CLINICAL ISSUES

Night admission to the emergency department: a factor delaying time to

surgery in patients with head injury
Young-Ju Kim

Aim. To investigate factors influencing time from patient’s arrival at the emergency department to surgery in patients with head
injury.
Background. A better understanding of factors influencing variation in time from patient’s arrival at the emergency department
to surgery for patients with head trauma could reduce mortality and morbidity associated with injury.
Design. A cross-sectional study of secondary data.
Methods. The sample represented 493 patients with head injury requiring surgery from the 17 level I and II trauma centres.
Data were extracted from the National Trauma Data Bank version 4.0. Two-level hierarchical models were used to analyse data
at the patient level while incorporating a unique random effect for each trauma centre. Factors entered in the models included
patient characteristics and trauma centre characteristics.
Results. Patients with a Glasgow coma scale score of 3–8 in the first ED assessment had earlier time to surgery compared with
those with a Glasgow coma scale of 13–15 (b = 0Æ31, 95% CI = 0Æ43–0Æ18). Patients who arrived at the hospital during the
nighttime (6pm–8am) had a significantly delayed time to surgery than those who arrived during the daytime (8am–6pm)
(b = 0Æ15, 95% CI = 0Æ26 to 0Æ04).
Conclusions. The more severely the injured patients were the faster surgery was performed. The time, when patients arrived to
the emergency department was found to be a significant factor influencing time to surgery. Patients who arrived at emergency
department at night had longer time to surgery than those who arrived during daytime, despite they were more severely head
injured than those who arrived during the day.
Relevance to clinical practice. When surgical intervention in head-injured patients is anticipated, especially during the night
shift, time from patient’s arrival at emergency department to surgery should be consistently assessed to identify opportunities
for improvement in the structure and process of trauma care.

Key words: admission, head injury, night, nurses, nursing, surgery

Accepted for publication: 18 April 2009

mately 1Æ5 million cases of head injury and an estimated

Introduction
Trauma is the fourth leading cause of death among all
Americans (O’Phelan et al. 2008) and accounts for 50–75%
of deaths for those 15–34 years of age (Anderson & Smith
2003). Among unintentional and intentional injuries, head
trauma is the leading cause of trauma-related deaths and
disability (Blank-Reid & Reid 2000). There are approxi-
50,000 deaths in the USA (O’Phelan et al. 2008).
Almost all injuries are treatable by available interventions,
however, the interval between injury and definitive care is
critical (Chesnut 2004). Trauma as a time-sensitive injury is
promulgated by the golden hour, the time when provision of
appropriate resuscitation and definitive care could save
those who otherwise would die (Maull et al. 1994). With

Author: Young-Ju Kim, PhD, RN, ACNP, Assistant Professor, Dongseon-dong, Seongbuk-gu, Seoul 136-742, Korea. Telephone:
College of Nursing, Sungshin Women’s University, Seoul, Korea +82 2 920 7510.
Correspondence: Young-Ju Kim, Assistant Professor, 249-1 E-mail: yjkim727@sungshin.ac.kr

 2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770 2763
doi: 10.1111/j.1365-2702.2009.03024.x
Y-J Kim
injury to the brain, unlike injury elsewhere in the body,
there is no room for expansion to accommodate the
increased fluid. As swelling of the brain progresses,
decreased blood flow to the remaining viable neurons
extends injury and causes further swelling. Therefore the
mainstays of head injury therapy are rapid diagnosis,
aggressive surgical evaluation and prevention of secondary
insults (Silver et al. 2005). Time is particularly of the
essence in preventing secondary insults mostly caused by
ischaemia, hypoxia, or intracranial mass. Whether a mass
lesion or severe swelling of brain is revealed through
diagnostical examinations, emergent surgical management
must be performed (Chesnut & Servadei 1999, Silver et al.
2005). Delays in diagnosis and treatment may result in
irreversible secondary injury and increasing morbidity and
mortality.
From a systems perspective, it is critical to identify factors
that predict a need for immediate neurosurgery in patients
with head injury (Esposito et al. 2005). This will allow for
identification of such patients in the hospital phase of care,
making it appropriate to bypass hospitals without neurosur-
gery services.
Patient characteristics identified as important factors
affecting time to surgery can serve as neurosurgical triage
criteria in the emergency department (ED), e.g., needing
immediate evaluation and surgical intervention by a trauma
surgeon or neurosurgeon. The triage system categorised by
physiological derangements and mechanism of injury resulted
in an increase in the early involvement of trauma service
while decreasing ED time and minimising inappropriate
triage (Kaplan et al. 1997). Accordingly, the recognition of
factors influencing time to surgery could lead to the devel-
opment of triage for emergency care which could reduce
mortality and morbidity associated with head injury. The
purpose of this study was to investigate factors influencing
time from patient’s arrival at the ED to surgery in patients
with head injury.
Methods
The research question for this study was: what trauma
centre and patient characteristics are related to time from
patient’s arrival at ED to surgery in patients with head
injury?
Setting and sample
Selection of trauma centres
The data for the study were extracted from the National
Trauma Data Bank (NTDB) version 4.0, which contains
2764  2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770
over 1,126,000 records from the 405 trauma centres in the
43 states, territories and the District of Columbia of the
USA from 1994–2003 (American College of Surgeons
Committee on Trauma 2004). Level I and II trauma centres
that treated patients with traumatic head injury were se-
lected for the study because level III, IV and V trauma
centres generally have limited or no neurosurgical coverage
(Committee on Trauma American College of Surgeons
1998). Table 1 explains what Level I through V trauma
centres are in the USA. To minimise the variation in criteria
by designation agencies, only trauma centres designated by
either American College of Surgeons Committee on Trauma
(ACSCOT) or state were selected. Because the time from ED
arrival to surgery was critical to this study, only trauma
centres where these data were available and had more than
five patients with the selected neurosurgeries were selected.
The final study sample represented 17 trauma centres across
the nation.
Table 1 Classification of trauma centre level in the USA
Level Descriptions
Level I Generally serves large cities or population-dense areas
Expected to admit at least 1200 trauma patients a year,
and of those, 20% with an Injury Severity Score (ISS)
‡15, or 35 patients per surgeon with an ISS ‡15
24-hour in-house availability of the attending surgeon
Establishes the trauma performance improvement
programme
Expected to conduct trauma research and be a leader in
education, prevention and outreach activities
Level II Provides comprehensive trauma care as a supplement to
a level I centre in a population-dense areas or the lead
trauma facility in less population-dense areas
Must meet the same criteria as level I except volume
performance
Volume performance standards depend on geographical
area served, population density, resources available
and the maturity of system
Level III Has continuous general surgical coverage
Has the capability to manage the initial care of the
majority of injured patients and transfer agreement
with other trauma hospitals
The general surgeon must be promptly available for all
major resuscitation
Must be involved with prevention and have an active
outreach programme
Level IV/V Usually located in a rural area
Supplements care within a larger network of hospitals
Provides initial evaluation and assessment of injured
patients
Have 24-hour emergency coverage by a physician
Specialty coverage may or may not be available
Clinical issues Nighttime admission to emergency department

Selection of patients independent variables including dummy variables was 366

The target sample for the study was 4412 patients with a patients (Cohen 1988). The final study sample included 493
head injury diagnosis and a neurosurgical procedure admitted patients from 17 trauma centres that met the inclusion and
from 2002–2003. The diagnosis and procedure codes related exclusion criteria. This was adequate to reach a desired
to head injury were recorded using the International Classi- power to detect an effect size. Based on data supplied in a
fication of Disease 9th Revision Clinical Modification (ICD- public use format, patients and trauma centres cannot be
9-CM) (Centers for Medicare and Medicaid Services 2003). identified. An exemption was approved by the University
Through the review of textbooks, literature and clinical Institutional Review Board.
experts, the ICD-9-CM diagnosis codes of 800Æ0–801Æ9,
803Æ0–804Æ9, 850Æ0–852Æ5, 853Æ0–853Æ1, 854Æ0–854Æ1 and
Measures
ICD-9-CM surgical procedure codes of 1Æ20–1Æ25, 1Æ39 were
selected (Table 2). The focus of the study was patients who Hospital teaching status, ownership, designation type and
had acute lesions in the head caused by traumatic injury level of trauma centre were selected as hospital characteristics
and who subsequently required a surgical procedure, there- that might relate to the outcomes in this study. Age, gender,
fore, only patients who underwent surgery within 72 hours race, diagnosis, Glasgow coma scale (GCS) score in the first
after ED arrival were included in this study. Intracranial ED assessment, injury type, injury severity score (ISS) and ED
haematoma lesions were considered to be acute, when they arrival time were patient characteristics associated with
require evacuation within 72 hours of trauma in previous outcomes to be examined in the study.
studies (Haselsberger et al. 1988, Massaro et al. 1996). Time to surgery, as an outcome variable, was defined as the
Patients with fractures of face bones (ICD-9-CM Code 802) total length of time from the patient’s arrival at the ED until
were excluded on assumption that injuries of this sort are the patient received a neurosurgical procedure. Cases, whose
unlikely to give rise to serious brain damage (Jennett & time from ED arrival to surgery was larger than total time of
MacMillan 1981, Sosin et al. 1995, Richardson 2000). Also, hospital stay or whose total time to surgery was zero were
patients were excluded if they had any burn injury, were excluded due to the unreliability of data.
transferred from another facility, died on arrival, or died in
the ED before surgery. The required sample size for the effect
Data analysis
size of 0Æ5 with alpha = 0Æ05, desired power = 0Æ80 and 14
Descriptive analyses were conducted for all variables to
Table 2 Diagnoses and surgical procedures used to select patients describe the characteristics of the sample and to assess
with head injury
statistical assumptions for subsequent analysis. Assumptions
ICD-9-CM code Diagnosis of multicollinearity, linearity and normality and presence of
800 – Fracture of skull outliers were also examined.
800Æ0–800Æ9 Fracture of vault skull The hierarchical models, also called multilevel models were
801Æ0–801Æ9 Fracture of base of skull employed for the regression analysis. Misestimated standard
803Æ0–803Æ9 Other and unqualified skull fracture errors occur with multilevel data, when the dependence
804Æ0–804Æ9 Multiple fractures involving skull or face
among individual patients in the same hospital is not taken
with other bones
into account (Bryk & Raudenbush 1992). Hierarchical
850 – Intracranial injury
850Æ0–850Æ9 Concussion models resolve this problem by incorporating into the
851Æ0–851Æ9 Cerebral laceration and contusions statistical model a unique random effect for each hospital.
852Æ0–852Æ5 Subarachnoid, subdural, and extradural Also, heterogeneity of regression occurs, when the relation-
haemorrhage following injury ships between patient characteristics and outcomes vary
853Æ0–853Æ1 Other and unspecified intracranial
across hospitals (Bryk & Raudenbush 1992). Therefore,
haemorrhage following injury
854Æ0–854Æ1 Intracranial injury of other and hierarchical models allow simultaneous examination of the
unspecified nature impact of patient characteristics and trauma centre charac-
teristics on time to surgery.
ICD-9-CM code Surgical procedure
Two-level hierarchical models were formulated. At level 1,
01Æ20–01Æ25 Craniotomy and craniectomy, including the unit of analysis was the individual patient and each
cranial decompression, removal of outcome was represented as a function of a set of patient
extradural haematoma
characteristics. In other words, the within-hospital variability
01Æ39 Drainage of intracerebral haematoma
of time to surgery was explained by patient characteristics.

 2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770 2765
Y-J Kim

At level 2, the unit of analysis was the hospital. The regression
coefficients in the level 1 model for each hospital were
included as outcome variables that were presumed to depend
on specific hospital characteristics. Thus, the level 2 model
examined whether hospital characteristics explained variabil-
ity of time to surgery between hospitals. Hierarchical multiple
regression analysis was used because time to surgery was a
continuous dependent variable. To examine model fit, the 2
restricted log likelihood and Akaike’s information criterion
(AIC) were 489Æ63 and 497Æ63, respectively. The intraclass
correlation coefficient (ICC) was calculated to measure the
extent to which observations are not independent of trauma
centres. The ICC for the model time to surgery was close to
zero, which means that most variability in time to surgery
is because of differences between patients, not differences
between trauma centres. The software programmes used for
multivariate analyses were the Linear Mixed Model in the
SPSS 12 (SPSS Inc., Chicago, IL, USA).
Results
Bivariate analysis
The median time from patient’s arrival at ED to neurosurgery
was 155 minutes (2Æ6 hours) and the overall mean time was
480 minutes (six hours). Table 3 illustrates trauma centre
characteristics by the four time to surgery categories. Nearly
half of surgeries (38–48%) were performed within two hours
after arrival to the ED.
Table 4 compares the patient characteristics by the four
time categories. Half of patients with penetrating type of
injury underwent surgery within two hours, while 38Æ6%
of patients with blunt injury did. Using the GCS score, 57Æ2%
of severely head-injured patients (3–8 of GCS score) had
surgery within two hours, compared with only 22Æ4% of
mildly head-injured patients (13–15 of GCS score). In the
four time categories, patients who arrived to the ED during
daytime (8am–6pm) underwent surgery significantly sooner
compared with patients who arrived to the ED at night
(6pm–8am).
Multilevel analysis
Table 5 describes the results of hierarchical multiple regres-
sion model for time to surgery. The severely head-injured
patients (3–8 of GCS) had surgery, significantly, sooner than
the mildly head-injured patients (13–15 of GCS) (b = 0Æ305,
95% CI = 0Æ434 to 0Æ177). Among the ISS groups, those
who had a ISS of less than 16, which indicates least severely
injured, were likely to have a longer time to surgery than the
reference group who had an ISS of more than 34 (b = 0Æ379,
95% CI = 0Æ095–0Æ663). Even after controlling for severity,
patients who arrived to the ED during the night were likely to
have surgery, significantly, later than those who arrived
during the day (b = 0Æ147, 95% CI = 0Æ258 to 0Æ036).
Discussion
Significant factors delaying time to surgery
In the early care of the head-injured patient, after a good
clinical assessment has been conducted with appropriate

Table 3 Distribution of four time categories by trauma centre characteristics

Time to surgery (%)

Valid
Trauma centre patient Less than Two to Four to Greater than
characteristics (n) two hours four hours six hours six hours

Ownership
Public 248 37Æ9 27Æ0 6Æ0 29Æ1
Private 189 41Æ8 26Æ4 8Æ5 23Æ3
Teaching status
University TC 264 40Æ2 25Æ4 8Æ7 27Æ7
Non-university TC 173 38Æ7 28Æ9 4Æ6 25Æ8
Designation type
ACS 362 38Æ4 27Æ6 8Æ0 26Æ0
State 71 47Æ9 21Æ1 1Æ4 29Æ6
Level of TC
Level I 245 40Æ0 26Æ1 6Æ9 26Æ9
Level II 192 39Æ1 27Æ6 7Æ3 26Æ0

Chi-square test found no significant differences.

TC, trauma centre; ACS, American College of Surgeons.

2766  2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770
Clinical issues Nighttime admission to emergency department

Table 4 Distribution of four time categories by patient characteristics

Time to surgery (%)

Valid Less than Two to Four to Greater than

Patient characteristics (n) two hours four hours six hours six hours

Gender
Male 323 40Æ9 25Æ1 7Æ1 26Æ9
Female 114 36Æ0 31Æ6 7Æ0 25Æ4
Age
<18 year 45 33Æ3 22Æ2 8Æ9 35Æ6
18–64 year 256 43Æ0 25Æ8 5Æ5 25Æ8
‡65 year 52 25Æ0 34Æ6 11Æ5 28Æ8
Diagnosis
Skull fracture 183 45Æ4 23Æ0 7Æ7 24Æ0
Multiple IH 62 38Æ7 22Æ6 3Æ2 35Æ5
SDH 101 33Æ7 31Æ7 7Æ9 26Æ7
Other single IH 77 36Æ4 32Æ5 6Æ5 24Æ7
Race or ethnic group
White, non-hispanic origin 268 42Æ5 26Æ1 5Æ6 25Æ7
Black, non-hispanic origin 31 22Æ6 45Æ2 3Æ2 29Æ0
Hispanic 32 43Æ8 18Æ8 6Æ3 31Æ3
Other 16 25Æ0 25Æ0 6Æ3 43Æ8
Injury type
Blunt injury 404 38Æ6 27Æ2 6Æ9 27Æ2
Penetrating injury 31 51Æ6 19Æ4 9Æ7 19Æ4
First GCS score in ED
Severe (3–8) 194 57Æ2 18Æ6 5Æ2 19Æ1**
Moderate (9–12) 40 30Æ0 37Æ5 12Æ5 20Æ0
Mild (13–15) 156 22Æ4 34Æ0 8Æ3 35Æ3
Injury severity score
<16 24 12Æ5 29Æ2 4Æ2 54Æ2**
16–24 130 30Æ0 30Æ0 10Æ0 30Æ0
25–34 221 45Æ2 25Æ8 6Æ3 22Æ6
>34 60 51Æ7 20Æ0 5Æ0 23Æ3
ED arrival time
Daytime (8am–6pm) 203 36Æ5 32Æ0 8Æ9 8Æ4*
Nighttime (6pm–8am) 234 42Æ3 22Æ2 5Æ6 9Æ8

*p < 0Æ05; **p < 0Æ01.

IH, intracranial haemorrhage; SDH, subdural haemorrhage; GCS, glasgow coma scale; ED, emergency department.

diagnostic studies, patients can quickly be placed into non-
operative and operative groups (Ducker & Aryanpur 1990).
Treatment strategies for head injury include medical and
pharmacological measures as well as surgical interventions
such as craniotomy, craniectomy, ventriculostomy and intra-
cranical haematoma removal (Soukiasian et al. 2002). This
study investigated patients who underwent a faster surgery
after they were placed into a group needing surgical inter-
vention. Not surprisingly, the GCS score and ISS initially
measured in the ED were found to be significant in predicting
time to surgery. The severely head-injured patients with
a GCS score of 3–8 were more likely to have a faster surgery.
Almost 76% of severely head-injured patients had surgery
within four hours after arrival at ED. Patients who require
surgery initially are those with intracranial haematoma that
is either causing deterioration or interfering with recovery
(Ducker & Aryanpur 1990). The GCS score provides a stan-
dardised approach to comprehensive neurological assessment
for patients with head injury. The GCS findings form the
basis for clinical management decisions, such as the necessity
for computerised tomography, surgical intervention and/or
drug modalities (Fischer & Mathieson 2001). Although there
is some controversy over whether accuracy is diminished by
summing the numerical scores of the three components (eye,
verbal and motor responses) and whether conditions or
treatments affect the measuring of GCS score, the lower GCS
score reflects severe deterioration in level of consciousness
and neurological dysfunction (Sternbach 2000).

 2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770 2767
Y-J Kim

Table 5 Multilevel linear regression for

Variables Beta SE 95% CI
time to surgery (n = 332)
Intercept 2Æ61* 0Æ18 0Æ25 2Æ97
Age (reference group: ‡65 years)
<18years 0Æ09 0Æ10 0Æ11 0Æ29
18–64 years 0Æ12 0Æ08 0Æ28 0Æ03
GCS (reference group: mild head injury)
Moderate head injury 0Æ11 0Æ09 0Æ29 0Æ06
Severe head injury 0Æ31* 0Æ07 0Æ43 0Æ18
ISS (reference group: >34)
<16 0Æ38* 0Æ14 0Æ09 0Æ63
16–24 0Æ01 0Æ06 0Æ18 0Æ21
25–34 0Æ08 0Æ03 0Æ08 0Æ23
ED arrival time
Day (vs. Night) 0Æ15* 0Æ56 0Æ26 0Æ04
Level I TC (vs. Level II TC) 0Æ04 0Æ01 0Æ19 0Æ28
ACS designated TC (vs. state designated) 0Æ09 0Æ12 0Æ41 0Æ23
University TC (vs. non-university TC) 0Æ00 0Æ04 0Æ21 0Æ21

*p < 0Æ01.
Time to surgery was log transformed.
GCS, glasgow coma scale; ISS, injury severity score; ED, emergency department; TC, trauma
centre; ACS, American College of Surgeons.

Interestingly, the study found that patient’s arrival time at
ED was significantly associated with time to surgery. Patients
who arrived at ED during the day (8am–6pm) had a faster
surgery than those who arrived at night (6pm–8am).
Moreover, despite patients who arrived at night having more
severe head injures (average GCS of 8), they had longer time
to surgery than those who arrived in the day (average GCS of
10). In the care of the critically injured, causes of delays
included a sequential approach to resuscitation and investi-
gation, limited staff and operating room availability and
failure to call the trauma team (McNicholl & Dearden 1992).
Optimal hospital care of head-injured patients requires
commitment from all emergency medical and nursing staff
and all support personnel necessary for the care of trauma
patients (Trunkey 1990).
Relevance to clinical practice
The key component of surgical care for head-injured patients
is the neurosurgeon. Given the paucity of neurosurgeons
relative to other surgical specialties, the availability of
neurosurgeons is often cited as a limiting factor in the
functioning of trauma systems (Meagher & Narayan 2000).
In a national survey of level I and II trauma centres in 2003,
22% of level I and II trauma centres had an in-house
neurosurgeon (Kim et al. 2007). Most of the neurosurgeons
(75%) were available through on-call systems, especially
during the nighttime. In 3% of level I and II trauma centres,
they were not available to trauma care at all (Kim et al. 2007).
Rural hospitals reported more difficulty in ensuring adequate
neurosurgical coverage in their emergency room than urban
facilities (Meagher & Narayan 2000). One of factors cited by
neurosurgeons for this lack of availability included usually
inconvenient work hours and inadequate reimbursement for
difficult work (Meagher & Narayan 2000). Because of these
factors, the neurosurgical coverage is worse at night, partic-
ularly in a community where a neurosurgeon provides
simultaneous coverage of two or more hospitals.
Nurses are available at the time of patient’s arrival to ED in
most trauma centres, while neurosurgeons are not (Kim et al.
2007). This fact emphasises a significant role of nurses in
coordinating the trauma care process. A nurse who is able
to work independently and cooperatively with the profes-
sional staff, able to identify and set priorities and able to
work on several different units may be the most qualified
person for the coordination of trauma care (Sinclair 2006).
Having a nurse coordinator for special trauma care will lead
to less delay in transporting a trauma patient who needs
surgery to the operating room by alerting all involving
personnel simultaneously.
Another possible cause for delay of surgery is lack of
immediate availability of operating room staff, including
anaesthesiologists, circulating and scrub nurses and techni-
cians. In most trauma centres, operating room staffs are
available on-call at home at night (Kim et al. 2007).
Moreover, the challenge to the operating room staff is to
function quickly and efficiently in any emergency situation at
any time. Ross (2006) suggested a way to train nurses and

2768  2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770
Clinical issues
technologists for trauma surgery by making a video that
would show what it looks like if one is ready for a trauma
patient in surgery. Again, when surgical intervention in head-
injured patients is anticipated, based on neurological assess-
ment and injury severity, time from patient’s arrival at ED to
surgery could be reduced by earlier decision-making, alerting
key personnel such as surgeon and operating room staff
promptly and performing tasks in parallel. This reinforces the
need to assess the structure and process of trauma care
consistently during the night shift to identify opportunities
for improvement.
Study limitations
There were several limitations in conducting this study. First,
other clinical variables likely affecting outcomes in patient
with head injury were not assessed in the study because of
unavailability of them in the NTDB data set. A second
limitation relates to the generalisabillity of the findings. The
NTDB is the largest database of its kind, however, it is not
population based. Voluntary submission of data from only a
limited number of trauma centres may introduce selection
bias. Moreover, the findings for this head trauma population
may not be generalisable to other trauma populations. Lastly,
observed trauma centre characteristics in this study may not
fully account for the difference between centres. Other
variables related to organisational structure and process
may affect time to surgery and outcomes. These variables
should be assessed in future studies.
Conclusion
The severity of injury measured with GCS score and ISS was
a significant predictor of time to surgery. The more severely
injured the patients were, the faster surgery was performed.
The time when patients arrived to the ED was found to be
a significant factor influencing time to surgery. Patients who
arrived at the ED at night had longer time to surgery than
those who arrived during the day, despite being more severely
head injured than those who arrived during the day.
Acknowledgement
This work was supported by the Sungshin Women’s Univer-
sity Research Grant.
Contributions
Study design: YJK; data collection and analysis: YJK and
manuscript preparation: YJK.
 2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770
Nighttime admission to emergency department
Conflict of interest
None.
References
American College of Surgeons Committee on Trauma (2004)
National Trauma Data Bank Report 2004. American College of
Surgeons, Chicago, USA.
Anderson RN & Smith BL (2003) Death: leading causes for 2001.
National Vital Statistics Report 52, 1–85.
Blank-Reid C & Reid PC (2000) Penetrating trauma to the head.
Critical Care Nursing Clinics of North America 12, 477–487.
Bryk AS & Raudenbush SW (1992) Hierarchical Linear Model.
SAGE Publication, Newbury Park, USA.
Centers for Medicare and Medicaid Services (2003) ICD-9-CM
Expert for Hospitals-Volumes 1, 2 & 3. 6th edn. Ingenix, Salt Lake
City, UT, USA.
Chesnut RM (2004) Management of brain and spine injuries. Critical
Care Clinics 20, 25–55.
Chesnut RM & Servadei F (1999) Surgical treatment of post-trau-
matic mass lesions. In Traumatic Brain Injury (Marion DW ed).
Thieme, New York, pp. 81–99.
Cohen J (1988) Statistical Power Analysis for the Behavioral Science.
Lawrence Erlbaum, Hillsdale.
Committee on Trauma American College of Surgeons (1998)
Resources for Optimal Care of the Injured Patient. American
College of Surgeons, Chicago, USA.
Ducker TB & Aryanpur J (1990) Central nervous system. In Early
Care of the Injured Patient (Moore EE 4th ed.). B.C. Decker,
Philadelphia, USA, pp. 25–55.
Esposito TJ, Reed RL, Gamelli RL & Luchette FA (2005) Neuro-
surgical coverage: essential, desired, or irrelevant for good patient
care and trauma center status. Annals of Surgery 242, 364–370.
Fischer J & Mathieson C (2001) The history of the Glasgow Coma
Scale: implications for practice. Critical Care Nursing Quarterly
23, 52–58.
Haselsberger K, Pucher R & Auer LM (1988) Prognosis after acute
subdural or epidural haemorrhage. Acta Neurochirurgica 90,
111–116.
Jennett B & MacMillan R (1981) Epidemiology of head injury.
British Medical Journal 282, 101–104.
Kaplan LJ, Santora TA, Blank-Reid CA & Trooskin SZ (1997)
Improved emergency department efficiency with a three-tier trau-
ma triage system. Injury 28, 449–453.
Kim YJ, Xiao Y & Mackenzie CF (2007) Availability of trauma
specialties in Level I and II trauma centers: a national survey.
Journal of Trauma 63, 676–683.
Massaro F, Lanotte M, Faccani G & Triolo C (1996) One hundred
and twenty-seven cases of acute subdural haematoma operated on.
Correlation between CT scan findings and outcome. Acta Neuro-
chirurgica 138, 185–191.
Maull KI, Enderson BL & Frame SB (1994) Comprehensive
management of the trauma patient. In Current Practice of
Trauma Surgery (Levine BA, Copeland EM III, Howard R, Su-
german HJ & Warshaw AL eds). Churchill Livingstone, New
York, pp. 3–18.
2769
Y-J Kim
McNicholl BP & Dearden CH (1992) Delays in care of the critically
injured. British Journal of Surgery 79, 171–173.
Meagher RJ & Narayan RK (2000) The triage and acute manage-
ment of severe head injury. Clinical Neurosurgery 46, 127–142.
O’Phelan K, McArthur DL, Chang CWJ, Green D & Hovda DA
(2008) The impact of substance abuse on mortality in patients with
severe traumatic brain injury. Journal of Trauma 65, 674–677.
Richardson JTE (2000) Clinical and Neuropsychological Aspects of
Closed Head Injury. Psychology Press, Philadelphia, USA.
Ross C (2006) Training nurses and technologists for trauma surgery.
Journal of Trauma Nursing 13, 193–195.
Silver JM, McAllister TW & Yudofsky SC (2005) Textbook of
Traumatic Brain Injury. American Psychiatric Publishing, Inc,
Arlington, VA.
2770  2010 Blackwell Publishing Ltd, Journal of Clinical Nursing, 19, 2763–2770
Sinclair TD (2006) The role of the rapid response nurse: hospitalwide
and in trauma resuscitations. Journal of Trauma Nursing 13,
175–177.
Sosin DM, Sniezek JE & Waxweiler RJ (1995) Trends in death
associated with traumatic brain injury, 1979 through 1992. JAMA
273, 1778–1780.
Soukiasian HJ, Hui T, Avital I, Eby J, Thompson R & Kleisli T
(2002) Decompressive craniectomy in trauma patients with severe
brain injury. American Surgeons 68, 1066–1071.
Sternbach GL (2000) The Glasgow coma scale. Journal of Emergency
Medicine 19, 67–71.
Trunkey DD (1990) Trauma: a public health problem. In Early Care
of the Injured Patient (Moore EE 4th ed), B.C. Decker, Philadel-
phia, PA, pp. 3–11.