Dopamine Versus Norepinephrine for the Treatment of Vasopressor Dependent Septic Shock

Research Question/Hypothesis: The initial selection of the vasopressor norepinephrine in comparison to dopamine will result in a decrease in mortality for patients in septic shock.

Study Design/Source Population:

This trial is a single-center, prospective, randomized, open-label comparison of dopamine versus norepinephrine as initial vasopressor for patients presenting with septic shock. The study takes place at Rush University Medical Center, a 600 bed facility. Patients are transferred to our medical intensive care unit from the emergency room (ER), general medical floors, and from outside hospitals. Patients were eligible if they were greater than 18 years of age, and presented with a diagnosis of SIRS plus a suspected or documented source of infection. Patients were not eligible if they were found to have hypovolemic and/or hemorrhagic etiologies of their vasodilatory shock or another etiology of their SIRS.

Patients in the medical intensive care unit presenting with septic shock were randomized to receive either dopamine or norepinephrine as the first-line vasopressors for septic shock. Randomization was based upon whether the patient presented on an odd or even day of the week. For example, if the patient had presented on the third then they would be randomized to dopamine and if they had presented to the ICU on the fourth then they would be randomized to the norepinephrine treatment arm. The study investigators accept that the randomization scheme is not truly randomized, however a patient presenting with sepsis is not dependent on the day of the week. For example, the timeline for a patient to present with sepsis is unpredictable and the therapy for septic shock is started immediately upon their diagnosis, therefore the selection of a vasoactive agent (norepinephrine or dopamine) is really determined by the date of the patient's presentation, rather than the investigator.

All patients were treated according to recommendations by the Surviving Sepsis Campaign (early-goal directed therapy including fluid resuscitation, early and appropriate antimicrobial therapy, strict glycemic control, and consideration of steroid replacement for patients with relative adrenal insufficiency). Patients presenting with hemodynamic instability are first treated with initial fluid resuscitation which encompasses either 500ml to 1000ml of crystalloid or 300ml to 500ml of colloid, depending on a clinician's preference. The administration and titration of vasopressors was directed to achieve a mean arterial pressure ≥ 60 mmHg or a systolic pressure ≥ 90 mmHg. If the predetermined maximum dose was reached for the initial vasopressor, then an addition of vasopressin at a continuous dose of 0.04 units/min was initiated. Patients who still required hemodynamic support were then started on an infusion of phenylephrine.

The primary endpoint was all-cause 28-day mortality. Secondary endpoints included length of stay in the intensive care unit, organ dysfunction/failure, and the occurrence of dysrhythmia's. The study was approved by the Institutional Review Board (IRB) for human experimentation. Data to be collected includes baseline characteristics, laboratory parameters, microbiology, APACHE II score, occurrence of dysrhythmia's, and survival.

Definition of Outcome:

Primary outcome- All cause 28 day mortality

Secondary outcome- Length of stay in the ICU (days), organ dysfunction/failure (MODS and SOFA scores), and the occurrence of dysrhythmia's between dopamine or norepinephrine (sinus tachycardia > 20% increase in heart rate from baseline or the presence of an abnormal atrial or ventricular rhythm based on EKG)

Definition of Exposure:

Therapy with either dopamine or norepinephrine than they will be followed for primary and secondary outcomes until vasopressor therapy in no longer required for hemodynamic support.

Statistical Analysis: The primary outcome variable will be survival. The exposure variable will be whether the patient received dopamine or norepinephrine for hemodynamic support in the setting of septic shock. The primary outcome of survival will be compared both via a Chi-square test and utilizing the time-to-event model of the Kaplan-Meier test. We anticipate similar mortality rates between the two vasopressor treatment groups. We will analyze the baseline demographics of the two treatment groups. Comparison of baseline categorical and continuous data will be completed using a Chi-square and t-test, respectively. If there appears to be an imbalance in the baseline characteristics then the confounding variables will be addressed by utilizing a Cox-proportional Hazards model in the final analysis.

The occurrence of the secondary outcomes of length of stay and organ dysfunction/failure will be analyzed using a t-test. The main secondary outcome to be evaluated was the occurrence of dysrhythmia's. The dysrhythmia's will be compared in both groups utilizing a Chi-square test. One particular confounding variable for this secondary endpoint is the patient's prior cardiac history, which should be balanced based upon randomization of vasopressor therapy.

Unfortunately, septic shock has an expected mortality of approximately 40-60%. We are looking for 20% reduction in mortality rate which would require an n=252 for our sample size to achieve a power of 80%. The results will aid clinicians who treat patient with septic shock in their selection of initial vasopressor agents. The external validity of our study is limited since this is a single-center evaluation and a high proportion of our patients have underlying malignancy and/or immunocompromised co-morbid conditions, and this could increase our 28 day mortality in comparison to other institutions; however this strengthens our internal validity. Two potential ways to overcome these limitations are to perform sub-group analysis and/or a Gray's survival analysis.7