Evaluation of long-term infusion of dexmedetomidine in critically ill patients: A retrospective analysis.

BACKGROUND: Dexmedetomidine is an α2-receptor agonist used for sedation in the intensive care unit (ICU). It is currently FDA indicated for short-term use (i.e., less than 24 h).
OBJECTIVES: To compare the safety and efficacy of dexmedetomidine if given long- term (>24 h) to short-term infusion (up to 24 h) for mechanically ventilated critically ill patients.
MATERIALS AND METHODS: The medical records of 73 patients were evaluated. Primary outcomes were significant changes in blood pressure or heart rate. Secondary outcomes included hospital and intensive care unit (ICU) length of stay (LOS), ventilator time, rate of reintubation, and rate of death. STATISTICAL ANALYSIS: Pair wise comparisons were based on independent student t-test for continuous data and Chi-square test for categorical data. Statistical difference was defined as P value < 0.05.
RESULTS: Of the patients evaluated, 50 received dexmedetomidine for more than 24 h and 23 patients received this agent for 24 h. Patients were similar at baseline except for age. Patients who received dexmedetomidine for more than 24 h were similar to the short-infusion arm in terms of the rate of bradycardia (8.6% vs10%; P = 0.22), hypotension episodes (30.4% vs 28%; P= 0.2), requirement of treatment for those episodes (37% vs 42%; P= 0.43), hospital LOS (30 days vs 38 days; P = 0.45), ICU LOS (14 days vs 19 days; P = 0.44), ventilation days (8 days vs 14 days; P =0.58), rate of reintubation (4% vs 10%; P = 0.79) and mortality (P = 0.2).
CONCLUSION: Long-term dexmedetomidine infusion (> 24 h) had similar safety and clinical outcomes in patients receiving this agent for short-term. Due to the retrospective nature of our investigation, more well-designed studies are needed to confirm these findings.

INTRODUCTION

Procedures needed for invasive monitoring and support in critically ill patients can lead to anxiety, agitation, and pain. Therefore, the administration of sedative and analgesic medications is deemed necessary for the comfort and safety of these patients.[1] For the past decades, Gamma aminobutyric acid (GABA) receptor agonists (including propofol and benzodiazepines such as lorazepam and midazolam) have been the standard of care for sedation in the intensive care unit (ICU).[2] However, GABA-mimetic sedatives have significant limitations including delirium, respiratory depression, dependence and withdrawal.[3] Recent studies have shown that benzodiazepine carries the risk for development of delirium, which is an independent predictor of higher mortality, length of stay, and cost of care.[45] These issues raised a thought of considering other sedative agents as better choices.[6]

Dexmedetomidine, a central and peripheral α2-receptor agonist distinct from GABA receptor for benzodiazepines and propofol, has been approved by the US Food and Drug Administration only for use up to 24 h in mechanically ventilated patients.[7] Dexmedetomidine can be administered with an optional loading dose of 1 μg/kg over 10 min, followed by a continuous infusion between 0.2 μgto 0.7 μg/kg per hour.[78] It has shown promise as a sedative agent for ICU patients as it may overcome the limitations of the GABA-mimetic sedatives. Most common side effects associated with dexmedetomidine were bradycardia and hypotension especially during the administration of loading doses and bolus injections.[9-12] In the largest randomized controlled trial published to date, there were 103 reports of bradycardia, 91 of which resolved without treatment. Likewise, there were 137 reports of hypotension; however, only about half resolved without treatment.[13] While hypotension and bradycardia remain a concern, bradycardia has resolved without treatment in a majority of patients.

Additionally, dexmedetomidine use is not associated with significant respiratory depression; its sedative and anxiolytic actions are provided by reducing sympathetic outflow from the locus ceruleus, while its analgesic actions are mediated through the release of substance P from the dorsal horn of the spinal cord.[14]

The safety of prolonged use of dexmedetomidine can be evaluated based on the available literature, because many comparative studies included patients who were continued on this drug for longer than 24 h. These studies compared long-term infusion of dexmedetomidine versus other sedatives.[15]

Results of studies comparing traditional sedatives and dexameditomidine regarding efficacy and end point of delirium were promising. These studies revealed that long term dexmedetomidine had shorter duration on mechanical ventilation and lower incidence of delirium.[15]

To our knowledge, there are no studies that compared long-term infusion to short-term infusion of dexmedetomidine.

The goal of this retrospective analysis was to compare the safety and clinical impact of dexmedetomidine if given long- term (>24 h) compared to short-term infusion (up to 24 h) in mechanically ventilated critically ill patients. From a safety perspective, primary outcomes were significant changes in blood pressure or heart rate. Secondary outcomes included hospital and ICU length of stay (LOS), ventilator time, rate of reintubation, and rate of death. Patients in both groups received a loading dose of dexmedetomidine (1 mcg/kg over 10 min), followed by standard maintenance dosages as per product labeling guidelines (0.2 to 0.7 mcg/kg/h) to target Ramsay Sedation Score (RSS) of 2-4. Hemodynamic parameters and RSSs were collected until 24 h after cessation.

MATERIALS AND METHODS

The study was approved by the Institutional Board Review at our hospital. The investigation was a single-center retrospective cohort analysis of adult mechanically ventilated patients who received dexmedetomidine from June 2009 to December 2010 at our 20-bed intensive care unit (ICU). The electronic medical records of 73 patients were evaluated which showed that 50 patients received dexmedetomidine for > 24 h while 23 patients received it for 24 h. Exclusion criteria were pediatric patients (<18 years) with no mechanical ventilation. Acute Physiology and Chronic Health Evaluation II (APACHE II) scores were calculated for the 24 h preceding dexmedetomidine administration in both groups. This investigation incorporated best sedation practices including a light-to-moderate sedation level and daily arousal assessments in both study groups.

Data collected included demographics, baseline characteristics, APACHE II score and hemodynamic parameters on day 1 of dexmedetomidine treatment, RSS scores during treatment with dexmedetomidine as documented by the bedside nurse, adverse effects (hypotension and bradycardia) attributed to dexmedetomidine, and concurrent use of additional sedatives, fluids, vasopressors, and inotropic agents.

Primary outcomes were the incidence of significant changes in blood pressure or heart rate and requirement for treatment for these episodes. Bradycardia was defined as heart rate < 70 betas/min while hypotension was defined as systolic blood pressure (SBP) < 90 mmHg. For both adverse events, a change from baseline > 30% was also considered significant. Secondary outcomes included hospital and ICU length of stay (LOS), ventilator time, rate of reintubation at 48 h, and rate of death. Pair wise comparisons were based on independent student t-test for continuous data, Mann-Whitney U test for nonparametric data. Categorical data or proportions were analyzed using either the Chi-square test or the Fisher exact test when the expected frequency was < 5. Descriptive statistics were calculated for all variables of interest. All tests were 2-tailed and statistical difference was defined as a P value < 0.05.

To determine the incidence of cardiovascular adverse drug events, the baseline systolic blood pressure and heart rate (HR) were recorded for both groups, which were then compared with the SBP and HR after infusing the drug. Additionally, for the documented adverse events, interventions (e.g., intravenous fluids for hypotension) were recorded. None of the patients studied in both groups received concomitant analgesic (e.g., morphine, Fentanyl, and remifentanyl), sedative (e.g., lorazepam, diazepam, and midazolam), and/or antipsychotic (e.g., haloperidol, olanzepine, and ziprasidone).

RESULTS

Based on our electronic data base, a total of 73 patients who received dexmedetomidine were evaluated; 50 patients received dexmedetomidine for > 24 h and 23 patients received it for 24 h.

Demographical, clinical and admission diagnosis characteristics for the study groups are presented in Table 1. Both study groups were similar when comparing gender, severity of illness (APACHE II score), renal function, and admission hemodynamics and diagnosis within the 24 h prior to dexmedetomidine initiation. Additionally, both groups had similar target sedation range. However, there was a statistically significant difference in age. Median infusion time for dexmedetomidine in the long-term administration group was 70 (IQR: 48-118) h [Table 2].

Table 1

Base line demographics and admission hemodynamics and diagnosis

Table 2

Duration of dexmedetomidine in the long-term infusion arm

In the safety assessment, the incidence of hypotension and bradycardia was not significantly different between the two groups [Table 3]. Patients who received dexmedetomidine for more than 24 h were similar to the short-infusion arm in terms of the incidence of bradycardia (8.6% vs10%; P = 0.22), hypotension episodes (30.4% vs 28%; P= 0.2). The percentage of patients who experienced a significant adverse drug event requiring intervention was similar between both groups (37% vs 42%; P= 0.43).

Table 3

Episodes of bradycardia and hypotension

Clinical outcomes are presented in Table 4. There was a non significant difference in mortality between both groups (6% in short-term infusion vs 18% in long-term infusion, P= 0.2). Patients who received dexmedetomidine for > 24 h had an overall similar median hospital and ICU LOS when compared to short-term infusion. Likewise, patients on long-term dexmedetomidine infusion experienced similar days on mechanical ventilation (14 vs 8, P=0.58). Moreover, patients in both groups had similar rates of re-intubation.

Table 4

Clinical outcome variables

DISCUSSION

Dexmedetomidine, a relatively new drug, with a unique mechanism of action via acting on Alpha 2 receptor that is distinct from the GABA receptor for benzodiazepines and propofol. It also lacks suppression of the respiratory drive and does not depress the neurologic status, resulting in a state of cooperative sedation and preservation of neutrophil function.[910] However, stimulation of the central Alpha 2 receptors can lead to bradycardia and hypotension especially in volume-depleted patients. Of note, its sympatholytic action can blunt the stress response in critically ill patients.[13] Dexmedetomidine is currently labeled by the Food and Drug Administration for use for short-term sedation (no more than 24 h).[7]

While episodes of hypotension and bradycardia remain a concern in patients treated with dexamedotomidine, results of our retrospective analysis showed a non significant difference in these cardiovascular adverse events between patients receiving dexmedetomidine for less than 24 h and patients receiving it for longer time. This similarity in the incidence rates adds to the safety of infusing dexmedetomidine for more than 24 h. The effect of confounding variables on cardiovascular parameters was eliminated with the similarity between patients in both groups in terms of baseline characteristics. Moreover, for patients who developed these hemodynamic adverse events of hypotension and bradycardia, interventions as corrective measures were similar between both groups. Of note, there was no evidence of cardiovascular rebound (hypertension or tachycardia) 24 h after abrupt cessation of dexmedetomidine infusion in either of the groups. The results of our analysis are consistent with the study conducted by Pandharipande[3] where prolonged infusion of dexmedetomidine was compared with lorazepam as an active control group. The results showed that there was no significant difference in the incidence of cardiovascular adverse events between both groups. However, in the study conducted by Riker[13] there was a twofold higher incidence of bradycardia (42.2% vs 18.9%; P < 0.001) when prolonged infusion of dexmedetomidine was compared with midazolam infusion, but the incidence of the other cardiovascular adverse events were similar between both groups. In regards to the adverse events that required intervention, the incidence of bradycardia was higher in patients treated with dexmedetomidine than in patients treated with midazolam.

The results of our study are consistent with results of the trial conducted by Shehabi[16] which demonstrated that using dexmedetomidine> 24 h resulted in effective sedation and analgesic-sparing properties. Another study showed that long-term infusion > 24 h of dexmedetomidine was associated with similar cardiovascular and hemodynamic changes when compared with short-term infusion.[16] Overall, it does not appear that safety was compromised by prolonged use of the drug.[16]

In terms of side effects, our results were similar to those reported in the trial conducted by Ruokonen[17] which showed that there were no intergroup differences in the incidence of serious adverse events in patients treated with dexmedetomidine compared to patients treated with midazolam or propofol.

Based on our investigation, and from a clinical outcomes point of view, patients receiving long-term dexmedetomidine infusion had similar rates in terms of mortality rate at one month, hospital and ICU LOS, rate of re intubation when compared with the short-term infusion group. Moreover, patients in both groups had similar days on mechanical ventilation. This finding is supported by the fact that dexmedetomidine, via acting on Alpha 2 receptor that are distinct from the GABA receptor, is not associated with respiratory depression. Lirola T[18] conducted a pharmacokinetic study for prolonged infusion of high dose dexmedetomidine. The results showed that there was a statistically significant linear relationship (r∧2 = 0.95; < 0.001) between the areas under the dexmedetomidine plasma concentration-time curves (AUC) and cumulative doses of dexmedetomidine. The results support the fact that even with the accumulation of drug due to prolonged infusion, the safety of dexmedetomidine on the respiratory system is not jeopardized due to its novel mechanism of action.

Our findings are supported by Riker[13] study which showed that the time to extubation in patients treated with dexmedetomidine was even shorter than that in patients treated with midazolam (3.7 days vs 5.6 days; P = 0.01). Moreover, the ICU LOS for the dexmedetomidine-treated group seemed to be shorter than that for the midazolam-treated group. There was also no intergroup difference in 30-day mortality between both groups. The percentage of patients transferred alive from the intensive care unit was also similar in both groups, 81.5% for patients treated with dexmedetomidinevs 81.9% for patients treated with midazolam; P > 0.99.

Pandharipande[3] showed similar findings; no significant differences were noted between the dexmedetomidine-treated and lorazepam-treated groups for ventilator-free days (22 vs 18 days; P = 0.22), ICU LOS (7.5 vs 9 days; P = 0.92), and rate of mortality after 28 days (17% vs 27%; P = 0.18).

Ruokonen[17] showed that patients in dexmedetomidine group when compared to standard care groups were similar in terms of duration of mechanical ventilation, time required for weaning, number of ventilator-free days, or time to discharge from the hospital. Surprisingly, post hoc analysis showed that the duration of mechanical ventilation was even shorter in patients treated with dexmedetomidine (77.2 h vs 110.6 h; P = 0.025), and median duration of mechanical ventilation in patients with light-to-moderate sedation in the dexmedetomidine-treated group was shorter than that in the standard care group (70.2 h vs 93.7 h; P = 0.027).

Due to the design of this study, it is not without several limitations. The retrospective nature of analysis does not allow randomization of patients in either group so that the similarity of patients in groups is jeopardized. Of note, the design of these studies raises the question of selection bias as patients being in either of both groups are left at the discretion of the treating physician. Moreover, due to the fact that critically ill patients’ clinical status changes frequently, APACHE II score at baseline may not reflect the severity of illness at a later time as it is a static measure. Consequently, channeling bias can occur where physicians can elect severely ill patients into the prolonged infusion group due to their clinical status which can jeopardize randomization. Additionally, retrospective secondary data analysis carries the risk of information bias especially with the incomplete documentation in these kinds of research. The sample size may also be small to detect a difference between both groups. Consequently, the similarities in clinical outcomes evident from our investigation can be type II error. A well-designed, randomized, and controlled trial with prospective data collection and sample size calculation is needed to confirm the findings in our study and to examine the association of long-term infusion of dexmedetomidine with clinical outcomes.

CONCLUSION

The results of our analysis showed that when compared to short-term infusion, long-term dexmedetomidine is considered safe in terms of cardiovascular adverse events. Additionally, the percentage patients of who needed intervention to correct hemodynamic compromise were similar between both groups. Moreover, patients in the long-term infusion were similar to short-term infusion in terms of days on the mechanical ventilation, mortality at 30 days and hospital and ICU LOS. Due to the retrospective nature of our investigation, more well-designed studies are needed to confirm these findings.