Abstract
BACKGROUND: Ketamine has many recognized effects that may be beneficial in patients undergoing mechanical ventilation. While ketamine provides sedation and analgesia, it has additional sympathomimetic properties that may help support critically ill patients in shock. We hypothesized patients undergoing mechanical ventilation with continuous ketamine infusion as an adjunct to sedation agents would be associated with a lower vasopressor requirement. METHODS: We performed a retrospective cohort study on 200 mechanically-ventilated patients (205 hospital encounters) in two academic hospitals between 2015 to 2019. Patients on ketamine as an adjunct (K-G) to another sedative were utilized as the intervention group. Patients on both propofol and fentanyl (PF-G), a common sedation regimen, were used as the control group. The primary outcome was vasopressor requirements before and after initiation of ketamine or propofol and fentanyl. Secondary outcomes included all-cause mortality, 30-day mortality, ICU length-of-stay (LOS), hospital LOS, and ventilator-free days. RESULTS: The overall proportion of males was 63.4% (p-value =0.5016). The norepinephrine average dose (up to 48 hours after initiating sedatives) was lower in K-G (8.7 mcg/kg/min) when compared with PF-G (14.2 mcg/kg/min), p-value<0.0001. The ICU, 30-day or any time all-cause mortality was similar in both groups (22.0, 21.5 and 32.2%, p-value=0.8952, 0.9709, 0.8019, respectively). The average ICU and hospital stay overall were 8.8 (p-value=0.5174) and 16.6 (p-value=0.9280) days, respectively. The average ventilator-free days for K-G was 22.8 days compared to 23.2 days in PF-G (p-value=0.5567). CONCLUSIONS: In our study, ketamine as an adjunct sedation agent was associated with decreased vasopressor requirements in patients on mechanical ventilation when compared to the standard use of propofol and fentanyl. Further prospective research is necessary before ketamine can be broadly recommended as an adjunct to sedation in critically ill patients with shock.