took computer programming in high school and learned how to code on a TI-82 calculator and in Basic, and then went to college and became an American Studies major because I loved the blend of the humanities, between political science, sociology, history, and english.

At the same time, I was incredibly fortunate to get to spend a summer with Dr. Francine Kaufman at CHLA doing research and shadowing her in diabetes clinic, including a week on medical staff at Camp Conrad-Chinnook diabetes sleepaway camp.  I knew right away I wanted to be an Endocrinologist, but the techie inside me was still awake, and the following summer I was blessed to be able to work in the lab of Dr. Udo Hoss, who was developing the first continuous glucose monitor at MiniMed in Northridge.  These two passions came together in my senior thesis at Northwestern, studying Project Dulce in San Diego with Dr. Athena Philis-Tsimikas and writing about the Social and Cultural Barriers to the Treatment of Type 2 Diabetes in Urban Latinos.

In residency at UCSF, I spent a significant amount of time with a small group, including Drs. Read Pierce, Patrick Kneeland, Lucy Kalanathi, and several others learning Leadership and Quality Improvement, which was at the time of the 2007 HITECH Act and the 2010 Affordable Care Act, each having seismic shifts on the American healthcare landscape.  It was a time of profound change in health policy, and in particular around health IT policy.  I spent weekends during residency sitting on the couch, watching football with friends working at Google, talking all the time about the improvements, especially in technology, that were needed in healthcare.  Small projects during residency included hospital systems redesign for anticoagulation and redesigning insulin order sets for a future computerized provider order entry system (CPOE), a precursor to a full EHR.  I really found myself loving the confluence of policy, technology, quality improvement.

Over the Christmas holiday in 2009, I got the opportunity to “put my money where my mouth was.”  I was on the wards with Dr. Adrienne Green, at the time Associate Chief Medical Officer, and told her that I was interested in a career in health tech, and she said, “you should go talk to Dr. Russ Cucina, because we just signed a contract with a company called Epic” to build our new electronic health records at UCSF.  I did, was offered a job, and was on a plane to Verona, WI in May 2010 for Epic training, spending the 2010-2011 academic year helping to build UCSF’s initial ambulatory EHR implementation, develop training materials, train our physicians, and launch the EHR in primary care and specialty practices.

I went back to training in 2011 for my Endocrinology fellowship, since I had known for many years that I wanted to be an Endocrinologist, so I temporarily stepped away from Informatics.  However, in 2012, I decided to spend my research time in fellowship building a non-profit software company called Tidepool.  We had realized at the time that diabetes software was… terrible… and that nobody could access a complete set of device data for a patient using multiple diabetes devices, and that the patient could almost never easily access their own device data.  I was so fortunate to connect with Drs. Saleh Adi and Jenise Wong at UCSF, and Steve McCanne and Howard Look (two tech industry diabetes dads), and we were inspired by the Open mHealth movement.  So, I spent the year working with that incredible team, trying to get probono legal help, branding and design help (thank you Marco Flavio Marinucci!), applying for funding (and getting some from a Sanofi Data Design Diabetes competition), and generally trying to figure out how to launch a company, ultimately deciding to do so as a non-profit.

At the same time, Dr. Michael Blum at UCSF, who was CMIO at the time, was starting the Center for Digital Health Innovation, building upon the foundation of our new EHR to form partnerships with Silicon Valley and continue to rethink health IT, and he invited me to rejoin the faculty.  Since then, in 2013, I’ve been part of trying to build our CDHI team and programs – working on building out a more connected care ecosystem, with interoperability and APIs, and leveraging patient-generated health data.  Our team has grown from a small team of mostly physician faculty to a team of software developers, solutions architects, product managers, designers, health policy experts, business development and communications, data scientists, and more, working both internally as an innovations and R&D team within UCSF and partnering with external companies.

What do the different career paths look like for someone who is interested in “health tech?”

Informatics is a super broad field. It encompasses bits and pieces of leadership, change management, software development, data science, clinical research, health policy, quality improvement and systems change, technical architecture, and more. The profiles of each person who is an informaticist can be entirely different.  You can be a PhD and do bioinformatics and work on techniques for genome sequencing.  My good friend, Dr. Eli Van Allen, is an Oncologist and does computational cancer genomics. So, his profile and skillset looks quite different than mine.  The type of work you can do could very broadly from more the pure research end of things to more applied things.

Through another lens, there is the question of who can you go work for.  You could be in academics, or you could be at a big tech company like Apple or Google or Microsoft, or you work at a startup tech company, or you can go work for a digital-first clinic (like Virta, Steady Health, or Omada).  Or, you can go work for a pharma company or a med device company doing digital health.

As you think about skillset development, there are some people who do full stack software development, some who do more data science and analytics, some who do design, usability, human computer interaction.  There are people who take a more traditional CMIO role who help lead the use of the electronic health record within a care delivery system.

Another big difference is there are physicians who do informatics within their own medical specialty versus physicians who take a more enterprise, whole organization perspective, and just happen to practice within a clinical specialty.  For example, most of what our team works on at CDHI is completely unrelated to diabetes and endocrinology – it’s really institutional-level, enterprise focused interventions, or projects focused on a particular service line like Orthopedics, Cardiology, or Lung Transplant.  But there are a lot of people who choose differently for their informatics careers, where they say, I’m an Orthopedist and my informatics work is focused just on the Orthopedics department.  So, that’s a big decision to make whether you want to go broad vs deep, and there are pros and cons to both.  No right answer.  They’re both good options.  What’s great about the latter, and I feel this when I do work on diabetes-related digital health projects, is that it aligns with my clinical passion and aligns with an area that I’m very comfortable with and I understand deeply.  It can feel like a more natural place to work. On the other hand, from a career options perspective, if you want to have a more senior administrative role in a health system, you will have to be ready to take a broader perspective.

What is an Informaticist?

Our professional society, AMIA, has a set of competencies that it has defined for the profession.  My simple explanation is that being an informaticist requires a handful of broad skills, including: Leadership, change management, data science and statistics, clinical research, user experience and human-computer interaction, health policy, strategy, economics and business, project management, and clinical medicine.  How many of these, and how much of each of these, any given informaticist has is very different.  Each one of us is like assembling a Madden NFL player, who might be a 95 in speed and a 65 in hands and 70 in vertical leap.  You could imagine a spider matrix or a Player Card for each informaticist that could describe to each other where our interests lie.  This would be helpful in differentiating where we spend our time, and in helping to differentiate how Informaticist A differs from Informaticist B.

Here’s mine:

What do we do at the UCSF Center for Digital Health Innovation, and how is that different from other groups at UCSF who work on health tech?

Center for Digital Health Innovation Our Mission at CDHI is: Creating effective digital solutions that transform health and enable compassionate care delivery for all.

Our Vision is: We make better health accessible for everyone, everywhere.

We work with large tech company partners like Samsung, Cisco, and GE, with other industry giants like WeHealth, with startup tech companies like Voalte, and with internal clinical and operational partners. Our team’s methodology is agile product design as per the image below.

Some project examples include:

  1. Referrals Automation: SMARTonFHIR application improving efficiency in PCP-to-Specialist referral process, enabling faster care, improved staff experience, and improving operational efficiency by >40%

  2. COVID19 Employee Daily Health Screener: Daily employee health screener to rapidly screen the population for

  3. Virtual Lung Transplant Care: Automated care chats, driving deeper engagement with post-lung transplant patients

  4. CareWeb: Next-generation care-team messaging platform

I’m blessed at UCSF to be part of a vibrant, broad group of faculty and centers who work in informatics. Some of these groups include:

  1. Bakar Computational Health Sciences Institute, led by Dr. Atul Butte. Advancing computational health sciences in research, practice and education — in support of Precision Medicine for all. Building a foundation of faculty and knowledge assets in computational health sciences and bringing together a community of thinkers interested in this emerging field.

  2. The Center for Clinical Informatics and Improvement Research (CLIIR), led by Dr. Julia Adler-Milstein, a national leader in healthcare policy research. Created out of a critical need to understand how to use digital tools to improve quality and value of healthcare.  They don’t build digital tools, but rather study their use and effectiveness.

  3. Health Informatics, led by CHIO Dr. Russ Cucina. Interdisciplinary clinical leadership of the design, development, adoption, and application of information services in health care delivery. They provide clinical leadership for UCSF’s electronic health record (EHR), data and analytic capabilities, and related systems to serve patients, their families, and our healthcare providers.

  4. Clinical Innovation Center, led by Dr. Ralph Gonzales. They help accelerate innovation to solve critical delivery system challenges at UCSF.

  5. SOLVE, led by Drs. Urmimala Sarkar and Courtney Lyles. The first health equity accelerator dedicated to adapting health technology for Medicaid and other vulnerable populations.

  6. UCSF Innovation Ventures: Responsible for technology transfer and commercialization.

  7. CTSI, Informatics & Research Innovation Program, led by Dr. Mark Pletcher. They work to make data more accessible and usable for research, and provides infrastructure for EHR-based clinical trials interventions.

  8. School of Medicine Technology Services, led by Kristin Chu. They provide leadership, advocacy and management across the School of Medicine’s technology spectrum, from research to development to data security.  They’re a resource for consultation, discovery, and development of technology tools.

  9. There are so many more faculty like Dr. Ida Sim, Dr. Laura Esserman, Dr. Stefano Bini, and others at UCSF doing incredible work in informatics.

What happens for people after they leave UCSF?

People have left UCSF to go on to so many exciting and important jobs in digital health. Some of them are:

  1. Dr. Carolyn Jasik – Chief Medical Officer, Omada Health

  2. Dr. Ami Parekh – Chief Medical Officer, Grand Rounds

  3. Dr. Alvin Rajkomar – Product Manager, Google

  4. Dr. Greg Burrell – Co-Founder, VP Clinical Product Strategy of Carbon Health

  5. Dr. Priyanka Agarwal – Digital Health Director, Myokardia

  6. Dr. Seth Bokser – Chief Medical Officer, OneView

  7. Dr. Carmen Peralta - Chief Medical Officer, Cricket Health

There are many more than this, but this is a sampling.

What about an informatics fellowship training program?

UCSF has had one for four years now.  These did not exist when I started – I was grandfathered in and was able to sit for the Informatics boards exam in 2014 based on practice experience.  For someone who has gone straight through school and training, I think an Informatics fellowship makes a lot of sense.  There are other people who have not done so – who may have had jobs at McKinsey doing consulting or writing code for a software company, or who have built other similar experience over the years outside of their medical training.  For some of these people, a fellowship may be of less value than the first group, since one of the biggest values to a fellowship is being in the environment with opportunities to lead a project from start to finish. Having the time and space and environment to pick a project and go do it is so key – at the end of the day, your ability to get a job and to be effective is based on your ability to demonstrate effectiveness at leading projects and delivering outcomes.  And you can show that either by having been out in the workforce and executed or by doing a fellowship where you have the opportunity to go do the same.  Also, if you’re not sure yet which area of informatics you want to work in, and you want to explore different areas of opportunity, a fellowship is a great way to do that.  You can try data science, try app development, try research, try implementing a clinical decision support tool in the EHR.

Do most informaticists continue to have a clinical practice?

Yes, most still do. For the people who leave health systems and leave academia to work in industry, this gets harder… not impossible, but more difficult to do.  It is not too dissimilar from academic faculty who have a research career. Everyone thinks they want to do 50% clinical and 50% other, but that is rarely a reality.  You kind of have to pick 70-80% of one or the other.  It can be challenging – for me, now that I run a larger team of 20-30 people… every Tuesday I have clinic and I go poof.  Emails and Slack messages are coming in, people have questions to ask me, and I’m dedicated to seeing patients and I’m not available to answer.  But, I value doing this so highly.  I love my clinical practice.

Do you feel like learning to code is important for physicians who want to do this kind of work, or do you mainly need to become a leader without knowing the nitty gritty?

I think I got lucky in a way because I joined the profession early enough that I was grandfathered in without a “hard skill.”  I think as the field becomes more saturated that having a hard skill is a differentiator, and it can give you a huge amount of autonomy and flexibility.  If you have an idea you want to pursue, you can do it on your own, without having to go find someone with that skill to help you. Having said that, I don’t think there’s huge value in becoming the best at it because that’s probably not what you’re going to do long-term as a physician. But the ability to whip up a proof of concept is pretty powerful. Now that’s more from the full-stack app development perspective.  From the perspective of analytics and data science, I would go a step further. For physicians who can develop skills in Python right now… that is a highly valuable and marketable skill.  This is true both in the research world and I think also from the health system perspective.  It can sometimes save a lot of back-and-forth on metric development if you are a physician, and you already understand what the data means and the workflows that generate those data, and you can also develop the analytics or dashboards yourself.  My friend and former colleague Dr. Ari Robicsek, now at Providence, and my friend Dr. Sara Murray at UCSF… they have done some absolutely incredible work leading health system analytics teams.  It’s a very marketable skill in my opinion, that is on the upswing.  The modern currency is data.  So for people who have to wait around to get help… you’re stuck.  If you can access data and you know what to do with it and you can actually start to make sense of it and run queries… your speed in terms of ability to move projects forward is just massively enhanced.

Later in your career, is it the best use of your time to be the one doing it as opposed to directing the team of people doing it? It’s like being in a lab where you learn how to do a Western Blot in your training.  Probably not, but for sure early in your career, huge benefit to being able to do it yourself.

Do you feel like this kind of work is feasible within academia?

It’s getting better. When I started out, there were many people, including some very senior ones in powerful positions, advising me that informatics was not a viable career option in academic medicine, especially if I was not going to be doing primarily informatics research. “Don’t bother.  Don’t do it.”  I’m so fortunate that I chose to carry on.  At times it felt like I had to fight like hell to keep my career options afloat. But, I’m happy to say that I think this has changed.  Academic medicine has now realized that this is a critical piece of the future of medicine.  It also does not hurt to have a Department Chair like Bob Wachter!  If you’re early in your career and you’re thinking about this as a career option, you should be very open about it.  Look for the training programs where they foster and encourage these types of career paths.  As far as I’m concerned, we should be packing our training programs with people with technology skills because it is so important for the future of medicine. If we don’t have people with these skills, we’re in big trouble.  My opinion is if there’s a program who thinks that a trainee having these skills and interest is a bad thing, you probably don’t want to be there.

Would I choose this path again?

1000%, yes. If you’re interested in pursuing a career in health tech and innovation, you should keep on the path. A more successful healthcare future will require significant technology to help us drive changes. There will be many opportunities for creative, smart people willing to challenge the status quo and use their skills in technology to drive change. Medicine needs young physicians who know how to code, who know data science, who know product design, who know business and entrepreneurship, who know implementation science, who know QI.  If you’re a great physician, and you have one or several of these skills, I’m always eager to meet you and look for opportunities to work together.

1,897 views0 comments

Recent Posts

See All

Smart Insulin Pens are here… finally.

The first “smart” insulin pen has finally hit the market. This is a big moment for diabetes care, as the digital toolbox expands. (I wrote a post in 2013 about this topic, asking for someone to make a


bottom of page