Sensitivity of Multiple-Cause Mortality Data for Surveillance of Deaths Associated with Head or Neck Injuries

Summary

Problem/Condition: Multiple-cause mortality data was assessed as a source of information for surveillance of deaths associated with head or neck injuries.

Reporting Period Covered: 1985-1986

Description of System: Data on causes of death were abstracted from death certificates in New Mexico and coded according to criteria of the International Classification of Diseases, Ninth Revision (ICD-9). Deaths with an external cause-of-death (E) code as the underlying cause of death and one or more head or neck injury nature-of-condition (N) codes as contributing causes of death were considered head or neck injury deaths. These data were compared with data for head or neck injury deaths obtained from computerized records from the New Mexico Office of the Medical Investigator (OMI). Data for alcohol or drug use were abstracted from both systems.

Results: Of the 699 head or neck injury deaths coded by the New Mexico OMI system in 1985-1986, 536 were identified as head or neck injury deaths in multiple-cause mortality data (sensitivity = 76.7%). Firearms were the leading cause of head or neck injury deaths, followed by motor vehicles. Multiple-cause mortality data contained alcohol codes for only 3.7% of OMI records with blood alcohol concentrations greater than or equal to 0.10 mg/dL and contained drug codes for none of the OMI records with positive toxicology tests for drugs.

Interpretation: The sensitivity of multiple-cause mortality data was relatively high for surveillance of head and neck injury deaths. This information source may be useful for conducting statewide surveillance for mortality from head injuries; however, the sensitivity and positive predictive value of these data require further assessment. Multiple-cause data substantially underestimate the extent of alcohol and drug involvement for head or neck injury deaths.

Actions Taken: The findings in this investigation have prompted further assessment of the usefulness of death certificate data for head injury surveillance.

INTRODUCTION

Although the incidence of head injuries has been estimated for local jurisdictions in the United States (1-7), surveillance at the state and national levels is needed to monitor such injuries and evaluate the impact of preventive measures. Some researchers have suggested that CDC's multiple-cause mortality data from death certificates could be used for such surveillance because the data are readily available and require no new collection system (8). Multiple-cause mortality data include underlying cause of death and up to 20 associated medical conditions, such as type of injury. Although no true standard exists for evaluating the validity of death certificate diagnoses, a medical examiner system is probably the best standard available because the cause of death in cases investigated by medical examiners is subject to more thorough investigation than are routine cases. Using 1985-1986 data from the New Mexico Office of the Medical Investigator * (OMI) as the standard, we compared the sensitivity of multiple-cause mortality data with data from deaths investigated by the OMI for ascertaining deaths caused by head or neck injuries.

METHODS

In New Mexico, by law, the OMI investigates deaths occurring on nonfederal lands that are of unknown cause or are sudden, violent, or untimely. (The OMI investigates similar types of deaths on federal lands only when invited by federal authorities). The chief medical investigator is a forensic pathologist who is responsible for appointing judicial district medical investigators. All district medical investigators are physicians.

Computerized OMI records were reviewed for all injury deaths among U.S. residents occurring in New Mexico during the period 1985-1986. All records with codes for subdural hematoma or head and neck injury were classified as head or neck injury deaths. In addition, records that might have included head or neck injury (OMI records coded as multiple injury, stab wound, gunshot wound, or child abuse) were reviewed manually; these deaths were classified as resulting from head or neck injury if there was any mention of injury to a head or neck anatomic structure. All head or neck injury deaths were categorized by external cause: motor vehicle, intentional and unintentional firearm discharge, fall, or other. Three injury epidemiologists reviewed and classified 27 records for which it was unclear whether the deaths were attributable to injury.

Head or neck injury deaths investigated by the OMI were classified as alcohol associated if tests indicated that the decedent had a blood alcohol concentration (BAC) greater than or equal to 0.10 mg/dL; for records with more than one BAC, the highest recorded level was used for classification. Deaths were considered to be drug associated if toxicology tests indicated the presence of any narcotic, barbiturate, analgesic, sedative, antidepressant, stimulant, tranquilizer, cannabinoid, or other psychoactive substance.

Using a broad definition for head or neck injury-associated deaths (Table_1), we reviewed multiple-cause mortality tapes for deaths among U.S. residents who died in New Mexico during the period 1985-1986. Multiple-cause mortality data are abstracted from death certificates and coded onto tapes according to criteria of the International Classification of Diseases, Ninth Revision (ICD-9) (9). We included all deaths with an external cause-of-death (E) code as the underlying cause and one or more nature-of-condition (N) code for head or neck injury as a contributing cause (Table_1). Categories for external cause were motor vehicle, unintentional or intentional firearm discharge, fall, or other. Multiple-cause mortality head or neck injury deaths were defined as alcohol or drug associated if certain codes were present (Table_1).

To assess the sensitivity of multiple-cause mortality data for ascertaining head or neck injury deaths, we matched OMI and multiple-cause mortality records involving head or neck injuries by sex, race, age, and date of death. We included 30 records from the OMI system that matched with a corresponding record from the vital records system for each variable except age, after manual review showed matching birth dates but different ages.

Multiple-cause mortality data contain information about county of death, but do not indicate whether deaths occurred on federal or nonfederal land. (Federal lands are present in the majority of New Mexico counties.) As a result, the specificity and positive predictive value of multiple-cause mortality data for head or neck injuries could not be calculated.

RESULTS

Of 699 head or neck injury deaths included in the OMI system, 536 were also in multiple-cause mortality data, for an overall sensitivity of 76.7%. The sensitivity was highest for deaths coded for firearms and lowest for those coded for falls (Table_2). Deaths in the OMI system not identified by multiple-cause mortality data were similar by race and sex to those found in both systems, but differed significantly by age (p less than 0.02) (Table_3).

Most head or neck injury deaths included in multiple-cause mortality data were coded under five general rubrics (Table_4). Of the 536 deaths common to both systems, 486 (90.7%) were coded as due to intracranial injury (N850-N854) or open wound of the eye, ear, or head (N870-N873). (The majority of the latter were coded as N873.1 -- open wound of the scalp.) All 15 multiple-cause mortality records with neck injury codes also had head injury codes.

OMI records indicated that 460 (85.8%) of 536 persons had been tested for the presence of alcohol or drugs, 168 (36.5%) of whom had positive tests. The sensitivity of multiple-cause mortality data was 3.7% (5/134) for ascertaining deaths associated with BACs greater than or equal to 0.10 mg/dL and no evidence of drug use, 0% for 17 deaths associated with drugs, and 0% for 17 deaths associated with both alcohol and drugs.

DISCUSSION

In the OMI system, head and neck injuries could not be separated. Because all deaths with neck injury codes reported in multiple-cause mortality data also had head injury codes, our study may be more representative of deaths resulting from head rather than neck injury.

Multiple-cause mortality data had fairly high sensitivity for ascertaining deaths reported in the OMI system that involved head or neck injury, especially for deaths associated with firearms. However, multiple-cause mortality data were not useful for identifying alcohol- and drug-associated deaths due to these injuries. Underreporting of alcohol use on death certificates has been noted before (10), but the extent of underreporting of alcohol use on death certificates with head or neck injury codes has not been previously reported. Death certifiers should obtain drug and alcohol tests and report the results on death certificates, even if delays in obtaining test results make it necessary to file supplemental reports.

In contrast to previous findings (2,4-8), firearms were the leading cause of head injury deaths. This finding underscores the importance of including open wounds of the eye, ear, or head when these deaths are defined, since most head injury deaths from firearms had one of those codes. In a study in which multiple-cause mortality data were used without the category for open wound of the eye, ear, or head, 315,328 deaths from head injury were found in the United States from 1979 through 1986 (8). When we reanalyzed the data and included these codes, we found an additional 102,404 deaths due to head injury, of which 97.0% contained an underlying cause of death attributed to firearms (11). This finding may indicate that nosologists use codes for open wounds to the eye, ear, or head inappropriately for head injury deaths, since these codes should be used for wounds that were incidental to intracranial injury (9).

The ICD codes used by researchers to define head injury have varied widely (3,5-8 ). Although we searched multiple-cause mortality records for a wide range of ICD-9 codes corresponding to head injury, we found that greater than 90% of deaths due to head injury were coded as intracranial injury or open wound of the eye, ear, or head. Further exploration is needed to determine the adequacy of a surveillance system for head injury mortality that uses a simple case definition with these two categories.

Although multiple-cause mortality data underascertain head or neck injury deaths from falls, our results demonstrate that death certificate data may be appropriate and useful for monitoring head injury deaths. Determining the sensitivity and positive predictive value is needed in other localities so that the use of multiple-cause mortality data as a surveillance system for mortality from head injury can be fully assessed. If multiple-cause mortality data are found to have adequate sensitivity and positive predictive value, then, at a minimal cost, state health departments could use these data for head injury mortality surveillance to monitor patterns, set priorities, and evaluate the effects of efforts to prevent head injuries.

Acknowledgment

We thank Patricia Holmgreen, M.S., for assistance in data analysis and James Mercy, Ph.D., and Richard Sattin, M.D., for their editorial comments.

References

1. Kraus JF. Epidemiology of head injury. In: PR Cooper, ed. Head injury. Baltimore: Williams and Wilkins, 1987:1-19.

2. Annegers JF, Grabow JD, Kurland LT, Laws ER. The incidence, causes, and secular trends of head trauma in Olmsted County, Minnesota, 1935-1974. Neurology 1980;30:912-9.

3. Cooper KD, Tabaddor K, Hauser WA, et al. The epidemiology of head injury in the Bronx. Neuroepidemiology 1983;2:70-88.

4. Jagger J, Levine JI, Jane JA, Rimel RW. Epidemiologic features of head injury in a predominantly rural population. J Trauma 1984;24:40-4.

5. Klauber MR, Barrett-Connor E, Marshall LF, Bowers SA. The epidemiology of head injury: a prospective study of an entire community -- San Diego County, California, 1978. Am J Epidemiol 1981;113:500-9.

6. Whitman S, Coonley-Hoganson R, Desai BT. Comparative head trauma experiences in two socioeconomically different Chicago-area communities: a population study. Am J Epidemiol 1984;119:570-80.

7. Kraus JF, Black MA, Hessol N, et al. The incidence of acute brain injury and serious impairment in a defined population. Am J Epidemiol 1984;119:186-201.

8. Sosin DM, Sacks JJ, Smith SM. Head injury-associated deaths in the United States, 1979-1986. JAMA 1989;262:2251-5.

9. Manual of international classification of diseases, injuries, and causes of death, 9th revision. Geneva: World Health Organization, 1977.

10. Pollock DA, et al.: Underreporting of alcohol-related mortality on death certificates of young US army veterans. JAMA 1987;258:345- 8.

11. Sosin DM, Nelson DE, Sacks JJ. Head injury deaths: the enormity of firearms {letter}. JAMA 1992;268:791.

In New Mexico, medical examiners are known as medical investigators, and the state office responsible for investigating deaths is the Office of the Medical Investigator.
Table_1
Note: To print large tables and graphs users may have to change their printer settings to landscape and use a small font size.

TABLE 1. International Classification of Diseases, Ninth Revision (ICD-9), nature of
condition (N) and external cause of death (E) codes used for multiple-cause mortality
definitions
==================================================================================================
                                     Nature of condition (N) codes
------------------------------------------------------------------------------------------------
                                              Head injury
------------------------------------------------------------------------------------------------
800-804         Fracture of skull
830.0-830.1     Dislocation of jaw
850-854         Intracranial injury, excluding those with skull fracture
870-873         Open wound of ocular adnexa, eyeball, ear, other wound of head
905.0           Late effect of fracture of skull and face bones
907.0           Late effect of intracranial injury without mention of skull fracture
907.1           Late effects of injury to cranial nerve(s)
910             Superficial injury of face, neck, and scalp (except for eye)
918             Superficial injury of eye and adnexa
920             Contusion of face, scalp, and neck
921             Contusion of eye and adnexa
925.0           Crushing injury of face, scalp, and neck
950-951         Injury to optic nerve and pathways, injury to other cranial nerves
953.0           Injury to cervical nerve root
959.0           Injury, other and unspecified, to face and neck
------------------------------------------------------------------------------------------------
                                              Neck injury
------------------------------------------------------------------------------------------------
344.0           Quadriplegia
805.0-805.1     Fracture of cervical spine without mention of spinal cord injury
806.0-806.1     Fracture of cervical vertebral column with spinal cord injury
807.5-807.6     Fracture of larynx and trachea
839.0-839.1     Dislocation of cervical vertebra
847.0           Sprains and strains of neck
848.1-848.2     Sprains and strains of septal cartilage of nose, jaw, thyroid region
874             Open wound of neck
900             Injury to blood vessels of head and neck
905.1           Late effect of fracture of spine and trunk without mention of spinal cord injury
906.0           Late effect of open wound of head, neck, and trunk
907.2           Late effect of spinal cord njury
908.3           Late effect of injury to blood vessel of head, neck, and extremities
------------------------------------------------------------------------------------------------
                                              Alcohol use
------------------------------------------------------------------------------------------------
291             Alcoholic psychoses
303             Alcohol dependence syndrome
305.0           Alcohol abuse
790.3           Excessive blood level of alcohol
------------------------------------------------------------------------------------------------
                                              Drug use
------------------------------------------------------------------------------------------------
292             Drug psychoses
304             Drug dependence
305.2-305.9     Nondependent abuse of drugs
965             Poisoning by analgesics, antipyretics, and antirheumatics
967             Poisoning by sedatives and hypnotics
968.5           Poisoning by surface and infiltration anesthetics
969             Poisoning by psychotropic agents
970.0           Poisoning by analeptics
977.0           Poisoning by other and unspecified medicinal substances: dietetics
------------------------------------------------------------------------------------------------
                               External cause-of-death (E) codes
------------------------------------------------------------------------------------------------
Motor vehicle:  810-825
Falls:          880-888
Firearms:       922.0-922.9, 955.0-955.4, 965.0-965.4, 970, 985,0-985.4
Other:          800-809, 826-879, 889-921, 923-954, 955.5-955.9, 956-964, 965.5-965.9,
                  966-969, 971-984, 985.5-985.9, 986-999
------------------------------------------------------------------------------------------------
==================================================================================================

TABLE 2. Sensitivity of multiple-cause mortality data for identifying head or neck injury
deaths in the Office of the Medical Investigator (OMI) records, New Mexico, by external
cause of death -- 1985-1986
===========================================================================================
                                 No. records
                        -----------------------------
                                       Multiple-cause
External cause          OMI data       mortality data       Sensitivity
-----------------------------------------------------------------------
Firearms                  303               277                91.4%
Motor vehicle             247               172                69.6%
Other                      96                63                65.6%
Fall                       53                24                45.3%

Total                     699               536                76.7%
-----------------------------------------------------------------------
===========================================================================================

TABLE 3. Demographic characteristics of persons with head or neck injury deaths
identified in both the multiple-cause mortality and Office of the Medical Investigator
(OMI) systems, compared with persons with head or neck injury deaths identified only
by records -- New Mexico, 1985-1986
=============================================================================================
                   Identified in both systems   Identified only in OMI system
                   --------------------------   -----------------------------
Characteristic           N         (%)                  N         (%)           P-value *
-----------------------------------------------------------------------------------------
Median age
  (years)                30                             35                        0.02
Sex
  Male                  417       (77.8)               121       (74.2)
  Female                119       (22.2)                42       (25.8)           0.40
Race
  White                 471       (87.8)               138       (84.7)
  Black                  10       ( 1.9)                 2       ( 1.2)           0.35
  Other                  55       (10.3)                23       (14.1)
-----------------------------------------------------------------------------------------
* The Kruskal-Wallis and chi-square tests were used to compare a) median age and b) sex and
  race distributions, respectively.
=============================================================================================

TABLE 4. ICD-9 * Nature-of-condition codes for 536 persons with head or neck injuries
common to the multiple-cause mortality data and Office of the Medical Investigator
systems -- New Mexico, 1985-1986
================================================================================================
                        Number and percent of deaths with nature-of-condition code +
                -----------------------------------------------------------------------------
                                                                                  Other and
                                                  Open wound                     unspecified
                                 Intracranial   of eye, ear, or   Spinal cord   face and neck
                Skull fracture      injury           head           injury         injury
Cause             (800-804)       (850-854)        (870-873)        (952.0)        (959.0)
---------------------------------------------------------------------------------------------
Firearms
  (n=277)          1 ( 0.4)         4 ( 1.4)      276 (99.6)       0 ( 0.0)        0 ( 0.0)
Motor vehicle
  (n=172)          4 ( 2.3)       169 (98.3)        0 ( 0.0)       1 ( 0.6)       10 ( 5.8)
Other (n=63)       4 ( 6.3)        52 (82.5)       15 (23.8)       0 ( 0.0)        3 ( 4.8)
Falls (n=24)       2 ( 8.3)        23 (95.8)        0 ( 0.0)       0 ( 0.0)        1 ( 4.2)
---------------------------------------------------------------------------------------------
* International Classification of Diseases, 9th Revision.
+ Numbers in parentheses are row percents and total more than 100 because some death
  certificates had >1 nature-of-condition code; similarly, individual row numbers do not equal
  row totals because multiple codes were assigned to one event.
================================================================================================

Disclaimer   All MMWR HTML versions of articles are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the electronic PDF version and/or the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

Page converted: 09/19/98

HOME  |  ABOUT MMWR  |  MMWR SEARCH  |  DOWNLOADS  |  RSS |  CONTACT POLICY  |  DISCLAIMER  |  ACCESSIBILITY

Morbidity and Mortality Weekly Report Centers for Disease Control and Prevention 1600 Clifton Rd, MailStop E-90, Atlanta, GA 30333, U.S.A Department of Health and Human Services

This page last reviewed 5/2/01