Grand Rounds : Turning Policy into Practice

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

I'm writing today to share some slides that I recently presented at a grand rounds for a very talented group of Northwell Health Clinical Informatics (CI) Fellows, courtesy of CI leaders Anncy Thomas, DO FAMIA and Keriann Latten, DNP.

The topic : Turning Policy into Practice. While I've discussed task grammar (the 'cupcake test') in the past, I wanted to first convey how Applied Clinical Informatics is constantly building on the work of the past - not replacing it - and so it's helpful to start our discussion by framing a big-picture look at exactly where we are in #Healthcare history :

While #Healthcare has been on a roughly 2000+ year journey, what we think of as modern Western medicine mostly started about 250 years ago, gradually evolving and bringing us to the inflection point we have reached in the last 20 years : Technology, payment reform, pandemics, and now Artificial Intelligence (AI). The key points to highlight during this recent part of the journey : 

• Change is happening faster than ever.
• Managing that constant, ongoing, accelerating change requires dedication, time, people, and resources.

Along with the global adoption of Electronic Health Records also came a need for a higher degree of architecture and engineering for our clinical workflows - Hence, our Applied Clinical Informatics professionals, helping #Healthcare to adopt a culture of well-orchestrated, well-organized, and well-developed design and implementation : 

This brings us to our discussion about turning policy into practice : It's not about rewriting policies - it's about adding a task grammar to help them become more executable. :

... which helps to better understand the workflow details, and connect them to real EHR build.

So now, for teaching purposes, let's imagine a very simple example of something that one might want to achieve - A well-baked and safe cupcake. 

While serving well-baked, safely-prepared cupcakes is always an admirable goal (who doesn't like cupcakes?), the goal itself is not operational. It leaves many questions unanswered, such as : 

• WHO is making the cupcakes?
• WHAT kind of cupcakes are they making?
• WHEN are they expected to make these cupcakes
• HOW are the cupcakes supposed to be prepared, to be safe (for the baker and the consumer)?
• WHY are we needing to establish this standard?

While these admirable (and usually necessary) standards exist all across #Healthcare, the modern question then becomes : Could you configure this in your EHR?

Before we go on, a word of respect (and thanks) to all of the #Healthcare leaders (globally) that got us to here in 2026. We are all standing on their shoulders, and so we owe them a great deal of gratitude and respect. The policies they developed helped to protect patients, align clinicians, and meet regulatory expectations : 

... which worked well before Electronic Health Records (EHRs), but did not always contain the task grammar (who, what, when, where, why, how) necessary for configuring them today.

Usually, this discussion raises the question : Why does Clinical Informatics (CI) get stuck in the middle here? It's not uncommon for CI professionals to get funny looks the first time they begin to discuss policies and compliance, but this translation is a necessary part of the job : 

So now, let's look at how to operationalize a policy by applying task grammar (who, what, when, where, why, how) to a sample 'aspirational' (not operational) policy : 

We can use this task grammar to define a task, the most granular unit of work : 

TASK = [ WHO ] will/may [ WHAT ] { how } { when } { where } { why } 

where :

• WHO = Who will perform the task
• will/may = Use WILL for required tasks, MAY for optional tasks
• WHAT = Brief description of the task
• { how } = Optional, use only to clarify how the task will/may be performed
• { when } = Optional, use only to clarify when the task will/may be performed
• { where } = Optional, use only to clarify where the task will/may be performed
• { why } = Optional, use only to clarify why the task will/may be performed 

... and use it to augment the procedure for improved EHR configuration purposes (a more 'operational policy')  : 

This augmentation using the above task grammar can significantly elevate the clarity and granularity of your policy and procedure, making it much easier to configure in an Electronic Health Record (EHR) : 

It's important to note that this culture change should never occur in a silo - and comes with potential risks and benefits - that you should always review with your own Clinical Leadership, and other Legal, Regulatory, and Compliance professionals.

In general, this culture shift can shift organizational risk patterns, and so you will want to talk to your own Clinical Leadership and Legal, Regulatory, and Compliance teams before implementing this new task grammar.

This new task grammar can also help you to update and refine your organizational definition of a Procedure : 

Procedure (aka process, workflow, recipe, algorithm) (n.) = A series of ordered TASKS that uses people, time, and resources to achieve a desired outcome.

... which, again, should only be undertaken in collaboration with your Clinical Leadership and Legal, Regulatory, and Compliance teams :

With this new task grammar at hand, you can then 'strengthen' your workflows (procedures) by better supporting them in your EHR configuration. You can also help ensure that their costs are better estimated, and that they are fully vetted (understood, reviewed, and secondarily approved) by the necessary stakeholders (usually Directors, VPs, Chairs, and Chiefs) before they receive final approvals (usually from your Clinical and/or Operational Leadership) : 

Since this task grammar also helps you define each task in a more granular manner, you can also provide a pretty decent estimate of the cost of each task, as well as the total (annual) cost of the policy : 

• TASK COST = TASK LABOR + TASK MATERIALS
• TASK COST = TASK (Time * Hourly salary) + TASK MATERIALS

... which can also help you to design workflows that help reduce costs by keeping everyone operating at the top of their license/certification. A simple way to demonstrate this effect is to now put the task grammar to work in a spreadsheet, allowing you to estimate the cost of making simple macaroni and cheese if a Doctor, a Nurse, or an MA prepares it (*Note: These salaries are very inaccurate guesses, but used for discussion purposes only!): 

... which all brings us back to the key data elements that this new task grammar can build onto a policy, to help make it more operational and configurable in an EHR : 

... and coordinating this policy with all of the other tools that shape workflow helps to make your workflows more clear, smooth, and predictable :

This brings us to some helpful take-home messages : 

• Policy writers help to create inspiration and aspiration.
• Clinical Informatics helps to create operation.
• Academic Medical Centers are often required to manage additional layers of complexity (usually created by trainees, supervision models, systems/multiple hospitals, research workflow, and extensive subspecialty variations)

As well as some final thoughts : 

• Modern #Healthcare is going through exponential levels of change.
• Continuously managing and supporting this change is key to EHR success.
• Understanding how to critically read and write policies can help reduce costs, reduce risks, and update/streamline your workflows. 

Remember, Applied Clinical Informatics is not about policy enforcement - It is about workflow translation, advocacy, and alignment.

You will also want to respect the work that got us here (and avoid policy debates) by choosing your terminology carefully when discussing this with your Clinical, Legal, Regulatory, and Compliance Leadership : 

Finally, I suggest that new Applied Clinical Informatics professionals should start with something small - Pick one aspirational policy, try adding the task grammar, work to convert it to a more operational policy, and then test your build feasibility. 


Remember : 

• Policies help define belief;
• Task grammar helps define behavior;
• Blueprints describe what we build;
• Your EHR configuration helps to enforce reality.

I hope this is helpful to you and your own Clinical Informatics teams, and always remember to explore this with your own Clinical, Legal, Regulatory, and Compliance leadership before adopting at your own organization.

Remember - This blog is for educational and discussion purposes only - Your mileage may vary! Have any similar experiences with writing task grammar, or operationalizing pre-existing policies? Feel free to share in the comments section below!