I have only clinical experience and it's distant. There may be literature on mortality impact in certain sub-populations that has evolved since I last looked, perhaps in low cardiac output cases in children and adults. See link for a good review.
There is a paucity of studies on this subject, mainly because "catecholamine-resistant septic shock" i.e. hipoperfusion associated with low cardiac output (CI
This is very nice clinical observation as it might need also to be proven by animal model to tease out the effect of Milrinone and other PDE3 inhibitors.
Here is the path-physiology scenario you might consider: In septic patient severe pulmonary vasoconstriction might occur during the septic shock. As you trying to perform fluid replacement therapy, hemodynamically, you might cause severe RV overload and RV dilation limiting its function. If you, in fact, have not performed aggressive fluid replacement therapy (otherwise protecting drop of your MAP) and concentrated rather on decrease of your immediate RV Afterload caused by severe pulmonary vasoconstriction by using PDE3 inhibitors i.e. e.g.Milrinone or other PDE3 you might be very surprised to find out that pulmonary vasodilation that will occur as a consequence will be extremely helpful, protecting hemodynamically unstable patient. Aggressive overloading with fluids might be soon reconsidered if reports, observations, case studies like yours comes out.
As a side note Milrinone is know to protect cAMP from degradation in cardiac muscle cells. Contractility (not measured by EF%) but rather single beat elastance or end systolic elastance will further your claim i.e. that Milrinone has short term (DOUBLE BENEFIT) decreasing RV afterload and improving RV or overall cardiac contractility.
We use milrinone frequently in pediatric septic shock. The typical scenario is a child who with tachycardia, cool extremities, delayed capillary refill time and low central venous saturation with normal or elevated blood pressure. I typically start it at a low dose (0.25 mcg/kg/min) without a loading bolus and watch for hypotension. If the child becomes hypotensive, I may give additional fluids or stop the milrinone. If the milrinone is tolerated, I increase it gradually to as high as 1.2 mcg/kg/min. If the patient has renal dysfunction I use it more carefully and at a lower dose. I frequently use it in combination with epinephrine.
See also Figure 1 in this article: http://www.ncbi.nlm.nih.gov/pubmed/19325359