Clinician Beware: Sewage flows uphill
This is my first post, aside from the Introduction. It's not exactly on the main topic of this Substack, but it's the most recent thing on my mind...
We all know the expression, “sewage flows downhill”, or a similar variant. In the workplace, “sewage” refers to undesirable, unsatisfying work, delegated down the chain of command until it reaches someone who has no choice but to do it. It generally describes tasks that could be made more efficient, or eliminated altogether. They may be internal tasks that are obsolete, yet occupy a place on some managerial to-do list. Or, they may originate from an external source beyond an organization’s control. The latter is often the case in a Community Health Center (CHC), which is a hub of documentation and coordination for an entire healthcare ecosystem.
In the world of primary care, however, these same tasks give rise to an entirely different phenomenon to the downhill flow. We might term it “sewage flows uphill”. To understand this, it helps to understand the typical hierarchy of medical licensure in a CHC.
At the top stands the clinician, an umbrella term applied to doctors, physician assistants and nurse practitioners who see patients independently. Under them is a chain of nurses: Registered Nurses (RNs) at the top, Licensed Practical Nurses (LPNs) below them, perhaps followed by Nursing Assistants, who perform various complex tasks appropriate to their training. Medical Assistants come next, handling tasks like triaging patients, taking vital signs and gathering data for the clinician to use. Finally, we have sundry administrative staff who manage things like phone calls, scheduling, check-in, insurance, referrals, external communications, medical records and processing documentation. There are many other tasks, and these roles often overlap, but this should give a general idea.
When a clinician is working at the top of their license, they are interviewing and examining patients, synthesizing their stories, recording clinical perceptions and plans, and delegating care tasks to the appropriate channels: vaccines and treatments to nursing staff, referrals to administrative staff, follow-up visits to scheduling staff, medications to the pharmacy, etc. This process, as it occurs during actual office visits, works fairly well in most clinics.
Administrative follow-up work, on the other hand, is very different. This refers to things like completing documentation and billing for office visits; filling out prior authorization forms; filling out and signing routine authorizations for all types of services; answering questions from patients and caregivers; reviewing test results; signing prescriptions renewals; processing electronic documents, by the hundreds, from insurance companies, care managers, pharmacy benefit managers and countless other sources. Many of these items are non-clinical. Many more require very limited clinical knowledge to complete. A great number could be handled categorically by non-clinician staff, given a set of “standing orders” - rules for handling each type of item, including when to surface it to a clinician.
Yet this delegation often never happens in a CHC, or doesn’t happen enough. For example, clinicians complain that messages from pharmacy benefit managers are useless, and could be signed off by administrative staff. Or that most authorization forms could be handled by nursing level staff. Or that they would trust a nurse pharmacologist to sign off on routine prescriptions renewals. The opportunities for delegation are clear, and Electronic Medical Record (EMR) systems frequently provide the means for delegation and approval. But health center staff, like nurses, assistants, and administrative staff, all work by the hour. Their time is measurable, and the cost is budgeted. Out of the entire hierarchy, the only role whose time is not trackable is the clinician. Leaving the clinician to handle additional tasks is therefore, officially, “free”.
So who pays? Clinicians spend their free moments clicking on meaningless electronic documents instead of reviewing their next patient. They forward mis-routed documents from place to place and sign off negative COVID tests instead of reading up on a new treatment. They catch-up on documentation in the evenings, instead of spending time with their families. So clinicians surely pay, and it drives burnout and increases attrition.
But the eventual victim is not just the clinician; it is, of course, the patient. CHC Clinicians frequently get behind, and spend their days trying to slip out of exam rooms as quickly as they can, hoping to keep up with their admin work in between patients. It is poignantly reflected in key moments, such as when the clinician tells a parent that their child is fine, and the parent just isn’t convinced. They really think something might be wrong, and that their story hasn't been fully heard. But the clinician’s mind is already on the next task, their hand is on the doorknob to leave the exam room, and the opportunity slips away.
When problems are addressed with systemic improvements that increase efficiency, it should be great news! When a core electronic system gets faster, or a task delegation is established, we should feel things improve. But upper management, especially non-clinical management, measures improvement very differently from clinicians. In a CHC, clinicians are employees. They don’t run the show the way they might in a small private office. And the pervading leadership logic is that if clinicians are given back time, their productivity should increase. And it doesn’t. The gradual, incessant theft of time, by technical, administrative and system waste, has forced clinicians to push themselves toward burnout while providing inadequate care. When a slice of time is returned to them, their instinct is not to push themselves equally hard to expand inadequate care to more patients. They instinctively return the time to its rightful places: patient care (for their existing patients), and self-care.
Sadly, when improved efficiency does not improve throughput, it can be used as evidence that such measures aren’t effective. I have personally witnessed examples where a major, desperately needed improvement provided welcome relief, but failed to increase throughput. These examples went on to be cited, for years, in arguments against the value of expending time and money on similar efforts. The inference was that clinicians are not suffering so much as they claim, and will only absorb any and all the time returned to them, giving the institution nothing in exchange.
The hard fiscal realities of primary care in a CHC underlie this entire structure, and there is no easy fix. But our own leadership often does not thoroughly understand or admit the intangible losses that result from the uphill flow of administrative work. They fail to realize the criticality of restoring the time that is hardest to measure. And while specialist offices and hospital systems can hire the staff they need to protect their clinicians, and still manage to post profits, the CHCs continue to struggle. They persevere with inadequate staff, continue to lose clinicians, and to shoulder much of the blame for the declining patient experience of healthcare.


Great post James! I'm just not convinced payors, patients, clinicians, and the health care systems that employ us are aligned on what primary care is and the proper cost for the exchange of this service. We all know some patients have more pressing matters than lowering their blood pressure, no matter the cost savings to "the system" when they don't have a stroke. We all want $200 services and $10 copays and scowl at our colleagues who can charge $500 or more for the same work without batting an eye (and get paid for it!). We all work for systems that exist because a few tired thumbs continue to plug holes in dikes that should have long overflowed, if not exploded. Finally, I think many of us have lost faith in the structures and underlying assumptions that brought us to this place, abandoning the only real weapons we possess against the lobbies of insurance and pharmacy. Technology is merely a place where these larger pressures are showing themselves.
https://workweek.com/2022/12/10/the-future-of-natural-language-processing-in-healthcare/
Interested in your thoughts on the natural language processing improving through put
There is a physician that is already using ChatGPT for approvals from insurance companies