Richard_Wang_HW1

Richard_Wang_HW1

by Richard Wang -
Number of replies: 3

EPI 245 Assignment 1 – Identify a gap between evidence and practice: the problem statement.

 

I’m interested in the treatment of sepsis in resource-limited settings, like in sub-Saharan Africa. An interesting conundrum has recently been encountered with regards to the appropriate use of intravenous fluids for patients with sepsis. In well-resourced settings, it is widely accepted that early and aggressive administration of intravenous fluid is life-saving in patients with sepsis. This is thought to be because, in sepsis, leaky blood vessels lead to hypovolemia, distributive shock, and, ultimately, death; bolus resuscitation when sepsis is recognized helps to mitigate or even avert shock. Many randomized clinical trials in have established the benefit of early fluid resuscitation in the United States, Europe, and Australia.

 

Until recently, not much was known about fluid resuscitation practices for sepsis in sub-Saharan Africa. A few observational studies suggested that patients with sepsis received far less in fluids that they would have had they been treated in the United States. Then, in 2017, a study in Zambia published the results of a randomized clinical trial comparing an early fluid resuscitation strategy with usual care in patients with sepsis, and found, surprisingly, that patients who were treated with an early fluid resuscitation strategy had increased mortality compared to usual care.

 

The conundrum is this: why would early fluid resuscitation have a different effect on patients with sepsis in Africa than it would in the United States?

 

In some ways, this is not really a gap between evidence and practice; it’s a gap between evidence and evidence. But that gap is really rich, and there’s all kinds of hypotheses. To wit:

1.    Is there a difference in the epidemiology of sepsis in the United States and in Africa? For example, does sepsis from tuberculous mycobacteremia—never seen in the US but makes up a sizable proportion of septic patients in sub-Saharan Africa—behave differently than conventional gram negative or gram positive sepsis?

2.    Aggressive and early fluid resuscitation sometimes causes pulmonary edema and respiratory failure. In the US, respiratory failure is treated with invasive mechanical ventilation. In sub-Saharan Africa, invasive mechanical ventilation is unavailable, and respiratory failure has a very high mortality rate. Does the lack of invasive mechanical ventilation explain the difference in outcomes from early fluid resuscitation?

3.    In the United States, septic patients are often admitted to the ICU with 1:2 or even 1:1 nursing ratios. In sub-Saharan Africa, access to ICU level care is frequently unavailable, and nursing ratios on the wards vary from 1:30 to 1:50. Does lack of access to ICU-level nursing ratios account for the difference in outcomes from early fluid resuscitation?

 

There are multiple avenues of inquiry—differences in access to emergency care, differences in insurance coverage (or lack thereof), differences in availability of diagnostic tests, differences in microbial epidemiology and resistance patterns, differences in availability of antibiotics, and differences in access to adjunctive therapies like renal replacement therapy, to name a few. The options are overwhelming, and one challenge is ascertaining which might be promising to investigate.


In reply to Richard Wang

Re: Richard_Wang_HW1

by Dan Kelly -

Thanks, Richard. 

It's important to note that ~90% of the study population was HIV-infected and that 2/3rds of them had a suspect diagnosis of tuberculosis. In Figure 3, the HIV-positive group who were in the sepsis protocol had an increased relative risk of in-hospital mortality as compared to the HIV-positive group who received usual care. The numbers were too small to conclude anything from the decreased relative risk among the HIV-negative group who received the sepsis protocol, but I would argue that this would have been a better study if they restricted to HIV-positive individuals or stratified randomization on HIV-status and extended enrollment. As is, I find the title of the article misleading because it proposes broader external validity than I think this article deserves. Nonetheless, it's a very interesting study. 

My take home is that we need to be careful in our assessment of 'sepsis' among HIV-infected persons who mostly were suspected tuberculosis. Perhaps the pathophysiology is different (i.e., no leaky vessels) and doesn't require aggressive fluid hydration. 


Dan

In reply to Richard Wang

Re: Richard_Wang_HW1

by Timothy -

Hi Richard,

Sepsis care is certainly and important topic - however as you point out, this topic seems to be more of a gap in current evidence than an implementation gap. While I'm certain this will lead to many great lines of inquiry...for this course I wonder if there are components of sepsis care that we know are effective and could be improved upon in your setting (i.e. process measures such as time to first antibiotics) which you might be able to focus on? Later assignments will drill down into methods to implement evidence-based strategies into practice and I think you'll find them more valuable with an implementation gap in mind .

Tim

In reply to Richard Wang

Re: Richard_Wang_HW1

by Teresa Kortz -

Hi Ricky,

You have chosen a subject near and dear to my heart: fluid resuscitation in sepsis! I am not as familiar with the adult ICU literature and guidelines; what are the current international guidelines for sepsis and resuscitation in adults? Is it EGDT, which includes early, large volume fluid resuscitation? Or, has someone or a group, like the WHO, suggested a different approach to resuscitation in resource-constrained regions (I know that they have for children)? Also, what do you make of the PRoMISe, ARISE and ProCESS results from HICs that do not show a benefit of EGDT?

Fluid resuscitation is a big, controversial, and very interesting question. Do you know which implementation gap you are interested in and want to address?

Teresa