Modes of mechanical ventilation vary between hospitals and intensive care units within a university healthcare system: a retrospective observational study
Jabaley et al. BMC Res Notes
Modes of mechanical ventilation vary between hospitals and intensive care units within a university healthcare system: a retrospective observational study
Craig S. Jabaley 0 2
Robert F. Groff 0 2
Milad Sharifpour 0
Jayashree K. Raikhelkar 0 2
James M. Blum 0 1 2
0 Division of Critical Care Medicine, Department of Anesthesiology, Emory University , 1364 Clifton Road NE, Atlanta, GA 30322 , USA
1 Department of Biomedical Informatics, Emory University School of Medicine , Atlanta, GA , USA
2 Division of Critical Care Medicine, Anesthesiology Service Line, Atlanta Veterans Affairs Medical Center , Decatur, GA , USA
Objective: As evidence-based guidance to aid clinicians with mechanical ventilation mode selection is scant, we sought to characterize the epidemiology thereof within a university healthcare system and hypothesized that nonconforming approaches could be readily identified. We conducted an exploratory retrospective observational database study of routinely recorded mechanical ventilation parameters between January 1, 2010 and December 31, 2016 from 12 intensive care units. Mode epoch count proportions were examined using Chi squared and Fisher exact tests as appropriate on an inter-unit basis with outlier detection for two test cases via post hoc pairwise analyses of a binomial regression model. Results: Final analysis included 559,734 mode epoch values. Significant heterogeneity was demonstrated between individual units (P < 0.05 for all comparisons). One unit demonstrated heightened utilization of high-frequency oscillatory ventilation, and three units demonstrated frequent synchronized intermittent mandatory ventilation utilization. Assist control ventilation was the most commonly recorded mode (51%), followed by adaptive support ventilation (23.1%). Volume-controlled modes were about twice as common as pressure-controlled modes (64.4% versus 35.6%). Our methodology provides a means by which to characterize the epidemiology of mechanical ventilation approaches and identify nonconforming practices. The observed variability warrants further clinical study about contributors and the impact on relevant outcomes.
Positive pressure ventilation; Respiratory failure; Mechanical ventilators; Noninvasive ventilation; Ventilator weaning; Intensive care
Relatively scant evidence exists to guide clinicians in
their selection of a mechanical ventilation (MV) mode.
International epidemiological studies have identified
that tidal volume (Vt), positive end-expiratory pressure
(PEEP), and other parameters are beginning to align with
lung protective ventilation strategies [
]. These and
other characterizations of MV mode selection are
variable in their scope, often examining only certain modes
or patient subsets, and a clear picture of practice patterns
remains elusive [
Developing a means by which to assess variability in
the approach to MV offers an opportunity to identify
outlying or nonconforming practices for which
educational, quality improvement, or other such interventions
could be appropriate [
]. For example, routine use of
high-frequency oscillatory ventilation (HFOV) and
synchronized intermittent mandatory ventilation (SIMV)
have been called into question. Two prospective studies
of HFOV demonstrated largely equivalent outcomes in
patients with acute respiratory distress syndrome, and
SIMV has been associated with delayed separation from
In light of multiple barriers to the consistent and
evidence-based selection of a MV mode, we sought to
characterize the epidemiology of MV mode selection within
four hospitals affiliated with a university healthcare
system. We hypothesized that this approach could identify
outlying or nonconforming approaches to MV,
specifically the provision of HFOV and SIMV as test cases.
Furthermore, we aimed to characterize variability between
ICUs that treat similar patient populations and identify,
if present, any consistent patterns between types of ICUs
and between individual ICUs across hospitals in support
of subsequent hypothesis generation.
We conducted an exploratory retrospective observational
database study of routinely collected MV parameters to
examine mode utilization in 12 adult ICUs across four
hospitals between January 1, 2011 and December 31,
2016. Condensed reporting herein follows the
Strengthening the Reporting of Observational Studies in
Epidemiology statement and the Reporting of Studies Conducted
Using Observational Routinely Collected Health Data
statements where applicable [
]. Approval was
granted by the Emory University Institutional Review
Board (ID #IRB00095006) with a waiver of informed
consent owing to its retrospective design.
Setting and source population
Twelve ICUs in four standalone hospitals affiliated with
Emory Healthcare (Atlanta, GA USA) were studied,
which comprised 191/237 critical care beds, or 80.6%
of the system wide total. We categorized ICUs by their
primary mission whereby a: (a) cardiothoracic surgical
ICU (CTICU) treats patients after major heart, lung, and
vascular surgery; (b) neuroscience ICU (NSICU) treats
patients following cerebrovascular insults and
intracranial procedures; (c) surgical ICU (SICU) treats
postoperative patients not in one of the two prior categories, (d)
medical ICU (MICU) treats critically ill adults not
having recently undergone surgery, and (e) medical-surgical
ICU (MSICU) treats a mixture of critically ill adults. The
studied ICUs vary in their format and staffing (see
Additional file 1: Table S1 for further details) [
Ventilator management is the responsibility of the critical care
physician, or the admitting physician in ICUs without
critical care staffing. MV equipment has changed across
the system during the study period with increasing
availability of Hamilton (Hamilton Medical AG, Bonaduz,
Switzerland) ventilators (e.g. Galileo, G3, and G5) and a
gradual reduction in Puritan Bennett (Covidien LP,
Boulder, CO, USA) models (e.g. 840) as a result of
MV parameters and settings, including mode, are
routinely charted every 4 h in the electronic medical record
(EMR), or immediately following a setting change, by
respiratory therapists (RTs). EMR documentation is
consolidated nightly into the Clinical Data Warehouse (CDW)
via an extract, transform, and load process, which has
been internally validated by Information Services. The
CDW is indexed to support advanced analytics and
is accessed via structured query language and
MicroStrategy (MicroStrategy Inc., Washington, DC, USA)
Structured query approach
Each instance of recorded MV parameters was
considered a stand-alone epoch and extracted from the CDW
on a per-epoch, per-ICU basis during the period of
interest and then imported into MariaDB (MariaDB
Corporation AB, Espoo, Finland). No database linking was
required. MariaDB was used to aggregate the data and
identify nonsensical (e.g. numerical) values, which were
excluded from analysis. Data cleaning consisted of
identification and removal of elements with typographical
errors (e.g. letter transposition and misspelling), which
were rare as the EMR relies heavily on pre-populated
dropdown charting for MV mode. The two NSICUs in
Hospital 1 were considered in aggregate for inter-unit
comparisons as they are managed by the same critical
care team. Reported epoch counts represent all those
recorded during the period of interest.
Proportions of routinely recorded nominal
categorical mode counts were generated and initially visualized
in SAS JMP Pro version 13.1.0 (SAS Institute Inc., Cary,
NC, USA). Final analysis was done with R version 3.4.4
(R Core Team, Vienna, Austria) in RStudio 1.1.453
(RStudio Inc, Boston, MA, USA). Pearson’s Chi square test
for overall homogeneity was run prior to further
comparisons. Sub-analyses of nominal categorical MV mode
count proportions were conducted using Chi square or
Fisher’s exact test as appropriate on an inter-ICU basis.
Modes accounting for less than 2% of per-unit epochs
were excluded from this analysis as they may be less
clinically significant. Outlying proportions of HFOV and
SIMV (both pressure- and volume-controlled variants)
as test cases were initially identified through
examination of standardized residuals. These were confirmed
via binomial regression with post hoc comparison of
least-squares means and Bonferroni corrected pairwise
comparisons to examine inter-ICU variability. P
values < 0.05 were considered significant. Given the limited
dataset and goals of the study, characterization of
demographic variables and attempts to control for bias or
confounding were beyond the scope of the investigation.
Figures were generated using the ggplot2 and
RColorBrewer packages [
The CDW query identified 559,762 recorded MV epochs
from the ICUs of interest between January 1, 2011 and
December 31, 2016. Of those, the MV mode values for
28 epochs (0.005%) were nonsensical and excluded from
analysis (Additional file 2: Figure S1). Mode epoch
frequencies were significantly heterogeneous between
individual ICUs as depicted with aggregate counts in Fig. 1
and proportions in Fig. 2 (Table 1, P < 0.05 for all
comparisons). Heterogeneity was therefore evident between
like types of ICU (Additional file 3: Table S2) as depicted
proportionally on a per-unit basis in Additional file 4:
Figure S2 and aggregated in Additional file 5: Figure S3.
Similar variability was evident when aggregating units by
hospital as depicted proportionally in Additional file 6:
HFOV utilization was found to be nonconforming in
single MICU in Hospital 1, which accounted for 40.9% of
system wide epochs (P < 0.001 for all pairwise
comparisons). Three ICUs similarly accounted for a
disproportionate proportion of overall SIMV utilization: Hospital 3
CTICU with 48.4% of epochs (N = 14,049/29,001),
Hospital 3 MSICU with 26.8% (N = 7769/29,001), and
Hospital 1 CTICU with 14.5% (N = 4209/29,001; adjusted
P < 0.001 for all pairwise comparisons).
Hospital 1 accounted for 51.3% of total recorded MV
mode epochs (N = 287,026/559,734). Assist control
(AC) was the most commonly recorded MV mode
overall (N = 285,669/559,734, 51%) followed by adaptive
CTICU MICU NSICU NSICU 2 SICU
CTICU MICU NSICU SICU
Intensive Care Unit
Fig. 1 Mechanical ventilation mode epoch counts per intensive care unit by hospital. See list of abbreviations section
support ventilation (ASV, N = 129,341/559,734, 23.1%),
and pressure support ventilation (PSV, i.e. continuous
positive airway pressure with pressure support [CPAP
with PS], N = 56,822/559,734, 10.2%). When
examining AC, SIMV, adaptive pressure ventilation (APV),
and considering ASV to be a pressure-controlled mode
for passive patients, the overall prevalence of invasive
pressure controlled modes of ventilation was 35.6%
(N = 161,627/453,977) compared to 64.4% for volume
controlled modes (N = 292,350/453,977). In examining
graphical comparisons, the two studied MICUs
demonstrated the most consistent approach to mode
selection when examining like-type ICUs (Additional file 4:
Figure S2.) Within individual hospitals with greater
than one ICU, units in Hospital 2 were the most
consistent (Fig. 2).
Mode epoch proportions were significantly
heterogeneous between individual ICUs, contributing to variability
on an aggregate basis. Outlying utilization of HFOV in
one ICU and SIMV in three suggests that this approach
can be utilized to identify potentially
nonconforming practice patterns. Only one hospital demonstrated
a relatively consistent distribution of MV modes across
its studied ICUs, and the two studied MICUs were most
similar to one another; however, there were still
statistically significant differences.
Nonconforming MV mode utilization was readily
identified, and our approach may be able, more broadly, to
identify other clinically outlying or inappropriate MV
parameter selections. Utilization of HFOV appeared to
be rare, which aligns with findings from large prospective
evn VM em )w% 6 .)5% 7 .)0% ,597 )% .(168% ,316 )% 8 .)1% 4 .)0% 8 .)1% 7 .)1% .)3% 1 .)9% 3 .)5% 43 .)2%
la /CC lou (ro ,719 (25 ,864 (65 10 .641 6394 15 .20 ,907 (74 ,278 (71 ,632 (86 ,361 (66 848 (19 ,206 (47 ,959 (67 ,644 (47
ic A v N 1 3 ( (2 3 5 2 2 4 1 1 2
ceh rsseu ow% 7 .)12% 1 .)0% .(08% .(05% 9 .)2% 9 .)0% 3 .)3% .(06% 2 .)0% .(07% .(15% 1 .)9% 35 .)8% iton
em C re (r 55 (1 40 (4 41 71 57 (2 60 (5 16 (4 38 43 (4 96 08 13 (3 ,21 (3 sce
t A p N 7 2 5 1 1 2 3 1 1 1 3 1 2 s
g 2 io
regg CU ITCU ICU ISCU ISCU ICU ITCU ICU ISCU ICU ITCU ISCU ISCU ll irtvae
A I C M N N S C M N S C M M A b
1 1 1 1 1 2 2 2 2 3 3 4 A S
studies suggesting equivalent clinical outcomes with
greater risks than more conventional forms of MV [
]. Overall utilization of SIMV was lower than the
26% rate reported in a recent large observational study,
although we were able to identify outlying units . Its
routine application as a weaning modality has been
questioned, and global utilization may be experiencing an
associated decline [
In keeping with prior studies demonstrating an
approximate utilization rate of 50–60%, AC was likewise found
to be the most frequently recorded mode of ventilation in
our healthcare system [
5, 6, 13, 35
]. Recent international
MV practice studies have, however, suggested that novel
modes may be gradually driving a move away from AC
toward closed-loop approaches . Volume-controlled
ventilation was about twice as common as
pressurecontrolled ventilation, which is in contrast to a recent
international epidemiological study demonstrating the
]. From a speculative standpoint, this could
reflect an emphasis on monitoring of Vt, greater ease of
Vt restriction with volume-controlled modes,
mischarting of volume-targeted pressure-controlled modes, or
physician preference as no definitive difference in
compliance, gas exchange, or outcomes has been
As discussed subsequently, limited inferences can be
drawn about weaning approaches. Although one
hospital demonstrated marked adoption of ASV, proportional
assist ventilation (PAV) was very uncommon, which may
suggest heterogeneity in the willingness to adopt
closedloop MV modes consistent with the findings of other
epidemiological studies [
]. PSV was utilized relatively
consistently across the system and in excess of CPAP
alone, which is consistent with current MV weaning
Our study has important limitations. Owing to its
retrospective nature and use of a limited dataset, both
unmeasured confounding with associated confounding bias and
indication bias preclude inferences as to the etiology of
heterogeneity, which may have been clinically
appropriate. Rather, posited causes of heterogeneity herein are
purely speculative. Staffing, equipment, patient-specific
considerations, or any number of other factors may
influence MV mode selection. Attempts to control for bias or
confounding were beyond the scope of our
epidemiological study, and these considerations limit generalizability
of the findings.
Owing to our methodologic approach, we were unable
to determine the duration or time sequence of specific
MV modes utilization. As a simple example, although
MV charting is consistent every 4 h or with changes,
a patient on AC for 3 h and PSV for one would have a
single AC and PSV epoch recorded. As such, the actual
AC:PSV time ratio of 3:1 would appear to be 1:1 via this
study’s methodology. This limits the inferences that can
be drawn about approaches to ventilator separation and
As with any study that relies on routinely collected
data, inaccurate bedside charting cannot be excluded.
The incidence of nonsensical MV mode values was very
low, likely owing to the EMR’s use of predefined MV
mode selections. In that sense, our nominal
categorical data may be more accurate than similarly recorded
ordinal data. However, the potential for information bias
cannot be excluded. For example, pressure-controlled
volume-targeted AC modes could conceivably be charted
The highly heterogeneous nature of MV mode
distribution found in the current study suggests that mode
selection likely involves a complex interplay of factors, which
could include institution, ICU, or provider-specific
considerations in addition to patient or disease-related
factors. The relationship between MV mode selection, MV
parameters (e.g. Vt, PEEP, plateau pressure, or driving
pressure), and clinical outcomes also warrants further
investigation, especially as approaches to mode selection
appear highly variable.
Additional file 1: Table S1. Details of studied hospitals and intensive care
Additional file 2: Figure S1. CONSORT flow diagram.
Additional file 3: Table S2. Mechanical ventilation mode epochs per
intensive care unit type.
Additional file 4: Figure S2. Ratios of mechanical ventilation mode
epochs per hospital by intensive care unit type. Hospital 1 NSICUs
depicted in aggregate. See list of abbreviations section.
Additional file 5: Figure S3. Ratios of mechanical ventilation mode
epochs per intensive care unit type. See list of abbreviations section.
Additional file 6: Figure S4. Ratios of mechanical ventilation mode
epochs per hospital. See list of abbreviations section.
AC: assist control; APRV: airway pressure release ventilation; APV: adaptive
pressure ventilation; ASV: adaptive support ventilation; CDW: Clinical Data
Warehouse; CMV: continuous mandatory ventilation; CPAP: continuous
positive airway pressure; CPAP + PS: continuous positive airway pressure with
pressure support (i.e. pressure support ventilation [PSV]); CTICU:
cardiothoracic intensive care unit; EMR: electronic medical record; HFOV: high frequency
oscillatory ventilation; ICU: intensive care unit; MICU: medical intensive care
unit; MSICU: medical-surgical intensive care unit; MV: mechanical ventilation;
NIPPV: non-invasive positive pressure ventilation; NSICU: neuroscience
intensive care unit; PAV: proportional assist ventilation; PEEP: positive end-expiratory
pressure; Pplat: plateau pressure; PS: pressure support; PSV: pressure support
ventilation; RT: respiratory therapist; SICU: surgical intensive care unit; SIMV:
synchronized intermittent mandatory ventilation.
All authors meet criteria for authorship under ICMJE criteria. JMB conceived
the study with critical input from CSJ and RFG. JMB developed and conducted
the structured data queries with assistance from CSJ and RFG. CSJ and JMB
designed and implemented the analysis plan with input from RFG. RFG, MS,
and JKR critically reviewed the analysis plan and results. CSJ wrote the
manuscript with critical revisions and conceptual advice from RFG, MS, JKR, and JMB.
All authors read and approved the final manuscript.
Preliminary findings were previously presented at the International Anesthesia
Research Society and Society of Critical Care Anesthesiologists 2017 annual
meetings in Washington, DC, USA.
The authors declare that they have no competing interests. JMB has an equity
stake in Intensix (Netanya, Israel).
Availability of data and materials
The datasets used and analyzed during the current study are available from
the corresponding author on reasonable request. Aggregate counts are
presented in Table 1.
Consent for publication
Ethics approval and consent to participate
Approval for the study was granted by the Emory University Institutional
Review Board (ID #IRB00095006). A waiver of informed consent was granted
owing to the retrospective nature of the study and use of de-identified data.
Internal departmental funds were used to support the authors’ time to
conduct the present study.
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
1. Bellani G , Laffey JG , Pham T , Fan E , Brochard L , Esteban A , et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries . JAMA. 2016 ; 315 ( 8 ): 788 - 800 .
2. Metnitz PGH , Metnitz B , Moreno RP , Bauer P , Sorbo LD , Hoermann C , et al. Epidemiology of mechanical ventilation: analysis of the SAPS 3 database . Intensive Care Med . 2009 ; 35 ( 5 ): 816 - 25 .
3. Neto AS , Barbas CSV , Simonis FD , Artigas-Raventos A , Canet J , Determann RM , et al. Epidemiological characteristics, practice of ventilation, and clinical outcome in patients at risk of acute respiratory distress syndrome in intensive care units from 16 countries (PRoVENT): an international, multicentre, prospective study . Lancet Respir Med . 2016 ; 4 ( 11 ): 882 - 93 .
4. Wunsch H , Linde-Zwirble WT , Angus DC , Hartman ME , Milbrandt EB , Kahn JM . The epidemiology of mechanical ventilation use in the United States . Crit Care Med . 2010 ; 38 ( 10 ): 1947 - 53 .
5. Esteban A , Anzueto A , Frutos F , Alia I , Brochard L , Stewart TE , et al. Characteristics and outcomes in adult patients receiving mechanical ventilation: a 28-day international study . JAMA . 2002 ; 287 ( 3 ): 345 - 55 .
6. Hong SB , Oh BJ , Kim YS , Kang EH , Kim CH , Park YB , et al. Characteristics of mechanical ventilation employed in intensive care units: a multicenter survey of hospitals . J Korean Med Sci . 2008 ; 23 ( 6 ): 948 - 53 .
7. Jeong B-H , Suh GY , An JY , Park MS , Lee JH , Lee M-G , et al. Clinical demographics and outcomes in mechanically ventilated patients in Korean intensive care units . J Korean Med Sci . 2014 ; 29 ( 6 ): 864 - 70 .
8. Ye Y , Zhu B , Jiang L , Jiang Q , Wang M , Hua L , et al. A contemporary assessment of acute mechanical ventilation in Beijing: description, costs, and outcomes . Crit Care Med . 2017 ; 45 ( 7 ): 1160 - 7 .
9. Wenstedt EFE , De Bie Dekker AJR , Roos AN , Verberne JJM , Korsten HHM , Schultz MJ , et al. Current practice of closed-loop mechanical ventilation modes on intensive care units-a nationwide survey in the Netherlands . Neth J Med . 2017 ; 75 ( 4 ): 145 - 50 .
10. Rose L , Blackwood B , Egerod I , Haugdahl HS , Hofhuis J , Isfort M , et al. Decisional responsibility for mechanical ventilation and weaning: an international survey . Crit Care . 2011 ; 15 ( 6 ): R295 .
11. Esteban A , Ferguson ND , Meade MO , Frutos-Vivar F , Apezteguia C , Brochard L , et al. Evolution of mechanical ventilation in response to clinical research . Am J Respir Crit Care Med . 2008 ; 177 ( 2 ): 170 - 7 .
12. Esteban A , Frutos-Vivar F , Muriel A , Ferguson ND , Penuelas O , Abraira V , et al. Evolution of mortality over time in patients receiving mechanical ventilation . Am J Respir Crit Care Med . 2013 ; 188 ( 2 ): 220 - 30 .
13. Esteban A , Anzueto A , Alía I , Gordo F , Apezteguía C , Pálizas F , et al. How is mechanical ventilation employed in the intensive care unit? Am J Respir Crit Care Med . 2000 ; 161 ( 5 ): 1450 - 8 .
14. Kübler A , Maciejewski D , Adamik B , Kaczorowska M. Mechanical ventilation in ICUs in Poland: a multi-center point-prevalence study . Med Sci Monit . 2013 ; 19 : 424 - 9 .
15. Esteban A , Alia I , Ibanez J , Benito S , Tobin MJ . Modes of mechanical ventilation and weaning. A national survey of Spanish hospitals . The Spanish Lung Failure Collaborative Group. Chest . 1994 ; 106 ( 4 ): 1188 - 93 .
16. Koh Y , Lim CM , Koh SO , Ahn JJ , Kim YS , Jung BH , et al. A national survey on the practice and outcomes of mechanical ventilation in Korean intensive care units . Anaesth Intensive Care . 2009 ; 37 ( 2 ): 272 - 80 .
17. Rose L , Presneill JJ , Johnston L , Nelson S , Cade JF . Ventilation and weaning practices in Australia and New Zealand . Anaesth Intensive Care . 2009 ; 37 ( 1 ): 99 - 107 .
18. Demoule A , Chevret S , Carlucci A , Kouatchet A , Jaber S , Meziani F , et al. Changing use of noninvasive ventilation in critically ill patients: trends over 15 years in francophone countries . Intensive Care Med . 2016 ; 42 ( 1 ): 82 - 92 .
19. Antonelli M , Conti G , Esquinas A , Montini L , Maggiore SM , Bello G , et al. A multiple-center survey on the use in clinical practice of noninvasive ventilation as a first-line intervention for acute respiratory distress syndrome . Crit Care Med . 2007 ; 35 ( 1 ): 18 - 25 .
20. Bellani G , Laffey JG , Pham T , Madotto F , Fan E , Brochard L , et al. Noninvasive ventilation of patients with acute respiratory distress syndrome. Insights from the LUNG SAFE study . Am J Respir Crit Care Med . 2017 ; 195 ( 1 ): 67 - 77 .
21. Walkey AJ , Wiener RS . Use of noninvasive ventilation in patients with acute respiratory failure, 2000 - 2009 : a population-based study . Ann Am Thorac Soc . 2013 ; 10 ( 1 ): 10 - 7 .
22. Murphy DJ , Lyu PF , Gregg SR , Martin GS , Hockenberry JM , Coopersmith CM , et al. Using incentives to improve resource utilization: a quasi-experimental evaluation of an ICU quality improvement program . Crit Care Med . 2016 ; 44 ( 1 ): 162 - 70 .
23. Ferguson ND , Cook DJ , Guyatt GH , Mehta S , Hand L , Austin P , et al. Highfrequency oscillation in early acute respiratory distress syndrome . N Engl J Med . 2013 ; 368 ( 9 ): 795 - 805 .
24. Young D , Lamb SE , Shah S , MacKenzie I , Tunnicliffe W , Lall R , et al. Highfrequency oscillation for acute respiratory distress syndrome . N Engl J Med . 2013 ; 368 ( 9 ): 806 - 13 .
25. Esteban A , Frutos F , Tobin MJ , Alia I , Solsona JF , Valverdu I , et al. A comparison of four methods of weaning patients from mechanical ventilation . Spanish Lung Failure Collaborative Group. N Engl J Med . 1995 ; 332 ( 6 ): 345 - 50 .
26. Brochard L , Rauss A , Benito S , Conti G , Mancebo J , Rekik N , et al. Comparison of three methods of gradual withdrawal from ventilatory support during weaning from mechanical ventilation . Am J Respir Crit Care Med . 1994 ; 150 ( 4 ): 896 - 903 .
27. Luo J , Wang MY , Liang BM , Yu H , Jiang FM , Wang T , et al. Initial synchronized intermittent mandatory ventilation versus assist/control ventilation in treatment of moderate acute respiratory distress syndrome: a prospective randomized controlled trial . J Thorac Dis . 2015 ; 7 ( 12 ): 2262 - 73 .
28. Ortiz G , Frutos-Vivar F , Ferguson ND , Esteban A , Raymondos K , Apezteguia C , et al. Outcomes of patients ventilated with synchronized intermittent mandatory ventilation with pressure support: a comparative propensity score study . Chest . 2010 ; 137 ( 6 ): 1265 - 77 .
29. Benchimol EI , Smeeth L , Guttmann A , Harron K , Moher D , Petersen I , et al. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement . PLoS Med . 2015 ; 12 ( 10 ): e1001885 .
30. von Elm E , Altman DG , Egger M , Pocock SJ , Gøtzsche PC , Vandenbroucke JP . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies . Lancet . 2007 ; 370 ( 9596 ): 1453 - 7 .
31. Grabenkort WR , Meissen HH , Gregg SR , Coopersmith CM . Acute care nurse practitioners and physician assistants in critical care: transforming education and practice . Crit Care Med . 2017 ; 45 ( 7 ): 1111 - 4 .
32. Buchman TG , Coopersmith CM , Meissen HW , Grabenkort WR , Bakshi V , Hiddleson CA , et al. Innovative interdisciplinary strategies to address the intensivist shortage . Crit Care Med . 2017 ; 45 ( 2 ): 298 - 304 .
33. Wickham H. ggplot2: elegant graphics for data analysis . New York: Springer; 2009 .
34. Kacmarek RM , Branson RD . Should intermittent mandatory ventilation be abolished? Respir Care . 2016 ; 61 ( 6 ): 854 - 66 .
35. Rose L , Schultz MJ , Cardwell CR , Jouvet P , McAuley DF , Blackwood B . Automated versus non-automated weaning for reducing the duration of mechanical ventilation for critically ill adults and children . Cochrane Database Syst Rev . 2014 ; 6 : CD009235 .
36. Rittayamai N , Katsios CM , Beloncle F , Friedrich JO , Mancebo J , Brochard L . Pressure-controlled vs volume-controlled ventilation in acute respiratory failure: a physiology-based narrative and systematic review . Chest . 2015 ; 148 ( 2 ): 340 - 55 .
37. Ouellette DR , Patel S , Girard TD , Morris PE , Schmidt GA , Truwit JD , et al. Liberation from mechanical ventilation in critically ill adults: an official American College of Chest Physicians/American Thoracic Society clinical practice guideline: inspiratory pressure augmentation during spontaneous breathing trials, protocols minimizing sedation, and noninvasive ventilation immediately after extubation . Chest . 2017 ; 151 ( 1 ): 166 - 80 .