Portal de Periódicos da CAPES

Chaos, Turmoil, Hysteria

2006; Lippincott Williams & Wilkins; Volume: 28; Issue: 8 Linguagem: Inglês

10.1097/00132981-200608000-00049

1552-3624

Anne Scheck,

It was a dark and stormy night. The emergency department began filling with patients. A multi-car pile-up sent in one group; another two were near-drownings, plucked from a flooded creek. One elderly man hobbled in on legs that had slid out from under him, right onto a rain-slicked sidewalk. It was a perfect time to test the effects of chaos-related “drift,” the supposition that appropriate decision-making can be compromised during high-stress peak periods. But there was just too much going on to assess how the scenario might affect triage. After all, in the thrall of such time-driven demands, is there a way to do such a study in the emergency department? In fact, several recent investigations have done just that, and all of them seem to have reached a similar conclusion: Not only is there a need for triage improvement, but there are innovative ways to meet that need. The studies suggest that making determinations while waging a struggle against time puts triage itself at risk, despite the fact that the system has been enhanced over the years. The Emergency Severity Index (ESI), a five-level triage system, for example, has proven more reliable than its predecessor, a three-tier one. The expanded index, though, appears to have one shortcoming: Patients at risk for immediate deterioration may not be placed into the most urgent category right away, particularly during periods of crowding. As a result, medical centers in the United States, Europe, and Canada have been studying triage processing, and not surprisingly, human responses are implicated. During patient overload, phones ring, orders are issued, and family members hover, all distractions that can impinge on the process. The crux of the problem seems to be the need for decisive, accurate risk stratification. Some investigators have been showing that oxygen saturation, aside from Glasgow coma scale and age, is the best predictor of mortality risk. (Emerg Med J 2006;23:372.) Others have demonstrated that when physicians instead of triage nurses undertake risk assessment, processing times are substantially shorter. (Emerg Med J 2006;23:262.)FigureWith some already at crisis levels, EDs may face even tougher times ahead. Last year, the American Association of Medical Colleges issued an ominous forecast at the Physician Workforce Research Conference. While ambulatory visits are sharply rising for the middle-aged, the number of physicians practicing primary care medicine is dipping. This graying of American medicine and the patients served by it spells a possible physician shortage that can mean heavier reliance than ever before on the ED, as well as a future patient population with potentially higher acuity. Revisions for ESI? So a study to determine a better way to stratify high-risk patients seems prescient, and that is just what some large metropolitan hospitals have done. One of them, Northwestern Memorial Hospital in Chicago, is a Level I trauma center whose staff observed that a substantial number of patients in category two seemed to need immediate care, a category one designation. “That was a big deal to us,” said James Adams, MD, a professor and the chairman of emergency medicine at the Feinberg School of Medicine at Northwestern University. “You got level one when you came in dead,” Dr. James Adams. An investigation was launched to evaluate whether the ESI needed revision for high-acuity patients. Three other centers also participated in the study: Brigham and Women's Hospital in Boston, University of North Carolina Hospitals in Chapel Hill, and Lehigh Valley Hospital in Allentown, PA. The source of the problem? Not early decision-making. Patients were being wheeled back to the appropriate area in a timely way, but the researchers found that level two was being used for those who were somewhat responsive, and level one, in contrast, was reserved for those with no pulse. “You got level one when you came in dead,” Dr. Adams noted. Findings from the study, which included nearly 800 patients in a prospective investigation, suggested that level one criteria had to encompass all patients who appeared to need immediate cardiac, airway, or hemodynamic intervention. In short, the implication from the data was simple and straightforward, according to Dr. Adams. “Designate critical patients as critical. Don't underestimate severity,” he said. By identifying patients who were in need of immediate support as level one, such as those in severe respiratory distress or who complained of severe cardiac symptoms, the life-saving measures boosted survival and the data yielded a clearer picture of patient load in the ED. “That's important because if we are downrating, then we don't have an optimal understanding of our resource needs,” said Dr. Adams. Essentially, what had been occurring was a rationing of care, a sort of battlefield strategy in which the sickest of the sick were at times the only patients getting immediate intervention. “It is not that this was inconsistent. It was very consistent on the nurses' part,” Dr. Adams explained. The highest category was reserved for the most critically ill, the result of attempting to provide the best resource allocation. But if you don't triage at the right level, you don't get an accurate picture, and you are a lot less likely to get the resources that are needed. “Understanding the acuity demand tells you what you need,” he said. Results from Canada and Great Britain appear to be remarkably similar. Studies there have been done with the same objective in mind, to determine if triage was being done with efficiency that could be documented and whether tweaking the technique could improve scoring in circumstances in which gaps were found. The problem, as such studies imply, is not inter-rater reliability. Just as Dr. Adams observed, nurses in a single center often showed great consistency. But so much of what triage nurses do depends on individual memory, knowledge, and experience that some form of standardization has been suggested as one way to overcome potential variability. Reducing Drift At the University of Alberta, emergency staff are trying to reduce the possibility of drift during triage. To do so, they are instituting a computer-assisted method; two studies suggest that there is better agreement in triage when it is used. But what staff member wants to do “e-triage” when split-second recommendations are required? “Most of the feedback has been very positive,” said Sandy Dong, MD, an assistant clinical professor of emergency medicine at the university. The senior author of the two investigations, Dr. Dong noted that “e-triage,” as it is sometimes called, is demonstrating quite a bit of improvement over traditional approaches in direct comparisons. (Acad Emerg Med 2005;12:502.) “I think it needs to be user-friendly. Get buy-in. Train everyone.” Dr. Sandy Dong Triage nurses, who prioritize patients under very demanding conditions and with great compassion, nonetheless at times seem prone to the same kind of down-triaging to which Dr. Adams alluded, the study showed. “There was significant discrepancy by nurses using memory-based triage when compared with a computer tool,” according to the study. “Triage decision support tools can mitigate this drift, which has administrative implications for EDs.” Another advantage is data collection, which can lead to the kind of resource allocation suggested by Dr. Adams, noted Dr. Dong, an attending physician at the Royal Alexandra Hospital in Edmonton, Alberta, Canada, adding that the tool has shown good reliability. (Acad Emerg Med 2006;13:269.) Looking back over the data, the researchers could see when fever began cropping up in the ED and used that to predict flu season. None of this means that experienced nurses should trust a computer screen more than their clinical instincts, he said, and users were in fact encouraged to override the findings if they believed that was in the patient's best interest. But in a busy ED, a computer that helps define the criteria for each triage level at the touch of a keystroke is a welcome addition, at least at his center. “I think it needs to be user-friendly,” he said. And for anyone wishing to undertake e-triage, it is essential to follow this familiar refrain: “Get buy-in. Train everyone,” Dr. Dong advised, adding that his ED held half-day training sessions for every staff member who had patient contact, an effort that avoided the possible pitfall of having some staff acquire skills that they are responsible for passing on to others. Finally, it is essential to appreciate triage nurses, who are called upon to make immediate decisions in the face of other demands. As Dr. Dong and his colleagues observed, take a look at any ED, and the triage desk is the “chaotic area,” even on nights that aren't necessarily dark and stormy. “Telephone calls, inquiries from patients and family members in the waiting room, and other distractions put undue pressure on the triage nurse,” Dr. Dong and his colleagues stated in one of the studies. “By contrast, the study nurse using the electronic triage tool did not face any such distractions, and had more time to assess each patient.” Triage for Children Faster than for Adults The medical literature is rife with anecdotal accounts of how a high index of suspicion can save lives, and many of them chronicle pediatric cases in the ED. Do such common case histories mean kids get faster ED care after triage? In an intriguing study of 675 patients, emergency personnel at Queen Elizabeth Hospital in Bridgetown, Barbados, found that children from infancy to about 11 had the shortest waiting times for care in a crowded ED, aside from patients with high-acuity scores. Those patients also had relatively short lengths of stay, particularly when compared with adults over 49, who racked up the longest lengths of stay. The study showed, discouragingly, that mean total length of stay for most patients who went to the ED was nearly twice the U.S. benchmark of 90 minutes. So why did children get in and out of the ED with such relative speed? Was there preferential attention paid to them? Could there have been greater availability of specialists, such as pediatric nurses or pediatricians? Or, perhaps, was it because many injuries in children are either very grave or very minor, necessitating either immediate action or relatively little intervention? “We do not really know why pediatric patients had so much shorter waiting periods,” said Anne Carter, MD, an author of the study, conducted by the School of Clinical Medicine and Research at the University of the West Indies in Barbados. Any suggestion would be “complete speculation,” she said. But one possibility is that the two diseases for which her center has nurse-driven protocols — asthma and sickle cell disease — are particularly frequent in children. “These patients do not have to wait to see a doctor, but are treated immediately according to protocol,” said Dr. Carter, a senior lecturer at the School of Clinical Medicine and Research at the University of the West Indies in Barbados. “However, we would have to do another study to find out the real reasons.” — Anne Scheck Comments about this article? Write to EMN at [email protected].