Research Roundup (July 2023)

The pace of these literature summaries has decreased over the years, but perhaps that means that quality has increased? I think there is an interesting variety of papers this time around, from sepsis, to bullshit, to patient access to their own results online. A few huge papers dropped in the past few weeks, including RCTs of REBOA and a new TXA in trauma paper, but you will have to wait until next time for those ones. Hope you enjoy. 

Perhaps herbs are the cure to sepsis?

Liu S, Yao C, Xie J, et al. Effect of an Herbal-Based Injection on 28-Day Mortality in Patients With Sepsis: The EXIT-SEP Randomized Clinical Trial. JAMA Intern Med. 2023 May 1:e230780. doi: 10.1001/jamainternmed.2023.0780. Epub ahead of print. PMID: 37126332 

Are herbs the cure for sepsis? (Of course not, but this is an interesting and well done RCT worth discussing.) This is a double-blind placebo controlled trial of 5 days of an injectable herbal product (xuebijing) given to relatively healthy adult patients with sepsis in 45 ICUs in China. There was a 7% decrease in absolute mortality, which is phenomenal, but also perhaps a little too good to be true? Unlike many of the studies I cover, the methodology here is pretty good. There are a few nitpicks, including the fact that 25% of patients didn’t complete the 5 day treatment course, and the fact that they didn’t know if 57 patients were alive or dead at the end of the trial, but there isn’t a reason to immediately throw out the study. I think the two big topics to consider are pretest probability and the illogic of herbal remedies in general. In terms of pretest probability, the chance that anything (let alone a herbal product) will reduce all cause mortality in sepsis in modern medicine is pretty low, and so a single trial will never bring us to certainty. These results need replication before they change clinical practice. More importantly, the herbal approach just doesn’t make sense. Using hundreds of chemicals in uncontrolled and highly variable doses is a recipe for disaster. If these results are true, the next step really needs to be identifying the important active ingredients, so we can produce a safer and more consistent product for our patients.

There is a full write up here. 

Bottom line: Herbs will never be the answer in medicine, but they might be the first step on the road to an answer.

The straw men (AKA bullshit) of integrative health and alternative medicine

Caulfield T. The straw men of integrative health and alternative medicine. BMJ Best Practice. Available at: https://bestpractice.bmj.com/info/toolkit/discuss-ebm/the-straw-men-of-integrative-health-and-alternative-medicine/

To believe in science is to reject the many forms of snake oil that people use to scam and exploit vulnerable patients in the name of “complementary and alternative medicine”. (If an “alternative medicine” is science based, it instantly just becomes a medicine.) Unfortunately, because of the ‘Gish Gallop’, it is much easier to spout endless nonsense than it is to refute it. This short essay refutes 4 of the most common nonsense straw man arguments made by anti-science ‘alternative medicine’ practitioners. It is a short and witty read, so anyone interested in the topic is probably better off skipping my summary and reading the article itself. The 4 main points of the article are 1) of course we appreciate and understand the placebo effect (but that doesn’t justify ridiculous magical thinking); 2) of course we understand that many conventional therapies don’t work (but that doesn’t justify using unproven ‘alternative’ therapies); 3) of course we embrace patient-centered and preventive care (and embracing the ‘whole patient’ doesn’t give you the right to sell the snake oil); and 4) of course we embrace being open minded, because the core ethos of science is changing your mind in the face of new evidence. (In fact it is the “alternative medicine” crowd that is incredibly close minded, because they will never change their mind, no matter how much evidence that is presented.)

Bottom line: Doctors have a responsibility to protect our patients from snake oil salesmen, which means we need to be at least somewhat fluent in the bullshit they spew.

“I still partly think this is bullshit”: perceptions of cannabinoid hyperemesis syndrome

Collins AB, Beaudoin FL, Metrik J, Wightman RS. “I still partly think this is bullshit”: A qualitative analysis of cannabinoid hyperemesis syndrome perceptions among people with chronic cannabis use and cyclic vomiting. Drug Alcohol Depend. 2023 May 1;246:109853. doi: 10.1016/j.drugalcdep.2023.109853. Epub 2023 Mar 24. PMID: 36996524

This is a qualitative study, based on interviews of 24 adult patients who presented to the emergency department with a presumptive diagnosis of cannabinoid hyperemesis syndrome. I will admit that I largely selected the paper because of the title, but I think this is valuable research, and we should probably spend more time using qualitative methodology to understand our patients’ experiences in the emergency department. (We like to think that pure rational facts are the foundation of medicine, but most of our management plans rely on behavioral change, and we know that human behavior is anything but rational). My big takeaway from this paper, which I think is pretty obvious when talking to patients, is the baseline level of cognitive dissonance that this disease creates. Most people have a long history of marijuana use without any issues prior to developing symptoms. There is not an obvious personal connection between their symptoms and the drug use, and therefore the as a component of distrust or denial, especially as there is not a gold standard test to prove the diagnosis. On the other hand, almost everyone I have met with this condition recognizes rationally that daily marijuana use is somewhat excessive, and that there is a logical tie between drug use and side effects. Hence the title quote, “I still partly think that this is bullshit, but I know it’s not.” These patients were all researching this condition online, so I think smart clinicians will embrace that. I tell my patients, “I am pretty sure you have cannabinoid hyperemesis syndrome, but you know your symptoms better than I do, so feel free to read about it online and make sure you think I am right.” Of course, I try to counsel them a little bit on varied presentations, and that not all online resources are created equal, but I think endorsing the resources they are going to use anyway helps build a therapeutic alliance. The other main issues raised in this paper are the stigma these patients often feel, and the lack of quality advice for quitting marijuana. I think these are both very valid complaints. WIth opioids, alcohol, and most other drugs, we have fairly intensive management plans that include harm reduction, medical aides to quitting, and intensive follow-up. For patients smoking weed, we often just tell them to stop, as if it was that easy a choice.

Bottom line: I think there is a huge amount to be learned from these types of qualitative studies. This paper provides insight into the lived experiences of patients with cannabinoid hyperemesis syndrome, which are essential to understand if we are going to be effective in producing behavioral change as part of our management strategies.

Speaking of bullshit, is metformin the cure for long COVID?

Bramante CT, Buse JB, Liebovitz DM, et al. Outpatient treatment of COVID-19 and incidence of post-COVID-19 condition over 10 months (COVID-OUT): a multicentre, randomised, quadruple-blind, parallel-group, phase 3 trial. Lancet Infect Dis. 2023  Jun 8:S1473-3099(23)00299-2. doi: 10.1016/S1473-3099(23)00299-2. PMID: 37302406 

I have seen this paper circulated widely on social media, with the suggestion that we should now be prescribing metformin widely to prevent long COVID. However, this paper definitely doesn’t support that practice. If you just read the abstract of this paper, you will see that it is a high quality study (double blind RCT) in which the metformin group had a 5% absolute lower rate of long COVID diagnosis than the placebo group. However, if you read the paper itself, you will discover that this was actually a secondary outcome, and the original trial showed no benefit. You will discover that this secondary outcome was added after data collection started (because long COVID wasn’t even a known thing when the trial started). You will see publication bias, in that this trial also looked at ivermectin and fluvoxamine, but they are barely mentioned anywhere. At the end of the day, this is a secondary outcome (with an unclear definition, based entirely on self-report in a survey) in a trial with a negative primary outcome. The most likely explanation is that this is a chance finding. While it is hypothesis generating, it is definitely not practice changing, especially if you consider the significant cost of screening every single patient with a viral illness just so you could prescribe this.

Bottom line: Don’t just read the abstract, or change your practice based on Twitter discussions. The results of this trial are for researchers, not practicing clinicians. 

You can find a full write up here. 

2 ‘conflicting’ trials on steroids for community acquired pneumonia

CAPE COD: Dequin PF, Meziani F, Quenot JP, et al; CRICS-TriGGERSep Network. Hydrocortisone in Severe Community-Acquired Pneumonia. N Engl J Med. 2023 Mar 21. doi: 10.1056/NEJMoa2215145. Epub ahead of print. PMID: 36942789

ESCAPe: Meduri GU, Shih MC, Bridges L, et al; ESCAPe Study Group. Low-dose methylprednisolone treatment in critically ill patients with severe community-acquired pneumonia. Intensive Care Med. 2022 Aug;48(8):1009-1023. doi:  10.1007/s00134-022-06684-3. Epub 2022 May 13. PMID: 35723686

Two very similar RCTs looking at steroids for community acquired pneumonia were published this year (although, in an interesting example of publication bias, it seems like most people have only heard about the positive trial.) (Full appraisals of both papers can be found here.) CAPE COD is a multicenter placebo controlled trial that enrolled 800 ICU patients with severe community acquired pneumonia (excluding influenza), and randomized them to a hydrocortisone infusion or placebo. There was a significant decrease in their primary outcome of 28 day all cause mortality (6.2% vs 11.9%). ESCAPe is another multicenter placebo controlled RCT that enrolled 584 ICU patients with severe community acquired pneumonia. This time, there was not a significant difference in the primary outcome of mortality (16% vs 18%), although I think it is important to note that a 2% absolute reduction in mortality would still be important, is consistent with CAPE COD, and is probably a more realistic number than the 6% ARR seen in CAPE COD. There have long been hints that steroids might help in pneumonia, but the evidence has been so low quality that, despite a Cochrane review demonstrating a statistical improvement in mortality, this has not been part of most clinical practice. Although there still remains significant uncertainty, I think it is time to change practice in our sickest patients.The biggest risk is indication creep, as we know steroids cause many harms, and so using them in a less sick population is likely to cancel out any potential benefit. Therefore, my plan is to use steroids in patients with severe community acquired pneumonia going to the ICU who look like the CAPE-COD patients. 

Bottom line: We have 1 ‘positive’ and 1 ‘negative’ trial of steroids in community acquired pneumonia. Given the previous studies pointing at a potential benefit, and the fact that the point estimate from both of these studies is consistent with a mortality benefit, I think it is time to change practice (while acknowledging ongoing uncertainty and embracing the need for future research). 

Another huge TXA trial with negative results

Pacheco LD, Clifton RG, Saade GR, et al; Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal–Fetal Medicine Units Network. Tranexamic Acid to Prevent Obstetrical Hemorrhage after Cesarean Delivery. N Engl J Med. 2023 Apr 13;388(15):1365-1375. doi: 10.1056/NEJMoa2207419. PMID: 37043652

This is a huge study, but it isn’t all that applicable to emergency medicine, aside from helping us revise our position on TXA as a wonder drug, so I will just do a quick summary. This is an RCT comparing TXA to placebo in 11,000 women after cesarean section (half of which were scheduled). The primary outcome was a composite of death and blood transfusion, and there was no difference (3.6% vs 4.3%, p=0.19) There was only 1 death in the entire trial, so this outcome was entirely driven by the need for blood transfusion. This shouldn’t be surprising, because despite being widely recommended, the WOMAN trial looking at TXA for postpartum hemorrhage was a negative trial, and the TRAAP2 trial looking at TXA after c-section also looked very negative, despite having a statistically positive primary outcome. We will talk a lot more about TXA in the next edition, with the recently published PATCH-Trauma trial, but since the publication of CRASH-2, we have seen a long series of negative TXA trials, and I think we can now say pretty definitely that it is not a miracle drug, and its role is probably pretty limited in medicine.

When patients know their results before we do

Steitz BD, Turer RW, Lin CT, MacDonald S, Salmi L, Wright A, Lehmann CU, Langford K, McDonald SA, Reese TJ, Sternberg P, Chen Q, Rosenbloom ST, DesRoches CM. Perspectives of Patients About Immediate Access to Test Results Through an Online Patient Portal. JAMA Netw Open. 2023 Mar 1;6(3):e233572. doi: 10.1001/jamanetworkopen.2023.3572. PMID: 36939703

In many practice settings, patients can access their own test results online. This means that they might see abnormal test results before their physician, and this can even occur in the fast paced emergency department (as I can attest). This is a large survey asking how patients feel about this. I won’t do a full critical appraisal, but less than 20% of people responded to the survey, and there is evidence of response bias. Overall, patients generally like having access to their results (95% state they prefer immediate access). For the most part, access to results didn’t increase anxiety, but we have to keep in mind the minority (17%) who will receive abnormal results that cause anxiety. I thought it was interesting that among patients with ‘non-normal’ results, only 87% had been contacted by their health care practitioner. They don’t provide any details about these non-normal results, so it is possible that they are talking about a sodium of 134, which definitely doesn’t warrant a physician phone call. However, I think this is a reminder that patient access to their own results could act as an important safety measure for the mistakes we all make. Another interesting aspect of these results, which many of us might need to adopt into our practices, is that 92% of patients state they received pre-counselling about test results at the time they were ordered. Only 40% admit to performing internet searches for additional information about their test results, which seems low to me.

Bottom line: Patient access to their own chart is here to stay, and my guess is this is a net positive. However, we need to be mindful of this access, and account for it when educating and counseling patients, as well as in our documentation. 

The role of IV haldol: a debate

Kennedy JM, Kunzler NM, Hayes BD. Intravenous Haloperidol Has a Limited Role in the Modern Emergency Department. Ann Emerg Med. 2023 Jan;81(1):96-98. doi: 10.1016/j.annemergmed.2022.07.008. Epub 2022 Oct 19. PMID: 36270850

Lentz SA, Walsh K, Long B. Haloperidol May Be Safely Administered Intravenously in the Emergency Department. Ann Emerg Med. 2023 Jan;81(1):95-96. doi: 10.1016/j.annemergmed.2022.07.004. Epub 2022 Oct 19. PMID: 36270852

This is a pro/con debate series focused on the use of intravenous haloperidol, which seems timely, as the use of haloperidol has increased significantly in recent years, with the evidence of efficacy in cyclic vomiting. The argument against IV haloperidol centers around the potential increased risk of torsades de pointes as compared to oral administration of the same, and the potential dystonic reactions. They don’t present enough information about torsades to give a great sense of the absolute risk, but even if they did, I think they are making an incorrect comparison. If I am using IV haloperidol, it is because oral is inappropriate, so I need to see how the risk compares to other IV antiemetics. We use ondansetron daily, which also prolongs the QT. We need to know how these agents compare, and this paper doesn’t provide us with any data. As far as dystonic reactions go, they tell us that droperidol is a superior agent to haloperidol, which I totally agree with, if you can get it.

In a fun demonstration of EBM, the opposite side of the debate uses essentially the same data to present a better picture of IV haloperidol. There is an RCT of IV haloperidol in the ICU setting, which showed no QT prolongation and no cases of torsades (which is much better data than the case reports that earn these black bow warnings). There has also never been a reported case of torsades with less than 2 mg of IV haloperidol, and 80% of cases received doses higher than 10mg. They argue that haloperidol has plenty of evidence of benefit, and that benefit probably outweighs the risk, especially if you are cognizant of risk factors for torsades (other drugs, known cardiac disease, bradycardia, age >65, and electrolyte abnormalities), and you are liberal in ordering ECGs and or cardiac monitoring after use. 

Bottom line: I have used intravenous haloperidol hundreds of times over the last 5 years, with excellent effect (discharging patients with cyclic vomiting who used to be admitted), and with what appears to be a reasonably good safety profile. This debate series isn’t going to change my practice (although availability of droperidol probably would).

PESIT 2.0? PE incidence in acute exertional dyspnea

Prandoni P, Lensing AWA, Prins MH, et al; Pulmonary Embolism Dyspnea Italian Study (PEDIS) Investigators. Prevalence of pulmonary embolism among patients with recent onset of dyspnea on exertion. A cross-sectional study. J Thromb Haemost. 2023 Jan;21(1):68-75. doi: 10.1016/j.jtha.2022.09.007. Epub 2022 Dec 22. PMID: 36695397

This is the same lead author as the infamous PESIT study (which said every patient with syncope had a PE), and so that might be a red flag. It is a prospective study looking at the prevalence of PE in patients with severe acute onset exertional dyspnea. They include 417 patients, about half of whom had signs and symptoms of VTE. In the group of patients without any indication of PE, the PE rule in rate was 20%. This doesn’t surprise me at all: unexplained, severe exertional dyspnea is a classic PE presentation. I think the key is to recognize their definition of severe (unable to walk 100 yards on flat ground), and be careful not to over-extrapolate the results.

Bottom line: Although it is often discussed in the context of chest pain, the primary symptom of PE is really dyspnea, and this study provides a reasonable reassurance of current medical practice.

Full write up here

Cheesy joke of the moth

An apple a day actually does keep the doctor away… if you throw hard enough.