In my life, I have made mistakes. Some are obvious only in retrospect. (Though I stand by my high school sartorial choices.) Others are obvious in the moment. I remember presenting some data to a board—a presentation that should have been pro forma and innocuous—only to find myself in a heated argument with the CEO over whether it was easier to move inpatient scores or clinic scores. I have no idea how this started and even as it unfolded, I wondered how I got here. It was not moving the agenda forward, and I could not see an exit ramp. Were it not for someone else in the room who made an “in the interest of time, perhaps this could be taken off-line” comment, I might still be having that argument to this day. This experience highlighted two things. First, while I don’t ever think that I am the most intelligent person in the room, I do get triggered by people who claim to be, when clearly, they are not, at least on the topic I am at-that-very-moment presenting on. Second, people often get confused about what makes improving patient experience scores hard: volume or complexity. Simply put:
- Volume: The CEO’s position was that because the volume of surveys on the clinic-side was so massive (clinics can get 1000x more surveys back than hospitals get for inpatients) it was harder to move the entire dataset. Getting one positive survey would move the HCAHPS data more than one positive survey in the clinics would.
- Complexity: I agreed that mathematically he was right, the average inpatient stay had way more moving pieces, which meant managing dozens of staff with over twenty hand-offs. When compared to the receptionistà rooming nurseà physician hand-offs, it was harder to get all inpatient staff on the same page and therefore harder to identify the gaps and therefore harder to improve inpatient numbers.
Now, the biggest part of this conflict was that we were really talking past each other. We were defining “difficult” or “important” differently, which led us to different outcomes. Since PX involves a lot of squishy or ill-defined words, this happens a lot.1 The underlying part of the conflict, though, was that these words were papering over real issues in both spaces that make building consistent execution of positive PX difficult and, more importantly, DIFFERENT. The real issue is that just as PX must be tailored to the patient to be effective, it also needs to be tailored to the care space as well.
Before exploring this, let me make two quick points.
- To those with a dog in this fight, I ask that you lower your hackles. Discussing why one is hard is not implying the other is easy. This is not to say anyone has it easy, but instead to discuss why they are differently difficult.
- Part of the subtext here is on sample sizes and the impact of ‘one survey’ on the whole sample. As much as I love math, I won’t be addressing this math issue here, in part because it is a Vanishing Leprechaun problem2 but mostly because it is not relevant to the actual issue of improving scores.
Why Hospitals are Harder
Hospitals have two challenges that make it harder to move the PX meter. First, their target population is generally unexpected captives. While many inpatients have planned trips to the hospital (like planned births or surgeries), many arrive in a hospital bed having started their day thinking they would be doing just about anything other than this. Jarring a person out of their routine with the immediacy of an inpatient stay means that this patient is constantly playing catch-up, feeling a step-behind. They often have anxiety and that anxiety can be outsized from expectation. Most are not likely to die, but the act of being made an inpatient itself feels life-threatening. This anxiety creates a defensive shell, and it can be hard for staff to break through that forcefield to provide positive experiences.
Second, the average hospital stay hovers around five days. Even if a hospital’s nursing staff is running 12-hour shifts and has consistent assignments (the same nurse getting the same patients on their shift), at the very least, a patient is going to have four different nurses providing care. Between hospitalists and any specialists, they are likely to have a host of physicians visiting them as well as a pharmacist or two. Then factor in the various phlebotomists, imaging technicians, and other clinical support staff. After that, consider all the non-clinical staff such as food and nutrition services, housekeeping, transportation, facilities management, security, and everyone else that keeps a hospital running. It is not outrageous to imagine a patient seeing twenty or thirty different employees and this doesn’t even include everyone behind the scenes who support these folks and their ability to provide excellent experiences.3
Patient experience in this space combines a spirited game of whack-a-mole with a side project of chasing down ghosts. Because there are so many people touching the lives of so many patients, you have to have a broad focus. Further, since so many of these folks might have an interaction with a large volume of patients, albeit brief, a PX plan has to spend a large amount of time addressing all areas, clinical and non-clinical alike. While you can break data down by discharge unit to gain some granularity, this overlooks the fact that there are staff that cross multiple units and the fact that patients may be turfed to different units before they are discharged. This means that you cannot treat the data in-isolation as much as you might think. So, while one can start the process by looking at Med/Surg because their scores are low, to fully appreciate what is going on there means understanding the staffing flow on and off that one unit. Most hospitals, for example, have a next-up philosophy to food service, housekeeping, transport, and lab services, meaning that these folks are not assigned to a unit, but instead answer the bell, regardless of where it rings. Further, while many nurses have a home unit, some may prefer to work in another unit or be a floater, rather than being unscheduled if their unit’s census is low.
Second, half of a PX person’s job is dealing with the “you need to talk to…” crowd. I have talked to plenty of floor nurses who want to blame PX scores on the emergency department nurses. ED nurses blame the floor nurses. Imaging blames transport. Transport blames everyone. Doctors will dismiss or discount each other rather than blame each other, which is often worse. Everyone wants to say that the problem with PX scores is because of those people over there. It sometimes feels like most of my time speaking to audiences is to say, “Yeah, I will talk to them later. Right now, I am talking to you.”4
Why Clinics are Harder
Reading all of that might make it seem like clinics, with their smaller self-contained staff, where patients have virtually no encounters with non-clinical staff outside of registration, would be a cake walk. And you would be wrong. Clinics function in a different care model, in a different space with a different kind of patient. This means that they have different problems, but problems, nonetheless.
Unlike the inpatient setting, clinic patients generally know that this visit was part of their day.5 These patients can still be anxious about a nagging pain, an abnormal test result, or a sudden change in their body. But because they picked this appointment, they generally have less anxiety than if they were in a hospital bed with an IV in their arm and their butt exposed. This means they fell less like a “patient” and more like a “customer” which comes with its own service expectations.
The very fact that it is a scheduled appointment is high on the list of expectations. The average primary doctor can have over twenty appointments in one day. Even a small clinic can have over one hundred patients file through every day. This means that in order for schedules to run smoothly a bunch of things have to go like clockwork. A slow computer, a nurse calling in sick, or a receptionist having to field a heavy call volume can all throw the process for a loop. And this doesn’t even include the fact that patients are actually human beings with complex care concerns or a laundry list of complaints that they saved up to avoid multiple visit co-pays. So, a patient who is trying to fit an office visit into a work lunch hour can anxiously watch the clock tick by, doing the math on when they have to abandon the appointment in order to get back to work. Even if patients factor in some flexibility in their day can be frustrated as their appointment time comes and goes by, especially if they see people who arrive after them actually get called in before them. The expectation that appointment times have meaning can create a negative vibe even before the patient gets to meet any of the care team.
Further, the reason why primary care providers can see so many patients (and get so many patients backlogged), is because appointment times are very brief. It is not uncommon to have primary care visits scheduled in 15-minute blocks. So, the care experience includes about 45-seconds of registration time, followed by a wait in the waiting room, followed by three or four minutes with a rooming nurse, followed by ten minutes with the doctor, followed by “please follow the signs to the exit.” While inpatient staff may complain that they have a lot of time to screw stuff up, the clinic staff will complain that they have no time to leave a lasting positive impression. And if the team is backed up and everyone seems rushed, which is daily, it is virtually impossible. The care space has patients with heightened expectations dealing with staff that have a limited chance to shine.
Worse, while an individual clinic may be self-contained, there are dozens, perhaps even more than one-hundred clinics out there in a system all trying to deliver the same “brand promise” as the marketers would say. While there are some staff and doctors that will work at multiple clinics, it is still easier to ask, “Why does Clinic A suck and Clinic B shine?” since their staff and manager are easily identifiable. The problem is that systems don’t generally ask that question.
Clinics will often self-define as the “wild west” of healthcare. They sit in isolation, miles from the hospital, doing what seems best in the moment for whoever ambles into their field of vision. While they can appreciate this autonomy in the abstract, it is not uncommon for them to also feel like the red-headed stepchildren of a hospital system. No one offers them a helping hand or a pat on the back. Ask a clinic manager and they will say that the only time someone from the system shows up is if there is a problem. For example, I spent a lot of time visiting a hospital system in Richmond, Virginia, which, by the way, remains the best per-capita food city that I have ever been to. When I went there, I would dedicate a whole week to each visit. I would spend a day at every hospital on-site and then on the final day, when I was getting ready to catch the last flight out of town,6 I would get a couple of hours with ALL of the clinics. In a central location. That they had to drive to. In the middle of the morning. I was surprised that the system leadership was surprised to learn that they felt underappreciated when it came to patient experience.7 I have only partnered with one hospital system that worked hard to include their clinics in any patient experience work I assisted on.8 The rest often have the outpatient space siloed off so they rarely see the outpatient leadership, are primarily interested in only physician recruitment, and only are aware of the patient experience data once a quarter, if that. It is not that system leadership thinks that patient experience work in the clinic is the same; they just assume that it isn’t different and no one challenges them. At least until one idiot picks a fight with their CEO.
The conflict between me and that CEO was not about who was right. It was about a fundamental confusion about what patient experience looks like in various care settings. Approaching the PX messaging and structure uniformly means creating a system that becomes one-size-fits-none. We end up holding the wrong leaders accountable for the wrong work because it is targeting the wrong people. I will work on an essay series addressing obstacles to PX improvement, but that cannot even begin if we cannot agree what work goes into it.
1This is why one of my suggestions when talking about PX is to define the words you are using to avoid any confusion or annoying confrontations about connotations and denotations.
2I have seen a ton of different versions of this, but I first saw it when I was a child as the Vanishing Leprechaun. Look at the images below, count the leprechauns, and you will see the problem. It is a visual trick where arranging the cards changes the number of leprechauns (or eggs or heads or cigars or spies). The coolest version of this is an entire wall-sized version with hand-cranks at the Safehouse restaurant in Milwaukee, Wisconsin.
3One of my early essays was discussing the relative importance of all these encounters, but suffice to say, while they may have different weights, all are combined to inform a patient’s perception.
4Sometimes I am tempted to say, “Really? You should hear what they say about YOU” but I do know that this would not help anyone ever. But in my mind…
5As always, there are exceptions. After all, urgent care clinics are a thing.
6Moment of truth time, Richmond. You are a wonderful city. I love your food. (I still dream about Croaker’s Spot.) I love Plan 9 Records. Go Spiders! Go Rams! But I am not a fan of your airport. Not a single flight of mine has ever taken off on-time and it is the only airport where I was actually locked inside of the airport, which is another story for another time.
7OK, this happened only once, but only because every other time I went, I would work with the clinic leadership to make sure my visit dovetailed with a meeting they were already having where I could get some agenda time.
8When I was there, often for a week, I would actually go with the PX champion to a bunch of their clinics to talk with the staff on-site about the PX work. Their system goals were designed to foster collaboration and attention across the clinic system. Of course, this was a few years ago, so I am not sure they are still as committed to the work, otherwise I would name-check them.
PX Matters with Joseph Snipp 
Leave a comment