Category Archives: Guidelines

How To Craft A Clinical Practice Guideline

All US trauma centers verified by the American College of Surgeons are required to have clinical practice guidelines (CPG). Trauma centers around the world generally have them, but may not be required to by their designating authority. But don’t confuse a policy about clinical management, say for head injury, with a real CPG. Policies are generally broad statements about how you (are supposed to) do things, whereas a CPG is a specific set of rules you use when managing a specific patient problem.

  1. Look around; don’t reinvent the wheel! This is the first mistake nearly every center makes. It seems like most want to spend hours and hours combing through the literature, trying to synthesize it and come up with a CPG from scratch. Guess what? Hundreds of other centers have already done this! And many have posted theirs online for all to see and learn from. Take advantage of their generosity. Look at several. Find the one that comes closest to meeting your needs. Then “borrow” it.
  2. Review the newest literature. Any existing CPG should have been created using the most up to date literature at the time. But that could have been several years ago. Look for anything new (and significant) that may require a few tweaks to the existing CPG.
  3. Create your draft, customizing it to your hospital. Doing things exactly the same as another center doesn’t always make sense, and it may not be possible. Tweak the protocols to match your resources and local standards of care. But don’t stray too far off of what the literature tells you is right.
  4. Make sure it is actionable. It should not be a literature summary, or a bunch of wishy-washy statements saying you could do this or consider doing that. Your CPG should spell out exactly what to do and when. (see examples below)
  5. Create a concise flow diagram. The fewer boxes the better. This needs to be easy to follow and simple to understand. It must fit on one page!
  6. Get buy-in from all services involved. Don’t try to implement your CPG by fiat. Use your draft as a launching pad. Let everyone who will be involved with it have their say, and be prepared to make some minor modifications to get buy-in from as many people as possible.
  7. Educate everybody! Start a campaign to explain the rationale and details of your CPG to everyone: physicians, nurses, techs, etc. Give educational presentations. You don’t want the eventual implementation to surprise anyone. Your colleagues don’t like surprises and will be less likely to follow along.
  8. Roll it out. Create processes and a timeline to roll it out. Give everyone several months to get used to it.
  9. Now monitor it! It makes no sense to implement something that no one follows. Create a monitoring system using your PI program. Include it in your reports or dashboards so providers can see how they are doing. And if you really want participation, let providers see how they are doing compared to their colleagues. Everyone wants to be the top dog.

In my next post, I’ll pontificate a bit about guidelines vs protocols, and the difference between evidence-based vs evidence-informed.

Print Friendly, PDF & Email

Why Create Practice Guidelines?

Practice guidelines are everywhere. More and more organizations have developed processes to create high quality ones. But why should we care? Do they improve what we already do?

Here are my reasons for using practice guidelines:

  • They provide a consistent way of approaching a clinical issue. Everybody working with the patient knows how things will be done, so they don’t have to remember the nuances that particular doctors or providers like.
  • They (hopefully) use the best and most valid scientific data to address the care issue, thus giving trauma professionals the opportunity to provide the best care we know of.
  • They decrease errors and complications by narrowing the number of choices available to providers.
  • They decrease waste for the same reason. For example, drawing blood every 6 hours vs daily for solid organ injuries can add up to three unneeded tests every day.
  • They provide our trainees with one good way to deal with the clinical issue. This is important when they move on to independent practice, and sometimes when taking standardized tests (boards).

Bottom line: If 10 trauma professionals deal with a given clinical problem 10 different ways, then none of them are doing it right! Develop a guideline that all of them can live with, based on current literature (if any). That way they can all be right for once, and our patients will reap the benefits.

In my next post, I’ll describe how to craft a good practice guideline.

Print Friendly, PDF & Email

5 Guidelines For Diaphragmatic Injury

Today’s post is another review of some of the practice guidelines published by the Eastern Association for the Surgery of Trauma (EAST).  This one covers the evaluation and management of diaphragmatic injury.

Diaphragm injury is a troublesome one to diagnose. It is essentially an elliptical sheet of muscle that is doubly-curved, so it does not lend itself well to diagnosis by axial imaging. Addition of sagittal and coronal reconstructions to a thoraco-abdominal CT has been helpful, but still has a far from perfect diagnostic record.

From an evaluation standpoint, there are several possibilities:

  • Observation – not generally recommended. It is usually combined with imaging such as chest x-ray to see if interval changes occur that would indicate the injury.
  • Chest x-ray – this is not often diagnostic, but when herniation of abdominal contents is obvious the patient most assuredly has an operative problem.
  • Thoraco-abdominal CT scan – this technology keeps getting better, especially with thinner cuts and different planes of reconstruction. Sometimes even subtle injuries can be detected. But this exam is still imperfect.
  • Laparoscopy or thoracoscopy – this technique yields excellent accuracy when the injury is in an area that can be viewed from the operative entry point chosen.
  • Laparotomy or thoracotomy – this is the ultimate choice and should be nearly 100% accurate. It is almost the most invasive and has more potential associated complications.

EAST reviewed a large body of literature and selected 56 pertinent papers for their quality and design. They critically reviewed them and applied a standard methodology to answer several questions.

Here are the questions with the recommendations from EAST, along with my comments:

  • Should laparoscopy or CT be used to evaluate left-sided thoraco-abdominal stab wounds? First, these patients must be hemodynamically stable and not have peritonitis. If either is present, there is no further need for diagnosis; a therapeutic procedure must be performed.
    Left sided diaphragm injuries from stabs are evil. The hole is small, and since the pressure within the abdomen is greater that the chest, things always try to wiggle their way through this small hole. It can remain asymptomatic if the wiggler is just a piece of fat, but can be catastrophic if a bit of the stomach or colon pushes through and becomes strangulated. Furthermore, these holes enlarge over time, so more and more stuff can push up into the chest.
    EAST recommends the use of laparoscopy for evaluation to decrease the incidence of missed injury. However, if the injury is in a less accessible location (posterior), the patient has body habitus issues, or adhesions from previous surgery may lead to incomplete evaluation, laparotomy should be strongly considered.
  • Should operative or nonoperative management be used to evaluate right-sided thoraco-abdominal penetrating wounds? Note that this is different than the last question. All penetrating injuries are included (stabs and gunshots), and this one is for management, not evaluation. And the same caveats regarding hemodynamic stability and peritonitis apply. It applies to both stabs and gunshots.
    Unlike left-sided injuries, right-sided ones are much more benign. The liver keeps anything from pushing up through small holes, and they do not tend to enlarge over time due to this protection. For that reason, EAST recommends nonoperative management to reduce mortality and morbidity related to operation.
  • Should stable patients with acute diaphragm injury undergo repair via an abdominal or thoracic approach? This question applies to any diaphragm injury that requires operation, such as right-sided penetrating injury or any blunt injury. EAST recommends an approach from the abdomen to reduce morbidity and mortality. Since abdominal injury frequently occurs in these cases, an approach from the chest limits the ability to identify and repair abdominal injuries. Otherwise, you may find yourself doing a laparotomy in addition to the thoracotomy.
  • Should patients with delayed visceral herniation through a diaphragm injury undergo repair via an abdominal or thoracic approach?  For years, the preferred approach for delayed presentations has been through the chest because the injury is easier to appreciate and repair.  However, if ischemic or gangrenous viscera are present, it will be more difficult to manage and repair from the chest. EAST does not make a specific recommendation for this question and suggests the surgical approach be determined on a case by case basis.
  • Should patients with an acute diaphragm injury from penetrating injury without concern for other intra-abdominal injuries undergo open or laparoscopic repair? The quality and quantity of data addressing this question were very low, but EAST recommends laparoscopy for repair of these injuries to reduce morbidity and mortality. This includes blunt injuries, which tend to be larger. There were some conversions to an open procedure, especially in the blunt cases. The usual caveats on exposure, injury location, body habitus and previous surgery apply.

Reference: Evaluation and management of traumatic diaphragmatic injuries: A Practice Management Guideline from the Eastern Association for the Surgery of Trauma. J Trauma 85(1):198-207.

Print Friendly, PDF & Email

Massive Transfusion: What Ratios Are People Using?

This is the first of a two-part series on massive transfusion protocol (MTP) ratios. Today, I’ll write about what ratios trauma centers around the country are using. Tomorrow I’ll review the literature we have to date on what the correct ratio should be. Are we all doing the right thing or not?

Back in the old days (which I remember fondly), we didn’t pay too much attention to the ratio of blood to plasma. We gave a bunch of bags of red cells, then at some point we remembered that we should give some plasma. And platelets? We were lucky to give any! And to top it all off, we gave LOTS of crystalloid. Turns out this was not exactly the best practice.

But things have changed. Some good research has shown us that a nice mix of blood component products is good and too much crystalloid is bad. But what exactly is the ideal mix of blood products? And what is everybody else doing? I’ll try to answer these questions in this series.

So first, what are all the other trauma centers doing? An interesting medley of anesthesia and pathology groups from the University of Chicago, a Dallas-based anesthesia group, and a blood center in my home base of St. Paul, conducted a survey of academic medical centers in 2016. They wanted to find out how many actually had a MTP and to scrutinize the details.

They constructed a SurveyMonkey survey and sent it to hospitals with accredited pathology residencies across the US. There were 32 questions in the survey, which asked for a lot of detail. As you can probably personally attest, the longer and more complicated the survey, the less likely you are to respond. That certainly happened here. Of 107 surveys sent out, it took a lot of nagging (initial email plus two nags) to get a total of 56 back.

Here are the factoids:

  • Most were larger hospitals, with 74% having 500 or more beds
  • All had massive transfusion protocols
  • Trauma center level: Level I (77%), Level II (4%), Level III (4%), Level IV (2%), no level (14%)
  • Nearly all (98%) used a fixed ratio MTP; very few used any lab-directed (e.g. TEG/ROTEM) resuscitation
  • Target RBC:plasma ratio: 1:1 (70%), 1.5:1 (9%), 2:1 (9%), other (9%)
  • Only 58% had the same RBC:plasma ratio in each MTP cooler
  • More than 86% had thawed plasma available (remember, these were generally large academic centers)
  • Half stored uncrossmatched type O PRBCs outside the blood bank, usually in the ED; only 1 stored thawed plasma in the ED
  • A total of 41% had more than one MTP (trauma, OB, GI, etc.)
  • 84% had some type of formal review process once the MTP was complete
  • About 68% had modified their MTP since the original implementation. Some increased or decreased ratios, expanded MTP to non-trauma services, decreased the number of units in each pack, changed to group A plasma from AB, or switched from ratio to TEG/ROTEM or back.

Bottom line: This is an intriguing snapshot of MTP practices around the country that is about four years old. Also remember, this is a somewhat skewed dataset. The survey was directed toward hospitals with academic pathology programs, not trauma centers. However, there is enough overlap that the results are probably generalizable. 

Most centers are (were) using MTP packs containing six units of PRBCs, and were attempting to achieve a fixed 1:1 ratio. Half of hospitals had the same number of units in each cooler, half varied them by cooler number. Nearly half had multiple flavors of MTP for different specialties. Very few used TEG/ROTEM during the initial phased of MTP. Most modified their MTP over time.

I’ve written quite a lot on most of these issues. See the links to my “MTP Week” series from earlier this year, below.

Tomorrow, I’ll review what we know and don’t know about the proper ratios to use in your MTP.

Reference: Massive Transfusion Protocols: A Survey of Academic
Medical Centers in the United States. Anesth & Analg 124(1):277-281, 2017.

MTP week series:

Print Friendly, PDF & Email

Best Practice: Use of CT Scan In Trauma Activations – Part 2

In my last post, I described how the unscheduled and random use of CT scan in trauma activations can interfere with normal radiology department workflow, creating access problems for other emergency and elective patients. Today, I’ll detail a project implemented at my hospital to analyze the magnitude of this problem and try to resolve it.

We started with a detailed analysis of how the scanner was being used for trauma activation patients. Regions Hospital has a single-tier trauma activation system, with no mechanism of injury criteria other than penetrating injury to the head, neck, and torso. There are usually about 850 activations per year, and traditionally the CT scanner has been “locked down” when the activation is announced. The CT techs would complete the current study on the table, then hold the scanner open until called or released by the trauma team.

Since we are a predominantly blunt trauma institution, we scan most stable patients. Our average time in the trauma bay is a bit less than 20 minutes. Add this time to the trauma activation prenotification time of up to 10 minutes, and the scanner has the potential to sit idle for up to half an hour. And in some cases when scan is not needed (minor injuries, rapid transport to OR) the techs were not notified and were not aware they could continue scanning their scheduled cases.

A multidisciplinary group was created and started with direct observation of the trauma activation process and a review of chart documentation and radiology logs. On average it was calculated that the scanner was held idle for an average of 17.9 minutes too long. This is more than enough time to complete one, or even two studies!

A new process was implemented that required the trauma team leader to call out to the ED clerk placing orders for the resuscitation 5 minutes before the patient would be ready for scan. I still remember the first time this happened to me. I was so used to just packing up and heading to scan, I got a little irritated when told that I hadn’t made the 5-minute call. But it’s a good feedback loop, and I never forgot again!

We studied our workflow and results over a 9-week period. And here are the factoids:

  • The average CT idle time for trauma activations before the project was 17.9 minutes
  • This decreased to an average idle time of 6.4 minutes during the pilot project
  • Total idle time for all activations was 8.3 hours, but would have been 36 hours under the old system
  • A total of 28.6 hours were freed up, which allowed an additional 114 patients to be scanned while waiting for the trauma activation patients

This was deemed a success, and the 5-minute rule is now part of the routine flow of our trauma activations. We rarely ever have to wait for CT, and if we do it’s usually due to the team leader not thinking ahead.

Bottom line: This illustrates the processes that should be used when a quality problem surfaces in your program:

  • Recognize that there is a problem
  • Convene a small group of experts to consider the nuances
  • Generate objective data that describes the problem in detail
  • Put on your thinking caps to come up with creative solutions
  • Test the solutions until you find one that shows the desired improvement
  • Be prepared to modify your new systems over time to ensure they continue to meet your needs
Print Friendly, PDF & Email