Hemodynamic optimization in critically ill patients

Abstract

Background: Intravenous fluids are one of the most common treatments in the health care setting and are usually given in response to hemodynamic instability or shock. However, this rests on the assumption that the cause is lack of intravasal volume, and that any response will be lasting. Unfortunately, many patients are non-responders, and while there is monitoring that can help discern them, these are highly limited in their utility, and there is consequently no agreed upon standard. The standards of treatment have been based on goal-directed therapy, where advanced monitoring and protocolized treatment have been used. These standards have been re-evaluated as newer trials have not been able to verify their efficacy. As a consequence, research has focused on signs that more restrictive fluid usage might be beneficial, but no trials in a Western setting have seen any benefit or harm from a restrictive protocol either. As such, fluid resuscitation usage varies greatly, and it is still difficult to give clear evidence-based recommendations on its use.

Aims: The overall aim of the research project is to explore ways to optimize hemodynamics in critically ill patients, including whether our current practice pattern for fluid therapy in the intensive care unit is effective. Studies IV and V focuses on the hemodynamics of the pulmonary circulation and aims to investigate the effects of treatment with inhaled nitric oxide for vasodilation in critically ill patients suffering from acute respiratory distress syndrome due to COVID-19 pneumonia.

Results: In Study I, our analysis of 13 papers investigating if mortality in the intensive care unit was affected by protocolized intervention based on hemodynamic measurement found no difference in mortality. Mortality was 23.8% in the intervention group versus 23.5% in the control group, odds ratio (OR) of 1.00 (95% CI 0.89–1.12). Results remained the same when analyzing only trials with low risk of bias or a validated measurement. Study II investigated if the temperature of a fluid bolus changed its effect. We found that there was a small increase in cardiac index, 0.05 L/min/m2 (0.02 to 0.09) p<0.001 and decrease in mean arterial pressure, -3.42 mmHg (-3.97 to -2.86) p<0.001 with warm compared to cold fluid over the 15-minute infusion. During the same timeframe, 80% and 65% had a response to fluid in the warm and cold groups, respectively. When researching clinicians’ expectations of fluid boluses and their accuracy compared to results, Study III showed a 22% accuracy at the end of the bolus and 29% after one hour for the primary reason to give fluid. Almost half of the boluses were given due to hypotension, and expectations for mean arterial pressure were accurate in 25% at the end of the bolus. In Study IV our research into the effects of inhaled nitric oxide in intubated patients with COVID-19 pneumonia showed that the ratio of partial pressure of arterial oxygen to fraction of inspired oxygen ratio of the group receiving high-dose treatment had a 39.1 mmHg (95% credible interval (18.1 - 60.3) greater change from baseline in 48 hours compared to the control group. Absolute mortality at 90 days was 34.0% vs 32.3% for the treatment and control group respectively, but analysis adjusted for baseline characteristics showed a RR of 0.87 (95% CrI 0.52–1.43) in favor of treatment. Study V investigated echocardiographic aspects of the patients in Study IV and showed that there were no statistically significant changes in right ventricular function over time measured by tricuspid annular plane systolic excursion or right ventricular global longitudinal strain from treatment with inhaled nitric oxide. Pulmonary pressures were high and >90% of patients had pulmonary hypertension. Rates of systolic right or left ventricular dysfunction were higher when measured by global longitudinal strain than with tricuspid annular plane systolic excursion or ejection fraction, and these measures were feasible in 1/3 of patients.

Conclusion: The overall findings of this thesis regarding intravenous fluids include the fact that using protocolized treatment based on hemodynamic monitoring does not confer a mortality benefit. It also shows that this type of monitoring is rarely used to direct fluid bolus therapy in practice as well, and that goals set for fluid bolus therapy based on parameters clinicians choose are rarely met. Lastly, in the experimental setting, it seems that the response to fluids may be confounded by temperature, as the effect on cardiac index and mean arterial pressure differs between warm and cold boluses. Regarding treatment with inhaled nitric oxide, treating critically ill patients suffering from COVID-19 with high-dose inhaled nitric oxide for 48h improved PaO2/FiO2, but did not reduce mortality or duration of mechanical ventilation. However, there was a steeper reduction in viral load. Pulmonary hypertension was common in this cohort and did not significantly decrease with inhaled nitric oxide. Inhaled nitric oxide did not affect the rate of left or right ventricular dysfunction either. Both right and left ventricular global longitudinal strain could be valuable but difficult methods to reveal patients with systolic ventricular dysfunction treated in the intensive care unit.