Tag Archives: PI

Trauma PI: The 7 Deadly Sins

Trauma performance improvement is the backbone of any trauma center. And it’s the most common reason that a center runs into problems or deficiencies during a site review. Today, I’ll review the seven most common problems encountered during site surveys and provide some possible solutions for them.

  1. No loop closure. Closing the loop is talked about all the time. One would think that this should never be a problem, but it is. It becomes evident to the reviewers in two ways: reviewing PI meeting minutes, and sin #2 below. This topic is complicated, so see my four part series on loop closure here.
  2. Repeat offenses. The same problem keeps coming up again and again. This usually happens because the problem was never really solved in the first place. See the link above on loop closure for the solution to this one.
  3. Superficial peer review discussions. This means that the minutes don’t reflect any in-depth discussion of PI issues. There are two possible reasons: there wasn’t any meaningful discussion, or the documentation just wasn’t that good (see point 4 below). The discussion must include a summary of the case, identification of the significant quality issues, and a description of what will be done to avoid the problem in the future and who is responsible for carrying it out. If issues are referred to other committees (trauma operations or hospital PI), then this should be stated and it should be possible to follow the PI trail in those minutes as well.
  4. Poor peer review meeting minutes. This is a carry-on from the last point. Sometimes there is robust discussion on an issue, but the minutes don’t reflect it. This occurs due to concern for discoverability by the public or outside legal counsel in some states, but most frequently happens because the person charged with documenting the minutes is not very good at it. Minutes need not mention specific names, but do need to detail the gist of any discussion, including specific points of concern and remedies. Most discussions will run several paragraphs long; a single brief one just won’t do.
  5. Poor record organization. All PI activity and documentation regarding a specific patient needs to be organized in a single location. A paper or electronic folder is recommended. A similar folder is recommended for each system issue that involves multiple patients. During a site review, be sure to include the appropriate folder of information with each of the patient charts that are inspected by reviewers. Don’t scatter your records across several file cabinets or notebooks. And make sure several people in the trauma program understand the organization system in case a key person gets sick or quits.
  6. No cooperation from other hospital services. Your trauma program sends a case to be reviewed out to another department. Two months pass and you finally notice you never received a reply. Repeat a couple of times with the same chart and at that point, no one remembers anything about the case. This is a sure-fire way to keep making the same patient care mistakes. Create the expectation of quick turnaround (2 weeks is reasonable) and start nagging when time is up. Escalate to your trauma medical director if it continues to be a problem. In more extreme cases, you may need to select another liaison to deal with, or enlist hospital administration or hospital PI to help put pressure on them from above.
  7. No cooperation between trauma program manager and trauma medical director. These two people must work very closely for the trauma PI program to function efficiently. Regular meetings (weekly) are essential so they can review and process the various items that must be addressed. The TMD must deal with any physician related items, such as counseling, verbal discussions, memos and letters. The TPM deals with items involving nursing and other personnel. Dysfunction at this level is somewhat common at Level II trauma centers and quickly drags the program down.
Please feel free to comment or ask questions below!

Related posts:

AAST 2011: Autopsy Reports and Performance Improvement

Autopsy reports have traditionally been used as part of the trauma performance improvement (PI) process. They are typically a tool to help determine preventability of death in cases where the etiology is not clear. Deaths that occur immediately prior to arrival or in the ED are typically those in which most questions arise.

The American College of Surgeons Trauma Verification Program includes a question on what percentage of deaths at a trauma center undergo autopsy. Low numbers are usually discussed further, and strategies for improving them are considered. But are autopsies really that helpful?

A total of 434 trauma fatalities in one state over a one year period were reviewed by a multidisciplinary committee and preventability of death was determined. Changes in preventability and diagnosis were noted after autopsy results were available. The autopsy rate was 83% for prehospital deaths and 37% for in-hospital deaths. Only 69% were complete autopsies; the remainder were limited internal or external only exams.

Addition of autopsy information changed the preventability determination in 2 prehospital deaths and on in-hospital death (1%). In contrast to this number, it changed the cause of death in about 40% of cases, mostly in the prehospital deaths.

Bottom line: From a purely performance improvement standpoint, autopsy does not appear to add much to determining preventability of death. It may modify the cause of death, which could be of interest to law enforcement personnel. I would still recommend obtaining the reports for their educational value, especially for those of you who are part of training programs.

Reference: Dead men tell no tales: analysis of the utility of autopsy reports in trauma system performance improvement activities. AAST 2011 Annual Meeting, Paper 63.

Timed PI Audit Filters: When Does The Clock Start?

Several of the performance improvement (PI) audit filters typically used at trauma centers include a time parameter. These include:

  • Craniotomy > 4 hrs
  • Laparotomy > 4 hrs
  • OR for open fracture > 8 hrs
  • Compartment syndrome > 2 hrs

The question that needs to be asked is: 2 or 4 or 8 hours after what?

There are several possible points at which to start the clock:

  • Arrival in the ED
  • When the diagnosis is made
  • When the decision to operate occurs

The answer is certainly open to interpretation. Here is my opinion on it:

The purpose of a PI filter is to measure system performance. There are a myriad of system problems that can delay taking a patient to the OR. These include care delays in the ED, delays in getting or interpreting diagnostic tests, delays in contact or response for consultants, delays in diagnosis, delays in OR scheduling or availability, and more. Does it make sense to limit the evaluation of your system by setting one of the later decision points as your start time?

Bottom line: I recommend starting the audit filter clock at the time of patient arrival in the ED. This enables the PI program to evaluate every system that can possibly enable or impede your patient’s progress to the OR.

Undertriage Revisited

I’ve updated my original post on trauma undertriage when activating your trauma team. The initial post gave a general approach that was reasonably accurate as long as the number of missed activations was low. Here’s the new and improved version!

Trauma centers look at over- and undertriage rates as part of their performance improvement programs. Both are undesirable for a number of reasons. I’ll focus on undertriage today, why it happens and what can be done about it.

Undertriage in trauma care refers to the situation where a patient who meets criteria for a trauma activation does not get one. First, calculate your “magic number”, the number of patients who should have been trauma activations.

If you track the exact triage criteria met at your hospital, it is calculated as follows:

 Magic Number = (Number of ED trauma patients who met activation criteria
                                           but were not trauma activations)

If you don’t track the triage criteria, you can use ISS>15 as a surrogate to identify those patients who had severe enough injuries that should have triggered an activation. This is not as accurate, because you can’t know the ISS when the patient comes in, but it will do in a pinch. In that case, the magic number is:

Magic Number = (Number of ED trauma patients with ISS>15
                                           but were not trauma activations)

Your undertriage rate is then calculated as follows

                                        Magic Number
        ———————————————————–    x 100
           (Total number of trauma activations) + Magic Number

Undertriage is bad because patients who have serious injuries are not met by the full trauma team, and would benefit from the extra manpower and speed possible with an activation.

The most common causes for undertriage are:

  • Failure to apply activation criteria
  • Criteria are too numerous or confusing
  • Injuries or mechanism information is missed or underappreciated

Undertriage rates can range from 0% to infinity (if you never activate your trauma team). A general rule is to try to keep it below 5%.

If your overtriage rate is climbing past the 5% threshold, identify every patient who meets the ISS criterion and do a complete ED flow review as concurrently as possible. Look at their injuries/mechanism and your criteria. If the criteria are not on your activation list, consider adding them. If the criterion is there, then look at the process by which the activation gets called. Typically the ED physicians and nurses will be able to clarify the problem and help you get it solved. 

Trauma Overtriage: Why Is It Bad?

Back in December I talked about the dangers of undertriaging trauma patients (click here to review). What about the opposite problem, overtriage?

First, how do you calculate your overtriage rate? It’s pretty simple. Use your trauma registry to count how many patients arriving in the ED were trauma activations but didn’t meet any criteria:

(Number of ED trauma patients who were trauma activations
                         but did not meet activation criteria)

        ——————————————————–           x 100
                  (Total number of trauma activations)

This can only be accurately determined if the activation criteria are recorded on each patient. If not, use the following equation:

 (Number of ED trauma patients who were trauma activations
                                     with ISS <= 15)
       ———————————————————           x 100
                  (Total number of trauma activations)

Values can range from 0% to 100%. The usually acceptable overtriage rate is 50-80%. What happens when the overtriage rate is too high? You wear out your trauma team. They are being called for patients with injuries that don’t warrant it.

The solution for overtriage? Change your activation criteria, or add a second level trauma response that doesn’t require as many people to respond. This requires a thoughtful analysis of your existing criteria so you can decide what needs to be changed or discarded.

The danger? More undertriage. Over- and undertriage go hand in hand. As overtriage decreases, undertriage increases. You need to strike a balance so that the undertriage rate stays below 5%. This makes an excellent performance improvement (PI) program project!