Tuesday, October 31, 2017

An Opinion : What exactly are "Protocols" and "Standing Orders"?


Hi fellow CMIOs, CNIOs, Informatics leaders, and other #clinicaljedi,

Protocols and standing orders. Next to order sets, these are two of the most ubiquitous tools in modern healthcare, used to create predictable routines and outcomes in clinical care. So what exactly are they? And what exactly is the difference between a "protocol" and a "standing order"?

For the Informaticist, these are not easy questions to answer. The confusion starts with the search for regulations and definitions, where there is a curious paucity of information. As of this writing, most major regulatory bodies have somewhat vague or conflicting information. 

Historically, it seems many regulatory bodies simply frowned upon the use of "protocols" or "standing orders". Why? Probably because of their function - If a doctor writes an order like "Vent liberation per protocol", he/she is actually asking someone else to take on the responsibility of managing a ventilator on his/her behalf. So it's actually a tool of delegation.

So I'm guessing the concern was, if the 'protocol' was a tool of delegation, then it's someone else providing that care at the direction of, and on behalf of the doctor. This raises some valid operational questions : 
  1. By using "per protocol", does the order actually refer to a approved, documented set of well-definedclearreasonableevidence-based, and agreed-upon instructions?
  2. If a nurse, pharmacist, or someone else can follow those instructions - What if something doesn't go according to plan? Would the nurse, pharmacist, or other care team member have the same skills and training as the doctor to manage any unexpected outcomes or scenarios?
  3. As the nurse, pharmacist, or other ancillary team member follows the written instructions - Are there any key points in the patient's care that the ordering doctor should at least be aware of? (e.g. if the protocol keeps asking the nurse to give  higher levels of oxygen, is that OK?)
  4. If the doctor is effectively unaware of the minute-by-minute details of what the nurse or pharmacist is doing, or the patient status, will the doctor still be responsible for the outcome of their pre-defined instructions? Or will the nurse or pharmacist be responsible?
  5. How do we know this delegation agreement was clear, effective, and appropriate?
So for many years, many of those regulatory agencies were concerned and understandably frowned upon the use of protocols and standing orders. Not enough was known about them, and the risks seemed to outweigh the benefits. And let's face it - If it's a tool of delegation, what's to stop a doctor from writing the "Dr. ______-is-away-this-weekend-protocol", making the nurses shoulder all of the responsibility for care? 

But then, around 2008, after more rigorous discussion and a few safety incidents where nurses were unable to initiate common life-saving treatments because their patients needed them and no doctor was immediately available - it seems like some agencies may have re-looked at protocols and standing orders, had a change of heart. In 2011, CMS issued this communication : 
... which says : 
"Standing orders 
Hospitals may adopt policies and procedures that permit the use of standing orders to address well- defined clinical scenarios involving medication administration. The policies and procedures must address the process by which a standing order is developed; approved; monitored; initiated by authorized staff; and subsequently authenticated by physicians or practitioners responsible for the care of the patient. The specific criteria for a nurse or other authorized personnel to initiate the execution of a particular standing order must be clearly identified in the protocol for the order, i.e., the specific clinical situations, patient conditions or diagnoses in which initiating the order would be appropriate. Policies and procedures must address the education of the medical, nursing, and other applicable professional staff on the conditions and criteria for using standing orders and the individual staff responsibilities associated with their initiation and execution. An order that has been initiated for a specific patient must be added to the patient’s medical record at the time of initiation, or as soon as possible thereafter. Likewise, standing order policies and procedures must specify the process whereby the physician or other practitioner responsible for the care of the patient acknowledges and authenticates the initiation of all standing orders after the fact, with the exception of influenza and pneumococcal polysaccharide vaccines, which do not require such authentication in accordance with §482.23(c)(2). 
The policies and procedures must also establish a process for monitoring and evaluating the use of standing orders, including proper adherence to the order’s protocol. There must also be a process for the identification and timely completion of any requisite updates, corrections, modifications, or revisions." 
This was a big step forward in creating some clarity around the subject of "standing orders", but does this also apply to "protocols"? It seems there is still a great deal of confusion over this issue. Articles like these : https://www.medscape.com/viewarticle/775617  suggest that people are still trying to understand if these are legal, and if so, how to design them safely.

So from a terminology standpoint, most Informaticists routinely have to struggle with questions like : 
  1. "What exactly is a protocol?" and 
  2. "Is it the same as a standing order?" or 
  3. "Is it the same as a clinical protocol?" and 
  4. "Is it the same as an oncology protocol?"
These are not easy questions to answer when the regulations and definitions don't guide you very well. 

So what is an Informaticist to do to help resolve the issue? Focus on the function, and work backwards to redefine the archetype and definition!

Let's look at a few things we *do* know : 
  1. The terms "Protocol" and "Standing Order" are almost used interchangeably - But not quite - So they still must have some kind of relationship.
  2. They seem to allow someone else than the provider to enter, modify, or stop an order, on behalf of the provider - so they appear to be some kind of tool of delegation.
  3. They seem to follow some kind of documented instructions, explaining exactly WHICH order(s) to start, modify, or stop, and when.
  4. For safety, the reasons that someone else is starting, modifying, or stopping an order should be very concrete and clear, without need for interpretation.
So if we can accept the above as true, then we can start drafting a definition :
DRAFT ] DEFINITION - PROTOCOL (n.) - A tool of delegation that allows a __________ to INITIATE, MODIFY, or DISCONTINUE an order on behalf of a Licensed Independent Practitioner (LIP), Advanced Practice Registered Nurse (APRN) or Nurse Practitioner (NP), Resident Physician, or Physician Assistant (PA).
Now, who exactly can start/modify/stop an order on behalf of a provider? From the patient care standpoint, you want the person following the protocol to have the training and clinical understanding to follow the protocol properly, and know how to navigate when things don't go as planned. Commonly, there are four roles that could likely fill this role : 
  1. Registered Nurses
  2. Registered Pharmacists
  3. Registered Dietitians
  4. Registered Respiratory Therapists
But if you want to include other care team members (like Medical Assistants), it is probably in an organization's best interest to make sure all team members expected to follow the protocol have a solid system of certification, training, and supervision. Either way, you'll need to work this into your organization's definition of a protocol : 
DRAFT ] DEFINITION - PROTOCOL (n.) - A tool of delegation that allows a Registered Nurse, Registered Dietitian, Registered Pharmacist, Registered Respiratory Therapist, (or certified and trained Medical Assistant) to INITIATE, MODIFY, or DISCONTINUE an order on behalf of a Licensed Independent Practitioner (LIP), Advanced Practice Registered Nurse (APRN) or Nurse Practitioner (NP), Resident Physician, or Physician Assistant (PA).
This is a pretty good start, but let's see if we can help craft some additional functionality and safety into this definition. 

If protocols should address common, well-understood clinical scenarios, then we need to consider what types of common, well-understood scenarios we might need protocols for : 
  1. Scenarios that only apply to a specific patient (e.g. Heparin titration protocols, Vent Liberation Protocols), and
  2. Scenarios that apply to a population of patients (e.g. Nurse vaccination protocols, pharmacy substitution protocols, etc.)
And so it seems like there is a need for two different kinds of protocols : 
  1. Protocols that only apply to a specific patient (e.g. Heparin titration protocols, Vent Liberation Protocols), and
  2. Protocols that apply to a population of patients (e.g. Nurse vaccination protocols, pharmacy substitution protocols)
And so, this suggests that the term "Protocol" might actually comes in two types : 
DRAFT ] DEFINITION - PROTOCOL (n.) - A documented tool of delegation that allows a Registered Nurse, Registered Pharmacist, Registered Dietitian, Registered Respiratory Therapist, (or certified and trained Medical Assistant) to INITIATE, MODIFY, or DISCONTINUE an order on behalf of a Licensed Independent Practitioner (LIP), Advanced Practice Registered Nurse (APRN) or Nurse Practitioner (NP), Resident Physician, or Physician Assistant (PA). All protocols are categorized as either : A. Protocols that apply only to a specific patient, or B. Protocols that apply to a defined population of patients. 
So if a protocol only applies to a specific patient, there must be some way that a provider can specify which patient(s) to use the protocol on - A way of turning the protocol "on-and-off", to tell a nurse when to follow the protocol, and when NOT to follow the protocol.

Likewise, for those protocols that apply to a defined population of patients, there must be some way to define which population of patients the protocol should be applied to.

And this brings us to the question of how to initiate/activate, or 'trigger' a protocol - If you need to activate/deactivate the protocol, then a handy trigger would be an ORDER, e.g. :
  • "Initiate/Follow heparin titration protocol" and 
  • "Discontinue/stop heparin titration protocol".
But if it's a protocol that is 'always on' for a defined group of patients (say, adult inpatients), then a handy 'trigger' could be a POLICY, e.g. "All adult inpatients will be on the pharmacy PPI substitution protocol" that allows a pharmacist to STOP one PPI and START another PPI (to replace one for the other).

So if we can accept that there are probably two ways to activate/'trigger' a protocol : 
  1. ORDERS - For those 'on-off'-type protocols, that need to be initiated/discontinued for a particular patient
  2. POLICY - For those 'always-on'-type protocols, that are always in effect for a defined patient population
... then we can work this into our drafted definition of a PROTOCOL (click below to enlarge): 
... and for additional clarity, we can bring in a few common, real-world examples below : 
So this is a reasonable starting point. But does this help us, yet, with a definition for "Standing Order"? I think it does - The term "Standing Order" is commonly used to describe scenarios where the provider has granted pre-approved, written, delegated authority to perform an action without their input or awareness. This is especially helpful in common scenarios where the risks/benefits of administration outweighs the risks/benefits of getting a provider order - E.g. For Public Health reasons, many states allow Registered Nurses to order and administer (low-risk) vaccinations without a provider's input.

So is it possible that the term "Standing Order" is actually a synonym for all or part of this protocol definition? I think it works pretty well for part 1.b below : 
And so, a "Standing Order" could then be defined, simply, as a PROTOCOL that is initiated/triggered by a POLICY (See section 1.b above.)

This leads me to ask about other common terms/synonyms - E.g. "Nurse-driven protocol", "Pharmacy-Driven Protocol", "Respiratory Therapy Protocol", etc. What exactly are these?

The problem with these other synonyms is that the terminology overlaps a bit, and when referring to a protocol, it's important to note both the method of initiation (e.g. Policy or Provider), and the team member(s) expected to follow the protocol, to initiate orders on behalf of the attending or ordering provider (e.g. Registered Nurse, Registered Dietitian, Registered Pharmacist, etc.)

So when referring to a protocol, it's always important to consider both : 
  • "____________-INITIATED protocol" - Describes the mechanism for initiating/triggering the policy (e.g. Provider-initiated, Policy-initiated)
  • "____________-DRIVEN protocol" - Describes the team member(s) who is/are expected to follow the protocol (to start/modify/stop orders, on behalf of a licensed prescriber) (e.g. Nurse-driven, Pharmacy-driven, Dietary-driven, etc.)
We can use these terminology concepts to help fill out our definition of a protocol / standing order even more - see 'synonym : Provider-initiated Protocol' in 1.a below :

And if we want to give some examples of each type of trigger, to help make the definition even more clear, we could include them too  - See 1.a.i and 1.b.i below : 
This is a pretty good start, but we'll want to work in some features of attribution, for the orders that are initiated, modified, or discontinued by these protocols. Since we now have a definition with two types of protocols : 
  1. Order-Initiated (aka Provider-initiated protocols)
  2. Policy-Initiated (aka Standing Orders)
It makes sense that for the resulting 'child' orders : 
  1. Order-Initiated (aka Provider-initiated protocols) - Attributed to the ORDERING provider
  2. Policy-Initiated (aka Standing Orders) - Attributed to the ATTENDING provider
And we can work this into our definition, too - See 1.a.ii and 1.b.ii below : 
Finally, for safety, we should consider the circumstances in which someone else other than a doctor might assume responsibility for INITIATING, MODIFYING, or DISCONTINUING an order - We want it to be very clear about WHENexactly, to start/modify/stop that order - In other words : 
  1. ACCEPTABLE = Clear, discrete data elements
  2. UNACCEPTABLE = Vague, ambiguous, or complex data elements
So we can work this safety feature into our definition too - See #2 below : 
This definition is much more robust than many regulatory agencies currently offer or publish. Some might see the adoption of such a definition as risky ("You don't want to paint yourself in a corner!") - However, it does provide a great deal of clarity and predictability, and if it exceeds the expectations of the regulatory agencies, then you are still meeting their expectations while simultaneously creating clarity and predictability - Which creates more predictable outcomes, which can lead to faster development time, higher development standards, and more standardized care. Before deciding whether or not to adopt such a definition and approach in your organization, your legal counsel, senior leadership, and informatics leadership will need to discuss the risks and benefits in detail.

However, if after rigorous examination and debate, you do adopt a similar definition, then it could help you answer questions like : 
  • Q : "What exactly is a protocol?"
  • A : See the [ DRAFT ] definition below (click to enlarge) : 
  • Q : What exactly is a standing order?
  • A : It is a PROTOCOL which is activated/triggered by a POLICY. The provider is not required to initiate action, and child orders are attributed to the ATTENDING provider. See the definition of PROTOCOLsection 1.b above.  
  • Q : Do I always need to activate a PROTOCOL with an order?
  • A : No - See the definition of PROTOCOL, section 1.b (aka 'Standing Order'). 
  • Q : Is a PROTOCOL the same as a 'research protocol'?
  • A : Without a solid definition of 'research protocol' it is not easy to answer this, but research protocols are typically used to guide the screening of research subjects, plan their data collection and management, with the goal of studying a subject. A PROTOCOL is only used to define a common clinical scenario where a licensed prescriber is delegating the authority to start, modify, or stop an order on his/her behalf.
  • Q : Is a PROTOCOL the same as an 'EMS protocol' or 'Emergency Protocol'?
  • A : Without a solid definition of 'EMS protocol' or 'Emergency Protocol', this is difficult to answer concretely - But in fact, many state EMS protocols have the same sort of functionality as a clinical PROTOCOL - Allowing a trained medical professional (paramedic or EMT) to initiate care in the field, on behalf of a supervising Emergency provider or Medical Director.
In my next blog post, I'll show how such a [ DRAFT ] definition could help you develop a protocol template that supports your protocol definition by creating an easy way for your protocol-builders to plan and build a professional-looking protocol document that supports your desired workflow and EMR configuration.

Remember this blog is for educational discussions only - You should consult your own legal counsel, senior leadership, and informatics professionals before considering adoption of any of the above approaches or definitions. Have any good definitions or regulations to share, or other ideas or comments? Leave them in the comments box below!

No comments: