close

Вход

Забыли?

вход по аккаунту

?

injuryprev-2017-042435

код для вставкиСкачать
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
IP Online First, published on October 22, 2017 as 10.1136/injuryprev-2017-042435
Original article
Epidemiology of paediatric trauma presenting to US
emergency departments: 2006–2012
Jacob B Avraham,1 Misha Bhandari,2,3 Spiros G Frangos,4 Deborah A Levine,5,6
Michael G Tunik,5,6 Charles J DiMaggio4,7
1
Department of Surgery, New
York University School of
Medicine, New York, NY, USA
2
Department of Surgery, New
York University School of
Medicine, New York City, New
York, USA
3
Department of Emergency
Medicine, New York
Presbyterian, The University
Hospital of Columbia and
Cornell, New York, NY
4
Department of Surgery, Division
of Acute Care and Trauma
Surgery, New York University
School of Medicine/Bellevue
Hospital Center, New York City,
New York, USA
5
Department of Pediatrics,
New York University School
of Medicine/Bellevue Hospital
Center, New York City, New
York, USA
6
Ronald O Perelman Department
of Emergency Medicine, New
York University School of
Medicine/Bellevue Hospital
Center, New York City, New
York, USA
7
Population Health, New York
University School of Medicine,
New York, NY, USA
Correspondence to
Dr Charles J DiMaggio,
Department of Surgery, Division
of Trauma and Acute Care
Surgery, New York University
School of Medicine/Bellevue
Hospital Center, New York
NY 10016, USA; ​Charles.​
[email protected]​nyumc.​org
Received 17 April 2017
Revised 6 October 2017
Accepted 10 October 2017
Abstract
Background Traumatic injury is the leading cause
of paediatric morbidity and mortality in the USA. We
present updated national data on emergency department
(ED) discharges for traumatic injury for a recent 7-year
period.
Methods We conducted a descriptive epidemiological
analysis of the Nationwide Emergency Department
Sample Survey, the largest and most comprehensive
database in the USA, for 2006–2012. Among children
and adolescents, we tracked changes in injury
mechanism and severity, cost of care, injury intent and
the role of trauma centres.
Results There was an 8.3% (95% CI 7.7 to 8.9)
decrease in the annual number of ED visits for traumatic
injury in children and adolescents over the study period,
from 8 557 904 (SE=5861) in 2006 to 7 846 912
(SE=5191) in 2012. The case-fatality rate was 0.04%
for all injuries and 3.2% for severely injured children.
Children and adolescents with high-mortality injury
mechanisms were more than three times more likely to
be treated at a level 1 trauma centre (OR=3.5, 95%
CI 3.3 to 3.7), but were more no more likely to die
(OR=0.96, 95% CI 0.93 to 1.00). Traumatic brain injury
diagnoses increased 22.2% (95% CI 20.6 to 23.9)
during the study period. Intentional assault accounted
for 3% (SE=0.1) of all child and adolescent ED injury
discharges and 7.2% (SE=0.3) of discharges among
15–19 year-olds. There was an 11.3% (95% CI 10.0
to 12.6) decline in motor vehicle injuries from 2009 to
2012. The total cost of care was $23 billion (SE=0.01), a
78% increase from 2006 to 2012.
Conclusions This analysis presents a recent portrait
of paediatric trauma across the USA. These analyses
indicate the important role and value of trauma centre
care for injured children and adolescents, and that the
most common causes and mechanisms of injury are
preventable.
Introduction
To cite: Avraham JB,
Bhandari M, Frangos SG,
et al.Inj Prev Published Online
First: [please include Day
Month Year]. doi:10.1136/
injuryprev-2017-042435
Injury is the leading cause of paediatric mortality
in the USA.1 Annually, 9.2 million children present
to US emergency departments (ED) for non-fatal
injuries, resulting in significant healthcare resource
utilisation2 and morbidity. The CDC estimates
that childhood injuries account for 225 000 yearly
hospitalisations and cost $87 billion,3 impacting
long-term quality of life and functional outcomes.4
Protecting society’s most vulnerable requires
accurate injury data. We know that motor vehicle
crashes (MVCs) are the leading cause of death
between 5 and 19 years of age, while drowning
surpasses MVCs in children aged 1–4 (most deaths
under age 1 result from congenital anomalies).5
According to the US Centers for Disease Control
and Prevention (CDC), during the first decade of
this century, childhood mortality rates fell 29%
among those aged 1–19 years. 6 Risk factors associated with mortality are both direct (ie, injury mechanism) and indirect (ie, lack of health insurance),7 8
and much focus has been placed on mitigating
injury risk.9 But beyond preventive strategies, how
we treat injured children is also of vital importance.
Several studies have examined the impact of
trauma centre designation on outcomes. A literature review concluded that injured children treated
at dedicated paediatric centres had improved
mortality and functional outcomes compared
with those treated at adult trauma centres and
non-trauma centres.10 However, dedicated paediatric trauma centres are scarce,11 thus the question
of optimal management remains, despite a greater
focus in recent years on paediatric readiness across
US EDs.12
In this study, we update the epidemiology of
paediatric ED trauma care in the USA by analysing
the Agency for Healthcare Research and Quality’s
(AHRQ) Healthcare Cost and Utilization Project
(HCUP) Nationwide Emergency Department
Sample Survey (NEDS) database. We track changes
in injury mechanism and severity, cost of care,
injury intent and the role of trauma centres. We aim
to renew current understanding of paediatric traumatic injury to optimise interventions.
Methods
Data sources and designs
Data were obtained from the U.S. Agency for
Healthcare Research and Quality (AHRQ) Healthcare Cost and Utilization Project (HCUP) NEDS
(2006–2012), the largest single publicly available
ED database in the USA, based on a 20% stratified single-cluster sample of hospital-based EDs. In
2013, thirty states participated in NEDS accounting
for 66% of all national ED visits, with core files
containing 100% of visits from sampled hospitals.
The most recent NEDS database contains about
30 million ED records. 13 Hospitals are defined
as non-federal general and specialty hospitals
including public hospitals and academic medical
centres. Additional stratification variables include
geographic area, urban/rural, ownership, trauma
centre and teaching status, and bed size.
Annual core files were read into an R data frame.
14
Survey-adjusted point estimates and standard
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Copyright Article author (or their employer) 2017. Produced by BMJ Publishing Group Ltd under licence.
1
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
errors SEs were verified against estimates obtained from a
publicly available HCUP online query system. Traumatic injury
discharges were identified using principle ICD-9 (International
Classification of Diseases) diagnosis codes for acute injury:
800–904.9, 909.4, 909.9, 910–994.9, 995.5–995.59 and
995.80–995.85. Discharges with codes for ‘late effect’ primary
diagnoses (ICD 905.0–909.9), insect bites (910.4, 910.5, 911.4,
911.5, 912.4, 912.5, 913.4, 913.5, 914.4, 914.5, 915.4, 915.5,
916.4, 916.5, 917.4, 917.5, 919.4, 919.5), poisonings (960.0–
964.9, 965.00–965.02, 965.09, 965.1, 965.4, 965.5, 965.61,
965.69, 965.7–969.0, 969.00–969.09, 969.70–969.73, 969.1–
969.7, 967.0–967.9, 969.79, 969.8–980.9, 970.81, 970.89,
981, 982.0–985.9, 986, 987.0–989.7, 989.81:989.89, 989.9,
990, 991.0–995.2, 995.20–995.29, 995.3, 995.4), anaphylaxis
(995.60–995.69, 995.7) and some additional miscellaneous diagnoses (malignant hyperthermia, systemic inflammatory response
syndrome, malfunctioning cardiac devices, 995.86–996.00)
were removed. Traumatic brain injury (TBI) was defined using
ICD-9 codes specified in the Barrel matrix (https://www.​cdc.​gov/​
nchs/​data/​ice/​final_​matrix_​post_​ice.​pdf). The R code to recreate
our classification system can be found at http://www.​injuryepi.​
org/​resources/​Misc/​knitHCUPCode.​pdf, pp 99–103.
The R ‘ICD-9’ package was used to apply descriptors to
ICD-9 codes and external cause of injury E-codes. All patients
under the age of 20 were included in the analysis, consistent with
the WHO’s definitions of infant, child and adolescent.15 While
the American College of Surgeons uses a maximum age of 15
to define ‘pediatric’ trauma,16 we chose a more inclusive definition consistent with the WHO to allow for comprehensive analysis, including comparisons with the older adolescent subgroup.
As such, any reference to ‘children’ or ‘pediatric’ implies this
broader definition.
Injury severity was quantified using the ICD-derived
ISS (ICISS)17 and categorised as severe versus non-severe. ICISS
was proposed in 1996 to estimate injury severity using ICD
administrative codes in hospital discharge data. It is calculated
in two steps. First, survival risk ratios for each injury diagnosis
in a data set are ‘calculated as the ratio of the number of times
a given ICD-9 code occurs in (surviving patients) to the total
number of occurrences of that code.’ Second, the ICISS for an
individual patient is calculated as ‘the product of all the survival
risk ratios for each of an individual patient’s injuries (for as many
as ten different injuries).’18 The ICISS is then defined as the probability of patient surviving their injuries and ranges from 0 to 1.
ICISS values below 0.94 were used to categorise patients into
those most severely injured,19 with ≥6% probability of death.
ICISS has been used previously, yielding an OR of 6.75 (95%
CI 6.48 to 7.03) in a multivariate logistic regression analysis of
trauma mortality.20 We dichotomise injury severity into severe
versus less severe both to capture injury acuity and to address
inherent problems with ISS as a continuous variable. 21 Yet, this
can result in the loss of important information about intercategorical differences, and our choice of 0.94 as the cut-off for
severe versus non-severe injury, while not arbitrary, is subject to
debate.22 Thus, in our study, an ICISS of 0.93 is ‘equivalent’ to
0.01. Injury severity scoring itself is inherently statistically problematic. Both the traditional and ICD-derived ISS behave poorly
as a continuous variable, with some authors recommending, ‘that
for statistical or analytical purposes the ISS/NISS should not be
considered a continuous variable’ [65]. While ICISS has been
reported to generally perform as well as ISS [66], it also behaves
similarly poorly as a continuous variable. We found statistical
manipulations as log, square root and inverse logit transformations to be unhelpful in this regard and chose to dichotomise.
2
This approach also allowed us to calculate informative statistics
based on probabilities, such as ORs.
To control for chronic conditions and other non-traumatic
injury diagnoses, a Charlson comorbidity index (CCI) score23
was calculated for each discharge using the listed ICD-9 codes.
Because of the high proportion of patients without CCI conditions, the score variable was heavily skewed, thus categorised
into an indicator variable for patients with a CCI greater than
2. Trauma centre designations were based on AHRQ ‘HOSP_
TRAUMA’ identifiers in the data set. Teaching hospitals were
similarly identified. We used HCUP Clinical Classification Software system to categorise procedure codes. Primary ICD-9 codes
were categorised by the Barell matrix, an injury diagnosis tool
used internationally to standardise the classification of ICD-9
injury codes by nature of injury and body location.24
Costs were based on charges for each discharge. Unlike the
HCUP National Inpatient Sample (NIS) files, cost-to-charge files
are not available for NEDS. We conservatively calculated costs
as 42% of charges 25 and adjusted for inflation, standardising
to 2012 US$ (the final year of the study) based on an all-item
average yearly consumer price index obtained from the Bureau
of Labor Statistics.
Analyses
Analyses were based on weighted data adjusted for the complex
survey design of HCUP NEDS using the R packages ‘survey’
and ‘sqlsurvey’. The R ‘rms’ package was used for models not
supported by ‘survey’ and ‘sqlsvy’, with robust covariances
specified to account for clustering by strata within the NEDS
survey design. Because they are not available in ‘survey’, ‘sqlsvy’
or ‘rms’, relative risks RRs were calculated using simulations
based on survey-adjusted counts and SEs for the numerator and
denominator of the ratio. Each simulation consisted of 1000
random normal draws. Year-to-year variability of rates was
assessed with the robust linear model function in the R package
‘MASS’ and bootstrap simulations for confidence intervals (CIs)
using the ‘Boot’ function in the R package ‘car’. Descriptive
statistical analysis consisted of survey-adjusted counts, proportions, means, with associated SEs, and 95% CIs. Annual total
and age-specific rates were calculated using US census data
obtained from AHRQ. We analysed yearly data with overlying
loess smoothing lines and present tables of the proportion of
common injury categories and causes. As the analyses involved
comparisons and assessment of trends over time, we tested for
the assumption of linearity of the year-to-year variability in the
survey results using an approach appropriate to multiyear survey
data.26 We assessed the association of level 1 trauma centres with
the risk of injury mortality from common preventable mechanisms using unadjusted risk ratios and survey-adjusted logistic
regression controlling for age group, and indicator variables
for gender, injury severity (severe vs not severe) and teaching
hospital versus non-teaching hospital. The control variables
were selected based on univariate associations with fatality and
previous analyses [64].
Additionally, we calculated the age group-specific proportion
of paediatric ED injury discharges classified as ‘intentional’ by
NEDS variables available for 2009–2012. We also analysed an
indicator variable for ‘intentional self-harm’ available in the
data set for 2006–2012, and for ‘intentional assaults’ for the
years 2009–2012. We calculated the total and mean cost of the
most common external cause of injury codes discharged from
US EDs, as well as those codes most frequently resulting in ED
paediatric and adolescent traumatic injury deaths in the USA
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
Table 1 Emergency department visits for traumatic injury, US
hospitals, 2006–2012
Table 2 Emergency department visits for traumatic injury by
paediatric age group and year, US hospitals, 2006–2012
Year
Year
Age group
Frequency
SE
2006
8 557 904 (5861)
2006
0–4
2 005 581
3101
2007
8 462 884 (5882)
2007
2 035 411
3122
2008
8 136 649 (5524)
2008
2 002 452
3009
2009
7 932 646 (5729)
2009
2 025 641
3113
2010
7 759 111 (5429)
2010
1 965 521
3038
2011
7 772 993 (4965)
2011
2 016 284
2972
2012
7 846 912 (5191)
2012
2 014 199
2997
Total
56 469 098 (2309)
2006
1 654 396
2825
2007
1 639 682
2817
2008
1 556 495
2666
during the study period; case-fatality ratios for those mechanisms were then performed.
A complete set of notes and code to reproduce or adapt the
study methods are available upon request.
2009
1 522 182
2713
2010
1 487 311
2667
2011
1 539 211
2623
2012
1 614 906
2711
2 174 792
3228
Results
Descriptive statistics
2007
2 123 751
3200
2008
2 016 378
3031
2009
1 943 041
3060
2010
1 933 069
3028
2011
1 959 446
2931
2012
1 996 348
2988
2 723 133
3599
2007
2 664 038
3560
2008
2 561 322
3398
2009
2 441 781
3431
2010
2 373 209
3348
2011
2 258 049
3126
2012
2 221 457
3149
Count (SE)
In the USA, there were 56 469 098 (SE=2309) ED visits for
traumatic injuries in children under age 20 between 2006 and
2012 (table 1, figure 1). There was an 8.3% (95% CI 7.7 to 8.9)
decrease in the annual number of such visits to children over the
study period, from 8 557 904 (SE=5861) total visits in 2006 to
7 846 912 (SE=5191) in 2012. A more notable rate decline was
observed in the 15–19 year-old group, from 12 487 per 100 000
population in 2006 to 10 403 per 100 000 in 2012 (table 2,
figure 2). Females Women represented 40.1% (SE=0.01) of all
paediatric traumatic injury in the USA, and 32.7% (SE=0.14) of
patients who are severely injured. Table 3 lists the most common
external cause of injury codes for all paediatric and adolescent
trauma during the study period.
Figure 1 Emergency department (ED) visits for traumatic injury ages
0–19 per 100 000 population, US hospitals, 2006–2012.
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
2006
2006
5–9
10–14
15–19
Figure 2 Emergency department (ED) visit rate for traumatic injury
per 100 000 population by paediatric age group and year, US hospitals,
2006–2012.
3
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
Table 3 Most common external cause of injury codes and associated costs, US hospitals, 2006–2012
E-code
Description
Frequency
Lower CI
Upper CI
Total cost (US$) Lower CI
Upper CI
Mean cost
(US$)
Lower
CI
Upper
CI
E9179
Hit by object
4 914 963
4 905 774
4 924 152
1 668 158 229
1 663 124 863
1 673 191 595
385
384
386
E9170
Sports accident
3 155 178
3 147 668
3 162 687
1 331 836 056
1 326 683 173
1 336 988 939
485
483
486
E8859
Slip and fall
3 150 718
3 143 124
3 158 311
1 209 795 036
1 204 914 109
1 214 675 963
458
457
460
E8889
Fall
3 069 425
3 062 092
3 076 758
1 207 196 969
1 202 345 240
1 212 048 699
456
454
457
E9208
Laceration
2 521 358
2 514 597
2 528 118
777 270 957
773 815 336
780 726 579
353
351
354
E9289
Accident
2 084 549
2 078 423
2 090 674
717 535 466
713 929 286
721 141 645
396
395
398
E8881
Fall striking object
1 799 153
1 793 460
1 804 845
599 574 014
596 439 080
602 708 948
384
383
386
E927
Overexertion
1 705 919
1 700 349
1 711 489
452 136 747
450 098 015
454 175 478
309
309
310
E915
Foreign body
1 324 338
1 319 381
1 329 294
411 737 787
408 591 652
414 883 921
369
366
371
E8121
MVC—passenger
1 273 405
1 268 522
1 278 288
770 846 053
764 485 592
777 206 515
701
696
706
Severity and mortality
From 2006 to 2012, there were 10 684 (SE=234) paediatric
traumatic injury-related deaths in the ED and 12 596 (SE=254)
deaths after admission, yielding a total case-fatality rate of 0.04%.
The case-fatality rate for all traumatic injury visits decreased
36.9% from 0.05% in 2006 to 0.03% in 2012. Among severely
injured children, the case-fatality rate was 3.2% over the study
period, declining 27%, from 3.7% in 2006 to 2.7% in 2012.
There was graphical evidence of a decline in injury sev rity in
the final 3 years of the study period for the youngest age group
(figure 3).
Overall, there were 24 145 824 (SE=1174) traumatic injury-related ED visits to non-teaching hospitals in metropolitan
areas, and 20 776 205 (SE=1631) visits to teaching hospitals in
metropolitan areas. A remaining 11 461 618 (SE=1138) visits
were to non-metropolitan hospitals. The risk ratio for the rate of
traumatic injury ED visits characterised as severe presenting to
metropolitan teaching compared with non-teaching metropolitan hospitals was 2.4 (95% CI 2.2 to 2.7).
There were 11 178 311 (SE=1861) paediatric traumatic injury-related ED visits to level 1 or 2 trauma centres during the
study period, of which 247 161 (SE=1092) or 2.2% (95% CI
2.0 to 2.1) were categorised as severe. By contrast, there were
31 703 413 (SE=1951) paediatric trauma discharges from
616 (SE=909) or 0.6%
non-trauma centres, of which 179 (95% CI 0.5, to 0.7) were classified as severe. The risk ratio for
severely injured children being discharged from a level 1 or 2
trauma centre compared with non-trauma centres was 3.9 (95%
CI 3.6 to 4.2).
Table 4 presents the frequencies and case-fatality ratios for
the external cause of injury codes most often associated with
childhood ED fatalities during the study period. Together, these
injuries accounted for nearly 11 000, or 47%, of all childhood
trauma deaths in the USA between 2006 and 2012. Firearm-related injuries alone accounted for 20% of such deaths during
that time. MVC-related injuries accounted for 25% of all fatalities, with pedestrian injuries a particularly lethal mechanism,
accounting for 29% of all MVC-related deaths. An injured
child with one of these most common fatal causes of injury was
over three times more likely to be cared for at a level 1 trauma
centre than a non-level 1 trauma centre (OR=3.5, 95% CI 3.3 to
3.7). A survey-adjusted multivariable logistic regression model
restricted to these injury mechanisms and controlling for age,
gender and injury severity, indicated no statistically significant
benefit to being treated in a level 1 trauma centre (table 5).
Injury types
Figure 3 Median and IQR. ICD-9-derived ISS (ICISS) by age group
and year, total scale (A) and focused scale (B). Paediatric emergency
department visits for traumatic injury, US hospitals, 2006–2012.
4
There was a 22.2% (95% CI 20.6 to 23.9) increase in diagnoses
classified as TBI according to the Barell matrix, led mainly by
the older age groups (>9 years), and a sharp increase in the rate
of diagnoses classified as ‘systemic’ for the youngest age group
(0–4 years) (figure 4). Across all paediatric age groups, there was
a notable decline in diagnoses of spinal cord injury (figure 5).
Table 6 lists the most common Barell injury categories presenting
to US EDs between 2006 and 2012.
Injury visits due to MVC and other transportation-related
mechanisms declined 11.3% (95% CI 10.0to 12.6) from 2009
to 2012, the years for which data were available. There was
little change in the rate of paediatric ED visits due to falls,
from 2920 per 100 000 in 2009 to 2954 per 100 000 in 2012.
Notably, there were 59 450 (SE=1096) firearm-related paediatric ED discharges in the USA between 2009 and 2012, a
10.2% decrease in the population-based rate of firearm-related
visits from 19.8 per 100 000 in 2009 to 17.8 per 100 000 in
2012. The 15–19 year-old age group accounted for the majority
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
Table 4 Mortality counts and case-fatality ratios (CFR) for the top 10 most fatal external causes of injury codes in US hospital emergency
departments, 2006–2012
E-code
Description
Deaths
SE (deaths)
Lower CI
Upper CI
Total injuries
Case-fatality
ratio
SE (CFR)
E985.4
Firearm, undetermined intention
536
52
433
638
7439
7.19
0.73
E965.4
Assault, firearm
2451
113
2230
2671
39 909
6.15
0.30
E965
Assault, handgun
724
60
606
842
17 191
4.22
0.35
E922.9
Firearm accident NOS
949
70
811
1087
22 796
4.18
0.31
E814.7
MVC with pedestrian
1653
92
1472
1834
196 057
0.84
0.05
E966
Assault by cutting/piercing instrument
570
52
468
673
97 970
0.58
0.05
E816.1
MVC, passenger
789
64
663
915
196 388
0.40
0.03
E816
MVC, driver
703
61
585
822
229 533
0.31
0.03
E812
MVC or other traffic accident
988
73
846
1131
699 893
0.14
0.01
E812.1
MVC NOS, passenger
1620
90
1445
1796
1 273 405
0.13
0.01
NOS, not otherwise specified.
of paediatric firearm-related ED visits and was the only group
to experience a relative decrease in the rate of such injuries
between 2009 and 2012 (figure 6). (Note that certain AHRQ-assigned injury mechanism codes are only available from 2009,
whereas most data presented are available from 2006, including
external cause of injury codes (E-codes).)
Injury intent
A plot of the proportion of paediatric ED discharges classified as intentional indicates a gradual decline across all age
groups (figure 7). Between 2006 and 2012, there were a total
of 155 424 (SE=885) paediatric ED discharges characterised as
self-harm, representing 0.3% (SE=0.002) of all paediatric injury
discharges. There was no appreciable change in this percentage
over the study period. The 15–19 year-old age group experienced
the greatest percentage of ED discharges classified as self-harm
(0.7%, SE=0.005). Between 2009 and 2012, a total of 937 385
(SE=2122) or 3% (SE=0.1) of all paediatric ED discharges for
which data were available were classified as intentional assault.
The percentage of all paediatric ED visits characterised as intentional assault ranged from a low of 0.5% (SE=0.001) for children under age 5 up to 7.2% (SE=0.3) for teens aged 15–19.
most expensive paediatric injury diagnoses. Of note, four of five
of these most expensive diagnoses involved TBI.
Discussion
Childhood mortality is a proxy for social development—with
advanced societies suffering lower average childhood mortality
rates27—and, therefore, capturing current and accurate injury
trends can have wide-ranging implications. To this end, our
study is unique in harnessing the power of NEDS, the largest
and most comprehensive database of its kind, to track paediatric
trauma epidemiology across the USA.
EDs are the front lines of US healthcare, influencing public
policy and resource allocation, and our analysis is intended as
an important update to the epidemiology of paediatric injury
presenting to US EDs. Here, we highlight a decline in both the
annual rate of traumatic injuries, as well as the case-fatality rate
associated with those injuries. High-mortality injury mechanisms were more likely treated at level 1 trauma centres, with
Causes and costs
The total inflation-adjusted costs of paediatric ED injury care in
the USA between 2006 and 2012 were $23.3 billion (SE=0.01).
This accounted for 23.4% of the total $99.75 billion (SE=0.03)
spent on all ED traumatic injury costs in the USA during that
period. As displayed in table 3, the top 5 external injury codes
involved falls, being struck by an object and lacerations, which
together accounted for an estimated $6.2 billion in ED costs over
the study period. Paediatric ED injury costs increased 77.5%
from 2006 to 2012, and the average cost for a paediatric trauma
visit was $487 (SE=0.2). Table 7 lists the mean costs of the five
Table 5 Survey-adjusted logistic regression results. Factors
associated with injury mortality. Emergency department visits for
paediatric traumatic injury, US hospitals, 2006–2012
Variable
OR (95% CI)
Age group
1.03 (1.01 to 1.05)
Female sex
0.67 (0.65 to 0.68)
Teaching hospital
1.19 (1.13 to 1.24)
Level 1 trauma centre
0.96 (0.93 to 1.00)
Severe injury
22.80 (21.75 to 23.90)
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Figure 4 Barell matrix injury types by age group and year. Emergency
department (ED) visits for traumatic injury, US hospitals, 2006–
2012. TBI, traumatic brain injury.
5
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
Figure 6 Rate per 100 000 population, paediatric emergency
department (ED) visits for firearm-related injuries by age group. USA,
2009–2012.
Figure 5 Barell matrix injury types by age group and year. Emergency
department (ED) visits for traumatic injury, US hospitals, 2006–2012.
equivalent mortality rates. TBI increased 22.2% across all studied
age groups; 0.3% of injuries were classified as self-harm and 3%
as intentional assault; MVC-related injuries decreased 11.3%,
while the total cost of trauma care increased 78%, to $23 billion.
Importantly, every one of the leading causes of mortality is
considered a ‘preventable’ injury mechanism, highlighting the
importance of good public policy on this topic, which begins
with sound epidemiology and awareness.
Our trauma centre findings are notable. While several recent
studies note that specialised paediatric trauma centres provide
optimal care to injured children,28 29 most are treated at centres
without such designation (and, with respect to our study, NEDS
data do not specify paediatric trauma centres) [72]. Thus, understanding whether community trauma centres are indeed better
equipped than non-trauma centres to treat severely injured children is key.30 We demonstrate that level 1 and 2 trauma centres
treat four times more severely injured children than other hospitals, with level 1 trauma centres in particular three times more
likely to care for children with top 10 fatal injury mechanisms.
While reports in the adult literature suggest that level 1 trauma
centre care confers a mortality benefit,31 our analysis in table 5
showed that, after controlling for other variables, including
injury severity, treatment at a level 1 trauma centre, while
providing no such additional mortality benefit in these data, was
not associated with the kind of increased mortality one might
expect for treating such a large proportion of severely injured
children and adolescents. While our data do not capture time to
hospital presentation (ie, whether these patients arrived within
the so-called ‘golden hour’) or any prehospital care that was
received, these remain important areas of research in the trauma
care of severely injured children and adolescence.
With respect to specific injury mechanisms, our data are
encouraging regarding MVCs, the dominant cause of traumatic injury mortality, which declined 12.2%, consistent with
other published reports [50]. As mentioned, the The interplay
between technological improvements in safety, legislative initiatives and improved injury care may—individually or collectively—be driving these findings, and further characterisation is
needed. Here we feel that it is important to note that definitions of specific injury types based on diagnostic discharge codes
may vary widely. 32 The CDC, for example, has defined TBI
using ICD-9 diagnostic codes which include, importantly, code
959.01—‘head Injury, unspecified’.33 The Barell matrix, which
we used, does not include that code [68]. We based our decision
to use the Barell matrix on its acceptance internationally, and as
a way to identify severe cases of injury by body region using the
most widely accepted ICD-9 diagnostic codes.
Firearm-related violence remains a topic of contentious social
debate in the U.S. We report a 10.2% decline in firearm-related
Table 6 Barell matrix categories presenting to paediatric emergency
department visits for traumatic injury, US hospitals, 2006–2012
Injury category
Count (SE)
Superficial contusions
12 624 352 (6977)
Sprains and strains
1 850 698 (2984)
Open wounds
14 987 692 (7400)
Fractures
8 126 846 (5880)
Unspecified
5 146 048 (4765)
Internal organs
1 634 391 (2822)
Crushing
Dislocation
Burns
6
9 523 918 (6247)
System-wide
206 970 (1029)
1 304 653 (2520)
975 795 (2208)
Blood vessels
10 041 (227)
Nerves
11 827 (245)
Amputations
65 860 (577)
Figure 7 Per cent of paediatric emergency department (ED) visits
classified as intentional by age group, USA, 2006–2012.
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
Table 7 Mean costs of five most expensive diagnoses in 2012
US$. Paediatric emergency department visits for traumatic injury, US
hospitals, 2006–2012
Diagnosis (ICD-9 code)
Mean cost US$ (95% CI)
1. Open skull fracture prolonged LOC (804.65)
11 602 (11 602, to 11 602)
2. Closed skull fracture moderate LOC (803.33)
10 659 (10 659, to 10 659)
3. Open skull fracture moderate LOC (804.83)
9960 (9960 to 9960)
4. Open fracture pisiform (814.14)
8371 (2522 to 14 219)
5. Extradural haemorrhage (803.43)
8144 (8144 to 8144)
LOC, loss of consciousness.
ED visits, driven by decreases in the 15–19 year-old age group.
Yet, rates increased in younger children, as noted in other
reports.34 In total, there were nearly 60 000 firearm-related injuries, or 1 every 30 min and, despite our reported overall decline,
these injuries disproportionately affect African-American and
other minority communities, irrespective of socioeconomic
standing.35 36 37
TBIs are potentially devastating in children, able to extinguish decades of productive life. Over 600 000 children visit
US EDs annually with TBI, though most are mild.38 Even
minor TBIs, however, are associated with significant economic
losses [82], thus small epidemiological changes may have
significant repercussions. We report increased TBI diagnoses
across all age groups, consistent with previous findings [78].
That study also showed TBI associated with a concomitant
14.7% decline in injuries to the torso over the same study
period—a similar trend suggested here in figure 4. Another
recent report demonstrated a 25.9% increase in TBI over the
same study period (that report did, however, also note a 4%
decrease in that trend from 2012 to 2013, a year we do not
capture here).39 Our results also reflect the relative nature of
primary diagnosis categorisation, where increased awareness
of TBI and its effects40 41 may decrease the relative importance
of other diagnoses, leading to an observed—but not actual—
TBI uptrend. For example, there has been an increase in the
passage of state youth sports concussion laws since 2009 [85];
also, it is possible that the introduction of electronic medical
records has improved access to ICD-9 codes for concussion
and other TBIs.42 Regardless of the precise cause (may also
be multifactorial), we highlight a topic of great social and
economic consequence.
The paediatric population is perhaps society’s most vulnerable, therefore, injury intent data are important. Overall, 0.3%
of all paediatric trauma is classified as self-harm, without significant change over the study period. This number includes only
those reaching the ED alive, and thus excludes all suicides. We
also report that 3% of all paediatric ED trauma results from
intentional assault. Here, rates varied widely by age range,
with adolescents aged 15–19 demonstrating the highest rate
of intentional assault (7.2%), somewhat less than the roughly
10% rate reported by the CDC during the same period [49].
The cost of paediatric ED trauma care increased 78%—
nearly a quarter of all US ED trauma costs. As cited previously,
the CDC reports that childhood injuries cost $87 billion annually; we place the figure around $3.9 billion. This apparent
discrepancy highlights a critical factor: indirect costs, including
missed school and parental wage losses, which are a considerable social and financial burden related to paediatric traumatic
injury. The top 5 most common external injury codes alone
accounted for 26.5% of total paediatric and adolescent trauma
costs.
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Limitations
There are limitations to using cross-sectional observational
data to capture injury trends. For example, our results do not
account for non-ED urgent care facilities that divert patients
who are less critically ill away from hospital EDs. Also, the
data set is available from 2006, so long-term trends analysis
is not yet possible. Furthermore, NEDS relies on administrative ICD-9-CM codes that, while reliable indicators of
injury classification in hospital admissions,43 have not been
widely validated with respect to ED visits. Data may also
be subject to individual coding variations and coder error.44
Additionally, NEDS lacks unique patient identifiers that
would allow for longitudinal analysis, so the rate at which
patients re-presented with traumatic injuries could not be
determined. Also, NEDS data do not include patients who
died on the scene, or those who were dead on arrival to the
ED; this is particularly limiting with respect to self-harm
and assault mechanism, including firearm injuries, where
high on-scene mortality may be expected. Finally, we use
ICISS to categorise injury severity and dichotomise based on
a >6% probability of death, limiting our ability to perform
intercategorical analyses. We dichotomise injury severity into severe versus less
severe both to capture injury acuity and to address inherent
problems with ISS as a continuous variable.21 Yet, this can
result in the loss of important information about intercategorical differences, and our choice of 0.94 as the cut-off for
severe versus non-severe injury, while not arbitrary, is subject
to debate.22 Thus, in our study, an ICISS of 0.93 is ‘equivalent’ to 0.01. Injury severity scoring itself is inherently statistically problematic. Both the traditional and ICD-derived ISS
behave poorly as a continuous variable, with some authors
recommending, ‘that for statistical or analytical purposes
the ISS/NISS should not be considered a continuous variable’. While ICISS has been reported to generally perform as
well as ISS, it also behaves similarly poorly as a continuous
variable. We found statistical manipulations as log, square
root and inverse logit transformations to be unhelpful in this
regard and chose to dichotomise. This approach also allowed
us to calculate informative statistics based on probabilities,
such as ORs.
In conclusion, our analysis is a comprehensive and recent
portrait of paediatric trauma across US EDs. We report on the
evolution of traumatic injury in children and highlight areas
for continued investigation and improvement, including the
value of trauma centre care, the impact of firearm-related
injuries, injury intent, TBI management and prevention,
motor vehicle injuries and ED costs associated with paediatric
injury care. In conjunction with others on this topic, our study
emphasises some encouraging trends with respect to childhood injury and suggests possible interval improvement in the
management of paediatric trauma, providing a framework for
future characterisation.
What is already known on the subject?
►► Common aetiologies of paediatric traumatic injury in
the USA.
►► Protective role of dedicated paediatric trauma centres.
Contributors CJD conceived the study, acquired and had full access to data,
obtained IRB approval, conducted all analyses, interpreted the results, made critical
revisions and had final approval of the version to be published. JBA and MB cowrote
7
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Original article
What this study adds?
►► Updates epidemiology of paediatric traumatic injury using a
comprehensive and recent database.
►► Reports on current injury mechanisms, including intention
and severity.
►► Questions the mortality benefit in children and adolescents or
non-paediatric-specialised trauma centres.
►► Reports on cost of trauma care in children and adolescents.
the initial draft of the manuscript and provided important revisions and edits. SGF,
DAL and MGT provided important revisions and edits, and approved the final version
of the report.
Funding The study was funded by the National Institute of Child Health and
Human Development (to CJD) (Grant 1 R01 HD087460). The study funder had no
role in the study design, analysis, drafting of the manuscript, or decision to submit
the study for publication.
Competing interests None declared.
Ethics approval The study was approved bythe New York University School of
Medicine Institutional Review Board, andconforms to the STROBE statement on
reporting of observational studies.
Provenance and peer review Not commissioned; externally peer reviewed.
© Article author(s) (or their employer(s) unless otherwise stated in the text of the
article) 2017. All rights reserved. No commercial use is permitted unless otherwise
expressly granted.
References
1 Borse N, Sleet DA. CDC childhood injury report: patterns of unintentional injuries
among 0- to 19-year olds in the United States, 2000-2006. Fam Community Health
2009;32:189.
2 Odetola FO, Gebremariam A. Paediatric trauma in the USA: patterns of emergency
department visits and associated hospital resource use. Int J Inj Contr Saf Promot
2015;22:260–6.
3 Centers for Disease Control and Prevention. National action plan for child injury
prevention. Washington, D.C, 2012.
4 Winthrop AL, Brasel KJ, Stahovic L, et al. Quality of life and functional outcome after
pediatric trauma. J Trauma 2005;58:468–74.
5 Centers for Disease Control and Prevention. Web-based injury statistics query and
reporting system(WISQARS), Leading causes of death reports, national and regional, ​
1999-​2014.​http://​webappa.​cdc.g​ ov/​sasweb/​ncipc/​leadcaus10_u​ s.​html. (accessed 12
Feb 2017).
6 Centers for Disease Control and Prevention(CDC). Vital signs: Unintentional injury
deaths among persons aged 0-19 years - United States, 2000-2009. MMWR Morb
Mortal Wkly Rep 2012;61:270–6.
7 Haider AH, Crompton JG, Oyetunji T, et al. Mechanism of injury predicts case fatality
and functional outcomes in pediatric trauma patients: the case for its use in trauma
outcomes studies. J Pediatr Surg 2011;46:1557–63.
8 Short SS, Liou DZ, Singer MB, et al. Insurance type, not race, predicts mortality after
pediatric trauma. J Surg Res 2013;184:383–7.
9 Kenefake ME, Swarm M, Walthall J. Nuances in pediatric trauma. Emerg Med Clin
North Am 2013;31:627–52.
10 Stylianos S, Ford HR. Outcomes in pediatric trauma care. Semin Pediatr Surg
2008;17:110–5.
11 Wesson DE. Pediatric Trauma Centers. Tex Heart Inst J 2012;39:871–3.
12 Gausche-Hill M, Ely M, Schmuhl P, et al. A national assessment of pediatric readiness
of emergency departments. JAMA Pediatr 2015;169:527–34.
13. Introduction to the HCUP Nationwide Emergency Department Sample(NEDS). 2015
https://www.​hcupus.​ahrq.g​ ov/​db/n​ ation/​neds/​NEDS2013Introduction.​pdf.
14. The R project for statistical computing. 2017 http://www.​r-​project.​org.
15 World Health Organization. Definition of key terms, 2013. http://www.​who.​int/​hiv/​
pub/g​ uidelines/​arv2013/i​ ntro/​keyterms/​en/.
16 American College of Surgeons. Committee on Trauma. Resources for optimal care of
the injured patient. 1990.
8
17 Osler T, Rutledge R, Deis J, et al. ICISS: an international classification of disease-9
based injury severity score. J Trauma 1996;41:380–8.
18 Seguí-Gómez M, Lopez-Valdes FJ. In: Li G, Baker SP, eds. Injury severity scaling,
in injury research: theories, methods, and approaches. Boston, MA: Springer US,
2012:281–95.
19 Bartolomeos KK, Cryer C, Fingerhut L, et al. Joint meeting of international
collaborative effort on injury statistics and the global burden of disease-injury expert
group.
20 DiMaggio C, Ayoung-Chee P, Shinseki M, et al. Traumatic injury in the United States:
In-patient epidemiology 2000-2011. Injury 2016;47:1393–403.
21 Stevenson M, Segui-Gomez M, Lescohier I, et al. An overview of the injury severity
score and the new injury severity score. Inj Prev 2001;7:10–13.
22 Palmer C. Major trauma and the injury severity score-where should we set the bar?
Annu Proc Assoc Adv Automot Med 2007;51:13–29.
23 Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic
comorbidity in longitudinal studies: development and validation. J Chronic Dis
1987;40:373–83.
24 Barell V, Aharonson-Daniel L, Fingerhut LA, et al. An introduction to the Barell body
region by nature of injury diagnosis matrix. Inj Prev 2002;8:91–6.
25 Hsia RY, MacIsaac D, Baker LC. Decreasing reimbursements for outpatient emergency
department visits across payer groups from 1996 to 2004. Ann Emerg Med
2008;51:265–74.
26 Bieler GS, Brown GG, Williams RL, et al. Estimating model-adjusted risks,
risk differences, and risk ratios from complex survey data. Am J Epidemiol
2010;171:618–23.
27 World Health OrganizationGlobal Health Observatory(GHO) Data. 2017. http://www.​
who.i​ nt/​gho/​en/.
28 Segui-Gomez M, Chang DC, Paidas CN, et al. Pediatric trauma care: an overview
of pediatric trauma systems and their practices in 18 US states. J Pediatr Surg
2003;38:1162–9.
29 Amini R, Lavoie A, Moore L, et al. Pediatric trauma mortality by type of designated
hospital in a mature inclusive trauma system. J Emerg Trauma Shock
2011;4:12.
30 Schappert SM, Bhuiya F. 2012. Availability of pediatric services and equipment in
emergency departments: United States, 2006. National health statistics reports; no
47. Hyattsville, MD: National Center for Health Statistics.
31 MacKenzie EJ, Rivara FP, Jurkovich GJ, et al. A national evaluation of the effect of
trauma-center care on mortality. N Engl J Med 2006;354:366–78.
32 DiMaggio C, Li G. Emergency department visits for traumatic brain injury in a birth
cohort of Medicaid-insured children. Brain Inj 2013;27:1238–43.
33 Colantonio A, Croxford R, Farooq S, et al. Trends in hospitalization associated
with traumatic brain injury in a publicly insured population, 1992-2002. J Trauma
2009;66:179–83.
34 DiMaggio CJ, Avraham JB, Lee DC, et al. The epidemiology of emergency department
trauma discharges in the United States. Wall SPAcad Emerg Med. In Press.(accepted
20 Feb 2017).
35 Kalesan B, Vyliparambil MA, Bogue E, et al. Race and ethnicity, neighborhood
poverty and pediatric firearm hospitalizations in the United States. Ann Epidemiol
2016;26:1–6.
36 Kalesan B, Vasan S, Mobily ME, et al. State-specific, racial and ethnic heterogeneity
in trends of firearm-related fatality rates in the USA from 2000 to 2010. BMJ Open
2014;4:e005628.
37 Srinivasan S, Mannix R, Lee LK. Epidemiology of paediatric firearm injuries in the USA,
2001-2010. Arch Dis Child 2014;99:331–5.
38 Graves JM, Rivara FP, Vavilala MS. Health care costs 1 year after pediatric traumatic
brain injury. Am J Public Health 2015;105:e35–e41.
39 Chen C, Shi J, Stanley RM, et al. U.S. Trends of ED visits for pediatric traumatic brain
injuries: implications for clinical trials. Int J Environ Res Public Health
2017;14:414.
40 Marin JR, Weaver MD, Yealy DM, et al. Trends in visits for traumatic brain injury to
emergency departments in the United States. JAMA 2014;311:1917–9.
41 Smith NA, Chounthirath T, Xiang H. Soccer-related injuries treated in emergency
departments: 1990-2014. Pediatrics 2016;138:e20160346.
42 . HealthIT.govBenefits of electronic medical records. https://www.​healthit.​gov/​
providers-p​ rofessionals/​benefits-​electronic-​health-r​ ecords-​ehrs. (accessed 25 Aug
2017).
43 LeMier M, Cummings P, West TA. Accuracy of external cause of injury codes reported
in Washington State hospital discharge records. Inj Prev 2001;7:334–8.
44 Hirshon JM, Warner M, Irvin CB, et al. Research using emergency department-related
data sets: current status and future directions. Acad Emerg Med
2009;16:1103–9.
Avraham JB, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042435
Downloaded from http://injuryprevention.bmj.com/ on October 25, 2017 - Published by group.bmj.com
Epidemiology of paediatric trauma presenting
to US emergency departments: 2006 −2012
Jacob B Avraham, Misha Bhandari, Spiros G Frangos, Deborah A Levine,
Michael G Tunik and Charles J DiMaggio
Inj Prev published online October 22, 2017
Updated information and services can be found at:
http://injuryprevention.bmj.com/content/early/2017/10/22/injuryprev2017-042435
These include:
References
This article cites 32 articles, 6 of which you can access for free at:
http://injuryprevention.bmj.com/content/early/2017/10/22/injuryprev2017-042435#BIBL
Email alerting
service
Receive free email alerts when new articles cite this article. Sign up in the
box at the top right corner of the online article.
Notes
To request permissions go to:
http://group.bmj.com/group/rights-licensing/permissions
To order reprints go to:
http://journals.bmj.com/cgi/reprintform
To subscribe to BMJ go to:
http://group.bmj.com/subscribe/
Документ
Категория
Без категории
Просмотров
2
Размер файла
839 Кб
Теги
injuryprev, 2017, 042435
1/--страниц
Пожаловаться на содержимое документа