This is the first publication from the New Jersey traumatic brain injury (TBI) surveillance system on the incidence of TBI. This report provides a profile on mortality and morbidity related to such injuries in New Jersey during 1994. The report was prepared by the Data Analysis and Evaluation section in the Center for Health Statistics (CHS) of the New Jersey Department of Health (NJDOH).

This project was funded through the Centers for Disease Control and Prevention (CDC) from a cooperative agreement (grant number: U59/CCU203367) between the Department of Health and the Disability Prevention Program of the New Jersey Department of Human Services (NJDHS).

We wish to extend special thanks to Deborah E.Cohen, Ph.D., Director of the Office for Prevention of Mental Retardation and Developmental Disabilities, NJDHS through whose office funds were made available for the development of the TBI system. We acknowledge cooperation from the administrative staff of CHS and NJDOH.

Inquiries regarding the content and use of this report should be addressed to the following:

Table of Contents
Traumatic Brain Injury Summary Highlights, New Jersey, 1994
Introduction
SECTION I:
Background
Scope of the Statewide TBI Surveillance System
TBI Surveillance Case Definition and Methodology
SECTION II:
Analysis of TBI Surveillance Data, New Jersey, 1994
Recommendations
References
List of Tables
Table 1A:	Distribution and rate of traumatic brain injuries by sex, age, cause, race and hospitalization status, New Jersey, 1994
Table 1B:	Distribution of traumatic brain injuries by sex, age, cause, race and severity, New Jersey, 1994
Table 2A:	Percent distribution and rate of traumatic brain injuries by age and sex, New Jersey, 1994
Table 2B:	Percent distribution and rate of traumatic brain injury fatalities by age and sex, New Jersey, 1994
Table 2C:	Percent distribution and rate of moderate to severe traumatic brain injuries by age and sex, New Jersey, 1994
Table 2D:	Percent distribution and rate of mild traumatic brain injuries by age and sex, New Jersey, 1994
Table 3A:	Percent distribution and rate of traumatic brain injuries by sex and county of residence, New Jersey, 1994
Table 3B:	Distribution of traumatic brain injuries by county of residence and severity, New Jersey, 1994
Table 3C:	Distribution of traumatic brain injuries by county of residence and cause, New Jersey, 1994
Table 3D:	Rate of traumatic brain injuries by county of residence and cause, New Jersey, 1994
Table 4A:	Distribution of traumatic brain injuries by age, sex, and cause, New Jersey, 1994
Table 4B:	Rate of traumatic brain injuries by age, sex, and cause, New Jersey, 1994
Table 5:	Average age (in years) at the time of traumatic brain injury by hospitalization status, New Jersey, 1994
Table 6:	Distribution of traumatic brain injury hospitalizations by severity, New Jersey, 1994
Appendix A
Table P1:	Population estimates by age and sex, New Jersey, 1994
Table P2:	Population estimates by county of residence and sex, New Jersey, 1994
Appendix B
Technical notes

• According to the traumatic brain injury (TBI) surveillance system developed by the New Jersey Department of Health (NJDOH), there were 10,479 incidents of traumatic brain injury in New Jersey during the calendar year 1994. These individuals were serious enough to require hospitalization or died with or without receiving any form of medical attention. Almost 90% of these occurrences involved non-fatal hospitalizations, 4.4% hospitalized deaths, and the remaining 5.9% involved non-hospitalized deaths.

• In 1994, there were 132.6 traumatic brain injuries per 100,000 New Jersey's estimated population. Almost 95% of the TBIs incidents that occurred in New Jersey occurred to residents of New Jersey, 3.6% involved out-of-state residents, and the remaining 1.5% were of unknown residence.

• Of the 10,479 TBI incidents, one in every 10 was fatal. Of the remaining 9,400 non-fatal hospitalizations, about one in five (22.5%) was classified as having a moderate to severe injury, while the majority were mild (74.0%). The remaining 3.4% were of undetermined level of severity (using the CDC injury severity scaling system).

• According to the most recent financial data available, the typical person who was hospitalized for TBI was approximately 40 years old and was hospitalized for about 6 days costing approximately $14,200 per hospitalization.

• Results indicated that TBI incidence rates were higher among males than females (a male to female ratio of 2.5 for fatalities; 2.0 for moderate to severe injuries; and 1.6 for mild injuries) at all levels of severity.

• Results indicated that TBI disproportionately affects the young (15-24 year olds), and the elderly (65 years and over). TBI incidence rates were substantially higher for males overall (174.9 vs. 92.9 for females per 100,000) and in each age group. The incidence rates ranged from a low of 52.1 per 100,000 for 5-14 year old females to a high of 719.3 per 100,000 for elderly males 85 and over.

• On an average day in 1994, about three persons (two males and one female) died and another 27 (17 males and 10 females) were hospitalized with TBI. The number of persons treated in emergency rooms and out-patient care settings for TBI are unknown.

• By intent, 85.4% of the traumatic brain injuries were classified as unintentional, 11.0% involved homicides/assaults; 1.9% involved suicide/self-inflicted injuries; and 1.8% were of other/undetermined/unknown intent.

• The leading cause of TBI was motor vehicle-related injury (37.7%), followed closely by falls (36.0%). These two causes account for almost three out of every four (73.7%) of the TBI incidents that occur in New Jersey.
• Traumatic brain injuries caused by firearms accounted for about 2.8% of the total (292 out of 10,479). Over half (156 or 53.4%) of these were self-inflicted. Firearms also were involved in one in three (32.9%) assault-related TBIs.

• Overall, TBI incidence rates were highest for Atlantic County residents (207.6 per 100,000) followed by residents of Hudson County (205.8 per 100,000). The lowest incidence of TBI occurred among residents of Hunterdon County (64.2 per 100,000) followed by residents of Morris County (74.6 per 100,000). The overall resident incidence rate was 125.8 per 100,000 population. These rates varied by age, sex and cause. These rates may be understated slightly due to unavailability of data on out-of-state hospitalizations to New Jersey residents.

• Motor vehicle-related TBI incidence rates ranged from a low of 24.4 per 100,000 in Morris County to a high of 85.1 per 100,000 in Camden County. The overall occurrence rate was 50.0 per 100,000 for New Jersey; 64.1 for males and 36.8 for females. Rates were higher for males than females for all age groups, particularly for those aged 45 years and under. Males and females in the age group 15-24 years, and males 85 years and over are at highest risk for motor vehicle-related TBI (137.7 per 100,000 males aged 15-24; 91.2 for males 85 and over; and 83.9 per 100,00 females aged 15-24). Motor vehicle-related injuries accounted for more than half (55.5%) of the total number (1,974) of traumatic brain injuries among 15-24 year olds.

• Fall-related TBI incidence rates ranged from a low of 22.6 per 100,000 in Hunterdon County to a high of 81.8 in Atlantic County. The overall rate for occurrences in New Jersey was 47.7 per 100,000. Both males and females in the age group 5 years and under and 65 and over had higher TBI incidence caused by falls. Males had substantially higher rates than females for all age groups for this category.

• The overall assault-related TBI occurrence rate for New Jersey was 14.6 per 100,000 population. The highest incidence of assault-related TBI occurred among residents of Hudson County (39.1 per 100,000); the lowest incidence was for Hunterdon county with no reported assault-related traumatic brain injuries in 1994.

• The rate of self-inflicted traumatic brain injuries, accounting for 1.9% of the total number of incidents, was 2.5 per 100,000 persons in 1994. This incidence was highest among the residents of Passaic County (4.8 per 100,000) and lowest among the residents of Gloucester County (1.2 per 100,000 residents). Self-inflicted rates tend to increase with advancing age with the highest incidence occurring among elderly males and females, 65 and over. Rates for males were higher than females in each age group.

• TBI was listed as the primary diagnosis in nearly three fourths of the TBI-related hospitalizations. Analysis of hospitalizations by disposition indicated that approximately 4.6% resulted in death. A vast majority (87.1%) of those who died had moderate to severe levels of traumatic brain injury and a majority of them were elderly. Other health conditions in addition to TBI may have contributed to these deaths. Most of those who were released to home (78.2% of 8,091) had relatively minor injuries.

• The race distribution indicates that more than three-fourths (76.9%) of those who died and/or sustained traumatic brain injuries were white. This was followed by blacks at 15.6%. The remaining 7.5% were mostly American Indians, Asians, and persons of unknown race.

This is the first annual incidence report from the New Jersey traumatic brain injury (TBI) surveillance system. Although developmental activities began as early as 1995, the data have been too incomplete until now to determine mortality and morbidity associated with traumatic brain injury. The New Jersey Department of Health (NJDOH) has served as primary developer of this surveillance system in collaboration with the Office for Prevention of Mental Retardation and Developmental Disabilities (OPMRDD) at the New Jersey Department of Human Services (NJDHS). Funds for these efforts were provided by the Centers for Disease Control and Prevention's (CDC) cooperative agreement # U59/CCU203367 with NJDHS. The contents of this report are solely the responsibility of the author and do not necessarily represent the views of the funding agency. This report, prepared at the Center for Health Statistics (CHS), is intended to serve as a guide for defining and understanding the magnitude and scope of TBI occurrence in New Jersey, and facilitating development of prevention programs.

TBI incidence* for the purpose of this report is defined as an event leading to either hospitalization or death with an injury to the brain as a sole or contributory cause. The TBI surveillance case definition and methodology are included in section I, page 5 of this report. The current surveillance system consists of data from more than one source which have been linked using probabilistic matching of records. The hospitalization (UB-92) and the single and multiple cause-of-death (SCD and MCD) files maintained by the NJDOH for the calendar year 1994 are the sources of information from which the TBI statistics are derived. Tables on the incidence of TBI are based on all three sources of information, while mortality statistics are based on a linked file of hospitalized deaths from the UB-92 and deaths attributed to TBI from the death files. Morbidity statistics are based on hospitalization records (UB-92) submitted to the NJDOH by all acute care facilities located in New Jersey. Due to extremely limited resources, validity and reliability of the data in the system have not been thoroughly studied. Some data problems and resolutions are discussed in Appendix B. Although some discrepancies exist among the various sources of data used to create the surveillance system, the impact on overall TBI incidence rates is expected to be minimal.

This report is divided into two sections. The first section provides background information on the development of TBI surveillance, case definition and methodology. Data analysis and detailed tables are presented in Section II. Population estimates are provided in Appendix A. Data sources, quality, problems, and resolutions are discussed in Appendix B.

* The term TBI incidence used in this report does not include TBI cases which were treated and released in hospital emergency departments or which receive some other form of superficial attention. Also strictly speaking, incidence rates should reflect the same at-risk population in the numerator and denominator, which these do not. The numerator is all incidents which occurred in New Jersey, while the denominator is an estimate of the resident population.

TBI incidence rate for a specified population =	Number of TBI incidents * 100,000
	Population at risk specific to the event

The estimated population figures (Appendix A) for New Jersey for the calendar year 1994 developed by the U.S. Bureau of the Census were used as denominators for computing approximate incidence rates. TBI incidence rates are expressed per 100,000 estimated population.

New Jersey, with an estimated population of 7.9 million in 1995, is the most densely populated state in the nation. Almost 80% of the state's population reside in one-half of the area, with dense urban and suburban neighborhoods extending across the state from the Delaware River at Philadelphia to the Hudson River which separates New Jersey from New York City. The remaining 20% of the population live in the mountains and farmlands of the northwestern part of the state and in the Pinelands forests, farms and ocean shore of the southern region.

New Jersey's population is older than the U.S. population as a whole, but also encompasses large number of residents in the young and young adult age groups. Approximately one-third of the population is under 25 years of age and another one-third is 25-44 years. One-fifth of the population is aged 45-64 years, and 13.6% are elderly. The population is 49% male and 51% female. Population estimates for 1994 suggest that 71.5% of the population is white non-Hispanic, 13.0% is black non- Hispanic, 4.4% is other races and non-Hispanic, and the remaining 11.0% is of Hispanic origin, of any race. Tables P1 and P2 of Appendix A summarize the population by county of residence, gender and age group.

Physical injury is a leading cause of morbidity and mortality throughout the United States. In New Jersey, injuries (i.e., intentional and unintentional injuries combined) are the fourth leading cause of death and a leading cause of premature death in particular, contributing about 20% of the total years of potential life lost (YPLL) before age 651. The average age at death of an individual from injury in New Jersey is only 44.8 years, twenty-six years younger than the average age of a death from all causes (71.2 years)12. In addition, it has been estimated that injuries contribute to about 20% of all disability in New Jersey, with an annual cost of over one billion dollars in lost output and earnings11.

Injuries are responsible for approximately 3,200 New Jersey resident deaths and 100,000 hospitalizations annually. The direct yearly costs of hospitalization in New Jersey are estimated to be between $800 million and one billion. About one-third of all injury deaths in New Jersey are classified as intentional (i.e., either homicide or suicide) and of these, about 40 percent involve firearms. Unintentional injury deaths in New Jersey most often involve motor vehicles (about 40%), followed by drug overdoses and other poisonings (about 25%), falls (about 15%), and fires (about 4%). While most injury-related deaths in general occur among younger adults, and homicides in particular disproportionately affect young adults, unintentional injury deaths (particularly those related to falls and motor vehicle crashes) and suicide deaths occur disproportionately among elderly individuals1.

Among all types of injury, those to the brain are most likely to have serious long term consequences, often resulting in death or permanent disability with a likelihood of developing secondary conditions according to CDC2 . Both the younger and oldest age groups are particularly vulnerable to morbidity and mortality from TBI 8. Nationally, about 70,000 head injuries are classified as moderate to severe; persons with moderate to severe TBI will live the rest of their lives with a combination of cognitive, behavioral, and emotional deficits and will require long-term rehabilitation services. Almost no data are available on so-called mild TBI or on the impact of mild TBI on a persons's capacity to function normally9. The costs for severe head injury treatment and rehabilitation are estimated to be $310,000 per patient10. Preventing such injuries is likely to result in an enormous savings in terms of both direct costs, including acute care, long-term care and rehabilitation, and indirect costs, such as lost productivity.

The New Jersey Brain Injury Association (NJBIA), a nonprofit organization, is primarily devoted to providing counseling services to individuals with head injuries and their families to enable them to return to their premorbid normal life, or to cope with possibly lifelong residual impairments. NJBIA contracted, through the Division of Developmental Disabilities of the Department of Human Services, to conduct a comprehensive study on the needs of New Jerseyans with severe head injuries. A legislative act (A-2299, signed in January 1986) required that the study specifically address the needs for housing, employment, specialized medical care, respite care and psychological evaluation and training among individuals who have experienced a TBI. The Needs Assessment study concluded that a statewide registry for brain injuries for accurate and timely case finding and follow-up to evaluate the outcomes of traumatic brain injury, efficacy of services, and the impact on patients and their families was one of the pressing needs of the population with head injuries3.

Currently, there is no legislation regarding TBI reporting requirements in New Jersey. Recently, a resolution on programs and policies for persons with head injuries was introduced in the state Senate. The purpose was to study and develop recommendations regarding the most effective means of improving the quality and scope of rehabilitative services provided or supported by the state government. These efforts were intended to address the needs of persons who have sustained head injuries, other than those which are developmental in nature, whether or not permanent disability results, and to define the role of state government regarding the provision of policies and programs to assist this population4. The outcome of this resolution is unknown at this time.

The ability to develop and evaluate prevention programs has been limited due to inadequate surveillance data specific to TBI. Using the existing sources of data within the NJDOH, a TBI surveillance system with 1994 as the baseline was developed. Data from 1994 and subsequent years will provide a source for public health surveillance, prevention and evaluation of information specific to traumatic brain injuries. Maintaining an ongoing surveillance system is essential to monitoring progress towards year 2000 objectives, developing prevention activities, and to studying secondary conditions resulting from brain injuries. The purpose of this report is to determine the incidence of mortality and morbidity related to traumatic brain injuries, and to describe differences among populations and geographic areas for the baseline year, 1994.

The TBI surveillance system accounts for all traumatic brain injury related hospitalizations and deaths which occurred in New Jersey during 1994. Fatal and non-fatal cases were identified through New Jersey's statewide hospital discharge reporting system (UB-92) and death files. The 1994 UB-92 has patient identifiers (including name) and nine diagnostic fields. The single cause-of-death (SCD) file has the underlying cause of death identified by nosologists in the NJDOH, as well as identifiers (including names) which facilitate linking. The multiple cause-of-death (MCD) file has 20 fields for contributory causes of death, as well as an underlying cause, coded independently by the National Center for Health Statistics (NCHS) using the software Automated Classification of Medical Entities (ACME). The records were linked using patient name, date of birth and other key fields, and then merged into one statewide TBI system. The population on which statistics are derived consists of records classified into one of three categories:

1. NON-FATAL TBI-RELATED HOSPITALIZATIONS: New Jersey residents and nonresidents admitted to one of the ninety-four acute care facilities located in New Jersey, and discharged to home, or transferred to another acute care facility or nursing home (disposition status other than 'deceased').

2. FATAL TBI-RELATED HOSPITALIZATIONS: All residents and nonresidents hospitalized for TBI-related causes, and discharged from New Jersey acute care facilities with a disposition status of 'deceased' are counted as TBI-related hospitalized deaths.

3. NON-HOSPITALIZED TBI-RELATED DEATHS: This category includes all deaths from TBI-related causes which occurred at the scene of the injury, while being transported, in emergency departments, outpatient treatment offices, and nursing homes.

   The specific ICD codes used in identifying TBI cases are:

   N800-N801: Fractures of vault of skull or base of skull

   N803-N804: Other and unqualified skull fractures and multiple fractures of the skull

   N850-N854: Intracranial injuries including concussion, contusion, laceration, and hemorrhage without skull fractures

   Additional cases of TBI deaths are identified from death certificates with a diagnosis of N873 (other open wounds of head) accompanied by an E (External cause-of- injury) code.

   The case definition excludes:
   hospitalizations with V codes (factors influencing health status) as the primary diagnosis; newborn admissions; medical injury (N996-N999) as the primary diagnosis; late effects of skull fractures (N905) or intracranial injury (N907); and out-of-state hospitalizations to New Jersey residents due to unavailability of such information.

   The New Jersey TBI surveillance system for the baseline year 1994 was created using the following methodology:

   1. From the 1994 UB-92 data file of about 1.4 million hospital discharges, an extract was created by selecting the records meeting the TBI surveillance case definition criteria.

   2. The software package AUTOMATCH5 was used to identify multiple hospital admissions using probabilistic matching within the TBI hospitalization file (about 10,000 records), and also to match hospitalized deaths (461 records) with death records (about 74,000 records) using probabilistic record linkage. The matching variables used in the unduplication and matching process were: Soundex of last name, name, date of birth, medical record number, residence code, and gender. All TBI hospitalizations with a deceased status at the time of discharge were counted as hospitalized deaths; cases who were discharged alive and died later prior to readmission were counted as non-hospitalized deaths (see paragraph c). All individuals were counted only once in the analysis except for individuals with admissions for multiple episodes of TBI. Only about 1.5% of the TBI surveillance system consists of such multiple admissions. Multiple admissions with TBI as primary and /or secondary diagnoses accompanied by other health conditions were more common among the elderly.

   3. The SCD and MCD files linked by New Jersey death certificate number were used to account for non- hospitalized deaths. The SCD has patient identifiers, underlying cause-of-death and demographic information, but no nature of injury information. The MCD file has both nature and external cause of injury information, but no patient identifiers. These two files were merged using the death certificate number and then records with a TBI diagnosis were selected according to the case definition criteria. After eliminating the hospitalized death records identified (paragraph b) and out-of-state deaths to New Jersey residents, the remaining records were counted as the TBI-related non-hospitalized death records.

The hospitalization records identified from the UB-92 file have information on demographics, nature and cause of injuries, disposition status, source of admission, source of payment, cost, and length-of-stay. Mortality records have demographics, geographic and nature and cause of TBI injury. The TBI system also includes additional computed variables such as age at the time of injury occurrence, CDC injury severity scores7, and the CDC proposed assignment of E codes for injury data (not shown). The current report does not address hospitalization cost analysis and the CDC proposed assignment of E codes to injury data. However, we intend to add hospitalization cost analysis in the subsequent reports. Information on functional outcome, circumstances and other risk factors will also be added when information becomes available from the New Jersey trauma registry.

In 1994, there were 10,479 total incidents (deaths and acute care hospitalizations combined) of traumatic brain injury in New Jersey-132.6 per 100,000 persons (Table 1A). Of these, 1,079 persons died, resulting in an annual crude incidence rate of 13.7 deaths per 100,000 persons (Table 2B). The remaining 9,400 incidents involved persons treated for traumatic brain injuries in one or more of the 94 acute care facilities located in New Jersey - resulting in an annual non-fatal hospitalization rate of 118.9 per 100,000 persons.

Approximately 1.5% of the hospitalizations were identified either as subsequent admissions to the same acute care facility for another episode of injury or transfers to or from one level to another for continued treatment. About 42.6% of the non- fatal traumatic brain injuries and a majority (84.1%) of the non-fatal severe traumatic brain injuries were treated in the eight designated trauma centers12.

Of the 9,400 TBI hospitalized patients who survived through discharge, 2,119 (22.5%) were diagnosed as having a moderate to severe TBI and 6,957 (74.0%) had a mild TBI. For an additional 324 (3.4%) cases, severity of the TBI could not be determined (Table 1B). Among the survivors, there were more males with a moderate to severe TBI than females (a ratio of 2 to 1). The modal age group for persons discharged alive with moderate to severe TBI was 65 through 84 (Table 2C), and the leading cause of these TBIs was falls (not shown). Motor vehicle-related injury was the leading cause of non-fatal mild TBI (Table 1B). There were many more whites with a non-fatal moderate to severe TBI than persons of other racial backgrounds (Table 1B).

The most serious long-term consequences of TBI are experienced by persons who are diagnosed with moderate to severe brain injuries. The highest rates were found in the 85 and over age group, with a rate of 132.7 per 100,000, followed by persons 65 through 84 years of age (a rate of 54.9 per 100,000). In each of these age groups, the male rates were considerably higher than the female rates (Table 2C). Although the rate of mild TBI in the population 85 and over was also higher than in any other age group (251.7 per 100,000), this rate was also relatively high in persons 15 through 24 (149.0 per 100,000 persons) (Table 2D).

The leading cause of TBI was motor vehicle-related injury (E810-E825) (37.7%). This was followed closely by falls (E880- E888) (36.0%). These two external factors combined accounted for nearly three out of four TBI incidents (Table 1A). A detailed distribution of external cause of TBI in 1994 (not shown) indicates that while most of the motor vehicle-related injuries were of occupants of motor vehicles (73.8%), a substantial number involved pedestrians (14.2%); the remaining were motorcyclists (4.3%), pedal cyclists (4.3%), and persons in unspecified circumstances (3.4%). Another major cause of TBI is persons struck by or against an object. The majority of these were assault-related (66.1%) and the remainder (33.9%) were unintentional injuries.

Geographically, the greatest number of TBI incidents occurred to residents of Hudson County (1,137 or 11.4%), followed by residents of Essex County (966 or 9.7%) (Table 3A). Essex County residents experienced the highest combined number of fatal and non-fatal moderate to severe TBIs (369) (Table 3B). TBI incidence rates were highest for Atlantic County (207.6 per 100,000), followed by Hudson County (205.8 per 100,000). The lowest incidence of TBI occurred among residents of Hunterdon County (64.2 per 100,000) followed by residents of Morris County (74.6 per 100,000). The overall resident incidence rate was 125.8 per 100,000 population (Table 3D). These rates may be slightly understated due to lack of data on out-of-state hospitalizations of New Jersey residents.

Camden County residents had the highest number of motor vehicle-related TBIs (431), followed by residents of Bergen County (308). Hunterdon and Warren Counties had the lowest number of motor vehicle-related TBIs (32 and 36 respectively). Hudson County residents experienced the greatest number of TBIs from falls (416) (Table 3C). Hudson County also had the highest number of assault-related TBIs (216). Motor vehicle-related TBI incidence rates ranged from a low of 24.4 per 100,000 in Morris County to a high of 85.1 per 100,000 in Camden County. The overall motor vehicle-related TBI rate for New Jersey residents was 46.0 per 100,000 for New Jersey (Table 3D).

With respect to age, the greatest number of motor vehicle-related TBIs occurred among persons 25 through 44 years, followed by 15 through 24 years (Table 4A). The highest rates, however, were in the 15 through 24 year olds, with a rate more than twice the rate in 25 though 44 year olds (111.3 and 51.3, respectively) (Table 4B). The second leading cause of TBI, falls, occurred at highest numbers among 65 through 84 year olds (Table 4A). Males 85 and over, however, experienced the highest rate of fall-related TBIs (503.2 per 100,000). The greatest number of TBIs from falls in any age-sex subgroup occurred in females 65 through 84. While the largest number of assault-related TBIs occurred in 25 through 44 year old males, the highest rate by far occurred in males aged 15 through 24 (59.1 per 100,000)(Tables 4A and 4B).

The mean and median age at the time of TBI occurrence was 40.2 and 34.5 years respectively. On average, males were approximately 10 years younger (mean age of 36.7 years) than females (mean age of 46.3 years) at the time of traumatic brain injury (Table 5).

Analysis of TBI-related hospitalizations indicated that an overwhelming majority (82.1%) were released to home; 11.1% were transferred to/or discharged under some form of medical care; 4.6% expired; and the remaining 2.2% left against medical advice (Table 6). A majority (266 or 68.9%) of those who were released to other institutions (mostly rehabilitation) had moderate to severe levels of TBI.

The primary goal of New Jersey's TBI surveillance system is to provide accurate and timely information to facilitate planning and evaluating prevention programs. Recommended areas for further action and study include:

• Develop quality control measures to ascertain the extent to which the information provided is accurate and reliable.

• Complete the analysis of the validation study of a sample of hospital records of TBI patients to provide an estimate of the validity of TBI diagnoses reported on the UB-92 file.

• Study the reasons for discrepancies in diagnoses and cause of death between UB-92 and death files.

• Investigate the possibility of obtaining additional funding to be able to expand the surveillance system and to complete project objectives in a timely fashion.

• Enhance the current system by adding more data elements, for example, concerning follow-up of patients affected by traumatic brain injuries.

• Develop strategies to increase the likelihood that the information provided will be utilized by public and private agencies and public health professionals.

1. New Jersey Department of Health, Center for Health Statistics. (1996). New Jersey Health Statistics: 1994. Trenton, NJ.

2. U.S. Department of Health and Human Services, CDC and National Center for Injury Prevention and Control. (1995). Guidelines for Surveillance of Central Nervous System Injury.

3. New Jersey Head Injury Association, Inc. (1989). Traumatic Brain Injury in New Jersey. A Needs Assessment.

4. Senate Concurrent Resolution No. 93 State of New Jersey, introduced March 2, 1995 by Senators DiFrancesco and Sinagra.

5. AUTOMATCH Ver 3.0, Record Linkage Technology System. Matchware Technologies, Inc. 1995.

6. New Jersey Department of Health, Center for Health Statistics. (1997). Evaluation of External Cause-of-Injury Code Compliance in 1994 Hospital Discharge Data. Trenton,NJ.

7. David Thurman, Brigette Finklestein, Steven Leadbetter, and Joseph Sniezek. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. (1996). A Proposed Classification of Traumatic Brain Injury Severity for Surveillance Systems. Atlanta,GA.

8. Kraus, JF. Epidemiology of head injury. (1993). In: Cooper PR, ed. Head injury. 3rd ed. Baltimore, MD: Williams and Wilkins. Page 1-25.

9. Traumatic brain injury, rehabilitation-where we are. Position papers from The Third National Injury Control Conference "Setting the National Agenda for Injury Control in the 1990's". (1992). Page 503.

10. Public Health Services, U.S. Department of Health and Human Services. (1990). Healthy People 2000. The Economics of Prevention. 1990. Washington, D.C. Page 5.

11. Rutgers University. Disability and Health Economics Research Section, Bureau of Economic Research. (1993) Injury and Impairment in New Jersey. Results from the New Jersey Demographics of Disability Survey.

12. New Jersey Department of Health, Center for Health Statistics, 1997. [1994 injury surveillance file]. Unpublished data.

These notes discuss the sources of data used in creating the TBI surveillance system, data limitations, and procedures for case ascertainment.

1. DATA LIMITATIONS, PROBLEMS AND RESOLUTION:
   1. It should be noted that the surveillance system is based on multiple sources of information, and each data file is subject to a certain unknown degree of error. We believe that the UB-92 may overstate the incidence of TBI and the death file may understate the incidence due to differences in administrative and coding practices. Resources were too limited to conduct a comprehensive investigation of data quality. However, efforts related to data validation are conducted whenever possible to minimize errors. Caution should be exercised in interpreting the results presented in this report.

   2. All resident death certificates are included on the SCD file regardless of the state of occurrence. However, the system does not include out-of-state hospitalizations which occurred to New Jersey residents, as this information is not available. Efforts are underway to obtain information on hospitalizations of New Jersey residents in the neighboring states of New York and Pennsylvania. The effect on overall incidence rates is expected to be minimal assuming an equal number of out-of-state hospitalizations to New Jersey residents and hospitalizations of out-of-state residents in New Jersey (approximately 3.6% of total occurrences are of out-of-state residents and 1.5% are of unknown residence).

   3. Non-hospitalized deaths obviously have no in-patient charges. Information on emergency department and outpatient costs/charges is currently not available, but acute care hospital charges are available for both fatal and non-fatal hospitalizations. This report does not include TBI-related hospitalization cost analysis.

   4. Another limitation is that both mortality and hospitalization files lack information on Abbreviated Injury Scores (AIS), Glasgow coma scores, use of protective devices and alcohol involvement. TBI cases were classified according to CDC's severity scale7 . Collecting information on some optional variables (AIS scores, Glasgow coma scores, etc) may be beyond the scope of the 1994 surveillance system project. The New Jersey trauma registry, which is expected to be implemented in the future, may also be used for case verification and for collecting additional information such as Glasgow coma scores, protective devices and alcohol involvement. It is planned to add these data elements from the NJDOH trauma registry, when implemented.
      
      

2. VALIDATION OF HOSPITAL DEATHS FROM THE DEATH FILES AND VICE-VERSA:

   The software package AUTOMATCH was also used to match records (461) between the UB-92 data file and the single-cause-of death file for verification of vital status. Death certificate records were located for almost all (99%) hospital patients with a discharge status of 'deceased'. The key variables used in the matching process included last name, soundex of last name, first name, date of birth, date of death, disposition status, municipality code of residence and place of death. Minor discrepancies in the vital statistics information, and major discrepancies regarding diagnoses information (about 38% of the hospitalized deaths identified through records on the UB-92 file did not have a TBI-related cause of death indicated on their death certificates) were observed. These certificates are also being reviewed for verification of TBI diagnoses.

   Some of the data problems that remain to be resolved include:

   1. Five hospital discharge records for deceased patients (out of 461 records) could not be matched with a corresponding death certificate. These records have been excluded from the TBI surveillance system.
   2. Of the UB-92 records with a TBI diagnosis as a listed cause of hospitalization and a discharge disposition of death which were matched with a death certificate, only 62% (283 records), also had a TBI diagnosis as a contributory cause of death on the matching death certificate. The reason the remaining 38% (173 records) did not have a TBI diagnosis on the death certificate remains to be investigated. The current analysis includes these 173 hospitalized deaths with a TBI diagnosis which are counted as TBI- related hospitalized deaths in the surveillance system. This inclusion may have resulted in an over-counting of TBI incidence by 1.7% (173) cases out of 10,479 records in the surveillance system.

   3. Approximately 80 death certificates indicating TBI as a cause of death did not have a matching TBI-related hospitalization record, although the death certificate indicated that the death occurred in a hospital. These deaths may have occurred in an emergency department or outpatient care setting in a hospital, and the death records may have been indicated incorrectly as in-hospital occurrences. These records are counted as non-hospitalized deaths.

3. E CODE EVALUATION OF 1994 UB-92 DATA FILE:

   The NJDOH mandated that acute care hospital discharge records related to injury (ICD-9-CM codes N800-N995 with some exclusions) be accompanied by at least one E (external cause of injury) code and a Z code (place of occurrence) in the designated fields, beginning January 1, 1994. A final evaluation report6 for completeness of E coding was prepared in July 1997. The results indicated a high (over 99%) level of compliance with E code regulations, specific to TBI.

4. VALIDATION OF TBI HOSPITALIZATIONS:

   A validation study to assess the 1994 UB-92 data related to TBI injuries was conducted during February through June 1997. This study involved reviewing medical charts by selecting a statistically valid sample of 650 medical charts for case ascertainment. The validation data are being analyzed at the CHS. Preliminary analysis suggests that minor discrepancies between hospital charts and UB-92 records exist with respect to diagnoses and demographic information. One of the reasons for these discrepancies according to hospital administrators is that the computer files are often updated, but these updates are often not reflected in medical charts, especially if records are old. The computer files are submitted to the NJDOH at the end of each quarter.