My Life as a Startup CMIO: Conversations with a Hospital CIO

Posted by Narath Carlile on February 02, 2018 · 7 mins read

(As a physician, computer scientist, entrepreneur and caregiver I care deeply about making healthcare better. So do most of the people that I meet in healthcare, and yet it is still so broken. These conversations are a chance to capture other points of view to make sure those working to make healthcare and caring better can come up with solutions which will have more impact. These start with a conversation in the style of Jane Jacobs and Douglas Hofstadter whose work I admire).

An innovative CIO and a startup CMIO are meeting. They have previously done innovation projects together within the CIO’s current institution. The CIO has recently been promoted to his current position. The organization recently completed transition to a new EMR. He is adapting to his new role and its multiple responsibilities. This is the first time they have met since he took on this new role.

Startup CMIO: How is the new position?

CIO: Very interesting, and incredibly busy. I have so many emails and so many meetings each day it is hard to do anything else.

Startup CMIO: It sounds like a really challenging environment, but one in which I think you will thrive. I appreciate you being able to take some time to talk about creating an innovation architecture for the hospital.

CIO: No problem, as you know I’m passionate about continuing to support needed innovations. I am however particularly frustrated right now, since I’m approached by so many vendors who want me to adopt their solution. It takes a lot of my time.

Startup CMIO: (Chuckles) I know that well, but from the other side. Interestingly the CIO is often the last person we (as innovators) want to speak to since they often just say no outright.

CIO: Ha, too true. You and I were on that other innovator side together not too long ago. Now I am the same person, but I feel very strongly the desire to say no to everything. I don’t have the time, staff or budget to get done the things we have already committed to, let alone anything new or “innovative”.

Startup CMIO: Sounds like a constraint-driven problem to me. In these problem spaces, knowing what is fixed and what is flexible — and then understanding prioritization — are the keys to optimization. Would you mind if we explored first what is fixed and what is flexible in your constraints? Second, we can explore how you prioritize within your organization. Finally we could see if we can come up with some lessons or guidance for informatics innovation that might be useful to others in your or my position. Is that a reasonable plan?

CIO: Sounds like a good plan to me. Let’s switch to document mode so that we can do a better job at creating rapidly digestible outlines for others to use.

Understanding the constraints of a hospital CIO

A hospital CIO is primarily constrained by:

  • Budget
  • Staffing
  • Heterogeneous infrastructure
  • Regulation
  • Organizational politics


The budget is determined by a hospital committee under the direction (in general) of the CFO. The CIO has significant input into the budget for the upcoming year, often helping to determine key technology investments. The budget is often further constrained by organizational strategic objectives (such as switching to a commercial EMR solution). In general, hospitals are also not-for-profit institutions with limited capacity for “visionary” or long term investments (even if they have significant future dividends). Hospitals also have live-or-die budget requirements, so it can be expected that submitted budgets will be cut by 30–50%. Investments with ROI greater than 5 years (unless aligned with core strategic objectives and then still often first to be cut) are unlikely to be funded.


Hospital IT departments (1) are usually understaffed since they are considered a cost center, rather than revenue generator; (2) do not easily attract top talent since they are typically not using the latest technology, and it is hard to compete with private sector commercial jobs; and (3) have a high turnover, especially in specialized areas (such as new EMR programming, interface development).

Heterogeneous infrastructure

Hospitals, as mid- to large-size enterprises which often provide multiple service lines (inpatient, critical care, surgeries, oncology, outpatient and others), have a series of technologies which have been bolted together. Some of these may be so old that they no longer have commercial support. A series of consultants may have recommended large strategic investments, but these are rarely applied uniformly, and given the budget limitations above, working infrastructure is rarely replaced to support simpler future maintenance. Hospital CIOs therefore inherit all of the previous — generally partially completed — projects of all prior CIOs.


There are a significant number of regulatory restrictions on hospital technology and information systems including ongoing certifications which require CIO staff time. Failure of certification can have significant impacts on the hospital ranging from (in the worst case) shutting down operations of some or all service lines, to fines, to public embarrassment or a decrease in ratings which can affect hospital revenue. There are also significant time consuming efforts that require a focus on policy requirements, such as ICD10 and Meaningful Use.


Hospitals are complex institutions often with multiple masters, including not just the CEO and CFO, but can also include the Board, as well as powerful physician groups within the hospital. Since informatics as infrastructure is often organization-wide, this work involves a very large number of stakeholders. In addition, many of these groups require custom work or custom innovations which tax the generally common pool of IT staff.


Startup CMIO: It’s a lot more complicated than I had thought, and really helpful for me to know about these issues, especially if my proposed solution has impacts in these areas. How on earth are you able to balance all of this?

CIO: It’s all about prioritization. Let’s talk about that next time.

Previously published on Medium on 02 Feb 2018.