Abstract
Rationale: Patients who receive invasive mechanical ventilation (IMV) are usually exposed to opioids as part of their sedation regimen. The rates of posthospital prescribing of opioids are unknown.
Objectives: To determine the frequency of persistent posthospital opioid use among patients who received IMV.
Methods: We assessed opioid-naive adults who were admitted to an ICU, received IMV, and survived at least 7 days after hospital discharge in Ontario, Canada over a 26-month period (February, 2013 through March, 2015). The primary outcome was new, persistent opioid use during the year after discharge. We assessed factors associated with persistent use by multivariable logistic regression. Patients receiving IMV were also compared with matched hospitalized patients who did not receive intensive care (non-ICU).
Measurements and Main Results: Among 25,085 opioid-naive patients on IMV, 5,007 (20.0%; 95% confidence interval [CI], 19.5–20.5) filled a prescription for opioids in the 7 days after hospital discharge. During the next year, 648 (2.6%; 95% CI, 2.4–2.8) of the IMV cohort met criteria for new, persistent opioid use. The patient characteristic most strongly associated with persistent use in the IMV cohort was being a surgical (vs. medical) patient (adjusted odds ratio, 3.29; 95% CI, 2.72–3.97). The rate of persistent use was slightly higher than for matched non-ICU patients (2.6% vs. 1.5%; adjusted odds ratio, 1.37 [95% CI, 1.19–1.58]).
Conclusions: A total of 20% of IMV patients received a prescription for opioids after hospital discharge, and 2.6% met criteria for persistent use, an average of 300 new persistent users per year in a population of 14 million. Receipt of surgery was the factor most strongly associated with persistent use.
Keywords: critical care, ICU, analgesics, patient outcomes
At a Glance Commentary
Scientific Knowledge on the Subject
Patients who receive invasive mechanical ventilation are usually exposed to opioids as part of their sedation regimen. The influence of this practice on posthospital opioid prescribing is unknown.
What This Study Adds to the Field
Among patients who received invasive mechanical ventilation in Ontario and survived to hospital discharge, 20% filled a prescription for opioids in the week after hospital discharge, and a little over 1 in 50 met criteria for persistent use during the year after the hospitalization. Surgery during hospitalization was the factor most strongly associated with persistent use.
Iatrogenic opioid dependence represents a major contributor to the epidemic of opioid-related harm in Canada and the United States (1), and a lot of attention has been paid to assessing opioid prescribing for individuals who are seen in emergency departments or receive surgery (2–4). As part of an “analgesia-first” approach to sedation during critical illness (5), many critically ill patients receive opioids (6), often before any other sedative agent. In particular, patients who receive invasive mechanical ventilation (IMV) almost always receive an infusion of opioids as part of a regimen of analgesia and sedation. Previous observational studies in Canada found that approximately 85% of critically ill patients receiving IMV were exposed to opioids, and the average daily dose for patients receiving IMV for 2–7 days was 63 mg of morphine equivalent (MME), increasing to 106 MME per day for patients receiving IMV for more than 7 days (7, 8). Similarly, in a point-prevalence study of practices in 50 ICUs in the United Kingdom, among patients receiving sedation, 86% had received analgesic (opioid) agents (9). More recently, a study of 12 community and academic hospitals in Pennsylvania reported 63% of patients admitted to the ICU received opioids, although this study included those with and without IMV (6).
Concerns have been raised that short-term exposure to opioids could lead to long-term use, physical dependence, and other associated harms (10–12). Given the frequent exposure to opioids, and in particular the use of opioid infusions in critically ill patients, there is a large concern that this population may be especially vulnerable to continued use of opioids after hospitalization. However, there is a major evidence gap regarding the frequency of opioid use after critical illness with IMV. We therefore used population-based data from Ontario, Canada to investigate the frequency of new opioid initiation and, particularly, persistent opioid use among critically ill patients who received IMV. To better understand the measured rates, we also compared these rates of prescriptions for opioids after hospital discharge with individuals who were hospitalized but were not critically ill and did not receive IMV.
Methods
Cohort Creation
The study was conducted using data from Ontario healthcare databases, including the Ontario Health Insurance Plan, the Canadian Institute for Health Information Discharge Abstract Database, the Narcotics Monitoring System, and the Registered Persons Database (see Table E1 in the online supplement for details of databases). These databases were linked using unique encoded identifiers and analyzed at ICES, a not-for-profit research institute providing secure access to Ontario’s health-related data. These data allow for longitudinal follow-up of individuals with regard to survival status and healthcare utilization.
Patients eligible for the study cohort included all “opioid-naive” patients admitted to a hospital between February 1, 2013, and March 31, 2015, (26 mo) who survived at least 7 days after hospital discharge. Patients had to have received IMV in an ICU during the hospitalization, defined using Ontario Health Insurance Plan physician billing codes for ventilation (G557-9 and G405-7) (13). To identify an opioid-naive population, patients were excluded if they filled an opioid prescription within the 90 days before hospitalization (14–16), had an opioid overdose as reason for hospitalization in the 1 year prior (T40.0–40.4 and 40.6), or had an opioid overdose as the reason for the index hospitalization. Patients were also excluded if they were younger than 18 years of age or older than 105 years of age, received cardiac surgery during the index hospitalization due to the very short duration of mechanical ventilation for most of these individuals (17), or had been admitted to an ICU in the year before the index hospitalization (see Figure E1 for a flow chart of exclusions).
Comparison Cohort
To allow for a comparison with rates of opioid use among hospitalized patients who were not critically ill, we matched IMV patients with patients who were hospitalized but did not receive either intensive care or IMV (non-ICU patients). To be eligible for comparison, hospital patients had to be admitted to an acute care hospital in Ontario during the same time period (February 1, 2013 through March 31, 2015), but not admitted to an ICU (identified using special care unit codes) (18), and survived for at least 7 days after hospital discharge. Patients also had to be opioid naive using the same criteria as above. All IMV patients were matched for comparison with these non-ICU patients on age, sex, and medical/surgical status using “greedy” matching (19). Greedy matching uses the nearest-available-pair matching method. The cases are ordered and sequentially matched to the nearest unmatched control. If more than one unmatched control matches to a case, the control is selected at random. Due to the large number of potential non-ICU patients for comparison, all IMV patients were able to be matched (see Figure E1).
Outcomes
The primary outcome was the percentage of patients who met criteria for persistent opioid use, defined as having filled a prescription within 7 days of hospital discharge, and filled 10 or more prescriptions in total or more than 120 days total supply within the first year after hospitalization (20). Secondary outcomes included 1) percentage of patients who filled a prescription for any opioid within 7 days after hospital discharge, 2) types of opioids dispensed in the first 7 days, and 3) dose (daily and total) in MMEs of opioid dispensed for initial filled prescriptions. Conversion of doses into MMEs were based on the suggested conversion table from the 2010 Canadian Guidelines for Safe and Effective Use of Opioids for Chronic Non-Cancer Pain (21).
Opioids were defined as any of the following medications: morphine, hydromorphone, fentanyl, codeine, oxycodone, meperidine, or tramadol. Buprenorphine and methadone were also included as opioids for the purposes of identifying prehospital opioid use but were not included in our definition of new opioid exposure, as these medications are primarily dispensed to treat opioid use disorder in Ontario. Of note, the Narcotics Monitoring System includes prescriptions for patients in skilled nursing facilities, but not in acute rehabilitation facilities, specialized acute care facilities (psychiatric), or chronic care facilities. For patients discharged to these latter locations, we commenced observation at the date of discharge from the rehabilitation or other facilities.
Statistical Analysis
For the cohort of IMV patients, we assessed baseline characteristics, including age, sex, type of patient (surgical vs. medical), Charlson comorbidities (22), prior diagnosis of a mental illness or substance abuse disorder (see Table E2 for details), neighborhood median income quintile (defined by postal code), rural residence, benzodiazepine use in the year before hospitalization, as well as location before index admission (home, long-term care, chronic care, rehabilitative care, or other), ICU and hospital length of stay, index hospital type (teaching, community, small, or pediatric), and discharge disposition (home, home with care, long-term care, chronic care, rehabilitative care, or other). We then assessed the same characteristics for patients stratified by whether they were medical or surgical. As P values are dependent on sample size, baseline comparisons were assessed using standardized differences, with an absolute value of <0.1 indicating balance. Next, we calculated opioid use after hospitalization using the outcome definitions described above. We then used multivariable logistic regression models to assess patient and hospital factors in the IMV cohort that were associated with immediate opioid use and persistent opioid use. We assessed each variable for inclusion in a final model using a combination of likelihood ratio tests and clinical judgment.
Sensitivity Analyses
Owing to the variability in estimates associated with different definitions of persistent opioid use, we performed two post hoc sensitivity analyses (23). First, we performed a sensitivity analysis defining persistent use without the requirement for a prescription within the first 7 days after hospital discharge. This allowed us to create a multivariable logistic regression model including the variables above, as well as type of opioid prescription within the first 7 days after hospital discharge as an exposure variable in the model. Finally, we estimated the rate of persistent opioid use, including prescriptions for buprenorphine and methadone.
Comparison with Non-ICU Patients
We first compared baseline characteristics between the IMV and non-ICU patients using standardized differences. We compared the distribution of types of opioids filled within 7 days after hospital discharge for IMV and non-ICU patients using chi-square tests to compare groups. We used conditional logistic regression to generate odds ratios (ORs) of opioid initiation and persistent opioid use, comparing the IMV cohort to the matched non-ICU patients, adjusting for Charlson comorbidity index score, prior substance abuse and mental illness diagnoses, use of benzodiazepines in the year before index admission, income quintile, rural residence, location before index admission, index hospital length of stay, and hospital type. For continuous outcomes, we used multiple linear regression models to compare the IMV cohort to non-ICU patients, adjusting for the same factors as for the binary outcomes. Because of differential death rates in the IMV and hospital cohorts, we performed two sensitivity analyses for the outcome of persistent opioid use. First, we used a Fine and Gray regression model with death as a competing risk (24). Next, we restricted the entire ICU and non-ICU cohorts to individuals who survived the first year after hospital discharge.
Ethics approval for the study was obtained from the Research Ethics Board of Sunnybrook Health Sciences Center. The need for written informed consent was waived. All significance testing was two-sided. Statistical analyses were performed with SAS Enterprise Guide 7.1 (SAS Institute Inc.).
Results
Patient Characteristics
Between February 1, 2013, and March 31, 2015, (26 mo), 25,085 opioid-naive patients were admitted to a hospital, received IMV, and survived at least 7 days after hospital discharge. Characteristics of these patients are shown in Table 1. The median age was 64 (interquartile range [IQR], 51–75) years, with 57.7% men. In the year before hospitalization, 18.5% had filled at least one prescription for a benzodiazepine, and 0.2% had a hospitalization for a nonopioid overdose. During the hospitalization, the median duration of IMV was 3 (IQR, 2–6) days, with a median ICU length of stay of 6 (IQR, 3–11) days, and hospital length of stay of 13 (IQR, 7–27) days. Discharge destination was directly home with or without home care for 68.4% of patients. There were differences in characteristics between medical and surgical patients (Table 1). In particular, more medical patients had three or more Charlson comorbidities (24.1% vs. 18.5%; standardized difference = 0.14) and were more likely to have filled a prescription for a benzodiazepine in the year before hospitalization (21.7% vs. 15.1%; standardized difference = 0.17) but had substantially shorter hospital lengths of stay (median 11 [IQR, 5–23] d vs. 15 [IQR, 8–33] d; standardized difference = 0.33).
Table 1.
Characteristics of Opioid-Naive Patients Surviving to 7 Days after an Index ICU Containing Hospitalization with Invasive Mechanical Ventilation
| ICU with IMV (n = 25,085) | Medical (n = 12,912) | Surgical (n = 12,173) | Standardized Difference | |
|---|---|---|---|---|
| Age, yr | ||||
| Mean (SD) | 61.7 (17.9) | 61.1 (18.7) | 62.3 (16.9) | 0.07 |
| Median (IQR) | 64 (51–75) | 64 (50–76) | 65 (53–75) | 0.05 |
| Age group, n (%) | ||||
| 18–44 yr | 4,330 (17.3) | 2,515 (19.5) | 1,815 (14.9) | 0.12 |
| 45–64 yr | 8,332 (33.2) | 4,132 (32.0) | 4,200 (34.5) | 0.05 |
| 65–74 yr | 5,707 (22.8) | 2,720 (21.1) | 2,987 (24.5) | 0.08 |
| ≥75 yr | 6,716 (26.8) | 3,545 (27.5) | 3,171 (26.0) | 0.03 |
| Sex, M, n (%) | 14,486 (57.7) | 7,229 (56.0) | 7,257 (59.6) | 0.07 |
| Surgical patient, n (%) | 12,173 (48.5) | NA | NA | NA |
| Charlson comorbidity, n (%) | ||||
| 0 | 10,676 (42.6) | 4,995 (38.7) | 5,681 (46.7) | 0.16 |
| 1–2 | 9,050 (36.1) | 4,804 (37.2) | 4,246 (34.9) | 0.05 |
| ≥3 | 5,359 (21.4) | 3,113 (24.1) | 2,246 (18.5) | 0.14 |
| Prior mental illness diagnosis, n (%) | 6,473 (25.8) | 3,797 (29.4) | 2,676 (22.0) | 0.17 |
| Prior substance abuse diagnosis, n (%) | 955 (3.8) | 685 (5.3) | 270 (2.2) | 0.16 |
| Income quintile, n (%)* | ||||
| 1 (lowest) | 5,687 (22.8) | 3,139 (24.5) | 2,548 (21.1) | 0.08 |
| 2 | 5,251 (21.1) | 2,817 (22.0) | 2,434 (20.1) | 0.05 |
| 3 | 4,823 (19.4) | 2,390 (18.6) | 2,433 (20.1) | 0.04 |
| 4 | 4,830 (19.4) | 2,396 (18.7) | 2,434 (20.1) | 0.04 |
| 5 (highest) | 4,315 (17.3) | 2,076 (16.2) | 2,239 (18.5) | 0.06 |
| Rural residence, n (%)* | 2,972 (11.8) | 1,400 (10.8) | 1,572 (12.9) | — |
| Drug use (year before index admission), n (%) | ||||
| Benzodiazepine | 4,639 (18.5) | 2,797 (21.7) | 1,842 (15.1) | 0.17 |
| Nonopioid overdose hospitalization | 50 (0.2) | 44 (0.3) | 6 (0.0) | 0.07 |
| Location before index admission, n (%) | ||||
| Home | 19,072 (76.0) | 9,789 (75.8) | 9,283 (76.3) | 0.01 |
| Long-term care | 466 (1.9) | 336 (2.6) | 130 (1.1) | 0.11 |
| Chronic care | 122 (0.5) | 86 (0.7) | 36 (0.3) | 0.05 |
| Rehabilitative care | 111 (0.4) | 60 (0.5) | 51 (0.4) | 0.01 |
| Other | 5,314 (21.2) | 2,641 (20.5) | 2,673 (22.0) | 0.04 |
| Dialysis during index admission, n (%) | 1,598 (6.4) | 866 (6.7) | 732 (6.0) | 0.03 |
| Duration of mechanical ventilation, d | ||||
| Mean (SD) | 6.5 (13.0) | 6.3 (13.0) | 6.6 (13.0) | 0.02 |
| Median (IQR) | 3 (2–6) | 3 (2–6) | 3 (2–6) | 0.07 |
| Hospital LOS, d | ||||
| Mean (SD) | 23.9 (36.8) | 20.1 (32.7) | 28.0 (40.3) | 0.22 |
| Median (IQR) | 13 (7–27) | 11 (5–23) | 15 (8–33) | 0.33 |
| ICU LOS, d | ||||
| Mean (SD) | 10.0 (16.3) | 9.5 (15.5) | 10.5 (17.0) | 0.06 |
| Median (IQR) | 6 (3–11) | 6 (3–10) | 5 (3–11) | 0.01 |
| Index hospital type, n (%)* | ||||
| Teaching | 10,665 (42.5) | 4,567 (35.4) | 6,098 (50.1) | 0.3 |
| Community | 14,092 (56.2) | 8,174 (63.3) | 5,918 (48.6) | 0.3 |
| Small† | 303 (1.2) | NA | NA | 0.01 |
| Pediatric† | 16 (0.1) | NA | NA | — |
| Discharge disposition, n (%) | ||||
| Home | 10,318 (41.1) | 5,599 (43.4) | 4,719 (38.8) | 0.09 |
| Home with care | 6,851 (27.3) | 3,275 (25.4) | 3,576 (29.4) | 0.09 |
| Long-term care | 1,642 (6.5) | 1,176 (9.1) | 466 (3.8) | 0.22 |
| Chronic care | 1,521 (6.1) | 741 (5.7) | 780 (6.4) | 0.03 |
| Rehabilitative care | 3,843 (15.3) | 1,554 (12.0) | 2,289 (18.8) | 0.19 |
| Other‡ | 910 (3.6) | 567 (4.4) | 343 (2.8) | 0.08 |
Definition of abbreviations: IMV = invasive mechanical ventilation; IQR = interquartile range; LOS = length of stay; NA = not applicable.
Missing data on income (n = 179), area of residence (n = 2), and index hospital type (n = 9).
Data for medical and surgical patients suppressed owing to small numbers, per ICES policy.
Other included: patients transferred to other ambulatory and inpatient care settings and patients who left the hospital against medical advice.
Opioid Prescriptions Filled after Hospital Discharge
For the IMV cohort of opioid-naive patients, 5,007 (20.0% [95% confidence interval [CI], 19.5–20.5]) filled a prescription within 7 days after hospital discharge. Among the patients in the IMV cohort who filled an initial prescription within 7 days, the mean total MME dose was 404 (SD, 785), and 25.7% filled four or more prescriptions over the course of the year (Table 2). Hydromorphone was the most common initial opioid prescription (42.6% of all initial prescriptions), followed by oxycodone and codeine. Stratified by medical or surgical status, a greater percentage of surgical patients filled an opioid prescription within 7 days (33.0% vs. 7.6%; P < 0.001).
Table 2.
Patterns of Opioid Use among Those Filling a Prescription within 7 Days for Patients with an Index Hospitalization with ICU Admission and Mechanical Ventilation
| ICU with IMV (n = 25,085) |
|||||
|---|---|---|---|---|---|
| n | Percentage (95% CI)* | Medical [n (%)] (n = 12,912) | Surgical [n (%)] (n = 12,173) | P Value | |
| Filled opioid prescription within 7 d of discharge | 5,007 | 20.0 (19.5–20.5) | 986 (7.6) | 4,021 (33.0) | <0.001 |
| Persistent opioid use | 648 | 2.6 (2.4–2.8) | 159 (1.2) | 489 (4.0) | <0.001 |
| Persistent opioid use (exclude requirement for prescription within 7 d after index) | 1,218 | 4.9 (4.6–5.1) | 437 (3.4) | 781 (6.4) | <0.001 |
| Persistent opioid use (including buprenorphine and methadone) | 669 | 2.7 (2.5–2.9) | 171 (1.3) | 498 (4.1) | <0.001 |
| Pattern of opioid use among those filling a prescription within 7 d |
n = 5,007 |
n = 986 | n = 4,021 | ||
| >1 Initial opioid prescription | 562 | 11.2 (10.3–12.1) | 136 (13.8) | 426 (10.6) | 0.004 |
| >1 Initial class of opioid | 189 | 3.8 (3.2–4.3) | 37 (3.8) | 152 (3.8) | 0.97 |
| Total dose MME (initial prescriptions) | |||||
| Mean (SD) | — | 404 (785) | 409 (893) | 402 (756) | 0.80 |
| Median (IQR) | — | 225 (135–375) | 195 (125–315) | 225 (140–375) | <0.001 |
| Daily dose MME (initial prescriptions) | |||||
| Mean (SD) | — | 59.6 (66.4) | 50.1 (63.8) | 62.0 (66.9) | <0.001 |
| Median (IQR) | — | 45 (29–75) | 34 (19–58) | 50 (30–75) | <0.001 |
| No. of prescriptions filled within the 7 d after discharge | 0.001 | ||||
| 1 | 4,368 | 87.2 (86.3–88.2) | 828 (84.0) | 3,540 (88.0) | — |
| 2 | 557 | 11.1 (10.3–12.0) | 136 (13.8) | 421 (10.5) | — |
| 3 | 72 | 1.4 (1.1–1.8) | 22 (2.2)† | 60 (1.5)† | — |
| >3 | 10 | 0.2 (0.1–0.3) | — | — | — |
| No. of prescriptions filled in year after discharge | <0.001 | ||||
| 1 | 2,257 | 45.1 (43.7–46.5) | 391 (39.7) | 1,866 (46.4) | — |
| 2 | 1,005 | 20.1 (19.0–21.2) | 200 (20.3) | 805 (20.0) | — |
| 3 | 459 | 9.2 (8.4–10.0) | 104 (10.5) | 355 (8.8) | — |
| >3 | 1,286 | 25.7 (24.5–26.9) | 291 (29.5) | 995 (24.7) | — |
| Persistent opioid use | 648 | 12.9 (12.0–13.9) | 159 (16.1) | 489 (12.2) | <0.001 |
Definition of abbreviations: CI = confidence interval; IMV = invasive mechanical ventilation; IQR = interquartile range; MME = milligrams of morphine equivalent.
Unless otherwise noted.
Includes the category of more than three prescriptions owing to small numbers, per ICES policy.
Persistent Opioid Use
In the IMV cohort, 2.6% (95% CI, 2.4–2.8) met criteria for persistent opioid use in the year after hospitalization (Table 2). Persistent opioid use was higher among surgical patients (4.0%) than medical patients (1.2%; P < 0.001). In sensitivity analyses, reclassifying patients without the requirement to have filled a first prescription within 7 days after hospital discharge, the rate of persistent opioid use was 4.9% (95% CI, 4.6–5.1), and including methadone and buprenorphine, the rate was 2.7% (95% CI, 2.5–2.9).
Factors Associated with Filling Opioid Prescriptions after IMV
In multivariable analysis of the IMV patients, the strongest predictor of filling an opioid prescription within 7 days of hospital discharge was being a surgical (vs. medical) patient (adjusted OR [aOR], 6.14 [95% CI, 5.67–6.65]; P < 0.001; Table 3). Age was also strongly associated, with older patients less likely to fill a prescription (aOR for ≥75 yr vs. 18–44 yr, 0.42 [95% CI, 0.37–0.47]; P < 0.001). Individuals with more comorbidities were less likely to fill an opioid prescription (aOR for ≥3 vs. 0 Charlson comorbidity index score, 0.69 [95% CI, 0.63–0.77]; P < 0.001), as were patients who spent longer in the ICU (aOR for >7 d vs. 0–3 d, 0.55 [95% CI, 0.50–0.60]; P < 0.001).
Table 3.
Multivariable Analysis of Factors Associated with Filling Opioid Prescriptions after Hospital Discharge among Patients on Invasive Mechanical Ventilation
| Characteristics | Total (n) | Filled a Prescription within 7 d of Discharge (n = 5,007) |
Persistent Use (n = 648) |
||||
|---|---|---|---|---|---|---|---|
| n (%) | OR (95% CI) | P Value | n (%) | OR (95% CI) | P Value | ||
| Age category | |||||||
| 18–44 yr | 4,330 | 1,105 (25.5) | Reference | 159 (3.7) | Reference | ||
| 45–64 yr | 8,332 | 1,893 (22.7) | 0.77 (0.70–0.85) | <0.001 | 276 (3.3) | 0.79 (0.64–0.97) | 0.03 |
| 65–74 yr | 5,707 | 1,143 (20.0) | 0.65 (0.58–0.72) | <0.001 | 120 (2.1) | 0.48 (0.37–0.61) | <0.001 |
| ≥75 yr | 6,716 | 866 (12.9) | 0.42 (0.37–0.47)) | <0.001 | 93 (1.4) | 0.30 (0.23–0.40) | <0.001 |
| Sex | |||||||
| F | 10,599 | 2,017 (19.0) | Reference | 259 (2.4) | Reference | ||
| M | 14,486 | 2,990 (20.6) | 0.99 (0.93–1.07) | 0.89 | 389 (2.7) | 0.99 (0.84–1.17) | 0.93 |
| Charlson comorbidity score | |||||||
| 0 | 10,676 | 2,722 (25.5) | Reference | 313 (2.9) | Reference | ||
| 1–2 | 9,050 | 1,510 (16.7) | 0.73 (0.67–0.79) | <0.001 | 208 (2.3) | 0.95 (0.79–1.15) | 0.62 |
| ≥3 | 5,359 | 775 (14.5) | 0.69 (0.63–0.77) | <0.001 | 127 (2.4) | 1.10 (0.88–1.38) | 0.39 |
| Prior mental illness | |||||||
| No | 18,612 | 3,811 (20.5) | Reference | 451 (2.4) | Reference | ||
| Yes | 6,473 | 1,196 (18.5) | 0.98 (0.90–1.07) | 0.67 | 197 (3.0) | 1.24 (1.03–1.49) | 0.02 |
| Prior substance abuse | |||||||
| No | 24,130 | 4,855 (20.1) | Reference | 609 (2.5) | Reference | ||
| Yes | 955 | 152 (15.9) | 1.00 (0.82–1.22) | 0.97 | 39 (4.1) | 1.70 (1.19–2.42) | 0.004 |
| Income quintile* | |||||||
| 1 (lowest) | 5,687 | 1,055 (18.6) | Reference | 128 (2.3) | Reference | ||
| 2 | 5,251 | 1,004 (19.1) | 1.03 (0.93–1.14) | 0.60 | 161 (3.1) | 1.43 (1.12–1.81) | 0.004 |
| 3 | 4,823 | 997 (20.7) | 1.07 (0.96–1.19) | 0.21 | 129 (2.7) | 1.20 (0.93–1.54) | 0.15 |
| 4 | 4,830 | 1,002 (20.7) | 1.07 (0.96–1.19) | 0.24 | 118 (2.4) | 1.13 (0.87–1.45) | 0.37 |
| 5 (highest) | 4,315 | 923 (21.4) | 1.08 (0.97–1.20) | 0.17 | 107 (2.5) | 1.10 (0.85–1.44) | 0.46 |
| Area of residence* | |||||||
| Urban | 22,111 | 4,346 (19.7) | Reference | 572 (2.6) | Reference | ||
| Rural | 2,972 | 661 (22.2) | 1.05 (0.94–1.17) | 0.37 | 76 (2.6) | 0.91 (0.70–1.19) | 0.49 |
| Location before index admission | |||||||
| Home | 19,072 | 3,959 (20.8) | Reference | 472 (2.5) | Reference | ||
| Long-term care | 466 | 47 (10.1) | 1.26 (0.88–1.82) | 0.21 | 6 (1.3) | 0.71 (0.29–1.69) | 0.44 |
| Chronic care† | — | — | — | — | — | — | — |
| Rehabilitative care† | — | — | — | — | — | — | — |
| Other | 5,314 | 982 (18.5) | 0.92 (0.84–1.01) | 0.07 | 163 (3.1) | 1.10 (0.91–1.33) | 0.32 |
| Hospital type* | |||||||
| Teaching | 10,665 | 2,702 (25.3) | Reference | 339 (3.2) | Reference | ||
| Community | 14,092 | 2,254 (16.0) | 0.76 (0.70–0.81) | <0.0001 | 298 (2.1) | 0.85 (0.72–1.00) | 0.05 |
| Other | 328 | 51 (15.5) | 0.66 (0.47–0.92) | 0.01 | 11 (3.4) | 1.44 (0.75–2.76) | 0.27 |
| Patient type | |||||||
| Medical | 12,912 | 986 (7.6) | Reference | 159 (1.2) | Reference | ||
| Surgical | 12,173 | 4,021 (33.0) | 6.14 (5.67–6.65) | <0.001 | 489 (4.0) | 3.29 (2.72–3.97) | <0.001 |
| Received dialysis | |||||||
| No | 23,487 | 4,794 (20.4) | Reference | 597 (2.5) | Reference | ||
| Yes | 1,598 | 213 (13.3) | 0.82 (0.69–0.96) | 0.02 | 51 (3.2) | 1.10 (0.81–1.48) | 0.55 |
| Cumulative days in ICU | |||||||
| 0–3 d | 7,230 | 1,981 (27.4) | Reference | 146 (2.0) | Reference | ||
| 4–5 d | 5,111 | 1,065 (20.8) | 0.80 (0.72–0.87) | <0.001 | 120 (2.3) | 1.18 (0.92–1.51) | 0.20 |
| 6–7 d | 3,323 | 598 (18.0) | 0.69 (0.62–0.77) | <0.001 | 75 (2.3) | 1.11 (0.83–1.48) | 0.48 |
| >7 d | 9,421 | 1,363 (14.5) | 0.55 (0.50–0.60) | <0.001 | 307 (3.3) | 1.32 (1.06–1.65) | 0.01 |
| Discharge disposition | |||||||
| Home | 10,318 | 2,416 (23.4) | Reference | 159 (1.5) | Reference | ||
| Home with care | 6,851 | 1,780 (26.0) | 1.40 (1.29–1.52) | <0.001 | 256 (3.7) | 2.63 (2.13–3.24) | <0.001 |
| Long-term care | 1,642 | 173 (10.5) | 0.65 (0.53–0.79) | <0.001 | 41 (2.5) | 2.35 (1.60–3.45) | <0.001 |
| Chronic care | 1,521 | 107 (7.0) | 0.35 (0.28–0.44) | <0.001 | 36 (2.4) | 1.71 (1.15–2.53) | 0.008 |
| Rehabilitative care | 3,843 | 464 (12.1) | 0.46 (0.40–0.51) | <0.001 | 143 (3.7) | 2.16 (1.69––2.78) | <0.001 |
| Other | 910 | 67 (7.4) | 0.30 (0.23–0.39) | <0.001 | 13 (1.4) | 0.86 (0.48–1.53) | 0.60 |
Definition of abbreviations: CI = confidence interval; OR = odds ratio.
Missing data on income (n = 179), area of residence (n = 2), and index hospital type (n = 9).
Suppression owing to small cell size required per ICES policy.
Being a surgical (vs. medical) patient was again the factor most strongly associated with persistent use (aOR, 3.29 [95% CI, 2.72–3.97]; P < 0.001; Table 3). Discharge home with additional care was associated with increased odds of persistent use (compared with discharge home without care; aOR, 2.63 [95% CI, 2.13–3.24]; P < 0.001). Longer time spent in the ICU had the opposite association with persistent use, with increased odds of persistent use for those in the ICU longer than 7 days compared with 0–3 days (aOR, 1.32 [95% CI, 1.06–1.65]; P = 0.01).
Reclassifying patients based on a definition of persistent opioid use that did not require a prescription filled within the first 7 days after hospital discharge demonstrated similar results. However, the effect of surgery was attenuated by the inclusion of the mediator of a prescription filled within the first 7 days (aOR for surgery, 1.17 [95% CI, 1.02–1.34]; P = 0.027), with high ORs associated with filling specific opioids in the first 7 days (highest for receipt of fentanyl or meperidine prescription [aOR, 32.75 (95% CI, 19.09–56.18); P < 0.001], followed by hydromorphone prescription [aOR, 7.40 (95% CI, 6.28–8.72); P < 0.001], and lowest for tramadol [aOR, 4.09 (95% CI, 2.80–5.98); P < 0.001]; Table E3).
Comparison with Non-ICU Patients
The non-ICU patients had fewer Charlson comorbidities, less benzodiazepine use in the year before hospitalization, and a greater proportion discharged directly home (Table E4). Patients in the IMV cohort of opioid-naive patients were less likely to fill a prescription within 7 days after hospital discharge (20.0% vs. 34.2% in the non-ICU hospital cohort; aOR, 0.59 [95% CI, 0.56–0.61]; P < 0.001 for comparison of IMV vs. non-ICU patients; Table E5). However, IMV patients were more likely to meet criteria for persistent opioid use of 2.6% versus 1.5% for the non-ICU cohort (aOR, 1.37 [95% CI, 1.19–1.58]; P = 0.01). To account for the 1-year mortality (13.4% for the IMV cohort and 5.7% for the hospital cohort), we performed two sensitivity analyses for the outcome of persistent use, first accounting for the competing risk of death using a Fine and Gray model (hazard ratio, 1.35 [95% CI, 1.18–1.55]; P < 0.001) and then excluding individuals who died during the year (OR, 1.23 [95% CI, 1.12–1.35]; P < 0.001) (Table E6).
Discussion
In this population-based cohort study of opioid-naive patients admitted to ICUs who received IMV, one in five patients filled a prescription for an opioid within 7 days of hospital discharge, and a little more than 1 in 50 were persistent opioid users at the end of 1 year, representing approximately 300 new persistent opioid users each year in Ontario. In comparison with non-ICU patients, the persistent use rate was slightly higher. Given that almost all patients who receive IMV are exposed to opioids, often at high doses, these findings suggest that use of opioids during acute critical illness does not lead to high rates of persistent use after hospital discharge. However, with over 200,000 patients admitted to ICUs in Canada (25), and millions in the United States each year (26), even these small percentages may be important from the public health perspective.
Opioids in the ICU are used both for pain control and also as part of sedative regimens (27, 28). Recent guidelines have recommended minimizing exposure to other traditional sedatives, such as benzodiazepines and propofol, owing to concerns of delirium, hypotension, and other side-effects, and, as a consequence, opioids are now used routinely to treat the discomfort associated with IMV (28). Although opioid infusions are typically discontinued before transfer out of an ICU, the frequency of continued use in oral form on the wards is not well described. One study by Donohue and colleagues (6) found that ICU patients were more likely to have a longer “opioid-free” period before hospital discharge compared with non-ICU hospitalized patients. In other work focused on other medications initiated during a critical illness, such as antipsychotics and gastric acid suppressants, there were high rates of unintentional continuation of these medications after hospital discharge (1.4% for antipsychotics and 6.1% for gastric acid suppressants) (29). However, opioids are inherently different, as many patients may experience on-going pain after a critical illness, particularly involving surgery (30, 31). It is notable that the largest driver of both early and more persistent opioid use was receipt of surgery during the hospitalization.
Our finding of approximately 2–3% persistent opioid use overall in the IMV population, and higher use in the surgical population, are consistent with studies of more general medical and surgical populations, which have found rates of 0.1–8% persistent use, depending on the cohort and definition of persistent use (6, 20, 32, 33), although it is important to note that comparison is challenging owing to differences in definitions chosen to identify persistent use (4). Estimates of persistent opioid use in the general population after any exposure to opioids are also similar, including an estimate of 2.6% examining individuals in the United States who had at least one opioid prescription and who were in a commercially ensured database (34), and up to 5% in an assessment of all residents in Oregon who filled any opioid prescription (35). It is also of interest that higher rates of persistent use were observed in individuals discharged to settings other than home. This may represent an opportunity for intervention, given that these are still formal care settings.
Strengths of this study include the ability to evaluate all adult patients admitted to ICUs who received IMV in a large geographic region; many previous studies of opioid prescribing in ICU patients had more limited focus owing to variability in prescription capture across databases (20, 32). We were also able to capture all prescriptions filled (including those obtained without insurance coverage) owing to the requirement for all opioid prescriptions to be recorded; many similar U.S. datasets only capture prescriptions filled by a subset of ensured patients paying for the prescription through the insurer (3, 20), or documented prescriptions provided at healthcare encounters (6).
The major limitation of this study is the inability to determine the exact exposure to opioids during the hospitalization. However, based on observational (8, 36, 37) and trial data (38, 39), we expect that at least 80% of patients receiving IMV would have received intravenous opioid during the ICU stay. We compared IMV patients with hospitalized patients who were not admitted to the ICU. Although we matched on a number of key variables, including age, sex, and whether or not they underwent a surgical procedure, heterogeneity remains with regard to diagnosis, and therefore potential need for analgesics after hospitalization. We also could not determine the posthospitalization pain of individuals or adjudicate the appropriateness of the opioid prescription. It is important to recognize that persistent pain (particularly after certain surgical procedures) is a known concern, and not all persistent use of opioids represents inappropriate use (30, 31). Finally, we did not assess patients for opioid use disorder, and persistent use is not the same as opioid use disorder.
In summary, we evaluated patients admitted to ICUs who received IMV to determine their rates of opioid use after hospital discharge. Overall, 20% of patients filled a prescription within 7 days after hospital discharge, and a little more than 1 in 50 had persistent opioid use over the subsequent year. These findings suggest that, although there is a small increase in the rate of persistent use compared with non-ICU hospitalized patients, overall the “analgesia-first” approach to sedation in ICUs does not result in high rates of opioid use after hospitalization.
Supplementary Material
Acknowledgments
Acknowledgment
The authors thank IMS Brogan Inc. for use of their Drug Information Database.
Footnotes
Supported by Canadian Institutes of Health Research grant 365432 (H.W., Principal Investigator) and NIH and Acute and Intensive Outcomes Research Network grant R01 DA042299, in part by University of Toronto Department of Anesthesia merit awards (H.W. and D.N.W.), and in part by Canadian Institutes of Health Research New Investigator Award and the Endowed Chair in Translational Anesthesiology Research at St. Michael’s Hospital and University of Toronto (D.N.W.); the study was also supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. The opinions, results, and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by ICES or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred. Funders had no role in the design or conduct of the study, collection, management, analysis, or interpretation of the data, preparation, review, or approval of the manuscript, or the decision to submit the manuscript for publication. Parts of this material are based on data and/or information compiled and provided by the Canadian Institute for Health Information. However, the analyses, conclusions, opinions and statements expressed in the material are those of the author(s), and not necessarily those of the Canadian Institute for Health Information.
Author Contributions: Full access to all of the data in the study and responsibility for the integrity of the data and the accuracy of the data analyses—L.F.; study concept and design—H.W., R.A.F., T.G., E.F., D.N.J., D.N.W., and D.C.S.; acquisition or interpretation of the data—all authors; statistical analyses—H.W., A.D.H., L.F., and R.P.; drafting of the manuscript—H.W.; review and revision of the manuscript for intellectual content—all authors; obtained funding for the study—H.W., R.A.F., and D.C.S.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Originally Published in Press as DOI: 10.1164/rccm.201912-2503OC on April 29, 2020
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1.U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. National action plan for adverse drug event prevention. Washington, DC: U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. 2014.
- 2.Hoppe JA, Kim H, Heard K. Association of emergency department opioid initiation with recurrent opioid use. Ann Emerg Med. 2015;65:493–499. doi: 10.1016/j.annemergmed.2014.11.015. [DOI] [PubMed] [Google Scholar]
- 3.Barnett ML, Olenski AR, Jena AB. Opioid-prescribing patterns of emergency physicians and risk of long-term use. N Engl J Med. 2017;376:663–673. doi: 10.1056/NEJMsa1610524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Neuman MD, Bateman BT, Wunsch H. Inappropriate opioid prescription after surgery. Lancet. 2019;393:1547–1557. doi: 10.1016/S0140-6736(19)30428-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Barr J, Fraser GL, Puntillo K, Ely EW, Gélinas C, Dasta JF, et al. American College of Critical Care Medicine. Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the intensive care unit. Crit Care Med. 2013;41:263–306. doi: 10.1097/CCM.0b013e3182783b72. [DOI] [PubMed] [Google Scholar]
- 6.Donohue JM, Kennedy JN, Seymour CW, Girard TD, Lo-Ciganic WH, Kim CH, et al. Patterns of opioid administration among opioid-naive inpatients and associations with postdischarge opioid use: a cohort study. Ann Intern Med. 2019;171:81–90. doi: 10.7326/M18-2864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mehta S, Burry L, Fischer S, Martinez-Motta JC, Hallett D, Bowman D, et al. Canadian Critical Care Trials Group. Canadian survey of the use of sedatives, analgesics, and neuromuscular blocking agents in critically ill patients. Crit Care Med. 2006;34:374–380. doi: 10.1097/01.ccm.0000196830.61965.f1. [DOI] [PubMed] [Google Scholar]
- 8.Burry LD, Williamson DR, Perreault MM, Rose L, Cook DJ, Ferguson ND, et al. Analgesic, sedative, antipsychotic, and neuromuscular blocker use in Canadian intensive care units: a prospective, multicentre, observational study. Can J Anaesth. 2014;61:619–630. doi: 10.1007/s12630-014-0174-1. [DOI] [PubMed] [Google Scholar]
- 9.Richards-Belle A, Canter RR, Power GS, Robinson EJ, Reschreiter H, Wunsch H, et al. National survey and point prevalence study of sedation practice in UK critical care. Crit Care. 2016;20:355. doi: 10.1186/s13054-016-1532-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dhalla IA, Persaud N, Juurlink DN. Facing up to the prescription opioid crisis. BMJ. 2011;343:d5142. doi: 10.1136/bmj.d5142. [DOI] [PubMed] [Google Scholar]
- 11.Beauchamp GA, Winstanley EL, Ryan SA, Lyons MS. Moving beyond misuse and diversion: the urgent need to consider the role of iatrogenic addiction in the current opioid epidemic. Am J Public Health. 2014;104:2023–2029. doi: 10.2105/AJPH.2014.302147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kissin I. Long-term opioid treatment of chronic nonmalignant pain: unproven efficacy and neglected safety? J Pain Res. 2013;6:513–529. doi: 10.2147/JPR.S47182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Scales DC, Guan J, Martin CM, Redelmeier DA. Administrative data accurately identified intensive care unit admissions in Ontario. J Clin Epidemiol. 2006;59:802–807. doi: 10.1016/j.jclinepi.2005.11.015. [DOI] [PubMed] [Google Scholar]
- 14.Rozet I, Nishio I, Robbertze R, Rotter D, Chansky H, Hernandez AV. Prolonged opioid use after knee arthroscopy in military veterans. Anesth Analg. 2014;119:454–459. doi: 10.1213/ANE.0000000000000292. [DOI] [PubMed] [Google Scholar]
- 15.Pugely AJ, Bedard NA, Kalakoti P, Hendrickson NR, Shillingford JN, Laratta JL, et al. Opioid use following cervical spine surgery: trends and factors associated with long-term use. Spine J. 2018;18:1974–1981. doi: 10.1016/j.spinee.2018.03.018. [DOI] [PubMed] [Google Scholar]
- 16.Holman JE, Stoddard GJ, Higgins TF. Rates of prescription opiate use before and after injury in patients with orthopaedic trauma and the risk factors for prolonged opiate use. J Bone Joint Surg Am. 2013;95:1075–1080. doi: 10.2106/JBJS.L.00619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gershengorn HB, Wunsch H, Hua M, Bavaria JE, Gutsche J. Association of overnight extubation with outcomes after cardiac surgery in the intensive care unit. Ann Thorac Surg. 2019;108:432–442. doi: 10.1016/j.athoracsur.2019.04.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Garland A, Yogendran M, Olafson K, Scales DC, McGowan KL, Fransoo R. The accuracy of administrative data for identifying the presence and timing of admission to intensive care units in a Canadian province. Med Care. 2012;50:e1–e6. doi: 10.1097/MLR.0b013e318245a754. [DOI] [PubMed] [Google Scholar]
- 19.Parsons L. Reducing bias in a propensity score matched-pair sample using greedy matching techniques. Presented at the 26th Annual SAS Users Group International Conference 2001. April 22–25, 2001, Long Beach, CA. Abstract 214-26.
- 20.Sun EC, Darnall BD, Baker LC, Mackey S. Incidence of and risk factors for chronic opioid use among opioid-naive patients in the postoperative period. JAMA Intern Med. 2016;176:1286–1293. doi: 10.1001/jamainternmed.2016.3298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Canadian guideline for safe and effective use of opioids for chronic non-cancer pain. National Opioid Use Guideline Group (NOUGG): Canada; 2010 [updated 2017 May 8; accessed 2019 Jan 24]. Available from: http://nationalpaincentre.mcmaster.ca/opioid/
- 22.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613–619. doi: 10.1016/0895-4356(92)90133-8. [DOI] [PubMed] [Google Scholar]
- 23.Kent ML, Hurley RW, Oderda GM, Gordon DB, Sun E, Mythen M, et al. POQI-4 Working Group. American Society for Enhanced Recovery and Perioperative Quality Initiative-4 joint consensus statement on persistent postoperative opioid use: definition, incidence, risk factors, and health care system initiatives. Anesth Analg. 2019;129:543–552. doi: 10.1213/ANE.0000000000003941. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. [Google Scholar]
- 25.Canadian Institute for Health Information. Care in Canadian ICUs. Ottawa, ON: CIHI; 2016. [Google Scholar]
- 26.Coopersmith CM, Wunsch H, Fink MP, Linde-Zwirble WT, Olsen KM, Sommers MS, et al. A comparison of critical care research funding and the financial burden of critical illness in the United States. Crit Care Med. 2012;40:1072–1079. doi: 10.1097/CCM.0b013e31823c8d03. [DOI] [PubMed] [Google Scholar]
- 27.Nasraway SA, Jr, Jacobi J, Murray MJ, Lumb PD Task Force of the American College of Critical Care Medicine of the Society of Critical Care Medicine and the American Society of Health-System Pharmacists, American College of Chest Physicians. Sedation, analgesia, and neuromuscular blockade of the critically ill adult: revised clinical practice guidelines for 2002. Crit Care Med. 2002;30:117–118. doi: 10.1097/00003246-200201000-00019. [DOI] [PubMed] [Google Scholar]
- 28.Devlin JW, Skrobik Y, Gélinas C, Needham DM, Slooter AJC, Pandharipande PP, et al. Clinical practice guidelines for the prevention and management of pain, agitation/sedation, delirium, immobility, and sleep disruption in adult patients in the ICU. Crit Care Med. 2018;46:e825–e873. doi: 10.1097/CCM.0000000000003299. [DOI] [PubMed] [Google Scholar]
- 29.Scales DC, Fischer HD, Li P, Bierman AS, Fernandes O, Mamdani M, et al. Unintentional continuation of medications intended for acute illness after hospital discharge: a population-based cohort study. J Gen Intern Med. 2016;31:196–202. doi: 10.1007/s11606-015-3501-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Johansen A, Romundstad L, Nielsen CS, Schirmer H, Stubhaug A. Persistent postsurgical pain in a general population: prevalence and predictors in the Tromsø study. Pain. 2012;153:1390–1396. doi: 10.1016/j.pain.2012.02.018. [DOI] [PubMed] [Google Scholar]
- 31.Kehlet H, Jensen TS, Woolf CJ. Persistent postsurgical pain: risk factors and prevention. Lancet. 2006;367:1618–1625. doi: 10.1016/S0140-6736(06)68700-X. [DOI] [PubMed] [Google Scholar]
- 32.Brummett CM, Waljee JF, Goesling J, Moser S, Lin P, Englesbe MJ, et al. New persistent opioid use after minor and major surgical procedures in US adults. JAMA Surg. 2017;152:e170504. doi: 10.1001/jamasurg.2017.0504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Alam A, Gomes T, Zheng H, Mamdani MM, Juurlink DN, Bell CM. Long-term analgesic use after low-risk surgery: a retrospective cohort study. Arch Intern Med. 2012;172:425–430. doi: 10.1001/archinternmed.2011.1827. [DOI] [PubMed] [Google Scholar]
- 34.Shah A, Hayes CJ, Martin BC. Characteristics of initial prescription episodes and likelihood of long-term opioid use: United States, 2006–2015. MMWR Morb Mortal Wkly Rep. 2017;66:265–269. doi: 10.15585/mmwr.mm6610a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Deyo RA, Hallvik SE, Hildebran C, Marino M, Dexter E, Irvine JM, et al. Association between initial opioid prescribing patterns and subsequent long-term use among opioid-naïve patients: a statewide retrospective cohort study. J Gen Intern Med. 2017;32:21–27. doi: 10.1007/s11606-016-3810-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wunsch H, Kahn JM, Kramer AA, Rubenfeld GD. Use of intravenous infusion sedation among mechanically ventilated patients in the United States. Crit Care Med. 2009;37:3031–3039. doi: 10.1097/CCM.0b013e3181b02eff. [DOI] [PubMed] [Google Scholar]
- 37.Shehabi Y, Bellomo R, Kadiman S, Ti LK, Howe B, Reade MC, et al. Sedation Practice in Intensive Care Evaluation (SPICE) Study Investigators and the Australian and New Zealand Intensive Care Society Clinical Trials Group. Sedation intensity in the first 48 hours of mechanical ventilation and 180-day mortality: a multinational prospective longitudinal cohort study. Crit Care Med. 2018;46:850–859. doi: 10.1097/CCM.0000000000003071. [DOI] [PubMed] [Google Scholar]
- 38.Mehta S, Burry L, Cook D, Fergusson D, Steinberg M, Granton J, et al. SLEAP Investigators; Canadian Critical Care Trials Group. Daily sedation interruption in mechanically ventilated critically ill patients cared for with a sedation protocol: a randomized controlled trial. JAMA. 2012;308:1985–1992. doi: 10.1001/jama.2012.13872. [DOI] [PubMed] [Google Scholar]
- 39.Riker RR, Shehabi Y, Bokesch PM, Ceraso D, Wisemandle W, Koura F, et al. SEDCOM (Safety and Efficacy of Dexmedetomidine Compared With Midazolam) Study Group. Dexmedetomidine vs midazolam for sedation of critically ill patients: a randomized trial. JAMA. 2009;301:489–499. doi: 10.1001/jama.2009.56. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.