The ICU Bounce Back

We’ve all experienced it. A seriously injured trauma patient is admitted to the ICU and begins the process of recovery. Everything looks well, and after a few days they are transferred to a ward bed. But then they languish, never really doing what we expect. Finally (and usually in the middle of the night), they begin to look bad enough where we have to transfer them back to the ICU.

What’s the problem here? A failure of the ICU team? Did we all miss something? Is there any way we can avoid this problem? The major issue is that these “bounce backs” tend to do poorly compared to patients who successfully stay in their ward bed. Estimates are that mortality for patients successfully and finally discharged from the ICU range from 4-8%, whereas the mortality in bounce back patients is 20-40%!

Researchers at the Medical University of South Carolina in Charleston looked at the characteristics that defined the bounce back patient. They looked at nearly 2000 patients discharged from the trauma ICU and analyzed the variables that predicted an unplanned bounce back. They noted the following interesting findings:

  • More than two thirds of bounce backs occurred within 3 days
  • Males, patients with an initial GCS < 9, transfer during the day shift were the major risk factors
  • More comorbidities was associated with a higher chance of bounce back
  • Mortality in the bounce back group was 20%
  • The most common factors causing bounce back were respiratory failure or bleeding

Bottom line: This is an intriguing single-institution study that supports my own personal observations. Fewer bounce backs occur at night because staffing tends to be lower and there is more resistance to transfers out. Both the ICU team and the ward team need to scrutinize every transfer carefully. Significant head injury or the presence of medical comorbidities should trigger a careful assessment to make sure that the transfer is appropriate. Otherwise, your patient may be placed in unnecessary jeopardy.

 

Reference: Intensive care unit bounce back in trauma patients: An analysis of unplanned returns to the intensive care unit. J Trauma 74(6):1528-1533, 2013.

What Is: A Trauma Performance Improvement (PI) Dictionary?

An anonymous reader posted this question. Let me first start by saying that I’m happy to answer any and all questions. However, it helps if you register with Tumblr or Disqus so I can reply directly to you as well. Sometimes I’ve answered a question in a previous post but am unable to explain why I’m not answering again. So please, register or email me directly so I can reply to your query!

Trauma PI is the most important part of any trauma program or center. Not only does the program have to provide high quality care, but it has to prove that it does this on paper. A performance improvement plan is important, as this outlines the specific methods used to self-assess clinical care. An important component of the plan is the PI dictionary.

A PI dictionary is simply the list of the clinical issues that are tracked by your PI program. This includes specific audit filters used to trigger PI review, as well as the list of issues and events that are routinely scrutinized. There is a core set of items that are found at every trauma center, like deaths and significant complications. However, no two centers’ dictionaries are identical because they must include local issues and problems as well.

Your trauma center should have a well-defined dictionary of PI review issues. And this dictionary must contain a “reasonable” number of items. Too many, and you will never be able to reasonably track everything down; too few and you will miss important problems that demand investigation. 

Related posts:

Bucket Handle Injuries Of The Intestine

Bucket Handle Injury

A bucket handle injury is a type of mesenteric injury of the intestine. The intestine itself separates from the mesentery, leaving a devascularized segment of bowel that looks like the handle on a bucket (get it?).

These injuries can occur after blunt trauma to the abdomen. The force required is rather extreme, so the usual mechanism is motor vehicle crash. In theory, it could occur after a fall from a significant height, and I have seen once case where a wood fragment was hurled at the abdomen by a malfunctioning lathe.

The mechanics of this injury are related to fixed vs mobile structures in the abdomen. Injuries tend to occur adjacent to areas of the intestine that are fixed, such as the cecum, ligament of Treitz, colonic flexures and rectum. During sudden deceleration, portions of the intestine adjacent to these areas continue to move, pulling on the nearby attachments. This causes the intestine itself to pull off of its mesentery.

The terminal ileum is the most common site for bucket handle tears. Proximal jejunum, transverse colon, and sigmoid colon are other possible areas. The picture above shows multiple bucket handle injuries in one patient. There are 3 injuries in the small bowel, and one involving the entire transverse colon. Note the obviously devascularized segment at the bottom center of the photo.

Always think about the possibility of this injury in patients with very high speed decelerationsSeat belt marks have a particularly high association with this injury. If your patient has an abnormal exam in the right lower quadrant, or if the CT shows unusual changes there (“dirty” mesenteric fat, thickened bowel wall, extravasation), I recommend a trip to the OR. In these cases, an injury will nearly always be present.

Related posts:

Source: personal archive. Not treated at Regions Hospital

Treating Numbers: Pulse Oximetry

How many times has this happened to you? You walk into a young, healthy trauma patient’s room and discover that they have nasal prongs and oxygen in place. Or better yet, these items appear overnight on a patient who never needed them previously. And the reason? The pulse oximeter reading had been low at some point.

This phenomenon of treating numbers without forethought has become one of my pet peeves. Somehow, it is assumed that an oximetry value less than the standard “normal” requires therapy. This is not the case.

In young, healthy people the peripheral oxygen saturation values (O2 sat) are typically 96-100% on room air. As we age, the normal values slowly decline. If we abuse ourselves (smoking, working in toxic environments, etc), lung damage occurs and the values can be significantly lower. Patients with obstructive sleep apnea will have much lower numbers intermittently through the night. 

So when does a trauma inpatient actually need supplemental oxygen? Unfortunately, the literature provides little guidance on what “normal” really is in older or less healthy patients. Probably because there is no norm. The key is that the patient must need oxygen therapy. How can you tell? Examine them! Talk to them! If the only abnormal finding is patient annoyance due to the persistent beeping of the machine, they don’t need oxygen. If they feel anxious, short of breath, or have new onset tachycardia, they probably do. Saturations in the low 90s or even upper 80s can be normal for the elderly and smokers.

Bottom line: Don’t get into the habit of treating numbers without thinking about them. There are lots of reasons for the oximeter to read artificially low. There are also many reasons for patients to have a low O2 sat reading which is not physiologically significant. So listen, talk, touch and observe. If your patient is comfortable and has no idea that their O2 sat is low, turn off the oxygen and toss the oximeter out the window. 

PAs and NPs In Level I Trauma Centers

Trauma service staffing is important to maintaining trauma center status. Teaching centers in the US have been grappling with resident work hour rules, and non-teaching centers have always had to deal with how to adequately staff their trauma service. What is the impact of staffing a trauma center with midlevel practitioners (MLPs) such as physician assistants and nurse practitioners?

A state designated Level I trauma center in Pennsylvania retrospectively examined the effect of adding MLPs to an existing complement of residents on their trauma service. They examined the usual outcomes, including complications, lengths of stay, ED dwell times and mortality. 

Here are the more interesting results:

  • ED dwell time decreased for trauma activations and transfers in, but it increased for trauma consults. Of note, data on dwell times suffered from inconsistent charting.
  • ICU length of stay decreased significantly
  • Hospital length of stay decreased somewhat but did not achieve significance
  • The incidence of most complications stayed the same, but urinary tract infection decreased significantly
  • There was no change in mortality

Bottom line: There is a growing body of literature showing the benefits of employing midlevel providers in trauma programs. Whereas residents may have a variable interest in the trauma service based on their career goals, MLPs are professionally dedicated to this task. This study demonstrates a creative and safe solution for managing daily clinical activity on a busy trauma service.

Reference: Utilization of PAs and NPs at a level I trauma center: effects on outcomes. J Amer Acad Physician Assts, July 2011.