Consequences of Flow

[Arcan57]

---

Members don't see this ad.

The post about the kid with the expanding neck hematoma got me thinking about some of the unintended consequence of delays or disruptions in ED flow. While the prior thread focused mostly on PG/economic reasons for improving flow, those aren't the only reasons why flow should be optimized.

Timely flow of information (especially as regards lab and x-ray results) is crucial to good patient care, and lack of flow can turn cases that are cut and dry under the retrospectoscope into self-perpetuating Charlie Foxtrots. In fact, one of the most crucial skills an attending has is the ability to quickly and correctly identify who can and can't tolerate their shops' normal pattern of flow. In residency, the extra man-power that can be brought to bear at any moment can, to some extent, alleviate initially poor planning. In most community gigs, there is little in the way of surge capacity. Initial mis-steps become self-compounding and bad things happen to patients in a way that is inexplicable to those outside the situation but almost inevitable under the initial conditions.

It would seem this cascade could be avoided by the EP pushing harder on the elements that are disrupting flow and it can. The problem is that you can't push on every patient (at least not at the level of working a shift) or else you end up having no effect (and pissing off everyone you work with). Some EPs try and make up for this by trying to brute-force the initial work-up (everyone gets troponins, d-dimers, cross-sectional imaging, lactate, ammonia, and serum osmolality) which is only adaptive in highly specialized scenarios (ie. seeing exclusively high acuity/tertiary care patients with multiple organ system dysfunction and a preponderance of non-specific chief complaints). For anyone not working in that environment, the lack of efficiency in whirling up the minigun's barrels (for when a shotgun isn't broad enough) will result in an opportunity to succeed elsewhere.

The key to meeting this challenge is to know what the average times are in your shop for the crucial steps (which I think we all internalize after a month or so at a new place), and to see if your patient can tolerate those times. At my shop, I've never been able to intubate someone using meds within ten minutes of informing the nursing staff of the need for intubation. So if someone rolls in via EMS in respiratory arrest, I'm bagging for 12 minutes or using brutane. I can't get a trauma pan-scan looked at in under an hour, so if I'm concerned about a trauma patient I have to read the CT myself or that expanding pneumo will be a tension by the time the radiologist calls me.

It's mentally taxing to pull information vs. having it pushed at you, and in general it's easier when you're first out of residency and are used to inefficient systems that require almost everything to be pulled. For the mid-career attending, identifying which patients are going to require pulling (will the nurse tell you she can't get an IV on the difficult stick with the HR in the 140s?) is our greatest day-to-day challenge.

For example: after 3 yrs of not seeing kids and coming from a pediatric mecca where almost everything was pushed at me ("I've worked up and gotten the kid admitted to their specialist, here's the orders to put in the computer"), it's not rare for me to be wrong-footed by kids that look simple but don't get better with treatment. Most of my nurses can't draw blood from sick kids under 1yo (although oddly they are pretty good at IVs). If I didn't think the kid needed labs when I first saw them, I'm often determining that the kid needs transferred prior to having any labs back to tell the receiving facility about. Which I am well aware makes me look like an a--hat to the peds attending on the other end of the line (especially if they ask how long the kid has been at my ED), but unless my radar gets more finely tuned I don't have a better solution.

Thoughts, comments, questions?

 

[EM_Rebuilder]

In fact, one of the most crucial skills an attending has is the ability to quickly and correctly identify who can and can't tolerate their shops' normal pattern of flow.

Great way to put this. I have never thought about it in this sense, but now that I reflect, I do this...

I remember being a resident, and sometimes the Attending comes down on your 'Make sure you call and get the CT right now, or why isnt that troponin back now, we need it now, now, now!"... and its not on every patient, and often espically as a lower level, you rolled your eyes at it...

Now, as an attending, I occasionaly go in a room and essentially go 'Crap'.. call CT, we need the scanner cleared, I want this guy there now. Or has the blood drawn, please be drawning while I am standing here, and dust off the i stat I need the troponin...

Your right, every shop has a basal 'flow' rate and every shop is slightly different in how their flow.. flows! The next layer of complexity is actual patient load...

Thanks for the insight on another way to think about true flow... basal flow... I like that term.

 

[docB]

It would seem this cascade could be avoided by the EP pushing harder on the elements that are disrupting flow and it can. The problem is that you can't push on every patient (at least not at the level of working a shift) or else you end up having no effect (and pissing off everyone you work with). Some EPs try and make up for this by trying to brute-force the initial work-up (everyone gets troponins, d-dimers, cross-sectional imaging, lactate, ammonia, and serum osmolality) which is only adaptive in highly specialized scenarios (ie. seeing exclusively high acuity/tertiary care patients with multiple organ system dysfunction and a preponderance of non-specific chief complaints).

Good points.

This concept of trying to push on everyone at once not being a viable strategy reminds me of something a frustrated partner said to me about a year ago. "When you have to treat everyone like a critical emergency on one gets what they really need."

He was talking about the proliferation of patients who we have to jump on by protocol. Code STEMI, Code Stroke, Code Sepsis, Cold Foot, Possible Torsion, R/O Meningitis, etc., etc., etc.

When you get stuck trying to meet the metrics on all of these stuff starts to fall through the cracks.

 

[Arcan57]

Good points.

This concept of trying to push on everyone at once not being a viable strategy reminds me of something a frustrated partner said to me about a year ago. "When you have to treat everyone like a critical emergency on one gets what they really need."

He was talking about the proliferation of patients who we have to jump on by protocol. Code STEMI, Code Stroke, Code Sepsis, Cold Foot, Possible Torsion, R/O Meningitis, etc., etc., etc.

When you get stuck trying to meet the metrics on all of these stuff starts to fall through the cracks.

There's actually some literature behind this, at least for ACS patients having worse outcomes if they present at the same time as a trauma activation. I know I've sent hyperkalemia to the cath lab at least once, a ESRD pt who swore he'd been dialyzed the day before. He was in the department about 8 minutes and as I was dictating his chart I realized the EKG (which had a new LBBB) was hyperK and not STEMI. By the time his K actually came back, he had arrested, been resuscitated and had a TV pacemaker floated.

 

[docB]

There's actually some literature behind this, at least for ACS patients having worse outcomes if they present at the same time as a trauma activation. I know I've sent hyperkalemia to the cath lab at least once, a ESRD pt who swore he'd been dialyzed the day before. He was in the department about 8 minutes and as I was dictating his chart I realized the EKG (which had a new LBBB) was hyperK and not STEMI. By the time his K actually came back, he had arrested, been resuscitated and had a TV pacemaker floated.

Let's not forget the famous case of John Ritter going to cath with an undiagnosed dissection. Now I think that was appropriate because we can't do the best thing for the STEMIs and catch all the dissections and STEMI is much more common but it's an example of how rushed some of these cases have gotten.

BTW if it makes you feel better one of my partners held a "STEMI" in the ER because he was convinced the guy was hyperK. He was right, K was >7 and the guy arrested in the ED but the chest pain center administration crucified him because of his "failure to follow protocol resulting in patient harm." He lost his privileges for that hospital. The worst part was that it was typical hospital scapegoating but the really didn't need a scapegoat because clinically everything was done right. It was only a misapplied metric that was "harmed."

On second though that story won't make anyone feel better.

 

[BJJVP]

There's actually some literature behind this, at least for ACS patients having worse outcomes if they present at the same time as a trauma activation. I know I've sent hyperkalemia to the cath lab at least once, a ESRD pt who swore he'd been dialyzed the day before. He was in the department about 8 minutes and as I was dictating his chart I realized the EKG (which had a new LBBB) was hyperK and not STEMI. By the time his K actually came back, he had arrested, been resuscitated and had a TV pacemaker floated.

Can you use an EKG alone to differentiate a LBBB with underlying cardiac ischemia from one associated with hyperkalemia?

 

[WilcoWorld]

BTW if it makes you feel better one of my partners held a "STEMI" in the ER because he was convinced the guy was hyperK. He was right, K was >7 and the guy arrested in the ED but the chest pain center administration crucified him because of his "failure to follow protocol resulting in patient harm." He lost his privileges for that hospital.

 

[RustedFox]

OMG, no.

LBBB vs. STEMI vs. HyperK vs. GodKnowsWhat

 

[deleted109597]

Can you use an EKG alone to differentiate a LBBB with underlying cardiac ischemia from one associated with hyperkalemia?

Considering the data doesn't support new LBBB as STEMI criteria, I would argue it doesn't matter. Unless you're using Sgarbossa criteria.
http://emcrit.org/podcasts/left-bundle-branch-block/
http://hqmeded-ecg.blogspot.com/2010/03/new-left-bundle-branch-block-is-poor.html

 

[Zanegray]

There's actually some literature behind this, at least for ACS patients having worse outcomes if they present at the same time as a trauma activation. I know I've sent hyperkalemia to the cath lab at least once, a ESRD pt who swore he'd been dialyzed the day before. He was in the department about 8 minutes and as I was dictating his chart I realized the EKG (which had a new LBBB) was hyperK and not STEMI. By the time his K actually came back, he had arrested, been resuscitated and had a TV pacemaker floated.

You know, I've often wondered about this. In my residency shop trauma bumped just about everything from the scanner. Now that should be nbd, but trauma accepted everything - so at times we were holding scans from the main ED to pan scan a stable low velocity mvc already worked up at an OSH.

 

[deleted109597]

But even trauma patients have worse outcomes if their head trauma happens on the weekend. Because although trauma is on, NSGY isn't.
http://www.ncbi.nlm.nih.gov/pubmed/22795343

 

[Arcan57]

But even trauma patients have worse outcomes if their head trauma happens on the weekend. Because although trauma is on, NSGY isn't.
http://www.ncbi.nlm.nih.gov/pubmed/22795343

I wonder if that's just a lack of surge capacity? The behaviors that lead to operative intracranial hemorrhages definitely pick up on the weekends. Most of the places I've worked that had NSx were 7 days a week but I have to refuse a head bleeds a weekend because they were already in the OR.

 

[docB]

Surge capacity is a big issue for us even though no one wants to talk about it. We and the hospital staff for low average volume. So average volume is busy and busy is really tough. But we all keep doing it because up staffing means docs making less per hour and the hospital paying more for nursing. I wish we'd just take the financial hit and staff at more humane levels.

 

[Arcan57]

Surge capacity is a big issue for us even though no one wants to talk about it. We and the hospital staff for low average volume. So average volume is busy and busy is really tough. But we all keep doing it because up staffing means docs making less per hour and the hospital paying more for nursing. I wish we'd just take the financial hit and staff at more humane levels.

We do the same in terms of staffing. The beauty is that once you demonstrate that you can do your job with current staffing level, that becomes your your new baseline. Of course that eventually reaches a level where you burn out your providers at which point you turn over your management staff for failure to meet metrics. The key if you want to be successful in administration is to jump to the next level before your key employees burn out.

 

[NinerNiner999]

The key if you want to be successful in administration is to jump to the next level before your key employees burn out.

Agreed - but this also has challenges, especially in the face of the limitations of your hospital medical staff, back door flow, and ED nursing staff motivation, all of which also have to be kept in play.

As the troponin core measure stands in question with CMS, chest pain accreditation will still continue to place a time expectation on troponin results. They may follow suit with CMS, but it will remain another arbitrary "number" that must be met.

I bring up this example because most hospitals, when confronted with the option to fix their underlying inefficiencies by measuring flow, improving individual provider metrics, and creating a group process to identify barriers of faster care, will opt for point of care testing for their critical metrics. This only serves as a band-aid to a bigger wound of inefficiency. Instead of exploring the wound, cleaning it, closing it, and checking it periodically to see if it heals, it is easier to apply bandages and hope the bleeding stops. The end result is to have everyone working harder, and nobody working smarter. Who cares why the lab can't turn a troponin around in 30 minutes - just put it in this new machine. There - problem solved. Or is it? Sadly, this can't be taught, but must be practiced.

When we suggested POC orders at my shop, I suggested we spend the extra money to purchase the entire spectrum of POC labs available, to incorporate this technology into a meaningfully useful system to be used as a adjunct for all of our major protocols - STEMI, Stroke, and Ice. Sadly, the turnaround time for troponin trumped the other key data points, leaving the patients to wait for the remaining time for their other lab results, effectively contributing no value to the flow of these patients, but allowing a single time requirement to be met. Then again, time is the limiting factor in all of healthcare. Time costs money, and the biggest investment to identify flow problems, fix them, train staff, and practice new work patterns requires time commitment greater than any hospital is ever going to be willing to spend. Hell - just ask your unit secretary how their job orientation for your ED worked out...

The sad reality is that we are going to have to do more with less, and the old adage of "you can't beat a dead horse" is pretty true. Nurses are in short supply and those that are still working will forever have ingrained in them a sense of triage being a location and not a process. This has to change. They have to be shown that when patients come straight back to their rooms, and physicians see them, flow drives itself. Discharges happen faster, the stress of the waiting room disappears, there is an empty waiting room, often without the need for a dedicated triage nurse who can then serve as an extra hand, patients turn around faster, and, while more patients are being seen, it does not "feel" like extra work. Suddenly, your work days are happier, and your happiness is reflected to your patients, who have better experiences and give better satisfaction scores, which in turn make you have even more pride for your job.

I do believe that everyone can be treated with the same initial sense of urgency - whether acute MI or toe pain, but the art of rapid assessment will quickly identify the true areas of need. There will be simultaneous critical patients, and there will be bad outcomes regardless of what we do, but we continue to set ourselves up for failure when we refuse to collectively adopt a system that lets us see patients as fast as possible. This includes buy-in from physicians, who are used to waiting for patients to be brought back, undressed, and have vitals before they go into the room. I don't know about the rest of you, but I can pretty much look at a person on the spot and tell when they're really sick (i.e. needs immediate intervention, or soon will). It frustrates me when this patient is #7 in line of several people who registered but have not been "triaged". It makes me angry inside when there are 8 empty beds in my department. With our EHR system, it takes the nurses 10 minutes to complete the entire triage process. This means that patient must wait 70 minutes before they can come back to a room - a room that was empty for them the entire time.

Nursing has given every excuse from "we need those rooms for EMS," to "they need vitals so we know which room to put them in," to "they came here to see us first (this one always makes me chuckle)", and "the nurses in the back are too busy to do the triage note." The reality is that "triage" is a process to "sort" patients who need immediate care (i.e. need to come back to a room) and when there are empty rooms for them to come straight back to, there is no use performing an exercise to identify who "needs" to come back first. Take them all back. Eliminate the triage process.

Then registration comes into play, and insists that they must collect all of their information up front, to have them entered in the system. Speaking only for myself, all I need is a name, and a date of birth, and, assuming there is no shenanigans, a gender, and I can start my workup. Registration needs to adopt universally practiced bedside registration processess as our industry standard. No questions asked.

Not to be selfish, as the attending emergency physician who will happily manage every patient in my department without worrying about which "room assignment I have", and what the patient's insurance information reveals, these patients came to see me. Get out of my way.

The nurses get frustrated when I take patients out of the waiting room and into an empty room to get their history and do a brief screening exam. They get frustrated when I discharge dental pains and minor cellulitis before the patients were "triaged." I don't make an issue of it, and happily hand them the chart, ask them to call me in the back if there is any vital sign out of whack, and keep a smile on my face. They should be mad at themselves and their process for allowing this to happen.

IMHO, using adjuncts such as POC testing, when applied to a lean system with good flow and buy-in from the hospital staff, can only be a good thing. There is no reason that all chest pain patients can't be admitted in 15 minutes. Stroke patients in 30. Code Ice straight admission to the ICU, and all other non-critical patients can have routine "STAT" workups, and faster turnaround of empty beds to be seen in, accommodating more volume.

At our shop, our door-to-balloon time is 43 minutes for last month for STEMI.
We have 40% admission rate
Our arrival to room time is 12 minutes
Our room to physician time is 4 minutes
Our total turnaround time (TAT) for discharged patients is 90 minutes
Our TAT for admitted patients (ED care complete) is 120 minutes.
Our care complete to depart ED time for admitted patients is 200 minutes. That's in addition to our workup of 120 minutes! It takes longer for a patient to leave the department than it does to have an entire four-month outpatient workup completed in the ED.

These numbers are consistent at both hospitals I am department chair for. Clearly, the hardest part of efficient flow, and I don't think my hospitals are unique, is breaking down the back door to the rest of the hospital. The players in this component of "patient care" are case management (who really doesn't manage much in my place), delayed discharges and extensive length of stay by our own medical staff (who have to be driven by case management and reviewed by hospital administration), and outpatient resources in the community. These are factors that won't change overnight, but must be identified, and collectively fixed by all parties involved.

In reference to the above quoted sentiment - I agree that the next level must be attained as soon as possible, but motivating the key players to take an active part in identifying the areas that need improvement, then empowering them to take action and try alternative methods can actually keep them from burning out, and make them gain pride in their work when things begin to improve. Then again, this is the challenge of leadership, and by far the hardest part of my job.

There. Rant over.