The Clinical Practice Guidelines for the Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep Disruption in Adult Patients in the ICU (PADIS Guidelines) suggest a protocol-based, stepwise approach for pain and sedation management that promotes treating pain over initiating a sedative agent (conditional recommendation, moderate quality of evidence).1 Opioid analgesics remain the most utilized agents to achieve this endpoint despite well-established adverse effects including but not limited to respiratory depression, constipation, delirium, and immunosuppression. Although there is substantial variation in the method of administration and agent selected, available data indicates that opioids are utilized in nearly 85% of mechanically ventilated patient-days.2

In 2016, guidelines published by the Centers for Disease Control (CDC) noted that long-term opioid use often begins when opioids are prescribed for acute pain.3 Since this time, the majority of every clinical practice guideline with recommendations related to pain management advocates for the use of opioid-sparing interventions in conjunction with non-pharmacologic approaches when available.4–9 In critically ill patient populations, the PADIS Guidelines suggest use of acetaminophen as an adjunct to opioids in critically ill patients and conditionally suggest use of ketamine for post-surgical pain and gabapentinoids for neuropathic pain to reduce opioid requirements and improve pain control.1 Nefopam is also recommended as an adjunct or alternatives to opioids but it is not commercially available in the United States or Canada.

The efficacy and safety of nonopioid analgesics for patients in an intensive care unit (ICU) setting was evaluated in a review of studies published up to October 1, 2019 by Wheeler et al. 10 The investigators found that the use of any adjunct agent compared to opioids alone was associated with decreased opioid use. However, the decrease in oral morphine equivalents over 24 hours was small (mean difference, 25.89 mg less; 95% CI, 19.97–31.81 mg) and of low certainty.10 Debate exists regarding what defines a substantial opioid-sparing effect, whether a reduction in opioids translates to improved patient outcomes, and what defines an acceptable benefit-to-risk ratio in critically ill patient populations at risk for adverse drug events.

In recent years, various groups and organizations including the Food and Drug Administration (FDA) have drafted guidance documents that address future research design for studies evaluating opioid-sparing medications and outcomes that would be considered clinically meaningful.11,12 The purpose of this scoping review is to provide an update to the review by Wheeler et al. and discuss if medications commonly recommended as opioid-sparing agents have evidence of an acceptable benefit-to-risk profile as elucidated in guidance documents to justify their use in heterogeneous populations of critically ill patients in ICU settings.

The PRISMA diagram for the scoping review is outlined in Fig. 1. A total of six randomized controlled trials were included for analysis.15–20 The patient populations evaluated in the trials included mixed medical/surgical ICU patients (n = 2), cardiothoracic ICU patients undergoing elective coronary artery bypass graft (CABG) surgery (n = 2), oral and maxillofacial surgical patients (n = 1), and neuroscience ICU patients (n = 1). One-half of the studies were completed in patients admitted to the ICU post-operatively. Two of the studies compared a non-opioid agent (ketamine in one study, gabapentin in the other study) to a control group, one study compared gabapentin to placebo, two studies compared dexmedetomidine to other agents (clonidine in one study, midazolam in the other study), and one study compared ketorolac to paracetamol. Full characteristics of the included trials are described in Table 1.

Of the trials published since 2019, the opioid-sparing effects of the intervention groups were commonly secondary endpoints with the majority demonstrating modest reductions in opioid use.15,16,19,20 Primary endpoints evaluated were commonly those that might be considered clinically important in critically ill patient populations (e.g., incidence and duration of delirium, duration of mechanical ventilation), but with no intent to relate them to opioid-sparing effects. Additionally, benefits with respect to the incidence of delirium and duration of mechanical ventilation are potentially confounded in the two studies that utilized benzodiazepines in the control group.15,20

The trials included in this scoping review demonstrated variable results with respect to opioid-sparing effects, with four trials15–17,20 revealing statistically significant reductions in opioid requirements and two trials18,19 reporting no differences in opioid requirements. Wang et al. reported a mean reduction in intravenous hydromorphone of 1.82 mg in mechanically ventilated patients undergoing maxillofacial surgery using a sedation regimen consisting of dexmedetomidine compared to midazolam.15 Another study by Shokri et al. reported a mean intravenous morphine reduction of 4.13 mg over 72 hours in mechanically ventilated patients undergoing CABG surgery that received dexmedetomidine compared to clonidine for sedation.16 Javaherforooshzadeh et al. evaluated morphine requirements with adjunctive paracetamol compared to ketorolac in postoperative CABG patients and reported a mean reduction 1.42 mg at 24 hours with paracetamol.17 Finally, Salarian et al. reported a mean difference in fentanyl infusion rate of 2.3 mcg/kg/hour in a mixed population of critically ill patients receiving mechanical ventilation with gabapentin added to standard care.20

The incidence and duration of delirium were evaluated in four out of six trials included in this scoping review. Two studies reported reductions in the incidence and duration of delirium with dexmedetomidine compared to midazolam or clonidine infusions for sedation.15,16 In one trial by Salarian et al. that compared the addition of gabapentin to a sedation regimen consisting of fentanyl and midazolam to a control group that received the same sedation regimen without gabapentin, the incidence of delirium was evaluated as a secondary endpoint.20 Patients that received gabapentin had lower rates of delirium compared to control (43% vs 17.3%, p = 0.03). Finally, a study using adjunctive ketamine reported no differences in the incidence of delirium compared to a protocolized, nurse-driven sedation regimen with daily spontaneous awakening trials and spontaneous breathing trials.19

Four out of the six randomized controlled trials compared duration of mechanical ventilation as either a primary or secondary outcome. Two of the trials reported reductions in duration of mechanical ventilation, both of which included dexmedetomidine in the intervention group.15,16 However, the absolute reductions in the duration of mechanical ventilation were small (6.3 minutes and 1.8 hours). The other two studies evaluating adjunctive ketamine and gabapentin reported no differences in the duration of mechanical ventilation.19,20

Of the trials included in this scoping review, none evaluated the impact of the intervention on opioid prescriptions at ICU transitions of care, hospital discharge, or long-term opioid utilization.

The results of this scoping review indicate that there have been a limited number of randomized controlled trials evaluating opioid-sparing agents in critically ill patient populations since a review by Wheeler et al. involving studies up to 2019.10 Additionally, there remains no consistent definition with respect to what absolute or percentage reduction in opioid requirements defines a clinically meaningful opioid-sparing agent.11,12 The review by Wheeler et al. concluded that opioid adjunctive medications in addition to opioids as compared to an opioids alone led to reductions in patient-reported pain scores at 24 hours (standard mean difference, –0.88; 95% CI, –1.29 to –0.47) and decreased opioid consumption (in oral morphine equivalents over 24 hour; mean difference, 25.89 mg less; 95% CI, 19.97–31.81 mg less), but both findings were of low certainty.

A recently published consensus document formulated by participants from universities, government agencies, industry, and patient advocacy organizations suggested that an opioid-sparing intervention must include the following elements without causing an unacceptable increase in pain: (1) prevents opioids from being started, (2) decreases opioid treatment duration, (3) decreases the cumulative opioid dose, or (4) reduces opioid-related adverse outcomes.11 In addition, the Food and Drug Administration (FDA) published a draft guidance document to address challenges of developing non-opioid medications to manage pain in light of the opioid crisis which questioned whether the term “opioid-sparing” was sufficient to justify labeling regarding reduced opioid requirements.12 While these guidance documents do not specifically address critically ill patient populations, the basic tenets generally remain applicable. Of the six included studies in this scoping review, none prevented an opioid from being started, none decreased opioid treatment duration, four trials showed a modest reduction in cumulative opioid dose, and four studies addressed delirium and mechanical ventilation as described below. None of the trials addressed all elements recommended in guidance documents.

Preventing opioid initiation in postoperative/trauma-related pain or in mechanically ventilated patients using the protocol-based, stepwise approach for pain and sedation management suggested in the PADIS guidelines might be desirable but is challenging with currently available non-opioid therapeutic options in patients with more severe pain states.1 However, minimizing duration in critically ill patients, particularly with respect to discontinuation of therapy at transitions of care out of the ICU or prior to discharge would likely represent a clinically important outcome. Several studies involving critically ill patient populations opioid-naïve prior to ICU admission and that received opioids as part of an analgosedation or analgesia-first regimen were ultimately discharged from the hospital with a prescription rates for an opioid varying between 20% to 47.1%.21–24 In addition, persistent opioid use among opioid-naïve ICU survivors measured at various time points between 3 and 12 months after ICU discharge ranged from 2.6% to 23%.22,24–26 Given concerns related to adverse effects of opioids in general and long-term opioid use disorder in ICU survivors, opioid-sparing agents that reduce duration of opioid use are warranted. A useful discussion of opioid-sparing options for managing analgesia, sedation, and delirium in critically ill adult patients is provided in clinical practice guidelines compiled by the Pan-American and Iberian Federation of Societies of Critical Medicine and Intensive Therapy.27 These guidelines recommend decreasing sedation levels with the use of multimodal pain management strategies while recognizing the need for more research on alternatives to opioids. Agents such as dexmedetomidine, remifentanil, and ketamine are discussed in the guidelines. Of note, the terminology regarding sedation has implications for patient management. For example, analgesia-first sedation in which an analgesic is used before a sedative is different from analgesia-based sedation in which an analgesic is used instead of a sedative to manage agitation or facilitate mechanical ventilation.28

Potential risks of opioid sparing strategies include inappropriate pain control which may lead to prolonged duration of mechanical ventilation, ventilator dyssynchrony, increased ICU length of stay, and post-traumatic stress disorder. Additionally, interpretation of any opioid-sparing effects and reductions in duration of mechanical ventilation or delirium is often complicated by confounding variables (e.g., use of benzodiazepines) in published studies. Regardless, there is appeal to the use of agents such as dexmedetomidine that may reduce the duration of mechanical ventilation and delirium while providing sedation to critically ill patients.29 A useful paper for clinicians wishing to minimize the problems associated with oversedation was compiled by the Zero Oversedation Project of the Sedation, Analgesia and Delirium Working Group of the Spanish Society of Intensive and Critical Care Medicine and Coronary Units.30 The paper contains practical information with tools to help ensure optimal patient management including minimizing the risk of excessive sedation.

The study by Salarian et al.,20 which suggested that the addition of gabapentin to a sedation regimen led to a reduced rate of delirium, is not corroborated by the results of a recent study that demonstrated that perioperative gabapentin use was associated with increased risk of delirium among older patients after major surgery.31 Furthermore, prescription use of gabapentin has increased more than five-fold between 2006 and 2018 with a coexisting increase in gabapentin-related overdose death exacerbated by concomitant opioid use.32,33 Given the widespread use of opioids in critically ill patients and concerns about safety, adoption of gabapentin use as an opioid-sparing agent in this patient population requires additional study.

Reduced duration of mechanical ventilation represents another potential clinically important outcome of opioid-sparing agents assuming pain assessments are routinely employed. As noted by García Sánchez and Alcántara Carmona, 70% of patients receiving mechanical ventilation in Spanish ICUs receive opioids but only 52% of ICUs use routine pain assessment scales.34 The studies in our scoping review demonstrating that dexmedetomidine but not ketamine reduced the duration of mechanical ventilation are consistent with a previously published meta-analysis, which reported a reduction in the duration of mechanical ventilation with dexmedetomidine (low certainty of evidence) and no reduction in duration of mechanical ventilation with ketamine (moderate certainty of evidence).10 The cause and effect of the potential opioid-sparing effects of dexmedetomidine on duration of mechanical ventilation requires additional study.

Future directions for research should include a focus on the effectiveness and safety of non-opioid medications in critically ill patients and not rely on extrapolations from studies conducted in other patient populations. Critically ill patients are at increased risk for a number of organ dysfunctions that might preclude the administration of medications commonly administered in other patient settings. For example, critically ill patients are at risk for acute kidney injury and gastrointestinal bleeding, conditions that would necessitate the avoidance of nonsteroidal anti-inflammatory drugs (NSAIDs).

Future investigations also need to evaluate outcomes beyond opioid sparing effects. A medication’s benefit in terms of an opioid-sparing effect must outweigh safety-related concerns (i.e., benefit exceeding risks), which becomes complex in critically ill patients at substantial risk for complications such as bleeding, acute kidney injury (AKI), and respiratory failure. Potential clinically important endpoints that could be evaluated with opioid-sparing regimens include duration of mechanical ventilation (or ventilator-free days), incidence and duration of delirium, reduction in post-operative ileus, and incidence of respiratory failure leading to mechanical ventilation. Ideally, any improvements in these clinical outcomes would not be outweighed by adverse outcomes such as alterations in hemodynamics, AKI, bleeding complications, among others. Clinically meaningful outcomes related specifically to opioid sparing that should be investigated might include avoidance of opioid use, decreased duration or cumulative doses of opioid administration, or reductions in opioid-related adverse effects. Finally, there is a need for more research to identify novel analgesic agents that can either serve to eliminate the use of opioids or serve as adjunctive therapies.

Utilization of adjunct agents to limit opioid exposure is a widespread practice amongst clinicians in the ICU setting, though it comes with difficulties and barriers. However, few studies have shown benefits (beyond decreased opioid consumption) of utilizing non-opioid medications in patients who are critically ill and requiring analgesics. Further studies are needed to determine the implications of opioid sparing amongst critically ill patients admitted to the ICU.

All authors contributed to conception and design, acquisition, analysis, or interpretation of data, drafted, and critically revised the manuscript.

No funding was used for this manuscript.

The authors have no conflict of interest to declare.

There is no shared data.