The opinion piece below was written by an emergency physician regarding a New York Times article by Jim Dwyer (picture at right). The author did not want to be identified due to fears of retribution from either the NY Times or from the hospital at which the physician is employed.
In addition to the points the author raises below, I would add these additional points of information:
1. The “Stop Sepsis” campaign cited in Mr. Dwyer’s article specifically stated that it is only to be used for tracking patients with severe sepsis and that “only those patients who are hypotensive after being given 2L of fluids or that have an elevated lactate should be entered in the data portal for this Collaborative.” Rory was not hypotensive and no lactate level was included in the labs pictured in Mr. Dwyer’s article. Mr. Dwyer never mentions any of these facts. The Collaborative does not allow access to links on this page describing its screening tools or to how it believes that a determination for ordering a lactate level should be made.
I will also note that Mr. Dwyer responded to some of the more than 1600 comments to his article, including some of the issues raised below, in this follow up article.
UPDATE JULY 22, 2012
Also see an important update to this debate at this link.
The New York Times published an incredibly sad story about a 12 year old boy named Rory who went into the NYU emergency department, was diagnosed with gastroenteritis (a viral stomach bug), and who was dead two days later from septic shock. Those are just about the only facts that are not in dispute. The rest of the New York Times article seems to build a mountain of evidence as to why the emergency physician screwed up. However, as is frequently the case, the truth is much more complicated than the media would have you believe. There are lots of comments from other doctors using the almighty retrospectoscope and so many clinical inaccuracies discussed that this sad story is turned into a piece of sensationalistic journalism.
This post is mostly for the non-medical people that read this blog to help you understand the medical issues a little better. This is a scientific discussion of the main inaccuracies of the article followed by what possibly could have been done better. I say “possibly” because I did not examine the patient and all of my information is through the New York Times article. If you have already read the article and decided that the ED doctor screwed up and nothing can change your mind, then stop reading. If you want a fair and evidence-based discussion of the article then read on.
- The article references the “Stop Sepsis” campaign and says that the vital signs that should have triggered an evaluation for severe sepsis. The article says that Rory had initial vital signs of a temperature of 102, a heart rate of 140, and then points to the Stop Sepsis guidelines. There are two problems with this: First, the Stop Sepsis guidelines are intended to be used in adults, not in children. Second, just because a patient has abnormal vital signs doesn’t mean that they have severe sepsis. Most patients in a pediatric ED waiting room would meet these criteria and yet they don’t have severe sepsis. The “Stop Sepsis” guidelines are a screening tool that can suggest sepsis but they have to be used in the right clinical context. Most physicians see pediatric patients every day who meet the “Stop Sepsis” criteria and who would best be described as having “the sniffles.”
- The article states that Rory’s temperature at home was “104, his highest ever.” The implication is that this high fever, in of itself, should have triggered a more thorough investigation. This is a misconception that must be dispelled. A temperature of 106.7 degrees and above is the only time a fever by itself is dangerous. Significant literature shows that a fever less than 106.7 degrees is not harmful. We see children in the ED all the time with “high fevers” but that look great and would not have been considered sick without their temperature being taken at home.
- After Zofran and IV fluids, Rory felt better. His vitals before discharge were a temp of 102 and a pulse of 131. The article continues to allege that these vital signs met sepsis criteria. However, a heart rate of 110 is the upper limit of normal for a 12 year old and, in general, a patient’s heart rate increases by about 10 beats per minute for every 2 degree increase in body temperature. In Rory’s case, a heart rate of 131 was appropriate or just above the upper limits of normal for his temperature. The persistence of a fever should also not cause worry just by itself. Physicians frequently discharge febrile children from the ED without any adverse consequences. When assessing patients for discharge, what matters most is how the patient looks. According to the documentation in this case, Rory looked better before he was discharged.
- The article alleges that the emergency physician didn’t see the vital signs before she wrote the discharge instructions. While this may be true, there are several more likely explanations. In order to be efficient, I sometimes write discharge orders on patients that I think are going home because I have a minute free to put the order in. Technically, the orders are entered before the vital signs are entered, but this is for the sake of efficiency. I still evaluate the patient prior to discharge. Another possibility is that it the nurse didn’t have the time to put Rory’s vital signs into the computer before he was discharged. It is likely that the ED physician saw Rory’s vital signs while she was in the room re-evaluating Rory and signed the discharge order before the nurse entered the vital signs. The way the article is written, it implies that the ED physician could never have seen the vital signs, but in reality, there is no way to tell for sure without asking the physician.
- The article implies that the white blood cell count of 14.7 should have triggered a more aggressive workup. There is a mountain of evidence to say that a high WBC count does not rule in or rule out an infection or severe sepsis.
In order to keep this balanced, here are some things that I believe could possibly have been done better in Rory’s case.
- Discharging the patient before labs were back. While this is sometimes done in cases of cultures or other “send out” tests, it is generally not a good idea to discharge patients before labs results are reported. The most concerning lab in Rory’s case was the elevated number of bands or immature white blood cells at 53%. High band count can be a red flag in the right clinical situation, but may also be a sign of a vigorous immune system response to a viral infection. In addition, Rory’s carbon dioxide was normal. In severe sepsis one would expect Rory to have acidosis and a low carbon dioxide level. A normal carbon dioxide level suggests that Rory may not have had severe sepsis at the time of his first ED visit. Instead, he was probably in the early stages of sepsis which can be very difficult to distinguish from a simple viral illness. A more thorough review of the labs may have prompted an admission for observation, but without having examined the patient no one can make that call.
- Rory’s vital signs at discharge were at the upper limits of what could be considered normal given his fever. While the vital signs did improve, when I read the article the first time I admit that the discharge vitals raised an eyebrow. However, as I previously noted, a child’s appearance is probably the most important indicator of severe illness and the ED physician is the only person who examined Rory.
I would like to end this article with a plea to the public to not crucify this ED doctor. The New York Times should not have published the doctor’s name. She is not a public figure and she has not been named in a lawsuit. It is egregious that the New York Times published her name and thus unleashed the public venom on a private citizen. I can guarantee that the ED doctor feels terrible about this case. We don’t need the rest of the world coming down on her as well. Even worse, due to federal patient privacy laws, the physician is prohibited from speaking about this case. In any other profession, if a newspaper published something condemning your professional abilities you would be able to give your side of the story.
This case is every ED doctor’s worst nightmare and it can happen to the best of us. As Greg Henry says, cases like this make you say to yourself “only by the grace of God go I.” Cases like this keep physicians up at night. The New York Times didn’t see it that way and wrote a sensational article condemning the ED doctor involved. The reality is that very rarely, kids get sick and die. Sepsis is a cruel disease and it can take a child that is otherwise healthy and looks great and kill them within hours to days. It is no one’s fault – it is just bad luck. Sometimes we catch that needle in the haystack and no one hears about it. Sometimes we don’t and then it becomes front page news. Let’s not make this situation worse by placing all the blame on this ED doctor. This is a terrible case and while we should always try to learn and be better for the next patient, sometimes bad things just happen.