Abstract
BACKGROUND: Sedation in intensive care is fundamental for optimizing clinical outcomes. For many years the world has been facing high rates of opioid use, and to combat the increasing opioid addiction plans at both national and international level have been implemented.1 The COVID-19 pandemic posed a major challenge for health systems and also increased the use of sedatives and opioid analgesia for prolonged periods of time, and at high doses, in a significant proportion of patients. In our institutions, the shortage of many drugs for intravenous (IV) analgosedation forces us to alternatives to replace out-of-stock drugs or to seek sedation goals, which are difficult to obtain with traditional drugs at high doses.2
METHODS: This was an analytical retrospective cohort study evaluating the follow-up of subjects with inclusion criteria from ICU admission to discharge (alive or dead). Five end points were measured: need for high-dose opioids (≥ 200 µg/h), comparison of inhaled versus IV sedation of opioid analgesic doses, midazolam dose, need for muscle relaxant, and risk of delirium.
RESULTS: A total of 283 subjects were included in the study, of whom 230 were administered IV sedation and 53 inhaled sedation. In the inhaled sedation group, the relative risks (RRs) were 0.5 (95% CI 0.4–0.8, P = .045) for need of high-dose fentanyl, 0.3 (95% CI 0.20–0.45, P < .001) for need of muscle relaxant, and 0.8 (95% CI 0.61–1.15, P = .25) for risk of delirium. The median difference of fentanyl dose between the inhaled sedation and IV sedation groups was 61 µg/h or 1,200 µg/d (2.2 ampules/d, P < .001), and that of midazolam dose was 5.7 mg/h.
CONCLUSIONS: Inhaled sedation was associated with lower doses of opioids, benzodiazepines, and muscle relaxants compared to IV sedation. This therapy should be considered as an alternative in critically ill patients requiring prolonged ventilatory support and where IV sedation is not possible, always under adequate supervision of ICU staff.
- analgesia
- sedation
- SARS-CoV-2
- COVID-19
- ARDS
- inhaled anesthetics
- intravenous sedation
- opioids
- benzodiazepines
- neuromuscular blockade
Introduction
Since March 11, 2020, the world has faced a new pandemic that has not only claimed a large number of lives but has placed an enormous strain on health care systems. SARS-CoV-2 infection has become one of the most important challenges to public health in the last 100 years, since the infectious process it generates in large swaths of the population due to its high contagious capacity brought health systems to a standstill in just a few months and caused a functional emergency.3-6
Due to the severity of the presentation of COVID-19, alternative respiratory care strategies were required such as the prolonged use of the prone position, sedatives, opioids, and muscle relaxation, lasting for up to several weeks.7-13
Sedation is one of the most frequently used measures in critical care patients undergoing mechanical ventilation. It aims to improve comfort, reduce anxiety and agitation, and help patient-ventilator synchrony.14 Benzodiazepines (especially midazolam) are among the drugs administered.15,16 In general, sedation seeks to achieve moderate hypnosis; during the pandemic, however, perhaps due to the severity of these patients, strategies seeking deep and prolonged sedation were used.
The inappropriate use of opioids in COVID17-20 entails a number of potential risks, especially adverse effects such as diarrhea, hyperalgesia, excitability, tolerance processes, delirium, and dependence. The prolonged use of opioids could be associated with immune system impairment.21 Additionally, opioid users represent a population at high risk of developing critical illnesses, especially the post-ICU group, in whom subsequent deprivation is associated with complications that can negatively affect prognosis, such as myocardial infarction, stroke, and infection.21
Neuromuscular blockade is a pharmacologic measure used in patients with severe ARDS.22 The aim is to improve endurance in the muscles of the rib cage and to abolish intractable effort and work of breathing.23,24 However, prolonged use of these drugs in critically ill patients has been associated with severe complications such as myopathy,25 prolonged hospital stay, increased mechanical release time, and muscle atrophy.26
In this scenario, inhaled anesthetics are drugs with a long history of use around the world. Although little is known about their mechanism of action and their pharmacologic properties, the use of molecular methods and pharmacologic profiles has shed some light on their characteristics and has expanded their use outside the anesthesia rooms into intensive care.27-29 These drugs have become a useful tool for providing sedation and analgesia to critically ill patients. The shortage of conventional drugs during the pandemic, and the complexity of pulmonary involvement in some patients, led us to use this alternative at our hospital and to share our experience with the scientific community.
QUICK LOOK
Current Knowledge
New sedation strategies have been published in recent years aiming to reduce delirium and decrease the use of benzodiazepines and opioids. Volatile anesthetics are a suitable and promising alternative to standard intravenous (IV) sedation. During the COVID-19 pandemic, these alternatives were frequently used due to the global shortage of commonly used drugs and the difficulty in achieving sedation goals in this patient population.
What This Paper Contributes to Our Knowledge?
These data suggest that in situations where the use of IV sedation is not possible alternatives such as inhaled sedation could be considered. In critically ill patients with ARDS, inhaled sedation was associated with less use of opioids, neuromuscular blockade, and benzodiazepines.
Methods
Design
A retrospective cohort study was conducted including subjects > 18 y of age admitted from July 1, 2020–December 1, 2021, to the ICU of the University Hospital of San José with a diagnosis of ARDS triggered by confirmed COVID-19 and requiring mechanical ventilation under sedation and analgesia.
Convenience sampling was performed. Subjects who met the inclusion criteria were followed from admission to discharge. The data collected were sociodemographic data, comorbidities, chronic treatments, symptoms of disease presentation, vital signs, which sedative and sedative dose, type of sedation (inhaled or intravenous [IV]), type of opioid analgesic used in infusion (only fentanyl was used in infusion), need for analgesia, dose of opioid analgesic, requirement of high or low dose (high dose ≥ 200 µg/h and low dose < 200 µg/h) based on the requirement of fentanyl infusion > 200 µg/h since admission and intubation to the ICU or during the stay for a period > 6 h, and need for neuromuscular blockade and type of agent for infusion (only used in non-depolarizing neuromuscular blockade infusion with cisatracurium). Finally, prognostic variables were included, such as days of mechanical ventilation; and presence or absence of delirium based on Confusion Assessment Method for the ICU (CAM-ICU) score was determined by examining the subject for (1) acute or fluctuating changes in mental status, (2) inattention, (3) altered level of consciousness, and (4) disorganized thinking. Subjects were considered delirious if they displayed acute or fluctuating changes in mental status and inattention, plus altered level of consciousness, and/or disorganized thinking on the CAM-ICU, and the result was a positive or negative dichotomous variable for delirium. Finally overall mortality was collected daily from the electronic medical record by the group of COVID-19 researchers from the CIMCA research group during follow-up.
Pressure controlled ventilation was used, and ventilator settings were set at target values of tidal volume 6–8 mL/kg body weight. Inspiratory peak pressures > 30 cm H2O were avoided. Tidal volume was calculated according to predicted body weight (predicted body weight [kg] = X + 0.91 × [height, cm] − 152.4 cm; with X male = 50 and X female = 45.5). Ventilator settings were adjusted according to blood gas analysis parameters (PaO2 [60–80 mm Hg], PaCO2 [35–45 mm Hg], SaO2 ≥ 88–93%, pH 7.35–7.45). Subjects in the IV sedation cohort were initially sedated with propofol or midazolam, in combination with an opioid analgesic drug (fentanyl or morphine) following our institutional protocol. The Richmond Agitation-Sedation Scale was used to monitor the depth of sedation throughout the ICU stay (target range −5 to −4 during the first days of acute illness and throughout the study period). For pain monitoring, physiological and behavioral indicators such as tachycardia, hypertension, diaphoresis, facial musculature, quietness, muscle tone, verbal responsiveness, and comfort were used, measured objectively using the Campbell scale, with our target score 0–3; the use of opioid analgesics was titrated to achieve this range, with measurements being performed every 4–6 h.
Sevoflurane or isoflurane was administered by a miniature vaporizing device (anesthesia conserving device [AnaConDa], Sedana Medical, Danderyd, Sweden), with anesthetic delivery provided using a modified heat and moisture exchanger (HME), which is incorporated into the breathing circuit between the Y-piece and the subject, instead of the usual HME for capturing exhaled gas for rebreathing. The inhaled gas was applied continuously using a syringe pump. Briefly, the syringe was filled with sevoflurane before connecting the device between the subject and the ventilator. It was filled with a 1.5 mL bolus, and additional 0.1 mL boluses were administered until gas was registered on the anesthetic gas monitor. Inhaled anesthetics were removed by connecting the ventilator exhaust air to an absorbent canister.
The decision to assign inhaled or IV sedation was made by the attending physician and also depended on the availability of the drug. Data were extracted and entered into a data collection instrument that fed an Excel database that served as an interface for subsequent use and analysis with the statistical package Stata 16 (StataCorp, College Station, Texas). The parameters recorded and calculated were described by mean (SD) or median (interquartile range); for categorical data, we calculated absolute and relative frequencies (count and %), and 2-proportion Z test was performed, and the P value was calculated. The Shapiro-Wilk test was used to determine the distribution of the data. Measurements before and after the change were compared using the paired Student t test for normally distributed data and the Wilcoxon rank-sum test for non-normally distributed data. The indicator of association chosen was relative risk (RR).
The need to obtain informed consent from subjects was waived because of the observational and retrospective nature of the study. All subjects’ personal data were anonymized for publication.
Results
From July 1, 2020–December 1, 2021, data were collected from 283 subjects who met the inclusion criteria and who had been assigned to IV (n = 230) or inhaled sedation management (n = 53). Baseline PaO2/FIO2 at ICU admission was 101. Demographic characteristics are summarized in Table 1. In the IV sedation group, the mean age was 60.9 (SD 14.5) y; Acute Physiology and Chronic Health Evaluation II (APACHE II) severity score was 15.3 (SD 3.1), and Sequential Organ Failure Assessment (SOFA) score was 6.1 (SD 2.3), whereas in the inhaled sedation group the mean age was 63.1 (SD 10.2); APACHE II severity scale was 11.3 (SD 2.8), and SOFA 5.8 (SD 2.8). The most frequent comorbidities were hypertension (45%), 17 subjects in inhaled sedation group versus 112 for IV group; type 2 diabetes (22%), 15 subjects in inhaled sedation group versus 48 for IV group; and obesity (23%), 11 subjects in inhaled group versus 55 for IV group.
Demographic Data for Inhaled Versus Intravenous Sedation Groups
Regarding the requirements of other drugs in sedation analgesia and neuromuscular blockade, in the inhaled sedation group the RR for the need for high-dose fentanyl was 0.5 (95% CI 0.4–0.8, P < .045), for the need for muscle relaxant 0.3 (95% CI 0.20–0.45, P < .001). There was no evidence of difference in doses in the group using relaxation and inhaled sedation, since a fixed infusion protocol was used for the first 24 h after intubation, and for the risk of delirium 0.8 (95% CI 0.61–1.15, P = .25) (Table 2).
Comparison of Use of Other Drugs Between Inhaled Versus Intravenous Sedation
The difference in median fentanyl doses between the inhaled and IV sedation groups was 61 µg/h or 1,200 µg/d (2.2 ampoules/d), P < .001, and the difference in midazolam dose was 5.7 mg/h. In the inhaled sedation group, the post-intubation loading dose was 0.2 mg/h.
Total time of mechanical ventilation was 10.5 d in the inhaled sedation group versus 12.7 d in the IV group, with a mean difference of −4.71 to −0.23. Total ICU length of stay was 13.1 d in the inhaled group versus 15.1 d in the IV group, with a mean difference of −3.71 to −0.28. Overall ICU mortality at 28 d was 55%.
Discussion
Our study, carried out with a large cohort with severe ARDS due to COVID-19 managed at 2 university hospitals in Colombia, corroborated the results of studies in other countries, namely that the use of inhaled sedation is a potential alternative in the management of critically ill subjects, able to reduce the use of opioids, muscle relaxants, and benzodiazepines.
COVID-19 pneumonia constituted a worldwide health crisis given that the massive and concomitant presentation of the disease saturated health care systems for long periods of time.27 During the pandemic, clinicians were confronted with shortages of drugs for daily use. In ICUs, the lack of resources for the most critically ill patients might have led to adverse and even fatal results,30 and so physicians were obliged to resort to unconventional sedation strategies. Initial reports demonstrate that the use of inhaled isoflurane was able to achieve the required deep sedation and reduced the need for IV sedation.31
Due to this concern, non-conventional alternatives emerged in intensive care such as inhaled sedation, which is frequently used in operating rooms. There are few reports on its routine use in critically ill patients; some beneficial effects in pain modulation have been reported and perhaps a reduction in analgesia requirement, particularly opioids. In Europe, this type of inhaled anesthetics has been used for approximately 12 years, which simplifies the role of the vaporizer and ensures safe use with conventional ventilators. Recently, the use of volatile anesthetics for ICU sedation has been authorized in several European countries and has been included in national guidelines.32,33
Data on the clinical benefit of inhaled anesthetics had already been reported prior to the COVID pandemic, especially with regard to the reduction of ventilation time in the ICU,34 which was close to 2 days in our study. Statistically significant differences were found both in length of stay and in duration of ventilation. There has also been evidence of a reduction in the use of neuromuscular blockade and a non-significant reduction in short- and long-term mortality, possibly due to the cardioprotective effects attributed to this therapy. Taking into account that many individual factors must be considered risks, the benefits of undergoing general anesthesia merit consideration on a case-by-case basis.34,35
In 2021, Meiser et al36 published their experience in 20 subjects diagnosed with ARDS due to COVID-19 comparing inhaled sedation with sevoflorane against propofol and the requirements of opioid doses (morphine) and the need for neuromuscular blockade. In our series with 283 subjects, the 53 with inhaled sedation received a lower fentanyl neuromuscular blockade; the contrast with Meiser et al study is even more striking: 0% versus 11%.
During the COVID-19 pandemic, the use of benzodiazepines increased, frequently at high doses with longer infusions than those previously used in daily practice. This practice has been related to a greater number of days of mechanical ventilation and episodes of delirium in ICU patients.37,38 Several hypotheses associated with the neurotropism of the virus due to episodes of hyperexcitation and agitation have been proposed,39 which explains the need for deep sedation in patients in the initial phases of ARDS in order to achieve protective ventilation goals. In our study we found a median dose difference between approaches of 5.7 mg/h, which compared with those who received inhaled sedation, who did not require benzodiazepines, suggests a possible decrease in these complications described; however, the reduction we found was not statistically significant in the inhaled sedation group, RR 0.8 (95% CI 0.61–1.15, P = .25).
Although in our study we did not evaluate other possible benefits of inhaled therapy, other effects that have been mentioned include physiological effects with a decrease in the systemic accumulation of the inhaled isoflurane and a lower risk for hepatic toxicity, which leads to a faster recovery from the anesthetic effects.40 Bronchodilation represents another potential benefit in patients with severe bronchospasm.41 Immunological phenomena have also been suggested, based on animal studies, regarding the effect on γ-aminobutyric acid type A receptors and via a modulation in the secretion of pro-inflammatory substances such as tumor necrosis factor-α, interleukin (IL)-6, and IL-1β, among others.
Our study has several limitations, above all its retrospective nature, the incompleteness of the records due to the limitations imposed by the pandemic, and the overload of the health system. In addition, titration of sedative doses inhalational anesthetics, were performed within a institutional protocol executed by the group of adjunct intensivists, most of them with additional training as anesthesiologists, however the care staff such as nursing, respiratory therapy and residents in training, lack such training and experience in handling this type of medication. The high mortality associated with viral pathology and related comorbidities should be considered in the interpretation of these results and in their extrapolation to other types of critically ill patients, especially those without COVID-19 infection.
Further studies are needed to demonstrate the usefulness of these drugs as an effective alternative. Sedation is commonly delivered using IV medications such as propofol, benzodiazepines, and analgesia with opioids. Data are from a local retrospective study, with which we do not intend to inform changes in the protocols and current clinical practices or replace IV sedation medications regularly used in critically ill patients. There are many considerations that must be taken into account before considering the use of this therapy, such as the training of ICU health personnel, safety studies of this therapy, cost-effectiveness, and define the doses and indications in patients without ARDS in whom the duration of mechanical ventilation is expected to be longer than 72 h.
This study shows the results of the use of this therapy at a specific moment of the COVID-19 pandemic, with a specific population, where we are forced to use alternatives to the medications that we traditionally use in our patients and that we believe should continue to study its benefits. Benefits and problems with the development of prospective studies that resolves the doubts that still revolve around inhaled sedation in ICU patients.
Conclusions
In conclusion, the sedation of patients with COVID-19 ARDS is a challenge for ICU staff. In situations where IV sedation is not available or demonstrates complications for its use, alternatives, including inhalation sedation, may be considered, especially for the management of sedation in critically ill patients requiring mechanical ventilation for a prolonged duration. Beneficial effects include reduced requirements for opioid analgesics and neuromuscular blockade.
Acknowledgments
We thank the residents and all the health care personnel of the critical medicine and intensive care program at Hospital San José for their dedication and commitment in the management of nearly 1,200 patients attended at our institution.
Footnotes
- Correspondence: Dr Daniel Molano Franco, Calle 18 N 10-57, 2 floor Intensive Care Unit, Hospital de San José, Bogotá, Colombia. E-mail: dalemofra{at}gmail.com
See the Related Editorial on Page 437
The authors have disclosed no conflicts of interest.
- Copyright © 2023 by Daedalus Enterprises
