|
Utah CODES History and Impact
(Introduction, Background, Current Activities, Future Directions, References)
Introduction
Motor vehicle crashes are the leading cause of
injury death for people aged 1 to 34 years (NCHS, 1998). Each year over 40,000 people die in a motor vehicle crash in
the United States and another 3.5 million suffer nonfatal injuries (NCHS, 1998;
Quinlan et al., 1999). Utah’s 2000 death rate from motor vehicle crashes of 16.7
deaths per 100,000 population is higher than the national average of 14.9 deaths
per 100,000 population. Both of these rates are higher than the goal of 9.0
deaths per 100,000 population outlined in Healthy People 2010 (U.S. Department
of Health and Human Services, January 2000). Utah’s nonfatal injury rate for motor vehicle crashes, 1,347
per 100,000 population, is even higher than the death rate. Motor vehicle crash
injuries may require treatment at the scene by emergency medical services, at an
emergency department, or result in the person being admitted as an inpatient to
the hospital.
For this reason the National Highway Traffic
Safety Administration (NHTSA) funded Utah, along with six other states, to
conduct probabilistic record linkage, a method for combining databases, to
study motor vehicle crashes in conjunction with other healthcare databases. This
project, known as the Crash Outcome Data Evaluation System (CODES) not only
demonstrated the feasibility of probabilistic record linkage using large,
statewide databases, but also quantified the risk of not wearing a seatbelt as
it pertains to being treated by emergency medical services, treated at the
emergency department, admitted to a hospital, and killed as a result of a motor
vehicle crash (Johnson and Walker, 1996). Since 1992 NHTSA has funded CODES projects in 26 states. The
availability of statewide, population-based datasets is a major strength of the
Utah CODES project. The Utah CODES project is based at the Intermountain Injury
Control Research Center, University of Utah School of Medicine and is directed
by J. Michael Dean MD, MBA and Larry Cook MStat.
Background
(back to top)
When first funded in 1992, the Utah CODES
project used four 1991 population-based databases: motor vehicle crash records,
completed by police officers at the scene, collected by Utah Department of
Transportation (UDOT); emergency medical services runs collected by the Utah
Department of Health, Bureau of Emergency Medical Services; emergency department
and hospital discharge records collected originally from individual hospital
organizations.
Included in the motor vehicle crash database are
several variables that are of interest for motor vehicle crash analysis,
including the time of day of the crash, weather conditions, type of crash, how
many people and vehicles are involved, speed of each vehicle involved, ages of
each participant, their seating location in the vehicle, and seatbelt use
information for each passenger. A police assessed injury score called KABCO is
assigned to each passenger at the scene of the crash. However, research has
shown that KABCO scores are an unreliable measure for assessing injury severity
(Agran, Castillo and Winn, 1990). More accurate measures of severity exist (i.e. ones
calculated from medical information) in other healthcare databases including
Glasgow Comma Score (GCS) assessed by emergency medical services, and
Abbreviated Injury Score (AIS) and Injury Severity Score (ISS) calculated from
emergency department and hospital discharge datasets. While there are injury
mechanism codes in the healthcare databases to identify motor vehicle crashes,
they include none of the crash characteristics mentioned above. Thus,
probabilistic linkage allows CODES researchers to access information from a
variety of sources to obtain a more complete and accurate assessment of the
crash event. Figure 1 shows how probabilistic linkage can aid in completing the
picture of motor vehicle morbidity, and that fatalities are the tip of the
iceberg when it comes to studying motor vehicle crashes.
Since these databases were collected
independently from different sources, there are no database keys that identify
person X in the crash file as person Y in the hospital file. For this reason,
techniques such as probabilistic record linkage are needed to combine these
databases. Record linkage is accomplished by
comparing common data fields in two different files, such as the date of birth
or the gender of a patient. The comparisons of numerous data fields lead to a
judgment that two records refer to the same patient event (and should be linked)
or that the records do not refer to the same patient event (and should not be
linked). This judgment is based on the cumulative agreement and disagreement of
field values. Probabilistic linkage software accomplishes this task
mathematically, rather than relying on the subjective impression of a human
clerical reviewer. More thorough treatments of the subject have been previously
published (Cook, Olson and Dean, 2001; Jaro, 1995; Newcombe, 1988).
Combining results from the Utah CODES linkage
with data from the other six states, researchers from NHTSA delivered a report
to congress in February 1996 on nearly 900,000 drivers and passengers of
vehicles that were involved in a motor vehicle crash during 1991. The report
shows that seatbelts are very effective at reducing injury following a motor
vehicle crash and effectiveness increases with the severity of injury studied.
For instance, while seatbelts were 20% effective at preventing injuries as a
result of a motor vehicle crash, they were 60% effective at preventing death
following a motor vehicle crash. For crash victims that are admitted to the
hospital, the use of a seatbelt lowers the average hospital charge by 55%
(Johnson and Walker, 1996).
Since the report to Congress, the staff at
Utah’s CODES project has continued to gather, probabilistically link and conduct
research on statewide databases. These databases include vital statistics
databases (death certificates, birth certificates, and fetal death certificates)
and driver license databases (moving citations and driver medical conditions) in
addition to the original motor vehicle crash, emergency medical services,
emergency department, and hospital discharge databases. Currently the Utah CODES
project has linked the motor vehicle crash records to hospital discharge records
for years 1992 – 1998, to emergency department and emergency medical services
records for years 1996 – 1998, to vital statistics data for 1992 to 1999, and to
driver license databases for years 1992 – 1996. Presentations of this research
have been given at a wide variety of national and international meetings
including: American Public Health Association, Association for the Advancement
of Automotive Medicine, International Forum for Traffic Records and Highway
Information Systems, and the National Congress on Childhood Emergencies.
Research on Utah CODES linked databases has also resulted in several
peer-reviewed publications. These studies have focused on subpopulations of
drivers at risk for a motor vehicle crash such as teenagers (Cvijanovich et al.,
2001b), older drivers (Cook et al., 2000), and drivers with specific medical conditions (Diller et al.,
1999; Vernon et al., 2001; Vernon et al., 2002); special types of vehicles including all terrain vehicles
(ATV) (Cvijanovich et al., 2001a; Cvijanovich et al., 2000) and snowmobiles (Skokan et al., 2001); the consequences of not properly wearing a seatbelt (Knight
et al., 2001); the effectiveness of safety restraints for children (Berg et
al., 2000; Corneli, Cook and Dean, 2000); the effect of known over reporting of seatbelt use in police
crash records (Dean, Reading and Nechodom, 1995); and general probabilistic linkage methodology (Cook, Olson
and Dean, 2000; Cook, Olson and Dean, 2001; Dean et al., 2001).
The Utah CODES project personnel have been
active advocates for traffic safety in the state. Data linked and analyzed by
Utah’s CODES project has been presented at the Utah legislature in support of
traffic-related bills. Linked data were integral in educating Utah policy makers
on the importance of passing a primary seatbelt law, which requires all drivers
and passengers under the age of 19 years to wear a seatbelt and all children
four years of age or younger to be restrained in a child safety seat. Linked
data were also used in the campaign to promote graduated driver licensing laws
in our state. This program removes driving restrictions in stages as novice,
teenage drivers gain maturity and experience, until full licensure is granted.
In cooperation with the Utah Department of Public Safety, Highway Safety Office,
Utah CODES has compiled the Utah Crash Summary since 1994. This document
contains detailed tables and graphs of all Utah public-roadway reported motor
vehicle crashes and their participants plus sections on certain subpopulations
such as pedestrians, bicyclists, alcohol and drug related crashes, and teenage
drivers. As a companion to the crash summary, several fact sheets are produced
for each section. Both the crash summary and fact sheets are available at the
Utah CODES website, http://www.utcodes.org.
In addition to providing data for traffic safety
advocates, Utah CODES’ personnel have served as members of traffic safety
organizations. Utah CODES team members have participated in the Utah Safety
Management System, the Coalition for Utah Traffic Safety, and the Coalition for
Effective Traffic Safety Legislation. The goal of these organizations is to
reduce morbidity and mortality associated with motor vehicle crashes in Utah.
Many approaches have been taken in striving for this goal including: public
service announcements, child safety seat checkpoints, newspaper articles, and
proposed legislation. The role of Utah CODES personnel for these endeavors has
been to provide the necessary data and aid in its presentation.
Through involvement with these organizations and
state agencies such as the Utah Department of Transportation, Utah’s Driver
License Division, and the Department of Public Safety, Utah CODES was asked to
facilitate the formation of the Utah Traffic Records Coordinating Committee
(TRCC) in 2000. The overall goal of the TRCC is to work with motor vehicle crash
agencies in the state to improve Utah’s traffic records system. The TRCC
strategic plan calls for accomplishing this goal by streamlining the motor
vehicle crash information processing system, and ensuring timely and available
motor vehicle crash data.
The
use of probabilistic linkage to combine databases has been an effective tool in
motor vehicle crash research. Furthermore, the use of probabilistic linkage is a
promising tool for studying injury mortality and morbidity for all mechanisms of
injuries, not just crashes (Junkins et al., 2001; Knight et al., 2000; Suruda et
al., 2000; Suruda et al., 1999). Through research and its involvement with other state
agencies and organizations, Utah CODES as part of the Intermountain Injury
Control Research Center, will work towards meeting the goals outlined in Healthy
People 2010, to increase the quality and years of healthy life in Utah. For more
information on the Utah CODES project visit our website at
http://www.utcodes.org.
Agran, P. F., Castillo, D. N., and Winn, D. G. (1990).
Limitations of data compiled from police reports on pediatric pedestrian and
bicycle motor vehicle events. Accident Analysis & Prevention
22(4):361-70.
Berg, M. D., Cook, L., Corneli,
H. M., et al. (2000). Effect of seating position and restraint use on injuries
to children in motor vehicle crashes. Pediatrics 105(4 Pt 1):831-5.
Cook, L. J., Knight, S., Olson,
L. M., et al. (2000). Motor vehicle crash characteristics and medical outcomes
among older drivers in Utah, 1992-1995. Annals of Emergency Medicine
35(6):585-91.
Cook, L. J., Olson, L. M., and
Dean, J. M. (2000). Usefulness of Name Information in Probabilistic Record
Linkage. Pediatric Emergency Care 16(1):66.
Cook, L. J., Olson, L. M., and
Dean, J. M. (2001). Probabilistic record linkage: relationships between file
sizes, identifiers and match weights. Methods of Information in Medicine
40(3):196-203.
Corneli, H. M., Cook, L. J., and
Dean, J. M. (2000). Adults and children in severe motor vehicle crashes: a
matched-pairs study. Annals of Emergency Medicine 36(4):340-5.
Cvijanovich, N. Z., Cook, L. J.,
Mann, N. C., et al. (2001a). A population-based assessment of pediatric
all-terrain vehicle injuries. Pediatrics 108(3):631-5.
Cvijanovich, N. Z., Cook, L. J.,
Mann, N. C., et al. (2001b). A population-based study of crashes involving 16-
and 17-year-old drivers: the potential benefit of graduated driver licensing
restrictions. Pediatrics 107(4):632-7.
Cvijanovich, N. Z., Mann, N. C.,
Cook, L., et al. (2000). A Study of ATV Injuries in Utah. Pediatric Emergency
Care 16(1):66.
Dean, J. M., Reading, J. C., and
Nechodom, P. J. Over reporting and measured effectiveness of seat belts in motor
vehicle crashes in Utah. In T. R. Board, ed., Transportation Research
Records, 1485.Washington D.C.: National Academy Press, 1995.
Dean, J. M., Vernon, D. D.,
Cook, L., et al. (2001). Probabilistic linkage of computerized ambulance and
inpatient hospital discharge records: a potential tool for evaluation of
emergency medical services. Annals of Emergency Medicine 37(6):616-26.
Diller, E. M., Cook, L. J.,
Leonard, D. R., et al. (1999) Evaluating drivers licensed with medical
conditions in Utah, 1992 - 1996. National Highway Traffic Safety Administration.
Jaro, M. A. (1995).
Probabilistic linkage of large public health data files. Statistics in
Medicine 14(5-7):491-8.
Johnson, S. W., and Walker, J.
(1996) The Crash Outcome Data Evaluation System (CODES). National Highway
Traffic Safety Administration.
Junkins, E. P., Jr., Knight, S.,
Olson, L. M., et al. (2001). Analysis of school injuries resulting in emergency
department or hospital admission. Academic Emergency Medicine 8(4):343-8.
Knight, S., Cook, L. J.,
Nechodom, P. J., et al. (2001). Shoulder belts in motor vehicle crashes: a
statewide analysis of restraint efficacy. Accident Analysis & Prevention
33(1):65-71.
Knight, S., Junkins, E. P., Jr.,
Lightfoot, A. C., et al. (2000). Injuries sustained by students in shop class.
Pediatrics 106(1 Pt 1):10-3.
CDC National Center for Health
Statistics (NCHS). (1998) National Mortality Data. Hyattsville, MD: NCHS.
Newcombe, H. B. (1988).
Handbook of Record Linkage: methods for health and statistical studies,
administration, and business. New York City: Oxford University Press.
Quinlan, K. P., Thompson, M. P.,
Annest, J. L., et al. (1999). Expanding the National Electronic Injury
Surveillance System to monitor all nonfatal injuries treated in US hospital
emergency departments. Annals of Emergency Medicine 34(5):637-45.
U.S. Department of Health and
Human Services. (January 2000) Healthy People 2010 (Conference Edition, in Two
Volumes). Washington, DC.
Skokan, E. G., Olson, L. M.,
Cook, L. J., et al. (2001). Snowmobile injuries in Utah. Academic Emergency
Medicine 8(12):1173-7.
Suruda, A., Vernon, D. D.,
Diller, E., et al. (2000). Usage of emergency medical services by children with
special health care needs. Prehospital Emergency Care 4(2):131-5.
Suruda, A., Vernon, D. D.,
Reading, J., et al. (1999). Pre-hospital emergency medical services: a
population based study of pediatric utilization. Injury Prevention
5(4):294-7.
Vernon, D. D., Diller, E. M.,
Cook, L. J., et al. (2001) Further analysis of drivers licensed with medical
conditions in Utah. National Traffic Highway Safety Administration.
Vernon, D. D., Diller, E. M.,
Cook, L. J., et al. (2002). Evaluating the crash and citation rates of Utah
drivers licensed with medical conditions, 1992-1996. Accident Analysis &
Prevention 34(2):237-246.
|
