Abstract Study Objectives Following extubation in the intensive care unit (ICU), upper airway (UA) edema and respiratory depressants may promote UA dysfunction. We tested the hypothesis that opioids increase the risk of sleep apnea early after extubation. Methods Fifty-six ICU patients underwent polysomnography the night after extubation. Airflow limitation during wakefulness was identified using bedside spirometry. Correlation and ordinal regression analyses were used to quantify the effects of preextubation opioid dose on postextubation apnea–hypopnea index (AHI) and severity of sleep apnea and whether or not inspiratory airway obstruction (ratio of maximum expiratory and inspiratory airflows at 50% of vital capacity [MEF50/MIF50] ≥ 1) during wakefulness predicts airway obstruction during sleep. Data were adjusted for age, gender, body mass index, as well as a generalized propensity score balanced for APACHE II, score for preoperative prediction of obstructive sleep apnea, duration of mechanical ventilation, chronic obstructive pulmonary disease, and a procedural severity score for morbidity. Results Sleep apnea (AHI ≥ 5) was present in 40 (71%) of the 56 patients. Morphine equivalent dose given 24 hours prior extubation predicted obstructive respiratory events during sleep (r = 0.35, p = .01) and sleep apnea (odds ratio [OR] 1.17; 95% confidence interval [CI] 1.02–1.34). Signs of inspiratory UA obstruction (MEF50/MIF50 ≥ 1) assessed by bedside spirometry were strongly associated with sleep apnea (OR 5.93; 95% CI 1.16–30.33). Conclusions High opioid dose given 24 hours prior to extubation increases the likelihood of postextubation sleep apnea in the ICU, particularly in patients with anatomical vulnerability following extubation. sleep apnea, opioids, polysomnography, intensive care unit, spirometry Statement of Significance Sleep apnea is a highly prevalent disorder associated with significant morbidity, particularly in the perioperative period. Opioids are frequently used for postoperative pain therapy and are known to have respiratory depressant effects. We show that opioids administered during mechanical ventilation increase the risk for sleep apnea the night following extubation in intensive care unit (ICU) patients. This is particularly significant for perioperative clinicians who should be aware of adverse opioids effects and their potential consequences in order to minimize the risk for respiratory complications in critically ill patients. INTRODUCTION Obstructive sleep apnea (OSA) affects approximately 13% of men and 6% of women in the general population1 and is associated with respiratory complications following surgery2 and cardiovascular death.3 OSA is often unidentified in the surgical population.4 Underlying mechanisms of OSA are multifactorial and include decreased motor drive to the upper airway (UA) dilator muscles during sleep5 in the context of an anatomical vulnerability to UA collapse.6 The perioperative period adds numerous factors that promote respiratory muscle dysfunction and UA dysfunction.7–9 Following extubation, anatomical UA obstruction may occur as a consequence of inflammation,10 tissue trauma (endotracheal tube), fluid overload,11 and airway secretions.12 In addition, opioids, which are frequently administered to facilitate mechanical ventilation, impair UA dilator muscle activity in a dose-dependent fashion.13,14 We tested the hypothesis that opioid dose administered prior to extubation predicts OSA during the night following extubation. Additionally, we aimed to identify patients with anatomic vulnerability for inspiratory flow limitation based on spirometry measurements. In exploratory analyses, we aimed to investigate factors that may contribute to postextubation sleep apnea such as a volume status (total body balance). METHODS In this prospective pharmacophysiological interaction study, we conducted polysomnography (PSG) during the first night following extubation after mechanical ventilation in intensive care unit (ICU) patients. The study was registered under clinicaltrials.gov (NCT02112604) and approved by The Institutional Review Board of Partners HealthCare (Protocol #2010P001919). We recruited adult, nonpregnant patients who were endotracheally intubated and mechanically ventilated for ≥24 hours in the surgical ICU at Massachusetts General Hospital, Boston, MA, USA. Patients who showed signs of delirium (positive Confusion Assessment Measurement–ICU score) or received continuous positive airway pressure (CPAP) treatment after extubation were excluded. Between November 2012 and January 2016, a total of 80 patients fulfilled our restrictive a priori defined criteria and gave consent for study participation. PSG (electroencephalography [EEG], electrooculography, submental and limb electromyography, abdominal and respiratory effort measurements, oxyhemoglobin saturation, and nasal respiratory airflow13,15) was performed during the first night after extubation, using a type 2 out-of-center device (Alice PDx, Philips Respironics Inc, Murrysville, PA). A member of the study team was available via pager throughout the study night to assist with technical problems. We performed bedside spirometry measurements to identify UA obstruction via flow-volume loops during wakefulness the day following the PSG study. Bedside spirometry (EasyOne Spirometer; ndd-Medizintechnik, Zuerich, Switzerland) was measured for three repetitive breathing trials during daytime by study staff members who were blinded to PSG results. In accordance with the 2005 ATS/ERS consensus statement,16 we required patients not to show signs and symptoms of dementia or confusional state as these conditions are unlikely to produce optimal or reproducible results of pulmonary function testing. Before performing spirometry measurements, we tested study patients for sufficient alertness and attention by using the Richmond Agitation-Sedation Scale (clinical signs of sedation) as well as a series of 1- and 2-step instructions to ensure the patient’s ability to follow the instructions given during measurements, as previously described.17 UA obstruction during wakefulness was defined as the ratio of maximum expiratory and inspiratory flows at 50% of vital capacity (MEF50/MIF50) ≥ 1 calculated from flow-volume loops.18 Baseline demographics, clinical characteristics and drug doses (opioids as morphine equivalents [mg],13 propofol [mg], multiples of 95-effective dose of non-depolarizing neuromuscular blocking agents, and benzodiazepines as midazolam equivalent [mg]19) were retrospectively collected from patient’s charts by blinded members of the study team (ES and SF). Sleep stages, respiratory events, and arousals from sleep were scored according to the 2007 American Academy of Sleep Medicine Guidelines by a blinded member of the study staff (SZ). Hypopneas, defined as ≥30% decrease in nasal flow curve for ≥ two consecutive breaths associated with a decrease in SpO2 of ≥3% ± arousal, were identified as obstructive (flattening and decrease in the amplitude of the flow curve with unchanged amplitude of respiratory effort signal), central (no flattening but decrease in amplitude of flow curve and respiratory effort signal, without signs of paradoxical breathing or snoring), or mixed (flattening of flow curve and decreased amplitude in respiratory effort channels) events. We performed EEG spectroscopy during sleep (fast Fourier transformation; FFT), as previously reported.13 Briefly, FFT was performed on all artifact-free EEG epochs visually scored as nonrapid eye movement (NREM) sleep during the entire recording. The average EEG power was calculated for β (12–30 Hz), α (8–13 Hz), θ (4–8 Hz), and δ (0.5–4 Hz) frequency bands within each individual patient, and compared between patients with and without sleep apnea. Statistical Analysis Spearman and Pearson correlation analyses were used to evaluate associations between exposures and outcomes for nonparametric and parametric data, as appropriate. We analyzed correlations between total morphine equivalent dose applied within 24 hours prior to extubation and the following outcomes: apnea–hypopnea index (AHI) (main outcome) and the number of central and obstructive respiratory events (apneas and hypopneas). Logarithmic transformation of the opioid dose data was used to achieve a normally distributed sample. We also analyzed correlations between doses of propofol, neuromuscular blocking agents, benzodiazepines and AHI, as well as MEF50/MIF50 and AHI. The association of 24-hour opioid dose and severity of OSA was further studied using multivariable ordinal regression analyses including the Brant test of parallel regression assumption. We calculated the odds of having moderate-to-severe sleep apnea (AHI ≥ 15) vs. mild sleep apnea (AHI 5 to <15), and mild vs. no sleep apnea (AHI<5) the night following extubation. The regression model was adjusted for age, gender, and body mass index. In addition, the model included a generalized propensity score20 in order to reduce bias induced by discrepancies in patient characteristics across different values of the hereby investigated continuous treatment regime (opioid dose) while accounting for the degrees of freedom available.20,21 This propensity score was generated from covariates identified as influencing the exposure (opioid administration prior to extubation) and consisted of the following: duration of mechanical ventilation, history of chronic obstructive pulmonary disease, score for preoperative prediction of obstructive sleep apnea (SPOSA),8 APACHE II, and the surgical Procedural Severity Score for postoperative morbidity.22 Sensitivity Analyses To account for a higher susceptibility for sleep apnea in patients who were diagnosed with OSA prior to study participation, we additionally adjusted for a prediagnosis of OSA. Short sleep duration might result in a false high AHI in polysomnographic recordings. In a subgroup analysis, we excluded patients with a total sleep time (TST) of <60 minutes. We also tested whether oxygen supplementation during the study night affects our results and added this variable as covariate to our model. Additional sensitivity analyses are presented in the online Supplementary material. Exploratory Analyses Fluid overload (a positive fluid balance—the difference in a patient’s total fluid intake vs. output [urine, wound/pleura secretions] over a specified time frame) has been recently associated with sleep apnea in a patient cohort with end-stage renal disease.11 In exploratory analyses, we analyzed the correlation between AHI and total body balance prior to extubation using Pearson correlation analysis. In an exploratory intent, we also investigated the association between duration of mechanical ventilation and postextubation AHI. Based on previous data published by our group,13,15 we expected a correlation coefficient of 0.4 between pre- extubation opioid dose and AHI. Considering an α-error of 0.05 and power of 80%, we calculated the required sample size to be 47 patients. Previous studies using PSG reported high drop-out rates;13,15,23,24 thus, we aimed to enroll 80 ICU patients. Statistical analyses were performed with Stata (version 13; StataCorp, College Station, TX). A two-tailed p-value of <.05 was considered statistically significant. RESULTS Fifty-six out of 80 eligible and enrolled patients successfully completed overnight PSG after extubation and were included in the primary analysis (Figure 1). Reasons for mechanical ventilation included postoperative airway protection (50%), aspiration/pneumonia (17.9%), respiratory insufficiency (23.2%), pneumothorax (5.4%), and decreased mental status (3.6%). The median mechanical ventilation time in our study was 2 (1–5) days. Baseline clinical characteristics of the study cohort according to the severity of postextubation sleep apnea are shown in Table 1. Table 1 Clinical Characteristics of Study Population. Variable All patients (n = 56) Patients without sleep apnea (n = 16) Patients with mild sleep apnea (n = 19) Patients with moderate-to-severe sleep apnea (n = 21) Gender female, No (%) 23 (41.1) 8 (50) 8 (42.1) 7 (33.3) Age (years), mean (SD) 62.5 (15.3) 59.7 (17.8) 64.2 (16) 63.2 (12.6) Body mass index (kg/m2), mean (SD) 29.1 (8.2) 31.4 (10.5) 25.1 (5.4) 30.9 (7.3) Prediagnosed OSA, No (%) 6 (10.7) 2 (12.5) 1 (5.2) 3 (14.3) History of chronic obstructive pulmonary disease, No (%) 11 (19.6) 3 (18.8) 5 (26.3) 3 (14.3) History of congestive heart failure, No (%) 19 (33.9) 5 (31.3) 4 (21.1) 10 (47.6) APACHE II score, median (IQR) 15.5 (8–21) 14.5 (6.5–21) 15 (9–21) 17 (8–21) SPOSA score, median (IQR) 22.5 (19–27) 23.5 (18–26.5) 19 (18–23) 24 (21–29) ASA status of ≥3, No (%) 49 (87.5) 16 (100) 15 (78.9) 18 (94.7) Surgical Patients, No (%) 51 (91.1) 13 (76.9) 18 (94.7) 20 (95.2) Procedure Severity Score for Morbidity, median (IQR) 54.6 (33.1–74.8) 45.6 (14.3–73.2) 55.6 (34.5–78.4) 61.4 (42.9–78.4) Total morphine equivalent dose given 24 hours prior to extubation (mg), median (IQR) 7.7 (0–45.8) 5.5 (0–24.2) 7.3 (0–40) 25.6 (3.8–82.4 Total propofol dose given 24 hours prior to extubation (mg), median (IQR) 1605 (1067.5–2930) 1337.5 (1002.5–4175) 1320 (0–1940) 2400 (1175–2970) Total NMBA dose 24 hours prior to extubation (multiples of ED95), median (IQR) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–1.7) Total midazolam equivalent dose 24 hours prior to extubation (mg), median (IQR) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–0) Duration of mechanical ventilation (days), median (IQR) 2 (1–5) 4 (2–8) 2 (2–6) 2 (1–3) Reason for mechanical ventilation, No (%) Postoperative airway protection 28 (50) 9 (56.3) 9 (47.4) 10 (47.6) Aspiration/pneumonia 10 (17.9) 3 (18.8) 2 (10.5) 5 (23.8) Respiratory insufficiency 13 (23.2) 4 (25) 7 (36.8) 2 (9.5) Pneumothorax 3 (5.4) 0 (0) 0 (0) 3 (14.3) Decrease in vigilance 2 (3.6) 0 (0) 1 (5.2) 1 (4.8) Variable All patients (n = 56) Patients without sleep apnea (n = 16) Patients with mild sleep apnea (n = 19) Patients with moderate-to-severe sleep apnea (n = 21) Gender female, No (%) 23 (41.1) 8 (50) 8 (42.1) 7 (33.3) Age (years), mean (SD) 62.5 (15.3) 59.7 (17.8) 64.2 (16) 63.2 (12.6) Body mass index (kg/m2), mean (SD) 29.1 (8.2) 31.4 (10.5) 25.1 (5.4) 30.9 (7.3) Prediagnosed OSA, No (%) 6 (10.7) 2 (12.5) 1 (5.2) 3 (14.3) History of chronic obstructive pulmonary disease, No (%) 11 (19.6) 3 (18.8) 5 (26.3) 3 (14.3) History of congestive heart failure, No (%) 19 (33.9) 5 (31.3) 4 (21.1) 10 (47.6) APACHE II score, median (IQR) 15.5 (8–21) 14.5 (6.5–21) 15 (9–21) 17 (8–21) SPOSA score, median (IQR) 22.5 (19–27) 23.5 (18–26.5) 19 (18–23) 24 (21–29) ASA status of ≥3, No (%) 49 (87.5) 16 (100) 15 (78.9) 18 (94.7) Surgical Patients, No (%) 51 (91.1) 13 (76.9) 18 (94.7) 20 (95.2) Procedure Severity Score for Morbidity, median (IQR) 54.6 (33.1–74.8) 45.6 (14.3–73.2) 55.6 (34.5–78.4) 61.4 (42.9–78.4) Total morphine equivalent dose given 24 hours prior to extubation (mg), median (IQR) 7.7 (0–45.8) 5.5 (0–24.2) 7.3 (0–40) 25.6 (3.8–82.4 Total propofol dose given 24 hours prior to extubation (mg), median (IQR) 1605 (1067.5–2930) 1337.5 (1002.5–4175) 1320 (0–1940) 2400 (1175–2970) Total NMBA dose 24 hours prior to extubation (multiples of ED95), median (IQR) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–1.7) Total midazolam equivalent dose 24 hours prior to extubation (mg), median (IQR) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–0) Duration of mechanical ventilation (days), median (IQR) 2 (1–5) 4 (2–8) 2 (2–6) 2 (1–3) Reason for mechanical ventilation, No (%) Postoperative airway protection 28 (50) 9 (56.3) 9 (47.4) 10 (47.6) Aspiration/pneumonia 10 (17.9) 3 (18.8) 2 (10.5) 5 (23.8) Respiratory insufficiency 13 (23.2) 4 (25) 7 (36.8) 2 (9.5) Pneumothorax 3 (5.4) 0 (0) 0 (0) 3 (14.3) Decrease in vigilance 2 (3.6) 0 (0) 1 (5.2) 1 (4.8) SD = standard deviation; OSA = obstructive sleep apnea; IQR = interquartile range; ASA = American Society of Anesthesiologists; NMBA = neuromuscular blocking agents. View Large Figure 1 View largeDownload slide Study flow. Figure 1 View largeDownload slide Study flow. Sleep Apnea and Opioid Effects Sleep apnea (AHI ≥ 5/hour) was present in 40 patients (71.4%) during the first night after extubation (AHI 20.6 ± 16.2/hour). Nineteen (33.9%) patients had mild sleep apnea (AHI 5 to <15; AHI 9.2 ± 3.5/hour) and 21 patients (37.5%) had moderate-to-severe sleep apnea (AHI ≥ 15; AHI 28.6 ± 16.5/hour, Tables 1 and 2). Of the 40 patients with sleep apnea, 35 predominantly had obstructive events and three patients had central sleep apnea. A predominant apnea type could not be determined in two patients due to poor signal quality from the respiratory belt. The median numbers of total respiratory events (apnea and hypopneas) are shown in Table 2. Table 2 Respiratory and Sleep Parameters During Sleep the Night Following Extubation, by Categories of Severity of Sleep Apnea. Variable All Patients (n = 56) Patients without sleep apnea (n = 16) Patients with mild sleep apnea (n = 19) Patients with moderate-to-severe sleep apnea (n = 21) Mean SpO2, (%), mean (SD) 94.3 (3.7) 94.5 (4.9) 95.3 (2.4) 93.3 (3.7) Lowest SpO2, (%), median (IQR) 86.5 (82–90) 88 (82–90) 86 (79–89) 87 (83–89) Desaturation index, median (IQR) 1.7 (0.4–11.8) 1.7 (0.3–13) 1.2 (0.3–14.4) 2.7 (0.4–5) Apnea–hypopnea index, median (IQR) 11.4 (2.7–21) 2.2 (0.8–2.3) 9.2 (5.9–13.1) 23.2 (19–32.2) Obstructive respiratory event index, median (IQR) 7.3 (1.8–16.8) 0.9 (0.6–1.5) 6.1 (3.3–10.4) 10.5 (16.2–20.2) Central respiratory event index, median (IQR) 1.5 (0.4–3.3) 0.2 (0–0.7) 1.8 (0.7–2.7) 3.9 (1.9–8.3) Mixed respiratory event index, median (IQR) 0.3 (0–1.2) 0 (0–0) 0.3 (0–0.6) 1.3 (0.5–2.8) Total sleep time (minutes), mean (SD) 264.1 (126.4) 246.7 (126.2) 279.1 (130.7) 263.1 (127.3) Time spent in non-REM sleep stage 1 (% of TST), mean (SD) 27.6 (22) 29.7 (21.8) 22.7 (20.5) 30.4 (23.7) Time spent in non-REM sleep stage 2 (% of TST), mean (SD) 62.7 (19.8) 57 (20.4) 67.5 (18.2) 62.6 (20.4) Time spent in non-REM sleep stage 3 (% of TST), mean (SD) 8.8 (14.4) 13.1 (20.1) 8.8 (13.3) 5.6 (9.2) Time spent in REM sleep (% of TST), mean (SD) 0.8 (3.3) 0.2 (0.6) 0.9 (2.6) 1.3 (4.8) Arousal index (number of arousals per hour), median (IQR) 12.6 (5.8–20.2) 17 (8–31.2) 8.9 (3.6–14.3) 14.5 (6.3–28.8) Variable All Patients (n = 56) Patients without sleep apnea (n = 16) Patients with mild sleep apnea (n = 19) Patients with moderate-to-severe sleep apnea (n = 21) Mean SpO2, (%), mean (SD) 94.3 (3.7) 94.5 (4.9) 95.3 (2.4) 93.3 (3.7) Lowest SpO2, (%), median (IQR) 86.5 (82–90) 88 (82–90) 86 (79–89) 87 (83–89) Desaturation index, median (IQR) 1.7 (0.4–11.8) 1.7 (0.3–13) 1.2 (0.3–14.4) 2.7 (0.4–5) Apnea–hypopnea index, median (IQR) 11.4 (2.7–21) 2.2 (0.8–2.3) 9.2 (5.9–13.1) 23.2 (19–32.2) Obstructive respiratory event index, median (IQR) 7.3 (1.8–16.8) 0.9 (0.6–1.5) 6.1 (3.3–10.4) 10.5 (16.2–20.2) Central respiratory event index, median (IQR) 1.5 (0.4–3.3) 0.2 (0–0.7) 1.8 (0.7–2.7) 3.9 (1.9–8.3) Mixed respiratory event index, median (IQR) 0.3 (0–1.2) 0 (0–0) 0.3 (0–0.6) 1.3 (0.5–2.8) Total sleep time (minutes), mean (SD) 264.1 (126.4) 246.7 (126.2) 279.1 (130.7) 263.1 (127.3) Time spent in non-REM sleep stage 1 (% of TST), mean (SD) 27.6 (22) 29.7 (21.8) 22.7 (20.5) 30.4 (23.7) Time spent in non-REM sleep stage 2 (% of TST), mean (SD) 62.7 (19.8) 57 (20.4) 67.5 (18.2) 62.6 (20.4) Time spent in non-REM sleep stage 3 (% of TST), mean (SD) 8.8 (14.4) 13.1 (20.1) 8.8 (13.3) 5.6 (9.2) Time spent in REM sleep (% of TST), mean (SD) 0.8 (3.3) 0.2 (0.6) 0.9 (2.6) 1.3 (4.8) Arousal index (number of arousals per hour), median (IQR) 12.6 (5.8–20.2) 17 (8–31.2) 8.9 (3.6–14.3) 14.5 (6.3–28.8) No sleep apnea (apnea–hypopnea index, AHI < 5), mild sleep apnea (AHI 5–15), moderate-to-severe sleep apnea (AHI > 15). SD = standard deviation; IQR = interquartile range; REM = rapid eye movement; TST = total sleep time. View Large The median opioid dose administered in our study cohort within 24 hours prior to extubation was 7.7 (0–45.8) mg morphine equivalent. Patients with moderate-to-severe sleep apnea received the highest doses of opioids 25.6 (3.8–82.4) mg/24 hours, whereas patients with mild sleep apnea received 7.3 (0–40) mg/24 hours and patients without sleep apnea received 5.5 (0–24.2) mg/24 hours (Table 1). According to the SPOSA,8 18 (32.1%) patients were at high risk (SPOSA ≥ 24) for OSA, whereas 38 (67.9%) patients were in the low risk group (SPOSA < 24). Despite having a baseline low risk for sleep apnea, 27 developed sleep apnea (AHI ≥ 5) postextubation. Six patients were diagnosed with OSA prior to study participation, and four of those demonstrated sleep apnea postextubation in the ICU. Patients with known OSA received 50.3 (6.7–196.5) mg morphine equivalent and patients without a prediagnosis 6.7 (0–40) mg (marginally significant p = .08). The logarithm of opioid dose given within 24 hours prior to extubation significantly correlated with postextubation AHI (r = 0.34; p = .01, Figure 2) and number of obstructive (r = 0.35; p = .01), but not with central (r = 0.19; p = .15) or mixed (r = 0.13; p = .34) respiratory events during sleep Figure 2 View largeDownload slide Correlation between postextubation AHI and (A) total morphine equivalent dose (mg) within 24 hours prior to extubation; (B) ratio between maximum expiratory flow and maximum inspiratory flow at 50% of vital capacity (MEF50/MIF50). Figure 2 View largeDownload slide Correlation between postextubation AHI and (A) total morphine equivalent dose (mg) within 24 hours prior to extubation; (B) ratio between maximum expiratory flow and maximum inspiratory flow at 50% of vital capacity (MEF50/MIF50). In multivariable ordinal regression analyses, each 10 mg morphine equivalent dose increased the odds of having more severe sleep apnea (moderate-to-severe sleep apnea vs. mild sleep apnea, and mild sleep apnea vs. no sleep apnea) following extubation by 15% (odds ratio [OR] 1.15; 95% confidence interval [CI] 1.01–1.32). This effect remained stable when the regression model was additionally adjusted for a prediagnosis of OSA. In brant test, neither the exposure nor any covariates were significant, indicating no violation of the proportional odds assumption. Sensitivity Analyses After excluding four patients with TST < 60 minutes, we confirmed our findings of a significant correlation between log transformed 24-hour preextubation opioid dose and postextubation AHI (r = 0.37; p = .01), as well as obstructive respiratory events (r = 0.37; p = .01), but no correlation with central and mixed events. The odds for more severe sleep apnea were also increased by opioid dose (per 10 mg morphine equivalent: OR 1.16; 95% CI 1.01–1.34). The unique association between opioid dose and AHI remained stable when oxygen supplementation (OR 1.15; 95% CI 1.01–1.32) was added as a potential confounder to the model. Propofol, benzodiazepines, and neuromuscular blocking agents given prior to extubation did not correlate with AHI during the first night following extubation (p > .33). UA Obstruction and Postextubation AHI Pulmonary function was assessed in 31 patients using bedside spirometry measurements one day after extubation. The remaining patients were not able to complete spirometry due to difficulties performing that test, poor functional status, or interim development of delirium. Flow limitation consistent with UA obstruction (defined as MEF50/MIF50 ≥ 1) was present in 10 patients (32.3%) during wakefulness. MEF50/MIF50 strongly correlated with AHI (ρ = 0.59; p < .001, Figure 2) as well as obstructive (ρ = 0.46; p = .01), central (ρ = 0.52; p = .003), and mixed (ρ = 0.4; p = .02) respiratory events during sleep. In multivariable ordinal regression analysis, MEF50/MIF50≥1 was a strong, independent predictor for sleep apnea (OR 6.38; 95% CI 1.21–33.56). Exploratory Analyses Data on cumulative total fluid balance (mL) within 72 hours of extubation were available for 35 patients. Postextubation AHI was associated with total fluid balance within 72 hours of extubation (r = 0.42; p = .01). Duration of mechanical ventilation was inversely associated with postextubation AHI (ρ = −0.29; p = .03) and doses of sedatives given within 24 hours prior to extubation decreased with increasing duration spent on the ventilator (24-hour propofol dose: ρ = −0.35, p < .001; midazolam equivalent dose: ρ = −0.27, p = .04). Sleep Architecture On average, study patients slept 264.1 ± 126.4 minutes (sleep efficiency 59%) and predominantly in NREM sleep stage 1 (N1; 27.6 ± 22% of TST) and NREM sleep stage 2 (N2; 62.7 ± 19.8%). Only six study patients (10.7%) reached REM sleep. Sleep cycles were fragmented with an increased number of arousals from sleep (Table 2). Cortical EEG power spectrum during sleep did not significantly differ between patients with and without sleep apnea (Figure 3). Figure 3 View largeDownload slide Average cortical EEG power spectra during sleep in patients with vs. without postextubation sleep apnea. EEG = electroencephalography. Figure 3 View largeDownload slide Average cortical EEG power spectra during sleep in patients with vs. without postextubation sleep apnea. EEG = electroencephalography. DISCUSSION In this pharmacophysiological interaction study, we demonstrate that the opioid dose given within 24 hours prior to extubation predicts an increased risk of sleep apnea after extubation. Inspiratory flow limitation identified after extubation in the ICU during wakefulness was strongly associated with an increased risk of sleep apnea. To the best of our knowledge, we are the first to report data on sleep apnea measured by PSG in ICU patients early after extubation, compared with previous studies performed in cohorts of non-ICU patients. Sleep apnea (AHI ≥ 5) was present in 71% of ICU patients during the first night following extubation. Similar prevalences of sleep apnea were found in 15 surgical ICU patients mechanically ventilated for ≥48 hours25 and 25% of surgical patients without preoperative sleep apnea developed moderate-to-severe sleep-disordered breathing postoperatively.23 Our data support that undiagnosed sleep apnea frequently occurs after surgery4 and suggests that routine analgesics are associated with OSA early after mechanical ventilation. Most patients in our study were at low preoperative risk (SPOSA < 24) for OSA, according to a most recently developed SPOSA.8 Despite being at a baseline low risk, 27 patients developed sleep apnea following extubation, which underscores the impact of opioid effects on the UA during sleep in vulnerable ICU patients in our study. We observed that patients in our study with a prediagnosis of sleep apnea received a much larger amount of opioids within 24 hours prior to extubation (50.3 [6.7–196.5] vs. 6.7 [0–40] mg; marginally significant—p = .08). Data indicate that patients with sleep apnea have a lower pain tolerance and require high doses of analgesics and opioids to accomplish adequate analgesia, compared with patients without sleep apnea.26 Several factors during the perioperative period promote UA collapsibility and thus may increase the likelihood for sleep apnea.7,9 These factors include medication effects, limited tracheal traction, supine body position, and a hypervolemic state. The latter factor has been shown to be relevant in studies of patients with end-stage renal disease11 and is also supported by our finding that total fluid balance within 72 hours prior to extubation correlated with AHI. The pathophysiology of sleep apnea after mechanical ventilation in ICU patients may therefore be similar to sleep apnea during the perioperative period. Sleep apnea induced by acute opioid treatment has widely been attributed to the effects of opioid on central respiratory depression.27 However, we found that opioid administration during mechanical ventilation in ICU patients predicts obstructive rather than central respiratory events during sleep early after extubation. This could be the consequence of drug-induced decrease of UA dilator muscle and/or respiratory pump muscle activation.14,27 The latter would decrease expiratory lung volume and destabilize airway patency. In our cohort, opioids dose-dependently increased the risk for more severe sleep apnea per 10 mg morphine equivalent by 15%. Acute effects of opioids on airway patency have been reported during a 12-hour postoperative period in patients who were not critically ill but at high risk for OSA.28 Postoperative opioids—even when applied in low doses—increase pharyngeal collapsibility in patients with and without OSA and thus promote sleep-disordered breathing.13,23 Our group recently showed that AHI dose-dependently increases by opioid dose given for postoperative pain therapy in the early postoperative period.13 The degree of opioid-induced respiratory depression was successfully mitigated by CPAP.13 According to these data, we speculate that noninvasive ventilation after extubation in the ICU may also help us to decrease opioid-induced obstructive events. Opioid-induced adverse effects on UA patency are most relevant during sleep.13 Our data suggest that a combination of opioid-induced airway dysfunction and enhanced anatomic vulnerability for airway obstruction synergistically leads to sleep apnea after extubation in ICU patients, which could be treated with CPAP.13 We found that the observed anatomical vulnerability for increased pharyngeal collapsibility during wakefulness predicts collapsibility during sleep.29 This may in part be the consequence of inflammatory responses from UA injury in the setting of prolonged mechanical ventilation.30 A recent study that was conducted 5 and 60 minutes following extubation in a pediatric population reported that postextubation UA obstruction was associated with high rates of subsequent reintubation (13.6%).31 Future studies are needed to understand the mechanisms of UA obstruction on extubation failure in the adult population. Propofol dose-dependently increases UA collapsibility.32 In our study, propofol dose was not associated with AHI. Propofol sedation may be a good alternative to opioid therapy to achieve patient’s comfort during mechanical ventilation in patients at risk of postoperative sleep apnea. Our findings support clinical guidelines that suggest minimizing opioid use in OSA patients.33 However, current data about the clinical implications of sleep apnea on postoperative outcomes are equivocal.8,34–36 Longer duration of mechanical ventilation was associated with lower rates of postextubation sleep apnea in our study. Patients who spent more time on the ventilator received lower doses of sedatives prior to extubation. Based on these data, we speculate that patients with longer intubation times may have tolerated the endotracheal tube at lower sedation levels. Disrupted and nonrestorative sleep is common in ICU patients24,37 and believed to be associated with adverse outcomes.38 In our study, average TST was 264 ± 126 minutes spent mainly in sleep stages N1 and N2, whereas slow-wave-sleep and REM sleep were drastically reduced, which is in accordance with previously reported finding.24,37 Contributing factors include the noise of the ICU environment, medication, and postoperative pain.36,39 OSA is associated with delirium40 and both can be induced by opioid infusions.38 We excluded patients with a positive CAM-ICU since we believed that delirium-associated agitation would be critical to the successful completion of the PSG study. Future studies are needed to investigate the interplay among opioid administration, delirium, and sleep apnea.41 Strengths and Limitations A strength of this study was the combined use of spirometry during wakefulness and PSG during sleep in a cohort of highly vulnerable and critically ill ICU patients. PSG in the ICU after mechanical ventilation is difficult to conduct as EEG, electromyography electrodes, and chest belts induce discomfort. In addition, sedatives and/or postoperative hyperactive delirium may affect polysomnographic and spirometric measurements. We used rigorous quality control criteria to minimize these effects on the data quality presented in this study. Spirometry was performed in a standardized sitting position with an angle of legs and upper body between 75 and 85 degrees. Flow-volume curves of all patients were visually displayed and checked for complete performance of inspiratory and expiratory maneuvers. The device that was used in the present study is also equipped with an internal quality assessment function that requires at least three acceptable tests and the difference between the best two FEV1 and FVC values to be equal to or less than 150 mL. In order to discriminate expected impairment of pulmonary function testing from obstruction, we used the MEF50/MIF50 ratio, which is thought to provide information on inspiratory flow limitation independent from restrictive pulmonary impairment.18,42 We therefore believe that a patient’s physical limitation potentially influencing the spirometry maneuver in our study cohort is not a bias but represents the true finding of poor spirometry results. We performed PSG for the assessment of sleep and sleep- disordered breathing during the nighttime. Study patients slept on average for about 4.5 hours mostly in non-REM sleep stages and only one of 10 patients reached REM sleep. In addition, patients showed a high number of arousals from sleep and fragmented sleep cycles, as has been observed previously in the ICU setting.43,44 However, sleep fragmentation and sleep architecture in the ICU were beyond the scope of this study. We evaluated the effects of sedatives on breathing and EEG during the study night.45,46 The observed δ-activity was higher than expected during normal sleep in both groups. Sedative agents increased cortical δ-EEG power in previous studies. However, in our study, cortical EEG power spectrum did not differ between patients with and without sleep apnea, suggesting that the unique association between opioid dose and AHI was not just a representation of different levels of EEG power.47 In our cohort of critically ill patients, supplemental oxygen was applied during the study night, which was shown to improve desaturation index and AHI by a recent study.48 Oxygen supplementation might have mitigated the association between opioid dose and AHI in a less extensive effect. Nevertheless, our results remained stable after adjusting our analysis for oxygen supplementation. CONCLUSIONS In this pharmacophysiological interaction study, we demonstrate that opioids received within the final 24 hours of mechanical ventilation increase the likelihood of sleep apnea in surgical patients during the first night after prolonged mechanical ventilation. Bedside spirometry identified patients with an anatomic vulnerability (inspiratory flow limitation) during wakefulness, which was associated with sleep apnea. SUPPLEMENTARY MATERIAL Supplementary material is available at SLEEP online. FUNDING This study was supported by an unrestricted research fund from Jeff and Judy Buzen to Matthias Eikermann. AUTHOR CONTRIBUTIONS FPT and SeZ contributed to the design and conception of the study, the acquisition, analysis and interpretation of data, and drafting of the first version of the manuscript. SDG, HNF, StZ, and CHS contributed in the acquisition, analysis, and interpretation of the data and provided valuable input in the preparation of the final version of the manuscript. JEM, SKR, and TK contributed valuable input to data analysis and helped us to prepare the final version of the manuscript. ES, SM, and SF contributed in the acquisition and provided valuable input in the preparation of the final version of the manuscript. ME is the guarantor of the paper. He takes responsibility for the integrity of the work as a whole, from inception to published article. All authors revised the work critically for important intellectual content. ME served as a clinical research mentor to FPT, SZ, and SDG. WORK PERFORMED Massachusetts General Hospital, 55 Fruit Street, Boston, MA 02114, USA. Clinical Trials Registration: NCT02112604 DISCLOSURE STATEMENT FPT, SeZ, SDG, HNF, StZ, ES, CHS, SM, SF, and JEM report no conflict of interest. SKR has received industry funding from Merck in 2014 for research on sleep apnea and early postoperative desaturation. TK has within the past 2 years received investigator-initiated research funding from the French National Research Agency and the US National Institutes of Health. Furthermore, he has received honorariums from the British Medical Journal and Cephalalgia for editorial services. ME has within the past 2 years received investigator-initiated research funding from Merck and the Buzen Foundation. Furthermore, he has received honorariums from Anesthesiology for editorial services. There are no other relationships or activities that could appear to have influenced the submitted work. ACKNOWLEDGMENTS The guarantor (Matthias Eikermann) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. FPT and SZ contributed equally to this work. REFERENCES 1. Peppard PE, Young T, Barnet JH, Palta M, Hagen EW, Hla KM. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol . 2013; 177( 9): 1006– 1014. Google Scholar CrossRef Search ADS PubMed 2. Memtsoudis S, Liu SS, Ma Yet al. Perioperative pulmonary outcomes in patients with sleep apnea after noncardiac surgery. Anesth Analg . 2011; 112( 1): 113– 121. Google Scholar CrossRef Search ADS PubMed 3. Punjabi NM, Caffo BS, Goodwin JLet al. Sleep-disordered breathing and mortality: a prospective cohort study. PLoS Med . 2009; 6( 8): e1000132. Google Scholar CrossRef Search ADS PubMed 4. Finkel KJ, Searleman AC, Tymkew Het al. Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center. Sleep Med . 2009; 10( 7): 753– 758. Google Scholar CrossRef Search ADS PubMed 5. Jordan AS, McSharry DG, Malhotra A. Adult obstructive sleep apnoea. Lancet . 2014; 383( 9918): 736– 747. Google Scholar CrossRef Search ADS PubMed 6. Schwab RJ, Pasirstein M, Pierson Ret al. Identification of upper airway anatomic risk factors for obstructive sleep apnea with volumetric magnetic resonance imaging. Am J Respir Crit Care Med . 2003; 168( 5): 522– 530. Google Scholar CrossRef Search ADS PubMed 7. Sasaki N, Meyer MJ, Eikermann M. Postoperative respiratory muscle dysfunction: pathophysiology and preventive strategies. Anesthesiology . 2013; 118( 4): 961– 978. Google Scholar CrossRef Search ADS PubMed 8. Shin CH, Grabitz SD, Timm FPet al. Development and validation of a score for preoperative prediction of obstructive sleep apnea (SPOSA) and its perioperative outcomes. BMC Anesthesiol . 2017; 17( 1): 71. Google Scholar CrossRef Search ADS PubMed 9. Zaremba S, Mojica JE, Eikermann M. Perioperative sleep apnea: a real problem or did we invent a new disease? [version 1; referees: 2 approved]. F1000Res 2016;5(F1000 Faculty Rev):48. 10. Vicente E, Marin JM, Carrizo SJet al. Upper airway and systemic inflammation in obstructive sleep apnoea. Eur Respir J . 2016; 48( 4): 1108– 1117. Google Scholar CrossRef Search ADS PubMed 11. Lyons OD, Inami T, Perger E, Yadollahi A, Chan CT, Bradley TD. The effect of fluid overload on sleep apnoea severity in haemodialysis patients. Eur Respir J . 2017; 49( 4). 12. Isono S, Tanaka A, Tagaito Y, Sho Y, Nishino T. Pharyngeal patency in response to advancement of the mandible in obese anesthetized persons. Anesthesiology . 1997; 87( 5): 1055– 1062. Google Scholar CrossRef Search ADS PubMed 13. Zaremba S, Shin CH, Hutter MMet al. Continuous positive airway pressure mitigates opioid-induced worsening of sleep-disordered breathing early after bariatric surgery. Anesthesiology . 2016; 125( 1): 92– 104. Google Scholar CrossRef Search ADS PubMed 14. Montandon G, Qin W, Liu H, Ren J, Greer JJ, Horner RL. PreBotzinger complex neurokinin-1 receptor-expressing neurons mediate opioid-induced respiratory depression. J Neurosci . 2011; 31( 4): 1292– 1301. Google Scholar CrossRef Search ADS PubMed 15. Zaremba S, Mueller N, Heisig AMet al. Elevated upper body position improves pregnancy-related OSA without impairing sleep quality or sleep architecture early after delivery. Chest . 2015; 148( 4): 936– 944. Google Scholar CrossRef Search ADS PubMed 16. Miller MR, Crapo R, Hankinson Jet al. General considerations for lung function testing. Eur Respir J . 2005; 26( 1): 153– 161. Google Scholar CrossRef Search ADS PubMed 17. Lee JJ, Waak K, Grosse-Sundrup Met al. Global muscle strength but not grip strength predicts mortality and length of stay in a general population in a surgical intensive care unit. Phys Ther . 2012; 92( 12): 1546– 1555. Google Scholar CrossRef Search ADS PubMed 18. Rotman HH, Liss HP, Weg JG. Diagnosis of upper airway obstruction by pulmonary function testing. Chest . 1975; 68( 6): 796– 799. Google Scholar CrossRef Search ADS PubMed 19. Barr J, Zomorodi K, Bertaccini EJ, Shafer SL, Geller E. A double-blind, randomized comparison of i.v. lorazepam versus midazolam for sedation of ICU patients via a pharmacologic model. Anesthesiology . 2001; 95( 2): 286– 298. Google Scholar CrossRef Search ADS PubMed 20. Hirano K, Imbens GW. “ The propensity score with continuous treatments” In: Gelman A, Ming XL, eds. Applied Bayesian modeling and Causal Inference from Incomplete-Data Perspectives: An Essential Journey with Donald Rubin’s Statistical Family . New York: John Wiley & Sons, Ltd; 2004: 73– 84. Google Scholar CrossRef Search ADS 21. Timm FP, Houle TT, Grabitz SDet al. Migraine and risk of perioperative ischemic stroke and hospital readmission: hospital based registry study. BMJ . 2017; 356: i6635. Google Scholar CrossRef Search ADS PubMed 22. Dalton JE, Kurz A, Turan A, Mascha EJ, Sessler DI, Saager L. Development and validation of a risk quantification index for 30-day postoperative mortality and morbidity in noncardiac surgical patients. Anesthesiology . 2011; 114( 6): 1336– 1344. Google Scholar CrossRef Search ADS PubMed 23. Chung F, Liao P, Elsaid H, Shapiro CM, Kang W. Factors associated with postoperative exacerbation of sleep-disordered breathing. Anesthesiology . 2014; 120( 2): 299– 311. Google Scholar CrossRef Search ADS PubMed 24. Elliott R, McKinley S, Cistulli P, Fien M. Characterisation of sleep in intensive care using 24-hour polysomnography: an observational study. Crit Care . 2013; 17( 2): R46. Google Scholar CrossRef Search ADS PubMed 25. Chishti A, Batchelor AM, Bullock RE, Fulton B, Gascoigne AD, Baudouin SV. Sleep-related breathing disorders following discharge from intensive care. Intensive Care Med . 2000; 26( 4): 426– 433. Google Scholar CrossRef Search ADS PubMed 26. Doufas AG, Tian L, Padrez KAet al. Experimental pain and opioid analgesia in volunteers at high risk for obstructive sleep apnea. PLoS One . 2013; 8( 1): e54807. Google Scholar CrossRef Search ADS PubMed 27. Mogri M, Khan MIA, Grant BJB, Mador MJ. Central sleep apnea induced by acute ingestion of opioids. Chest . 2008; 133( 6): 1484– 1488. Google Scholar CrossRef Search ADS PubMed 28. Blake DW, Yew CY, Donnan GB, Williams DL. Postoperative analgesia and respiratory events in patients with symptoms of obstructive sleep apnoea. Anaesth Intensive Care . 2009; 37( 5): 720– 725. Google Scholar PubMed 29. Malhotra A, Pillar G, Fogel R, Beauregard J, Edwards J, White DP. Upper-airway collapsibility: measurements and sleep effects. Chest . 2001; 120( 1): 156– 161. Google Scholar CrossRef Search ADS PubMed 30. Avramescu S, Wang DS, Lecker Iet al. Inflammation increases neuronal sensitivity to general anesthetics. Anesthesiology . 2016; 124( 2): 417– 427. Google Scholar CrossRef Search ADS PubMed 31. Khemani RG, Hotz J, Morzov Ret al. Evaluating risk factors for pediatric post-extubation upper airway obstruction using a physiology-based tool. Am J Respir Crit Care Med . 2016; 193( 2): 198– 209. Google Scholar CrossRef Search ADS PubMed 32. Simons JC, Pierce E, Diaz-Gil Det al. Effects of depth of propofol and sevoflurane anesthesia on upper airway collapsibility, respiratory genioglossus activation, and breathing in healthy volunteers. Anesthesiology . 2016; 125( 3): 525– 534. Google Scholar CrossRef Search ADS PubMed 33. Chung F, Memtsoudis SG, Ramachandran SKet al. Society of anesthesia and sleep medicine guidelines on preoperative screening and assessment of adult patients with obstructive sleep apnea. Anesth Analg . 2016; 123( 2): 452– 473. Google Scholar CrossRef Search ADS PubMed 34. Memtsoudis SG, Stundner O, Rasul Ret al. The impact of sleep apnea on postoperative utilization of resources and adverse outcomes. Anesth Analg . 2014; 118( 2): 407– 418. Google Scholar CrossRef Search ADS PubMed 35. Mokhlesi B, Hovda MD, Vekhter B, Arora VM, Chung F, Meltzer DO. Sleep-disordered breathing and postoperative outcomes after elective surgery: analysis of the nationwide inpatient sample. Chest . 2013; 144( 3): 903– 914. Google Scholar CrossRef Search ADS PubMed 36. Hardin KA. Sleep in the ICU: potential mechanisms and clinical implications. Chest . 2009; 136( 1): 284– 294. Google Scholar CrossRef Search ADS PubMed 37. Gabor JY, Cooper AB, Crombach SAet al. Contribution of the intensive care unit environment to sleep disruption in mechanically ventilated patients and healthy subjects. Am J Respir Crit Care Med . 2003; 167( 5): 708– 715. Google Scholar CrossRef Search ADS PubMed 38. Kamdar BB, Niessen T, Colantuoni Eet al. Delirium transitions in the medical ICU: exploring the role of sleep quality and other factors. Crit Care Med . 2015; 43( 1): 135– 141. Google Scholar CrossRef Search ADS PubMed 39. Elliott R, Rai T, McKinley S. Factors affecting sleep in the critically ill: an observational study. J Crit Care . 2014; 29( 5): 859– 863. Google Scholar CrossRef Search ADS PubMed 40. Bateman BT, Eikermann M. Obstructive sleep apnea predicts adverse perioperative outcome: evidence for an association between obstructive sleep apnea and delirium. Anesthesiology . 2012; 116( 4): 753– 755. Google Scholar CrossRef Search ADS PubMed 41. Flannery AH, Oyler DR, Weinhouse GL. The impact of interventions to improve sleep on delirium in the ICU: a systematic review and research framework. Crit Care Med . 2016; 44( 12): 2231– 2240. Google Scholar CrossRef Search ADS PubMed 42. Melissant CF, Lammers JW, Demedts M. Rigid external resistances cause effort dependent maximal expiratory and inspiratory flows. Am J Respir Crit Care Med . 1995; 152( 5 Pt 1): 1709– 1712. Google Scholar CrossRef Search ADS PubMed 43. Aurell J, Elmqvist D. Sleep in the surgical intensive care unit: continuous polygraphic recording of sleep in nine patients receiving postoperative care. Br Med J (Clin Res Ed) . 1985; 290( 6474): 1029– 1032. Google Scholar CrossRef Search ADS PubMed 44. Freedman NS, Gazendam J, Levan L, Pack AI, Schwab RJ. Abnormal sleep/wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit. Am J Respir Crit Care Med . 2001; 163( 2): 451– 457. Google Scholar CrossRef Search ADS PubMed 45. Eastwood PR, Platt PR, Shepherd K, Maddison K, Hillman DR. Collapsibility of the upper airway at different concentrations of propofol anesthesia. Anesthesiology . 2005; 103( 3): 470– 477. Google Scholar CrossRef Search ADS PubMed 46. Eikermann M, Eckert DJ, Chamberlin NLet al. Effects of pentobarbital on upper airway patency during sleep. Eur Respir J . 2010; 36( 3): 569– 576. Google Scholar CrossRef Search ADS PubMed 47. Boly M, Moran R, Murphy Met al. Connectivity changes underlying spectral EEG changes during propofol-induced loss of consciousness. J Neurosci . 2012; 32( 20): 7082– 7090. Google Scholar CrossRef Search ADS PubMed 48. Liao P, Wong J, Singh Met al. Postoperative oxygen therapy in patients with OSA: a randomized controlled trial. Chest . 2017; 151( 3): 597– 611. Google Scholar CrossRef Search ADS PubMed © Sleep Research Society 2017. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail firstname.lastname@example.org.
SLEEP – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera