The guiding physiological principles of resuscitation have for so long been based off the restoration of normal macroscopic hemodynamics in the hopes that the reversal of such circulatory perturbations will correct the underlying cellular injustices. And yet time after time such strategies have failed to show definitive benefits when empirically tested. The simple practice of administering a fluid bolus has gotten a scientific pass. Until a recent trial, published in JAMA by Andrews et al, called this very clinical certainty into question (1).
The authors randomized patients 18 years or older presenting to Zambia’s 1500-bed national hospital with sepsis (defined as a suspected infection and two or more SIRS criteria) and hypotension, to either standard care or the simplified severe sepsis protocol (SSSP). Patients randomized to the SSSP arm were given an initial 2-liter bolus followed by an assessment looking for signs of fluid intolerance, including a decrease in oxygen saturation, an increase in the respiratory rate, or an elevation in the jugular venous pressure (> 3 cm above the sternal angle). If any of these three findings were identified, fluid therapy was terminated. Otherwise, patients received an additional 2 liters over the next 4 hours. Following the initial fluid bolus, if the MAP remained less than 65 mm Hg, patients were started on peripheral dopamine, which was titrated to achieve a MAP of 65 mm Hg or greater. Finally, patients who were randomized to the SSSP arm with a hemoglobin under 7 g/dL, underwent a blood transfusion. In contrast patients in the usual care arm had their treatment determined by the treating physician.
Over a one year period the authors enrolled 212 of the 382 patients that presented to the Emergency Department with sepsis and hypotension. In general, patients randomized to the SSSP group received far more aggressive upfront care than their counterparts, receiving a median of 3.5 L (IQR, 2.7-4.0 L) of IV fluids compared with 2.0 L (IQR, 1.0-2.5 L) in the usual care group. 38.7% of patients in the SSSP arm received 4 L or greater of IV fluid within the first 6 hours of care, while in the usual care group, only 48.3% received an intravenous fluid bolus. More patients in the SSSP arm received dopamine infusions during the 6-hour period compared to their usual care counterparts (14.2% vs 1.9%). These patients were rewarded with a markedly faster rates of lactate clearance.
But these early surrogates of optimized resuscitation failed to translate into improvements in patients oriented benefits, and in fact resulted in an increase in in-hospital mortality. Patients in the SSSP group had an inpatient mortality of 48.1%, compared to only 33.0% in the usual care group. This 15.1 was statistically significant with a p-value of 0.03. 28-day mortality was also notably increased in the patients who underwent protocolized management (67.0% vs 45.3%).
There is certainly a question regarding the ability to extrapolate these results to the more generalized population of septic patients encountered in a high resource setting. The majority of the patients enrolled had HIV (89%), with a median CD4 well below 100. Approximately 50% of these patients were on antiretroviral therapy at the time of enrollment. 62.7% of the patients were treated for suspected tuberculosis, the mean hemoglobin was 7.8 g/dL and the median albumin level was 2.1 g/dL and 2.3 g/dL in the SSSP and control arms, respectively. Virtually none of these patients had access to ICU level care. How these difference translate are unclear. Additionally, the surrogate used as the marker of CVP was JVP which may have led to some discrepancies in the volume of fluid administered. Finally, the vasopressive agent of choice used in this study was dopamine, which has come under some criticism in recent years since the publication of the SOAP II trial demonstrated an increased rate of arrhythmias associated with the use of dopamine for the treatment of shock (2).
Despite the factors potential limiting this trial’s external validity, these results should certainly make us question our current practices. After all, this is not the first study to demonstrate the harmful effects of aggressive fluid administration. The FEAST trial published in the NEJM in 2010 by Maitland et al was stopped early for harm after enrolling 3141 African children in septic shock randomized to receive a bolus of normal saline, a bolus of albumin, or no bolus (3). The trial was stopped early after finding an increased mortality in children randomized to either the saline or albumin arms of the trial. At the same hospital in Zambia, Andrews et al performed a pilot study examining their SSSP protocol in 112 patients presenting with sepsis and hypotension (4). Similar to their recent publication, the authors found that patients randomized to the SSSP arm did significantly worse than their control group counterparts. All these results point in one direction, overzealous fluid resuscitation driven by markers of perfusion lead to worse outcomes. And while many would argue that these studies do not apply to managing septic patients in high income countries, they would also argue that these studies could not be performed in high income countries because of lack of equipoise in the administration of fluid. And so we are stuck in a cycle of perpetual denial. Unable to conduct the studies in the environment we want, and unwilling to believe the results when performed in the environment they are conducted.
Despite these shortcomings, this trial should remind us that our actions in the Emergency Department have consequences. We have discussed the hemodynamic follies of a fluid responsiveness guided resuscitation, concluding that the exact tool used to guide resuscitative efforts likely matters far less than how that tool is used. Whether it is JVP, measurements in a Zambian Emergency Department, or a non-invasive cardiac output monitor in a resource laden modern ICU, an empiric goal directive strategy that drives patients towards the flat portion of their Starling curve without consideration of their underlying pathology is likely to lead to over-resuscitation and downstream harms.
Sources Cited:
- Andrews B, Semler MW, Muchemwa L, Kelly P, Lakhi S, Heimburger DC, Mabula C, Bwalya M, Bernard GR. Effect of an Early Resuscitation Protocol on In-hospital Mortality Among Adults With Sepsis and HypotensionA Randomized Clinical Trial. Published online September 27, 2017.
- De backer D, Biston P, Devriendt J, et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med. 2010;362(9):779-89.
- Maitland K, Kiguli S, Opoka RO, et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. 2011;364(26):2483-95.
- Andrews B, Muchemwa L, Kelly P, Lakhi S, Heimburger DC, Bernard GR. Simplified severe sepsis protocol: a randomized controlled trial of modified early goal-directed therapy in Zambia. Crit Care Med. 2014;42(11):2315-24.diversive
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019
Agree with all you say. Single center, p=.03 make me think this could be the usual false positive result esp since it is in direction opposite hypothesized
Thanks Scott,
I guess that depends on for which hypothetical perspective your viewing things from. Yes the authors null hypothesis was their was no difference between usual care and a SSSP guided strategy, their alternative hypothesis being that the SSSP would lead to improved outcomes, but their initial study found the exact opposite. These findings in some ways validated the results of their initial pilot trial. Either way the data is still not robust enough to be used in a vacuum. A large multi-center RCT is needed, but certainly this and other data justifies doing such an study. Thanks again!
Obviously, patient demographic differences between this massive Zambian referral center and most centers forming the bulk of the EMCrit audience makes this paper not generalizable to our routine management of septic patients. It is, however, a triumph for EDs who see a very high prevalence of young HIV positive patients with AIDs and a high rate of tuberculosis. They were also malnourished and anemic. Comparing this trial to SOAP II in any way is difficult other than to say that I doubt a substantial number of patients in this trial had cardiogenic shock and would have benefited from Norepi. I… Read more »
Sorry for a typo:
*”Most of the patients in the sepsis protocol arm had to stop receiving fluids prior to receiving the protocolized 4L IVFs by 6 hours.”
Great comments as always Jason, I completely agree, we cannot apply these findings to the population of patients we typically treat in most American Emergency Departments. But I think it does speak to some underlying principles that often go undiscussed in our quixotic search for the optimal hemodynamic approach to resuscitation. Like you said this was a very specific population that had a lot of reasons to be specifically sensitive to fluid and over-resuscitation. But it doesn’t necessarily mean the more traditional patients we see in our Emergency Departments do not have some of those very same characteristics. In fact… Read more »
Thanks for the great reply, Rory. I hope that we can get more data on the utility of IVFs for septic patients across the world. As an EM researcher (like you), it would be difficult to perform an adequate study in the US…and I agree with what you’ve stated. This study, again, is a triumph for that specific patient population. that does exist in within some areas of the diverse American ED fabric of patients. Without question. more studies like this Zambia study need to be performed. As this study proved, small beginnings can have great impact of the future… Read more »
Very nice! Enjoyed the read and resources cited. Great reminder of how early and even pre-hospital fluid management can make a huge difference in long term patient outcomes. I know there is definite question in the external validity of the article. However, transporting patients from community to tertiary care facilities, sometimes it does not seem that far off…
“empiric goal directive strategy that drives patients towards the flat portion of their Starling curve without consideration of their underlying pathology”. Rory, that is one of the finest lines i have ever read (not quite as lovely as elena ferrante, “Neapolitan Novels”, but very fine). i just had the pleasure of reading dr marik’s more recent post, and commented too much there. just wanted to say “thank you”. tom ps: what perhaps is also so very disconcerting is the stark difference ( the horror, in a way) of resources in different parts of the same planet. also 89% HIV/AIDS, 63%… Read more »