Case of the week #7
Sepsis caused by UTI with iatrogenic pulmonary edema and volume overload
This female with recurrent UTIs presents to the ED with dysuria, a urinalysis consistent with a urinary tract infection as well as signs of sepsis. This is a very common presentation, with over 1 million visits to the emergency department for UTIs each year.
​
The surviving sepsis campaign recommends "rapidly administer[ing] a minimum of 30 mL/kg crystalloid solution intravenously in all patients with septic shock and those with elevated blood lactate levels."
​
Is this recommendation evidence based? A recent paper titled "Fluid resuscitation in sepsis: the great 30 mL per kg hoax" says no, arguing there is no credible evidence to support this level of fluid resusitcation and results in great harm.
​
Without knowing anything about the above review article, if you have POCUS in your tool kit, seems like a look at the heart, lungs and IVC will go a long way in deciding if a patient needs aggressive fluid replacement, or if they are volume overloaded.
Cardiac imaging
In this patient her heart function is good. Her systolic function/ejection fraction appears to be normal:
-
The left ventricular walls seem to be contracting symmetrically ​and sufficiently
-
The anterior leaflet of the mitral valve nearly hits the septum as it opens
​
There does not appear to be a pleural effusion or pericardial effusion. However this does not rule out diastolic heart failure of any kind.
Lung imaging
One method is to divide the lung into two zones anteriorly, lateraly and posteriorly, and scan through each zone for evidence of B-lines(learn more about B-lines). Three or more B-lines in one rib space in multiple lung fields is consistent with pulmonary edema with very high accuracy.
​
In this patient, 2 liters of normal saline were enough to develop evidence of pulmonary edema.
Confirming your suspicion for edema is a small right pleural effusion. For more on how to interpret pleural effusions, go here.
Inferior vena cava for volume status
Finally we reach the IVC to assess for volume status. This is a very tricky way to assess volume and it is riddled with exceptions. The evidence supporting IVC for volume assessment is shaky but in the extremes it can provide some useful data.
​
In this IVC, it appears to be filled with fluid without much variation with respiration. This is further evidence to support the above lung findings that maybe this patient is actually volume overloaded and should not receive anymore fluid, and in fact may need diuresis.
In this patient with sepsis related to a urinary tract infection not responding to levaquin, the initial uncontroversial management is broad antibiotics. While the surviving sepsis guidelines would suggest she also needs aggressive IV fluids, at least in her current state with a dilated IVC and diffuse B-lines, further IV fluids would not be indicated and should be held, and she may need diuresis in its place.
​
The authors of the review article argue that there is no good data supporting aggressive fluid resuscitation in sepsis (apart from clear dehydration) and significant evidence that it can be harmful, via various mechanisms including fluid induced vasodilation, direct cardiotoxicity, as well as increasing left atrial pressure with resulting pulmonary edema.
​
This is certainly contrary to the standard of care currently. With more trials suggesting aggressive hydration is not helpful, will the sepsis standard of care change? Time will tell. In the mean time, consider evaluating a sepsis patient with pocus before drowning them with fluids!
​