Practices and Outcomes Associated With Sublingual Buprenorphine Use in a Critically Ill Population

Abstract

IMPORTANCE:

Buprenorphine for opioid use disorder (OUD) is commonly used in the outpatient setting with increasing use in hospitalized patients. However, there is limited literature describing its use in critically ill populations.

OBJECTIVES:

The primary objective was to report the practices of buprenorphine prescribing among ICU teams. We also assessed the effect of a novel initiation strategy on opioid requirements in the ICU and the incidence of precipitated withdrawal.

DESIGN, SETTING, PARTICIPANTS:

Single-center, retrospective, descriptive study of patients receiving buprenorphine in the ICU.

MAIN OUTCOMES AND MEASURES:

The main outcome was to describe the use of buprenorphine in ICU patients through indication, initiation strategy, dosing information, and time from ICU admission to the first dose. We also detailed the incidence of precipitated withdrawal overall and the difference in opioid requirements before and after a low-dose induction strategy (buprenorphine initiated while receiving full agonist opioids [5-d titration from 150 μg to 4 mg four times daily]).

RESULTS:

A total of 153 patients were included. Most patients (86.3%) received buprenorphine for treatment of OUD. Of the 75 patients taking buprenorphine before admission, 46 (61%) had it restarted within 24 hours of ICU admission. Among 95 patients requiring buprenorphine induction, 57 (60%) underwent standard induction and 38 (40%) underwent low-dose induction, with only one instance of precipitated withdrawal. Median morphine milligram equivalents (MMEs) of concomitant full agonist opioids in patients completing low-dose induction decreased from 1057.5 mg to 262.5 mg in the 24 hours before initiation compared with the 24 hours after target buprenorphine dose was reached (p < 0.005).

CONCLUSIONS AND RELEVANCE:

Use of sublingual buprenorphine was most often in patients with OUD. Timely continuation of home buprenorphine in the ICU was suboptimal. Both standard and low-dose induction strategies appear to be safe with a low risk of precipitating withdrawal. When implemented appropriately, low-dose buprenorphine induction may lead to significant reduction in full agonist opioids in critically ill patients.

KEY POINTS

Question: What are the current indications and clinical outcomes associated with buprenorphine administration in critically ill patients?

Findings: Most patients received buprenorphine for opioid use disorder (OUD). Patients using buprenorphine before admission often did not have it continued at ICU admission or there were significant delays to continuation. Both standard and low-dose induction strategies appear safe with a low risk of precipitating withdrawal. A 5-day low-dose buprenorphine induction strategy for patients on scheduled full agonist opioids was associated with a significant reduction in opioid requirements 24 hours before and after induction.

Meaning: Buprenorphine was used most often in patients with OUD and timely continuation in the ICU should be optimized. In ICU patients with OUD, a low-dose induction may lead to significant reductions in full agonist opioid requirements when implemented appropriately.

Buprenorphine, a partial mu-opioid receptor agonist, is a common treatment for opioid dependence. However, much of the literature describes its use in the outpatient setting with less evidence in hospitalized patients. Specifically, there is little information regarding buprenorphine use in critically ill patients, a population that may be exposed to high amounts of opioids. For this reason, minimizing opioid withdrawal in ICU patients presents a challenging scenario, especially in patients with opioid use disorder (OUD). Although data exist evaluating buprenorphine use in hospitalized patients as a whole, there is currently a lack of published literature describing the use of buprenorphine in the ICU.

In the ICU, there are several potential indications for buprenorphine. These include continuing buprenorphine for patients who are already receiving buprenorphine as a medication for OUD (MOUD) in the outpatient setting, for those identified as having OUD or iatrogenic withdrawal that may benefit from buprenorphine induction during their ICU admission, or for those where buprenorphine may be used as part of a multimodal pain regimen (1, 2).

Patients with OUD or those with iatrogenic withdrawal may benefit from buprenorphine initiation during their ICU admission. As mentioned, patients with previously untreated OUD may have higher full agonist opioid requirements to manage their critical illness and potential withdrawal making mechanical ventilation weaning increasingly difficult or identifying an efficacious multimodal pain management strategy challenging. Separately, clinical practice guidelines for patients in the ICU recommend the use of opioids as first-line therapy to control pain before initiating additional sedative agents, a strategy known as analgosedation or analgesia-first sedation (3, 4). During mechanical ventilation, opioids are often given as continuous infusions, especially if intermittent dosing is frequent or if neuromuscular blockade is required. This can result in high cumulative opioid doses especially when patients are mechanically ventilated for an extended period of time (5). Increased rates of iatrogenic withdrawal have been detected in this patient population when opioids were abruptly discontinued, possibly resulting in longer durations of mechanical ventilation (4, 6, 7). One approach to manage iatrogenic withdrawal is to use full opioid agonists (e.g., methadone). However, dosing can have high interpatient variability because there is no established minimum methadone dose to prevent withdrawal. On the other hand, buprenorphine dosed on an average of 8 mg/d provides more predictable withdrawal management (7).

Current addiction medicine guidelines recommend an opioid washout period before buprenorphine initiation (8). This is because buprenorphine can precipitate a withdrawal syndrome secondary to its high binding affinity and partial agonism at the mu-opioid receptor. This results in the displacement of full opioid agonists, and thus, withdrawal (9). In critically ill patients requiring high doses of full agonist opioids, such as those requiring mechanical ventilation with continuous or scheduled intermittent analgesia, a washout period may not be feasible. Recent case reports describe buprenorphine low-dose induction protocols (commonly referred to as microdose or microinduction) which have successfully titrated patients to therapeutic doses of buprenorphine while concurrently receiving full agonist opioids, without precipitating withdrawal (10–12). There is also a report of a successful transition to buprenorphine via low-dose induction without precipitated withdrawal in a mechanically ventilated patient (4).

The present study aims to describe the use of buprenorphine in critically ill patients at a large academic medical center, with an emphasis on two unique induction strategies.

MATERIALS AND METHODS

This was a single-center, retrospective, descriptive study of adult patients who received buprenorphine while admitted to any ICU. The study site is an 886-bed academic medical center with approximately 100 ICU beds and is an American College of Surgeon’s Level 1 Trauma Center. The study was reviewed by the institution’s research subjects review board and deemed to be exempt from review, as it did not fall under the board’s guidelines as human subject research.

Patients were identified using the hospital’s electronic medical record (EMR) and screened for inclusion if they received at least one dose of a buprenorphine-containing medication in an ICU between January 1, 2016, and August 31, 2021. If patients had more than one admission to the ICU during a single hospitalization, only the index ICU admission was included. However, if a patient has more than one hospital admission during the study period, each hospital admission was treated as a separate encounter and multiple could be included. Patients were excluded if they received buprenorphine via IV or transdermal routes exclusively or if they received sublingual or buccal buprenorphine with a maximum daily dose of less than 1 mg/d. These exclusion criteria were created with the intent to exclude patients using low-dose buprenorphine for the sole indication of chronic pain.

The primary objective was to characterize the use of buprenorphine in critically ill patients. Specifically, we aimed to describe indications for buprenorphine, two unique initiation strategies (an institution-specific standard induction and low-dose induction), the final dose reached during induction (in the subset of patients not using buprenorphine before hospitalization), and the time from hospital presentation to first buprenorphine dose in the ICU. Secondary outcomes included incidence of precipitated withdrawal, identification of patients with outpatient buprenorphine use, and percentage of these patients with buprenorphine administration within 24 hours of ICU admission (as these patients would receive a new induction based on institutional practice). Reasons for delays in continuation of outpatient buprenorphine were detailed if not administered within 24 hours.

Data collection included patient demographic information such as age, sex, past medical history, location from which the patient was admitted to the ICU, primary reason for admission, prevalence of mechanical ventilation or extracorporeal membrane oxygenation at first dose of buprenorphine, and duration of mechanical ventilation after first buprenorphine dose. Medical history of OUD, substance use disorder, and alcohol use disorder were collected and determined based on chart documentation. Buprenorphine treatment characteristics including indication for buprenorphine use in the ICU, time from ICU admission to first buprenorphine dose, whether the Addiction Medicine Team was consulted for assistance with buprenorphine induction, and incidence of precipitated withdrawal were collected. Precipitated withdrawal was identified via diagnosis by the primary team and chart documentation of any of the following: sweating, tachycardia, chills, or body aches without documented fever, vomiting, or diarrhea (new onset of four or more liquid stools within 24 hr without another documented cause such as Clostridiodes difficile infection or new antibiotics). Additional data collection included if the patient was previously on buprenorphine as an outpatient determined by the before-admission medications list in the EMR documentation and in this subset of patients whether outpatient buprenorphine was continued within 24 hours of ICU admission.

Patients who underwent a buprenorphine induction were further categorized as having received a standard induction or low-dose induction (Fig. 1). At our institution, a standard buprenorphine induction consists of an 8–12-hour opioid-free washout period followed by a 2-mg test dose. If the test dose is tolerated without signs of precipitated withdrawal within 1 hour, a therapeutic buprenorphine regimen is initiated immediately. Standard inductions were chosen when, per the primary team’s assessment, a patient could withstand the washout period without experiencing severe opioid withdrawal. Low-dose induction consists of a 5-day titration at the following doses, each given every 6 hours for four doses: 150 µg, 300 µg, 1 mg, 2 mg, and 4 mg; often concomitantly with full agonist opioids. The 150 and 300 µg doses were delivered buccally, using a 300-µg buprenorphine buccal film (150 µg dose given as half of a 300-µg film). The 1-mg, 2-mg, and 4-mg doses were delivered via sublingual film, using a combination buprenorphine/naloxone sublingual film. Low-dose inductions were chosen for patients who would likely not be able to sustain a washout period, for example, patients requiring frequent intermittent opioids or receiving a continuous infusion. For all patients that underwent buprenorphine induction, the median daily dose before ICU discharge, rate of buprenorphine continuance at ICU discharge, and rate of buprenorphine continuation at hospital discharge with intent for indefinite treatment (determined via EMR progress note with documentation of appointment with buprenorphine prescriber) were collected.

Figure 1.

Buprenorphine induction strategies. SL = sublingual.

In the subset of patients that had low-dose buprenorphine induction, the total duration of buprenorphine induction, if full agonist opioids were concomitantly administered with buprenorphine (including type and route of administration), total days full agonist opioids were continued after the first dose of buprenorphine, total full agonist MME 24 hours before the first buprenorphine dose and 24 hours after reaching the target buprenorphine dose were collected. To minimize confounding of MME differences after low-dose induction, only patients who successfully completed low-dose induction were reported. Lastly, the buprenorphine dose at ICU discharge, total number of days receiving buprenorphine while hospitalized, final buprenorphine dose and frequency, patient disposition, and ICU and hospital length of stay were collected for all patients.

Two investigators, including one critical care pharmacist resident and one pharmacy student, collected data using a data dictionary and standardized electronic data collection tool to reduce variability in data collection. All data abstractors were trained in data collection procedures by one investigator using example patients to demonstrate the retrospective EMR review process and database input. The same investigator oversaw this process and served as the gold standard adjudicator in any instances of ambiguity.

Data are reported using descriptive statistics and measures of central tendencies to describe the population and outcome variables. Independent samples were analyzed using chi-square or Fisher exact test for dichotomous data, as appropriate. Wilcoxon rank-sum was used for continuous data, and a Wilcoxon signed-rank test was used for dependent samples of continuous data. A prespecified subgroup analysis was planned between standard and low-dose induction groups. There was no sample size calculation completed. Data analysis was conducted using Microsoft Excel (Office 2019) and R 3.6.2 (R Core Team [2019]. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/).

RESULTS

A total of 168 patients received at least one dose of buprenorphine while admitted to the ICU during the study period and 153 patients were included in the final analysis (Fig. 2). Most patients received buprenorphine for OUD (86.3%). Of note, three of these patients received buprenorphine in the ICU who did not have a previously diagnosed OUD before admission. Eighty patients (52.3%) used buprenorphine before ICU admission, 75 of which received buprenorphine via prescription with five patients reporting illicit use. Induction strategies were employed in 95 patients (62.1%). Standard induction was used for 57 of 95 patients (60%), whereas 38 of 95 patients (40%) received low-dose induction. Patient demographics are shown in Table 1.

Figure 2.

Consolidated Standards of Reporting Trials diagram.

TABLE 1.

Patient Demographics (n = 153)

Variables	Value
Age, median, yr (IQR, range)	39 (32–52, 19–78)
Sex, n (%)
Male	83 (54.2)
Female	70 (45.8)
Medical history, n (%)
Opioid use disorder	129 (84.3)
Substance use disorder	90 (58.8)
Alcohol use disorder	47 (30.7)
Location from which the patient was admitted to the ICU, n (%)
Emergency department	90 (58.8)
Outside hospital	26 (17.0)
Hospital floor	19 (12.4)
Operating room	18 (11.8)
Primary reason for ICU admission, n (%)
Medical	118 (77.1)
Overdose	16/118 (13.6)
Infection-related to IV drug use	17/118 (14.4)
Trauma	27 (17.6)
Emergent surgical (nontrauma)	6 (3.9)
Burn	2 (1.3)
Mechanical ventilation at first buprenorphine dose, n (%)	54 (35.3)
Days of mechanical ventilation after first buprenorphine dose, median (IQR)	4.5 (2–11.75)
Extracorporeal membrane oxygenation at first buprenorphine dose, n (%)	4 (2.6)
Patient disposition, n (%)
Home	97 (63.4)
Acute/inpatient rehabilitation	21 (13.7)
Long-term care, skilled nursing facility, long-term rehabilitation	12 (7.8)
Death	11 (7.2)
Other	12 (7.8)
ICU readmission, n (%)	14 (9.2)
ICU length of stay, median, d (IQR)	6 (3–15)
Hospital length of stay, median, d (IQR)	15 (7–30)

IQR = interquartile range.

The median time from ICU admission to the first buprenorphine dose was 2 days (interquartile range [IQR] 1–5). Among the 75 patients with previous prescription use of buprenorphine, 46 patients (61.3%) received their routinely prescribed dose within the first 24 hours of their ICU admission. Most delays in initiation were due to the ICU treatment team holding buprenorphine; 24 of 29 (82.8%). Buprenorphine was continued at ICU discharge in 131 patients (85.6%). Among patients who were alive at hospital discharge, 111 of 142 patients (78.2%) left the hospital with a buprenorphine prescription, and 101 of 142 patients (71.1%) were prescribed buprenorphine with the intent for indefinite treatment of OUD. Table 2 describes buprenorphine treatment characteristics among the entire population.

TABLE 2.

Buprenorphine Treatment Characteristics (n = 153)

Variables	Value
Indication for buprenorphine, n (%)
Opioid use disorder	132 (86.3)
Iatrogenic opioid dependence	10 (6.5)
Acute pain	10 (6.5)
Central storming due to anoxic brain injury	1 (0.7)
Time from ICU admission to first buprenorphine dose, median, d (IQR)	2 (1–5)
Receiving buprenorphine via prescription as an outpatient, n (%)a	75 (49)
Outpatient buprenorphine continued within 24 hr of ICU admission, n (%)	46/75 (61.3)
Outpatient buprenorphine NOT continued within 24 hr of ICU admission, n (%)	29/75 (38.7)
Reasons for delay of initiation of outpatient buprenorphine past 24 hr from admission, n (%)
Held by the ICU treatment team	24/29 (82.8)
Surgery	2/29 (6.9)
Held by the floor team before ICU	1/29 (3.4)
Refused by patient	1/29 (3.4)
Team unaware of outpatient prescription	1/29 (3.4)
Incidence of precipitated withdrawal with buprenorphine administration, n (%)	1 (0.7)
Addiction Medicine Team consult, n (%)	126 (82.4)
Final buprenorphine dose reached during induction, median mg (IQR)b	4 (2–8)
Buprenorphine continued as hospital discharge for patients previously receiving buprenorphine via prescription before admission, n (%)c	63/71 (88.7)
Buprenorphine continued at hospital discharge for patients receiving buprenorphine for a nonopioid use disorder indication in the ICU, n (%)c	10/18 (55.6)

IQR = interquartile range.

^aFive patients reported buprenorphine use without a valid prescription.

^bIn the subset of patients not taking buprenorphine before admission (n = 73).

^cIn those surviving hospital discharge.

When comparing patients receiving standard inductions versus low-dose inductions, reasons for admission and indications for buprenorphine between the groups were similar, however, a higher percentage of patients receiving low-dose induction were mechanically ventilated at the start of induction (75.9% vs. 38.6%, p < 0.005). During the induction period, continuous infusion sedatives were used less often for standard compared with low-dose induction (52.6% vs. 79.3%, p = 0.02). This held true for midazolam, propofol, and dexmedetomidine. For intermittent sedation, more patients in the low-dose induction group received phenobarbital. There was one patient who received buprenorphine while under neuromuscular blockade within the low-dose induction group. Details of comparisons between patients receiving standard versus low-dose inductions are in Table 3.

TABLE 3.

Standard Induction Versus Low-Dose Induction

Variables	Standard Induction (n = 57)	Low-Dose Induction (n = 38)	p
Mechanical ventilation, n (%)	22 (38.6)	50 (75.9)	< 0.005
Extracorporeal membrane oxygenation, n (%)	2 (3.5)	2 (5.3)	NS
Continuous infusion sedatives, n (%)	30 (52.6)	30 (79.3)	0.02
Midazolam	5 (8.8)	12 (31.6)	0.006
Propofol	17 (29.8)	20 (52.6)	0.03
Dexmedetomidine	15 (26.3)	22 (57.9)	0.003
Ketamine	4 (7)	6 (15.8)	0.19
Intermittent sedatives, n (%)	32 (56.1)	24 (63.2)	0.53
Benzodiazepines	12 (21.1)	8 (21.1)	NS
Phenobarbital	4 (7)	12 (31.6)	0.004
Haloperidol	5 (8.8)	6 (15.8)	0.34
Atypical antipsychotic	9 (15.8)	8 (21.1)	0.59
Clonidine	11 (19.3)	9 (23.7)	0.62
Addiction Medicine Team consult, n (%)	56 (98.2)	34 (89.7)	0.11
Buprenorphine started within 24 hr, n (%)a	8/22 (36.4)	0/13 (0.0)	0.02
Daily dose of buprenorphine reached before ICU discharge, median mg (IQR)	16 (8–16)	16 (8–16)	NS
ICU length of stay, median, d (IQR)	8 (4–17)	13 (8–22)	0.04
Buprenorphine continued at ICU discharge, n (%)	45 (79)	34 (89.7)b	0.37
Buprenorphine continued at hospital discharge with intent for indefinite treatment, n (%)b	32/49 (65.3)	26/37 (70.3)c	0.65
Buprenorphine continued at hospital discharge for patients previously receiving buprenorphine via prescription before admission, n (%)b	15/18 (83.3)	10/12 (83.3)	1.00

IQR = interquartile range, NS = not significant.

^aDenominator represents patients who were receiving buprenorphine before admission.

^bIn those patients that survived hospital discharge.

^cThose patients who were started on a low-dose induction but did not complete it due to being switched to standard induction are included in this number.

There was a single withdrawal episode that occurred during a standard induction (the patient experienced acute onset diarrhea which resolved following a full agonist opioid dose). The median ICU length of stay was 8 days (IQR 4–17) for the subset that underwent standard induction and 13 (IQR 8–22) for the low-dose induction group (p = 0.036).

Although low-dose inductions were initiated in 38 patients, there was successful completion in 29 (76.3%). The remaining nine patients had their induction interrupted for the following reasons: five patients switched to standard induction, three patients refused further buprenorphine therapy, and one patient’s goals of care moved to comfort measures only. There was a 795 milligram (75.2%) decrease (p < 0.005) in daily MME when comparing full agonist opioid requirements in the 24 hours before low-dose induction to the 24 hours following completion of buprenorphine induction. Details of completed low-dose inductions are in Table 4.

TABLE 4.

Description of Successfully Completed Buprenorphine Low-Dose Induction Patients (n = 29)

Variables	Value
Duration of buprenorphine low-dose initiation, median, d (IQR)	5 (5–5)
Scheduled full agonist opioid analgesic administered during buprenorphine induction, n (%)	24 (82.7)
Opioid analgesic administered during buprenorphine induction, n (%)
Continuous infusiona	24 (82.8)
Fentanyl	15 (51.7)
Hydromorphone	9 (31.0)
Intermittenta	25 (86.2)
IV	15 (51.7)
Oral	10 (34.5)
Duration of opioid continuous infusion after first buprenorphine dose, median, d (IQR)	4 (1.5–5)
Days receiving opioids after first buprenorphine dose, median (IQR)	8 (5–18)
Total MME 24 hr before first buprenorphine dose, median MME (IQR)	1057.5 (380–1470)b
Total MME 24 hr after reaching target buprenorphine dose, median MME (IQR)	262.5 (30–879)b
Buprenorphine dose at ICU discharge, median mg (IQR)	4 (2–8)
Buprenorphine total daily dose at ICU discharge, median mg (IQR)	16 (8–16)
Buprenorphine frequency at ICU discharge, n (%)
Two times daily	3 (10.3)
Three times daily	7 (24.1)
Four times daily	19 (65.6)
Days receiving buprenorphine while hospitalized, median, d (IQR)	19 (10–27)

IQR = interquartile range, MME = morphine milligram equivalents.

^aPatients may have received continuous infusion and intermittent opioids concurrently.

^bp < 0.005 for the comparison of MME 24 hr before the first buprenorphine dose compared with the 24 hr after reaching the target buprenorphine dose.

DISCUSSION

During our 5-year study period, we identified 153 critically ill patients who received buprenorphine, with most patients receiving buprenorphine for the indication of OUD.

Many patients were started on their home buprenorphine regimen, albeit with delays in initiation, and more than half of patients underwent standard or low-dose buprenorphine induction while in the ICU. Most low-dose inductions were performed in the setting of concomitant full agonist opioids and were associated with a significant reduction in MME requirements upon completion.

More than half of our patients with previous buprenorphine use continued treatment within 24 hours of ICU admission. In the patients that experienced a delay in therapy, this was most commonly because of the ICU treatment team’s choice to hold buprenorphine in favor of full agonist opioids. Over the past two decades, there has been much discussion on how to approach patients maintained on buprenorphine therapy when admitted to the hospital (1). Previously, it was thought that buprenorphine’s mechanism of action would prevent optimal analgesia with traditional agents. However, interruption in buprenorphine may complicate a patient’s course, leading to withdrawal while simultaneously removing a source of analgesia. In surgical patients, continuation of home buprenorphine has been associated with reduced MME requirements (1). Similarly, Culshaw et al (13) describe a significantly lower MME requirement among trauma patients who continued on home buprenorphine therapy compared with discontinuation. Although not in critically ill patients, this report suggests that continuing outpatient buprenorphine treatment may allow for decreased full agonist opioid use.

Eighty-six patients successfully completed some form of buprenorphine induction (57 standard inductions, 29 low-dose inductions). Standard inductions require patients to tolerate a washout period of opioid agonists, effectively leading to mild withdrawal, and may not be tolerated for patients requiring analgesia with full opioid agonists. Although withdrawal is required, standard inductions have proven useful in initiating buprenorphine in patients with OUD (14, 15). Conversely, low-dose inductions provide the benefit of avoiding withdrawal and cessation of analgesia by continuing full agonist opioid therapy during the initial stages of induction. During this time, buprenorphine, theoretically, will gradually occupy increasing amounts of mu-opioid receptors. Previous case reports detail the success of low-dose induction in patients with a variety of opioid requirements (10–12). The present study describes 29 successful low-dose inductions, adding to the evidence of the feasibility and effectiveness of this approach. To our knowledge, there is only a single published case of low-dose induction in a mechanically ventilated patient (10). Despite requiring mechanical ventilation, 22 patients in our study completed induction with buccal followed by sublingual buprenorphine, showing that the presence of invasive mechanical ventilation is not a limitation to the use of buprenorphine.

Low-dose inductions were more often used in patients with longer ICU stays and greater analgesia and sedation needs in the setting of mechanical ventilation. It is unclear whether the longer duration of induction contributed to the longer ICU length of stay or if this was a higher acuity population. We hypothesize that the differences in sedative use were due to patient acuity and higher incidence of mechanical ventilation. However, higher sedative use, especially dexmedetomidine, could have mitigated sympathetic withdrawal symptoms in this population and complicated our safety assessment. We also noted a higher proportion of patients receiving phenobarbital in the low-dose induction group. At our institution, phenobarbital is commonly used to wean off of long-term continuous benzodiazepines. Therefore, we expect this difference to be due to attempts to wean patients off continuous opioids and continuous benzodiazepines concurrently. One patient received a low-dose induction while under continuous neuromuscular blockade, which may be a safety concern given the limited ability of the patient to report symptoms of withdrawal and the potential for awareness with paralysis. Interestingly, patients reached similar doses and rates of continuation of buprenorphine at discharge whether they underwent standard or low-dose induction. Ultimately, it appears that buprenorphine can be titrated to an effective dose regardless of initiation strategy. Low-dose inductions were completed over a median of 5 days in our population, a shorter period than has been reported in previous cases (11). We propose that our strategy is simpler to implement compared with previous cases which were administering buprenorphine twice as often at an interval of every 3 hours as opposed to every 6 hours (10, 12). An added benefit of our interval is the reduced workload for nursing which may allow for optimized grouping of care. Of note, a lower proportion of patients receiving low-dose induction had a consult from the Addiction Medicine Team. This may be reflective of increasing comfort among the multidisciplinary ICU teams with this strategy as well as a standardized order set created in the EMR.

Among the 29 patients who completed low-dose induction over a median of 5 days, we found a statistically significant reduction in median full agonist MME requirements after completion by almost 800 mg. Additionally, starting patients on a partial opioid agonist like buprenorphine, without risk of respiratory depression compared with a full agonist opioid like methadone, may be a safer choice which can also be continued at hospital discharge, an especially important consideration for patients with OUD. In patients without OUD requiring prolonged ICU stay receiving large doses of continuous infusion opioids, there is currently no standard of care for opioid weaning. We hypothesize that buprenorphine may be a reasonable approach in these patients to allow for a quick taper off opioids after ICU stay.

There was only a single instance of precipitated withdrawal in our population (0.7%), which occurred in the standard induction group. This patient experienced acute onset diarrhea immediately after buprenorphine was administered. The Addiction Medicine Team was consulted and confirmed the diagnosis of withdrawal, which subsequently resolved after a dose of a full agonist opioid. Previous reports of buprenorphine initiation in the hospital recount much higher incidences of precipitated withdrawal, occurring in up to 39% of patients (16). Though unclear, the cause for the increased rates of withdrawal may be attributed to higher doses of buprenorphine used in those reports, compared with our dosing schema.

Patients with OUD that start buprenorphine in the hospital are more likely to both follow-up with a treatment center and self-report less illicit opioid use (17). Similarly, patients supplied buprenorphine prescriptions from the emergency department have been shown to have effective linkage to outpatient care (18). Approximately 70% of patients in our study were discharged from the hospital with the intent to continue buprenorphine indefinitely, suggesting the ICU can also serve as an effective arena to begin MOUD in patients who may require long-term treatment.

Limitations of this study include the retrospective study design and reliance on a complete and accurate EMR for data collection. Although we collected whether patients required either continuous or intermittent sedative agents, we did not record specific doses. There was also no specific conversion used in patients not continuing buprenorphine and requiring opioids during their ICU stay; opioid titration was at the discretion of the primary treatment team. Additionally, all buprenorphine inductions in this study use protocols specific to our institution with the availability of several buprenorphine formulations (our standard of care is to administer sublingual films), limiting the external validity of our results. Although we anticipate that other approaches to induction may not largely differ, there is the possibility that other institutions with different protocols may experience different results. While we set out to describe a variety of indications for buprenorphine, most buprenorphine use in our population was specifically treating OUD. Furthermore, we only aimed to capture aspects of care related to buprenorphine during inpatient admission and did not collect follow-up information regarding the continued treatment of patients with OUD after discharge, including dose and formulation. Reasons for buprenorphine not being continued at discharge were not collected. In patients with OUD previously receiving a buprenorphine prescription that did not receive a buprenorphine prescription at hospital discharge, we speculate that the care team referred the patient to their primary treatment providers for follow-up.

With increasing literature describing the use of buprenorphine in hospitalized patients, there remains a lack of dedicated research for use specifically in critically ill patients (19). A main strength of the study is the lack of extensive exclusion criteria, which allowed for a greater breadth of included patients for better generalizability to outside institutions. In our critically ill population, buprenorphine was well tolerated. When patients have a history of outpatient buprenorphine use, we believe that all patients should continue their home therapy in the ICU as soon as possible. In buprenorphine-naive patients, both standard and low-dose induction approaches can be safely implemented in the ICU. Buprenorphine low-dose induction may reduce overall full agonist opioid requirements. Further research into the use of buprenorphine and its initiation strategies in this population is warranted.

ACKNOWLEDGMENTS

The authors thank Timothy Wiegand, MD, Jade Malcho, MD, Benjamin McKinney, DO, and Trevor Clarkson, DO for their contributions to the study.