The investigators hypothesis is that for ICU patients with shock, the use of the vasoactive drugs phenylephrine and vasopressin will reduce tachydysrhythmias when compared to norepinephrine and epinephrine. To investigate this hypothesis, the investigators are conducting a randomized double blind controlled trial comparing phenylephrine and vasopressin vs. norepinephrine and epinephrine in ICU patients with shock that is not responsive to IV fluids. All patients admitted to the adult intensive care units at the University of Chicago will be screened for eligibility.

Shock, defined by inadequate tissue perfusion, is a common problem in critically ill patients. Most patients who have shock have hypotension and this is typically treated initially with intravenous fluid resuscitation in patients who are fluid responsive. If patients remain hypotensive, they are typically treated with vasoactive medications. Four of the commonly used FDA approved vasoactive medications are norepinephrine, phenylephrine, epinephrine, and vasopressin. Apart from a 2010 trial comparing norepinephrine to dopamine, there are no studies to date that have shown one of the four above-mentioned vasoactive medications to be superior to another. Accordingly, choice of vasoactive medication is based upon individual physician preference, without an outcomes-related evidence base.

Two of the four above mentioned vasoactive medications (norepinephrine and epinephrine) have chronotropic effects (i.e. the tendency to increase heart rate), while the other two (phenylephrine and vasopressin) have less of a propensity to chronotropy. The potential benefits of the chronotropic effects in patients with shock (increasing cardiac output) are offset by the potential detriments (predilection to tachydysrhythmias and myocardial ischemia).

Recent evidence suggests that tachydysrhythmias are associated with worse outcomes in ICU patients. One study demonstrated that administration of the beta blocking agent esmolol improved hemodynamic outcomes and survival in patients with septic shock. It is not clear if a vasoactive drug regimen that utilizes phenylephrine and vasopressin will be associated with lower heart rates compared to a regimen that utilizes norepinephrine and epinephrine.

The investigators hypothesis is that for ICU patients with shock, the use of the vasoactive drugs phenylephrine and vasopressin will reduce tachydysrhythmias when compared to norepinephrine and epinephrine. To investigate this hypothesis, we are conducting a randomized double blind controlled trial comparing phenylephrine and vasopressin vs. norepinephrine and epinephrine in ICU patients with shock that is not responsive to IV fluids. All patients admitted to the adult intensive care units at the University of Chicago will be screened for eligibility.

Patients will be randomized to receive either phenylephrine (0.3-3.0 mcg/kg/minute), with the addition of vasopressin (0.1-0.6 milliunits/kg/minute) if a second vasopressor is required, or norepinephrine (0.03 to 0.3 mcg/kg/minute), with the addition of epinephrine (0.03 to 0.3 mcg/kg/minute) if a second vasopressor is required. These drugs will be mixed and blinded by the research pharmacy. Only the research pharmacist will know the identity of the particular vasoactive drug. As per current standard practice, the medical team in charge of the patient will determine the target blood pressure.

In either group, if two vasoactive drugs are not adequate to raise the blood pressure to the target level, open-label norepinephrine will be added. If three vasoactive drugs are inadequate to raise the blood pressure to the target level, open-label epinephrine will be added.

There will be up to a twelve-hour period from initiation of standard, non-study vasoactive support during which the patient can be consented and enrolled. This will allow the research team to contact the patient and/or family in order to obtain informed consent. Once randomized, all patients will be initiated on study drug vasoactive support at 50 percent of the maximal infusion rate. The study drug will be titrated to maintain blood pressure and the initial non-study drug will be titrated off. The primary team will direct other aspects of patient care.

We plan to examine the following pre-specified sub-groups:

1. Patients who received corticosteroids during their ICU stay vs. patients who did not receive corticosteroids during their ICU stay
2. Patients with depressed left ventricular ejection fraction (< 40%) vs. patients with normal left ventricular ejection fraction
3. Patients with coronary artery disease vs. patients without known coronary artery disease
4. Patients with different etiologies of shock (i.e. septic, cardiogenic, hypovolemic)

Inclusion Criteria:

1. Age greater than or equal to 18 years old
2. Requirement for vasoactive drugs via a central venous catheter for the treatment of shock. Shock will be defined as mean arterial pressure less than 70 mmHg or systolic blood pressure less than 100 mmHg despite administration of at least 1000 mL of crystalloid or 500 mL of colloid, unless there is an elevation in the central venous pressure to > 12 mmHg or in the pulmonary artery occlusion pressure to > 14 mmHg coupled with signs of tissue hypoperfusion (e.g. altered mental state, mottled skin, urine output < 0.5 mL/kg body weight for one hour, or a serum lactate level of > 2 mmol per liter).

Exclusion Criteria:

1. Cardiopulmonary arrest
2. Pregnancy
3. Severe right heart failure