Increased Occurrence of Tracheal Intubation Associated Events during Nights and
Weekends in the Pediatric Intensive Care Unit
Kyle J Rehder MD FCCP1, John S Giuliano Jr MD FAAP2, Natalie Napolitano MPH RRT-NPS
FAARC3, David A Turner MD FCCM FCCP1, Gabrielle Nuthall MBChB FRACP CICM4,
Vinay M Nadkarni MD FCCM5, Akira Nishisaki MD MSCE5 for the NEAR4KIDS and PALISI
investigators
Department of Pediatrics, Division of Pediatric Critical Care Medicine, Duke Children’s
Hospital, Durham, NC.
2
Department of Pediatrics, Division of Critical Care Medicine, Yale University School of
Medicine, New Haven, CT.
3
Department of Nursing, Respiratory Care, and Neurodiagnostics, The Children’s Hospital of
Philadelphia, Philadelphia, PA.
4
Division of Pediatric Intensive Care, Starship Children’s Health Center, Grafton, Auckland,
New Zealand.
5
Department of Anesthesiology and Critical Care Medicine, The Children’s Hospital of
Philadelphia, Philadelphia, PA.
1
Drs. Rehder and Giuliano Jr shared first authorship.
Institution where study performed: Twenty centers in the National Emergency Airway for
Children Registry (NEAR4KIDS), data analysis performed at Duke University Medical Center
(Durham, NC) and Children’s Hospital of Philadelphia (Philadelphia, PA)
Corresponding Author / Address for reprints:
Kyle Rehder, MD
Division of Pediatric Critical Care
DUMC Box 3046
Durham, NC 27710
Funding: Endowed Chair, Critical Care Medicine, The Children’s Hospital of Philadelphia,
Unrestricted Research fund from Laerdal Foundation Acute Care Medicine, AHRQ
1R03HS021583- 01, AHRQ 1 R18 HS022464-01
1
Supplemental digital content is available for this article. Direct URL citations appear in the
printed text and are provided in the HTML and PDF versions of this article on the journal’s
website (http://journals.lww.com/ccmjournal).
Keywords: airway, intubation, pediatric, personnel staffing, in-house, patient safety
Copyright form disclosures: Dr. Napolitano served as a board member for AARC and AAN
(only compensation is travel to meetings), consulted for AAN (Qualitative project on Decisions
of Families for Asthma Care), lectured for Draeger Medical, and received support for travel from
AARC (for AARC Conference). Her institution has a Pending Patent for NIV Mask and received
grant support from the AHRQ, CVS Health, Aerogen, Nihon Kohden, and AHRQ. Dr. Nuthall is
employed by Auckland District Health Board. Her institution received other support from A+
Trust (Contribution to research nurses salary to collect ongoing data for the project). Dr.
Nadkarni’s institution received grant support from AHRQ (R18 Grant). Dr. Nishisaki received
support for article research from the Agency for Healthcare Research and Quality). His
institution received grant support from Improving the safety and quality of tracheal intubation in
pediatric ICUs (AHRQ R18HS022464) and from Evaluating safety and quality of tracheal
intubation in pediatric ICUs (AHRQ R03HS021583) (Supported by the Agency for Healthcare
Research and Quality [AHRQ]) and received grant support from Evaluation of Mainstream
Capnography (Nihon Kohden America). The remaining authors have disclosed that they do not
have any potential conflicts of interest.
2
Abstract
Objectives: Adverse tracheal intubation associated events (TIAEs) are common in PICUs. Prior
studies suggest provider and practice factors are important contributors to TIAEs. Little is known
about how the incidence of TIAEs is affected by the time of day, day of the week, or presence of
in-hospital attending-level intensivists. We hypothesize that tracheal intubations occurring during
nights and weekends are associated with a higher incidence of TIAEs.
Design: Retrospective observational cohort study
Setting: Twenty international PICUs
Subjects: Critically ill children requiring tracheal intubation
Interventions: None
Measurements and Main Results: We analyzed 5096 tracheal intubation courses from July
2010 to March 2014 from the prospective multicenter National Emergency Airway Registry for
Children (NEAR4KIDS). Incidence of a priori defined TIAEs was the primary outcome.
Occurrence of any TIAEs and severe TIAEs were more common during nights (19:00-06:59)
and weekends compared to weekdays (19% vs. 16%, p=0.01; 7% vs. 6%, p=0.05, respectively).
This difference was significant in emergent intubations after adjusting for site level clustering
and patient factors: adjusted odds ratio (aOR) for any TIAEs: aOR 1.20, CI951.02-1.41, p=0.03;
but not significant in non-emergent intubations: aOR 0.94, CI95 0.63-1.40,p=0.75. For emergent
intubations, PICUs with home call attending coverage had a significantly higher incidence of
TIAEs during nights and weekends (aOR 1.29: 1.01-1.66, p=0.04), and this difference was
attenuated in PICUs with in-hospital attending coverage (aOR: 1.12, 95% CI 0.91-1.39, p=0.28).
Conclusions: Higher incidence of TIAEs was observed during nights and weekends. This
difference was primarily attributed to emergent intubations. In-hospital attending physician
3
coverage attenuated this discrepancy between weekdays vs. nights and weekends, but was not
fully protective for TIAEs.
4
Introduction
Pediatric advanced airway management is challenging and can result in significant
morbidity or mortality1-4. Patients requiring emergent tracheal intubation (henceforth referred to
as ‘intubation’) rather than non-emergent intubation are at the greatest risk for complications1,5.
Some data also suggest worse outcomes for critically ill patients during nights and weekends
compared to traditional weekday staffing models1,6-9. Furthermore, intubation attempts by
inexperienced providers have been associated with an increased incidence of adverse events3.
Considering these factors and others, many children’s hospitals have employed 24-hour inhospital intensive care attending coverage with the intent to mediate these risks. However, the
presence of in-hospital coverage has not consistently demonstrated improved patient outcomes1012
.
The aim of this study was to evaluate the effect of time of day and 24/7 in-hospital
coverage on tracheal intubation associated events (TIAEs) and outcomes in the PICU using the
National Emergency Airway Registry for Children (NEAR4KIDS) database5. We hypothesized
that intubations during nighttime and weekends are associated with higher occurrences of
adverse TIAEs. Since night and weekend intubations are more likely emergent and at higher risk
for TIAEs, we further evaluated our hypothesis among the intubations for emergent indications.
We then evaluated the impact of in-hospital coverage on TIAEs during nights and weekends.
5
Materials and Methods
The previously described NEAR4KIDS registry is a multicenter quality improvement
collaborative to prospectively collect safety and quality data on intubations in PICUs5. Data
collection was either approved or declared exempt by the Institutional Review Board at each
participating center. Data collection included basic patient demographics, provider presence,
intubating provider specialty and level of training, equipment used, and details on the intubation
encounter. Centers also reported if the intubation was elective (non-emergent) or emergent. An
“encounter” was defined as one completed episode of advanced airway management
intervention, typically ending with successful intubation. Within each intubation encounter, a
“course” encompassed one method or approach to secure an artificial airway (e.g., switching
from direct laryngoscopy to fiber optic laryngoscopy was defined as two courses), while an
“attempt” was defined as a single distinct advanced airway maneuver (e.g., insertion of a device
such as laryngoscope, endotracheal tube, or laryngeal mask into patient’s mouth or nose) ².
Pediatric Index of Mortality (PIM2) was captured at the time of PICU admission13.
Adverse TIAEs were defined a priori. Severe TIAEs were defined as cardiac arrest,
esophageal intubation with delayed recognition, emesis with witnessed aspiration, hypotension
requiring intervention (fluid and/or pressors), laryngospasm, pneumothorax or
pneumomediastinum, or direct airway injury. Non-severe TIAEs included mainstem bronchial
intubation (confirmed by chest radiograph), esophageal intubation with immediate recognition,
emesis without aspiration, hypertension requiring therapy, epistaxis, dental or lip trauma,
medication error, arrhythmia, or pain and/or agitation requiring additional medication with delay
in intubation.
6
Intubation time was recorded on the data collection form for each intubation encounter.
‘Weekdays’ were defined as 07:00-18:59, Monday through Friday (i.e. excluding nights and
weekends). ‘Nights and weekends’ included (1) weeknights (defined as 19:00-06:59 the next
morning), and (2) weekends (defined as Friday at 19:00-06:59 Monday morning). For the data
collection period, centers identified themselves as having 24/7 in-hospital pediatric intensivist
coverage or home coverage and provided the start date for in-hospital coverage (to properly
identify intubations that occurred under in-hospital coverage).
Statistical analysis
The primary outcome was occurrence of any TIAE during an intubation course. The
primary exposure variable was weekday vs. nights and weekends. Non-emergent intubation
status was considered as an effect modifier and analyses were stratified by emergent status when
the association of TIAEs and weekday status was evaluated. For the analysis of in-hospital
coverage, only emergent intubations were included to focus on those intubations where inhospital coverage was most likely to affect outcomes. Summary statistics are provided as
percentages for categorical variables or median and interquartile ranges (IQR) for continuous
variables. Dichotomous categorical variables were analyzed using chi-square test, while
continuous variables were compared using Wilcoxon rank-sum test. Statistical analysis was
performed with JMP 11 (SAS, Cary, NC) and STATA 11.2 (Stata Corp, College Station, TX). A
random effect multivariate logistic regression model was developed with the site as a group
variable (20 sites) and occurrence of TIAE as a dichotomous outcome. Factors associated with
weekday vs. night and weekend status in the univariate analysis (p<0.1) and also known to be
associated with TIAEs in previous studies were included as covariates in this model. We also
7
conducted sensitivity analyses using two alternative definitions for nights (nighttime 16:00 –
06:59, and nighttime 18:00-06:59). P-values of < 0.05 were considered statistically significant.
Results
Patient demographics and provider characteristics
Five thousand ninety-six intubation courses (representing 4775 encounters) were reported
at 20 participating institutions during the study period (July 2010 to March 2014). A median of
162 (IQR: 81-347) courses were reported from each site. Patient demographics and indications
for intubation are reported in Table 1. Respiratory failure, shock status and emergent intubations
were associated with night and weekend intubations. Of note, admission PIM2 scores and history
of difficult airway were not associated with time of intubation. First attempt intubation providers
for the course are reported in Table 2.
Weekday vs. night and weekend intubations and outcomes
Overall, 2702 intubation courses (53%) occurred during nights and weekends. Adverse
TIAEs were reported in 898 (18%) of all intubation courses, and severe TIAEs were reported in
319 (6%).
Tracheal intubations during nights and weekends were associated with a higher
occurrence of TIAEs, as shown in Table 3. This was primarily attributed to emergent intubations.
This difference was significant in emergent intubations after adjusting for site level clustering
and patient factors: adjusted odds ratio (aOR) for any TIAEs: aOR 1.20, 95% Confidence
Interval (CI95) 1.02-1.41, p=0.03; but not significant in non-emergent intubations: aOR 0.94, CI95
0.63-1.40, p=0.75.,
8
Attending-level providers were more likely to be present at weekday intubations.
Pediatric residents were more likely to be involved as first intubators during nights and
weekends. This latter finding was consistent for both non-emergent and emergent intubations.
24/7 in-hospital intensivist coverage and attending physician presence at emergent intubations
Nine of 20 (45%) PICUs reported having in-hospital attending physician coverage during
the study period: seven PICUs had in-hospital coverage throughout and two PICUs transitioned
from home coverage to an in-hospital coverage system during the study period. Overall, 56% of
emergent, night and weekend intubations occurred in PICUs with in-hospital coverage. Table 4
displays the occurrence of TIAEs and severe TIAEs, as well as provider characteristics for
emergent intubations, based on the presence of in-hospital coverage. Emergent intubations at the
PICUs with in-hospital attending coverage were associated with a higher occurrence of TIAEs
during weekdays as well as nights and weekends.
For emergent intubations, the odds of having a TIAE were significantly higher in night
and weekend intubations vs. weekday intubations (OR 1.29, p=0.04, Table 5) in PICUs with
home call attending coverage, but not for PICUs with in-hospital coverage (OR: 1.12, p=0.28)
after adjusted for covariates associated with time of intubation (respiratory failure, shock status)
and age.
Attending physicians in in-hospital coverage were more likely than those in home
coverage to be present for both night and weekend and weekday intubations, although home
coverage attending physicians were commonly present at night and weekend intubations (68%).
Attending physician presence for emergent, night and weekend intubations were not significantly
9
associated with occurrence of any TIAEs (attending presence: 393/1923 (20%) vs. attending not
present: 70/319 (18%), p= 0.27). This was also the case for the occurrence of severe TIAEs (146/
1,923 (7%) vs. 29/ 389 (7%), p= 0.93). Attending physicians were more likely and pediatric
residents were less likely to be the first intubating providers for night and weekend intubations in
PICUs with in-hospital coverage (Table 4).
Sensitivity analysis with alternative definitions for nighttime hours
Alternative definitions of nighttime (18:00-06:59 and 16:00-06:59) were also used to
evaluate the association of TIAEs with nights and weekends. For the first definition (nighttime:
18:00-06:59), 2871 intubation courses were classified as intubations during nights and weekends.
The night and weekend hours were significantly associated with any TIAEs (weekday 16% vs.
night and weekend 19%, p=0.01) but not with severe TIAEs (6% vs.7%, p=0.11).
For the second definition (16:00-06:59), 3337 intubation courses were classified as
intubations during nights and weekends. The night and weekend hours were significantly
associated with both any TIAEs (weekday 16% vs. night and weekend 19%, p=0.004) and severe
TIAEs (weekday 5% vs. night and weekend 7%, p=0.03). Multivariate analyses with these two
alternative definitions revealed that in-hospital attending coverage and the occurrence of TIAEs
were significantly associated for weekday intubations but not for night and weekend intubations,
which was consistent with the original definition of nighttime (19:00-06:59) (Supplemental
Tables 1 and 2, Supplemental Digital Content 1, ).
10
Discussion
Our study using a large multicenter intubation quality improvement database
(NEAR4KIDS) demonstrated an increased occurrence of TIAEs for tracheal intubations
performed during nights and weekends. This remained significant after adjusting for patient
factors and excluding non-emergent intubations. When in-hospital attending coverage was
compared against home call coverage, the increased risk of TIAEs during nights and weekends
persisted in home coverage models but not in in-hospital coverage, suggesting a protective effect
of 24/7 in-hospital coverage. Because in-hospital coverage does not assure attending presence at
each intubation, we separately analyzed documented attending presence regardless of the
hospital model. Attending physician presence at the bedside was not associated with occurrence
of TIAEs among night and weekend emergent intubations.
This is the largest study to date evaluating the association between time of day and
TIAEs in multiple PICUs. Our study adds to the growing body of literature evaluating patient
care in intensive care units outside of traditional weekday work hours. Similar to much of the
other published literature, our study demonstrates inconsistency in outcomes by time and day of
week. However, this variability is likely multifactorial. In evaluating this finding, we must
consider the night and weekend differences in patient population and disease processes,
indications for the procedure, unit staffing, and provider experience level.
Several studies have demonstrated worse outcomes for patients admitted during nights
and weekends7,14-16, however, this discrepancy may be secondary to confounders such as severity
of illness or admitting diagnoses17,18. Less data exist regarding procedure outcomes by time of
day. In a single center pediatric study, Carroll et al demonstrated a three-fold risk of
11
complications for intubations occurring during nights and weekends1. Our data supports this
finding in a larger multicenter cohort.
Concern over differences in outcomes during off-hours have led to guidelines
recommending 24/7 in-hospital intensivist coverage of all Level I and Level II adult intensive
care units19. However, the published data regarding in-hospital coverage is unclear regarding its
effect on patient outcomes11,20-23. Furthermore, aggregate outcomes may not directly assess the
impact of in-hospital attending coverage, as these outcomes are also affected by care provided
during weekday hours. Investigating procedural outcomes during nights and weekends provides
the unique opportunity to measure the immediate effect of in-hospital attending coverage. In our
study, the association between nights and weekends and occurrence of TIAEs was somewhat
attenuated, but not eliminated, in the presence of in-hospital attending coverage. This result
suggests that the higher intubation risks associated with nights and weekends may be mainly
from patient factors.
While we may have been unable to fully adjust for patient-level factors in our
multivariate model, it is also possible that the occurrence of TIAEs were more susceptible to
non-attending physician ICU staffing (physician trainees, nurses, respiratory therapists), which
may have been quite different during nights and weekends. We may have also over-estimated the
clinical impact of having in-hospital attending physician coverage in our hypothesis. Our recent
publication demonstrated similar intubation skills among pediatric critical care medicine fellows
and attending physicians when analyzed with TIAEs as an outcome3, suggesting that attending
presence may not be the most important factor in minimizing TIAEs. It is important to note that
even when present at intubations, attending physicians were rarely the first attempt providers
(Table 4).
12
Our study also demonstrated increased occurrence of TIAEs among intubations with 24/7
in-hospital attending coverage during the weekday hours (OR 1.53), with a lesser, nonsignificant increase during nights and weekends (OR 1.33). It is likely that hospitals choosing to
provide in-hospital coverage represent a higher acuity population, which could explain this
discrepancy. While there may be other inherent differences in the patient populations or
provider characteristics as confounders for the increased TIAE incidence with in-hospital
coverage, it is improbable that differences attributed to the coverage model would lead to a larger
discrepancy during weekday hours.
This study has several limitations. It is an observational study relying on self-reported
data; however, reporting bias is minimized by prospective data collection and careful monitoring
of center compliance5. We are unable to account for unmeasured patient and institutional factors
which may affect the incidence of TIAEs, including prior institutional team training or the
presence of difficult airway algorithms. The data did not specify when the attending physician
arrived at the intubation scene; therefore we cannot report the degree of attending physician
involvement for each intubation. Also, our analysis did not include patient outcomes beyond the
intubation encounter, including length of ventilation or mortality. The data collection form does
not specify when the attending arrives, therefore we cannot report which intubations had
attending presence for the entire procedure. Finally, while this study reports a large number of
intubation courses from twenty academic centers, this cohort may not be representative of all
intubations that occur in PICUs, particularly intubations that occur in community PICUs.
While PICUs strive to provide equivalent care at any hour of the day, this study
demonstrates that there is variability in the incidence of TIAEs between traditional weekday
work hours and those occurring during nights and weekends. This variance remains present after
13
adjusting for patient factors and excluding non-emergent intubations, and was somewhat
attenuated in the presence of in-hospital attending coverage. Further investigation is needed to
elucidate factors which contribute to this variance, and to determine processes which may lead to
safer tracheal intubations for this high-risk population.
Conclusions:
Adverse TIAEs are more common during nights and weekends. This association remained
significant after adjusting for patient factors. Presence of 24/7 in-hospital attending coverage was
partially protective for TIAEs during nights and weekends. Further research is required to
identify explanatory factors.
14
Acknowledgements:
We would like to thank Hayley Buffman for her tireless efforts as the coordinator for the
multicenter NEAR4KIDS registry.
15
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Carroll CL, Spinella PC, Corsi JM, Stoltz P, Zucker AR. Emergent endotracheal intubations in
children: be careful if it's late when you intubate. Pediatr Crit Care Med. May 2010;11(3):343348.
Nishisaki A, Ferry S, Colborn S, et al. Characterization of tracheal intubation process of care and
safety outcomes in a tertiary pediatric intensive care unit. Pediatr Crit Care Med. Jan
2012;13(1):e5-10.
Sanders RC, Jr., Giuliano JS, Jr., Sullivan JE, et al. Level of trainee and tracheal intubation
outcomes. Pediatrics. Mar 2013;131(3):e821-828.
Easley RB, Segeleon JE, Haun SE, Tobias JD. Prospective study of airway management of children
requiring endotracheal intubation before admission to a pediatric intensive care unit. Crit Care
Med. Jun 2000;28(6):2058-2063.
Nishisaki A, Turner DA, Brown CA, 3rd, Walls RM, Nadkarni VM. A National Emergency Airway
Registry for children: landscape of tracheal intubation in 15 PICUs. Crit Care Med. Mar
2013;41(3):874-885.
Pronovost PJ, Angus DC, Dorman T, Robinson KA, Dremsizov TT, Young TL. Physician staffing
patterns and clinical outcomes in critically ill patients: a systematic review. JAMA. Nov 6
2002;288(17):2151-2162.
Cavallazzi R, Marik PE, Hirani A, Pachinburavan M, Vasu TS, Leiby BE. Association between time
of admission to the ICU and mortality: a systematic review and metaanalysis. Chest. Jul
2010;138(1):68-75.
Gajic O, Afessa B, Hanson AC, et al. Effect of 24-hour mandatory versus on-demand critical care
specialist presence on quality of care and family and provider satisfaction in the intensive care
unit of a teaching hospital. Crit Care Med. Jan 2008;36(1):36-44.
Wallace DJ, Angus DC, Barnato AE, Kramer AA, Kahn JM. Nighttime intensivist staffing and
mortality among critically ill patients. The New England journal of medicine. May 31
2012;366(22):2093-2101.
Gajic O, Afessa B. Physician staffing models and patient safety in the ICU. Chest. Apr
2009;135(4):1038-1044.
Kerlin MP, Small DS, Cooney E, et al. A randomized trial of nighttime physician staffing in an
intensive care unit. The New England journal of medicine. Jun 6 2013;368(23):2201-2209.
Rehder KJ, Cheifetz IM, Markovitz BP, Turner DA. Survey of in-house coverage by pediatric
intensivists: characterization of 24/7 in-hospital pediatric critical care faculty coverage. Pediatr
Crit Care Med. Feb 2014;15(2):97-104.
Slater A, Shann F, Pearson G. PIM2: a revised version of the Paediatric Index of Mortality.
Intensive Care Med. Feb 2003;29(2):278-285.
Sorita A, Ahmed A, Starr SR, et al. Off-hour presentation and outcomes in patients with acute
ischemic stroke: a systematic review and meta-analysis. European journal of internal medicine.
Apr 2014;25(4):394-400.
Arias Y, Taylor DS, Marcin JP. Association between evening admissions and higher mortality
rates in the pediatric intensive care unit. Pediatrics. Jun 2004;113(6):e530-534.
Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on weekends as
compared with weekdays. The New England journal of medicine. Aug 30 2001;345(9):663-668.
16
17.
18.
19.
20.
21.
22.
23.
McCrory MC, Gower EW, Simpson SL, Nakagawa TA, Mou SS, Morris PE. Off-hours admission to
pediatric intensive care and mortality. Pediatrics. Nov 2014;134(5):e1345-1353.
Peeters B, Jansen NJ, Bollen CW, van Vught AJ, van der Heide D, Albers MJ. Off-hours admission
and mortality in two pediatric intensive care units without 24-h in-house senior staff
attendance. Intensive Care Med. Nov 2010;36(11):1923-1927.
Haupt MT, Bekes CE, Brilli RJ, et al. Guidelines on critical care services and personnel:
Recommendations based on a system of categorization of three levels of care. Crit Care Med.
Nov 2003;31(11):2677-2683.
Arabi Y. Pro/Con debate: should 24/7 in-house intensivist coverage be implemented? Crit Care.
2008;12(3):216.
Nishisaki A, Pines JM, Lin R, et al. The impact of 24-hr, in-hospital pediatric critical care attending
physician presence on process of care and patient outcomes*. Crit Care Med. Jul
2012;40(7):2190-2195.
Rehder KJ, Cheifetz IM, Willson DF, Turner DA. Perceptions of 24/7 in-hospital intensivist
coverage on pediatric housestaff education. Pediatrics. Jan 2014;133(1):88-95.
Reriani M, Biehl M, Sloan JA, Malinchoc M, Gajic O. Effect of 24-hour mandatory vs on-demand
critical care specialist presence on long-term survival and quality of life of critically ill patients in
the intensive care unit of a teaching hospital. J Crit Care. Dec 13 2011.
17