Reports of Audiometric Testing in Service Medical Records of Military Veterans
One of the tasks for this study was to review the service medical records of military veterans to examine the compliance by the military services with regulations requiring audiograms. This chapter describes and presents results from the study to evaluate service medical records for the presence of audiograms performed when service members entered and left active duty (referred to here as entrance and separation audiograms). The Army, Navy, and Marine Corps have required such testing for all service members since at least the early 1980s, but some audiometric testing was being done in all of the services as early as the 1940s.
The specific language of the Statement of Task called for records to be “examined for regulatory compliance regarding audiometric surveillance (including reference, periodic, and termination audiograms).” Service members enrolled in a hearing conservation program are to receive “reference, periodic, and termination” audiograms at the beginning, annually, and at the end of their enrollment in the program. As noted in Chapter 5, only service members who meet certain noise-exposure criteria are placed in a hearing conservation program (see Table 5-1).
As framed, the stated task presented two problems. First, the hearing conservation programs of the Army, Navy, and Marine Corps did not have mandatory testing until the late 1970s and 1980. As a result, there is no basis for “compliance” with testing requirements for the period from World War II through the 1970s. Second, the services do not have central registries of personnel enrolled in their hearing conservation programs. As a result, it was not feasible to draw study samples limited to personnel who had participated in those programs.
The study that was conducted was based on data from service medical records of individuals who had served in the military, without regard to their enrollment in a hearing conservation program. In the committee’s view, this was a more appropriate basis for the study than considering only personnel who had been enrolled in hearing conservation programs. As has been noted throughout this report, it is critical that measures of hearing and tinnitus be obtained at entry and exit from military service for all personnel in order to document any changes in hearing or tinnitus that occur during military service. Although participants in hearing conservation programs may be exposed to known noise hazards, other military personnel may also be exposed to hazardous noise. Moreover, the Department of Veterans Affairs (VA) must consider claims regarding hearing loss or tinnitus from veterans, regardless of their enrollment in a service’s hearing conservation program. The majority of veterans receiving compensation for hearing loss at the end of 2004, for example, served in the military before the widespread implementation of hearing conservation programs in the late 1970s (Department of Veterans Affairs, 2005).
As conducted, this study assessed the extent to which audiometric test results were present in the medical records of 3,570 randomly selected service members who had separated from military service during one of five eras spanning the period from World War II to 2002. These records were reviewed, and the dates of all reports of audiometric tests were abstracted. The information on these tests was used to assess the proportion of service members who received audiometric tests at the beginning and the end of their military service. The study was to have included a review of records for Coast Guard personnel, but the Coast Guard did not grant the Institute of Medicine permission to have access to Coast Guard service medical records. Therefore, Coast Guard records were not included in this study.
The study protocol was approved by the Institutional Review Board of the National Academies.
Selection of Study Sample
Individual service medical records were selected for use in the study in the following manner. Random samples of service member identifiers were selected from listings of service members (referred to as rosters) available to the Medical Follow-up Agency (MFUA) of the Institute of Medicine. These rosters span the period from World War II to 2002. Five service eras were defined: 1940 to 1949, 1950 to 1969, 1970 to the year hearing conserva-
tion program regulations were implemented,1 the year hearing conservation program regulations were implemented to 1993, and 1994 to 2002. Lists of service members were generated from the available rosters for each branch of service (Army, Navy, Air Force, Marine Corps) and for each of the five service eras used. Records were assigned to one of the five service eras based on the service member’s date of separation from service (release from active duty). Additional detail about the sampling frames for these service eras is provided in Box 6-1.
Sample-size targets were set at 100 records for each service and era prior to 1970, and 200 records for each service and era after 1970. These targets were established based on a priori estimates of the percentages of files with audiograms and the degree of confidence sought in those numbers. On the basis of MFUA experience in obtaining service medical records, rosters were oversampled for each service branch–service era category to take into account files that were missing or otherwise unavailable. To reach the overall target of 3,200 records, 6,218 records were requested. When the desired sample size for a branch of service and service era had been met or exceeded, record review for that time period was discontinued and any remaining files were returned without review.
Record Access and Data Abstraction
It was possible for the service medical records that were sought for the study to be located at the National Personnel Records Center (NPRC) in St. Louis, Missouri, the VA Regional Management Center in St. Louis, or VA regional offices across the country (see Box 6-2). Lists of service members selected for the study were matched against VA’s Beneficiary Identification and Records Locator Subsystem (BIRLS) to ascertain the location of the records. Records were requested from the appropriate source. Records held by VA regional offices were mailed to the National Research Council office at the VA Regional Office in Washington, D.C., where they were abstracted under the supervision of MFUA staff. Records stored in St. Louis, at either the VA Regional Management Center or NPRC, were abstracted on-site under the supervision of MFUA staff.
Presence in the service medical record of any of the following forms was noted and recorded: report of medical examination (SF 88, DD Form 2808); medical history (SF 89, DD Form 2807); reference audiogram (AF 1491, DD Form 2215); or monitoring or termination audiogram (AF 1490,
No single comprehensive list exists of persons who served in the Armed Forces during the period from World War II to 2002, or of service members for whom a service medical record is available. Several different representative listings of service members were used to span this time period:
World War II through 1949: Medical Follow-up Agency’s (MFUA) World War II database (Roster #500) is derived from 1 percent or 2 percent samples of National Service Life Insurance policyholders. This insurance program for service members and veterans issued roughly 22 million policies from October 1940 through April 1951, and most of the participants served during World War II. Because Roster #500 differs from the other listings in that it is not based on year of separation from military service, the potential existed for persons selected from this roster to also be selected from one of the other listings. The records selected from Roster #500 were reviewed to identify any duplication, and none was found. Where available, entrance and separation dates for service members identified through this roster were ascertained using dates provided through the Beneficiary Identification and Records Locator Subsystem (BIRLS).
1950 through 1969: Two MFUA rosters were used to span the time period from 1950 through 1969. Roster #552 is a 0.1 percent sample of Armed Forces separations and spans separation years from 1950 to 1959 for some of the services, but it is most useful for the period 1950–1954. Roster #588 is a 1 percent sample of Armed Forces separations from 1955 through 1969.
1970 through 2002: The Defense Manpower Data Center (DMDC) of the Department of Defense maintains a computerized database of service records covering the period from the early 1970s to the present. DMDC has provided MFUA with a 1 percent sample of all the separations in its database. This listing provided sampling frames for the time periods “1970 through year of regulation,” “year of regulation through 1993,” and “1994 through 2002.”
Service medical records are either stored by the services at the National Personnel Records Center (NPRC) in St. Louis, Missouri, or held by the Department of Veterans Affairs (VA). For individuals who left military service before approximately 1994, VA has custody in one of its regional offices of those service medical records that have been used in support of a medical claim. Service medical records of those who left military service before 1994, but have not filed medical claims, are stored at NPRC. VA holds the service medical records of all service members who separated from military service since 1994, either at the VA Regional Management Center in St. Louis or at VA sites across the country, where they are in use in conjunction with claims.
DD Form 2216). From these forms, information was recorded about the date, duty occupation code, presence/absence of numerical data from an audiogram, type of audiometer, type of hearing protection issued/used, purpose of or reason for the test, and the presence/absence of a significant threshold shift (STS). Numerical reports of hearing thresholds were not abstracted.
Table 6-1 shows the number of medical records reviewed for each service and time period. As can be seen, except for the Air Force for the two earliest eras, sample sizes were at least 90 percent of the targeted size and, most often, exceeded the targets. The fact that the Air Force was established as a separate service only in 1947 may have influenced the availability of Air Force records for the earliest era.
Presence of Any Audiograms
Overall, 82 percent (95% CI 80–83 percent) of the records reviewed contained the report of at least one audiometric test.2 As shown in Table 6-2, the percentage of records containing audiometric data was lowest for those who separated from service in the earliest time periods. The Air Force was an early leader in audiometric testing. By the 1970s, however, at least 93 percent, and typically 98–100 percent, of the medical records sampled from each of the services had at least one audiogram.
Presence of Entrance Audiograms
Table 6-3 shows the percentages of service medical records with audiograms obtained within 60 days before or after the service member’s entrance into active duty. The committee decided on ±60 days as a reasonable, but arbitrary, window. Using this window reflects the committee’s recognition that not all audiograms will have been obtained for all personnel exactly on the dates of entry into and separation from the military while still being in reasonably close proximity to the dates of entry and separation. There is no question that the duration of this time window will have an impact on the percentage of service medical records considered to contain entry and separation audiograms. The wider the window, the higher the resulting percentages are. The maximum possible
TABLE 6-1 Number of Service Medical Records Reviewed and Abstracted
percentages for an unlimited time window are captured by the data shown in Table 6-2. Again, to be of value in assessing whether hearing loss or tinnitus are service connected, measurements must be obtained near the beginning and end of military service, and for this analysis, the committee considered 60 days to be an acceptable time frame for obtaining these measurements.
In general, although the effect varied considerably with the military service branch and era, additional analyses of these data indicated that doubling of the time window from ±60 days to ±120 days increased the percentages shown in Tables 6-3 through 6-5 by approximately 9–12 percentage points. That is, if 30 percent of the service medical records for a given branch and era had entrance audiograms when using a time window of ±60 days, then about 39–42 percent of the records contained entrance
TABLE 6-2 Percentages of Service Medical Records (95% Confidence Intervals) with Reports Containing Any Numeric Data from an Audiogram (n = 3,570)
TABLE 6-3 Percentages of Service Medical Records (95% Confidence Intervals) with Reports of Audiometric Examinations Within 60 Days of Entry into Active Duty (n = 3,212)
From Table 6-3, for personnel who left military service during the 1970s, the percentage of service medical records containing such audiograms was 30–37 percent across all branches of the military. For the two earlier periods, the percentages were appreciably lower for all branches. For both the Army and the Air Force personnel who left military service during the periods since the 1980s, the percentage of records with entrance audiograms was slightly lower than for the 1970s era, but the differences were not statistically significant. Percentages of Marine Corps and Navy records with audiograms, on the other hand, increased in each era since the 1970s and for the most recent time period were 69 percent and 70 percent, respectively.
Presence of Separation Audiograms
Separation audiograms were defined as audiograms recorded within 60 days of a service member’s release from active duty. As shown in Table 6-4, the percentage of records with a separation audiogram has ranged from 0 to 54 percent over the time periods considered. The percentage of records with separation audiograms has declined for all the services in the most recent period. The general trend across branches is for the percentage of files with separation audiograms to increase steadily or remain the same up to the 1970s, to remain fairly constant from the 1970s through the 1980s, and to decrease during the most recent era. From 1970 to 2002, however, the percentages for the Navy and Marine Corps are about twice as high as those for the Army or Air Force.
TABLE 6-4 Percentages of Service Medical Records (95% Confidence Intervals) with Reports of Audiometric Examinations Within 60 Days of Release from Active Duty (n = 3,226)
Presence of Entrance and Separation Audiograms
The percentages of records containing both an audiogram obtained within 60 days of entrance and an audiogram obtained within 60 days of separation from military service are shown in Table 6-5. As noted with regard to either of these audiograms alone, there is an increase in the percentages from the earliest era to the 1970s. After that, the percentages for the Army and Air Force decreased in the more recent periods, whereas the percentages for the Navy and Marine Corps increased slightly or held steady in each successive time period. Although the general trends over time are similar to those observed for each type of audiogram (at entry and at separation) alone (see Tables 6-3 and 6-4), the overall percentages of files having both audiograms are considerably lower, as expected.
TABLE 6-5 Percentages of Service Medical Records (95% Confidence Intervals) with Reports of Audiometric Examinations Within 60 Days of Entrance into and Release from Active Duty (n = 3,210)
COMPLIANCE WITH REGULATIONS
As noted in Chapter 5, the Army, Navy, and Marine Corps issued regulations by 1980 requiring that audiograms be performed at entrance into and separation from active duty (regulations issued in 1980 for the Army and 1979 for the Navy and Marine Corps). These regulations do not specify time windows for performing the audiograms, so, as noted, a ±60-day window was selected for this analysis.
A 1956 Air Force regulation required audiograms in conjunction with routine physical examinations at entrance into and separation from active duty, but by 1973 audiograms were no longer required immediately upon entrance. Standard practice, as of 1982, was to obtain an audiogram within 90 days of service members’ assignment to their first permanent duty stations (Department of the Air Force, 1982), which might typically take place up to 6 months after entry into service (Pluta, 2005). Separation audiograms were required only of those Air Force personnel routinely exposed to hazardous noise (Department of the Air Force, 1982). As of this writing, it is not clear that any Air Force regulation requires entrance or exit audiograms for all Air Force military personnel. As of 1996, however, the Department of Defense established a requirement that all military personnel receive a reference audiogram at basic training prior to noise exposure (DoD, 1996). The percentages in Table 6-3 for the era 1994–2002 suggest that this testing occurred for about 20–30 percent of the personnel in the Army or Air Force and about 70 percent of those in the Navy or Marine Corps. At present, the Department of Defense does not require an audiogram upon separation from service.
Thus, with regard to the Army, Navy, and Marine Corps, it appears that compliance with the regulations in place since the early 1980s has been and is incomplete, particularly with regard to audiograms obtained at the time of separation from active duty. Navy and Marine Corps percentages are consistently higher than those for the Army, however. The percentages of Air Force records with entrance and separation audiograms declined over the three later time periods and were the lowest of the four services for the most recent time period. However, because Air Force regulations for the two later periods did not call for entrance or separation audiograms for all personnel, the service’s testing practices may have been in better compliance with existing regulations than is reflected under the criteria used in the analysis reported here. Nevertheless, as noted, since 1996, the Department of Defense has required entrance audiograms before noise exposure at basic training for all military personnel.
Several factors should be taken into account in interpreting these data. One is that the data on the presence of an audiogram within 60 days of entrance into active duty are reported on the basis of the era in which the
service member was released from active duty. Thus, individuals who served for many years may have begun their military service during a previous era with requirements for entrance testing that were different from those in place at the time they left military service. Because the Air Force has had higher reenlistment rates (DoD, 1997) and a longer median time-in-service than the other services, this potential discrepancy was more likely for Air Force study participants. For example, for the current study, whereas the median length of service for the Army, Navy, and Marine Corps across all five time periods ranged from 2 to 4 years, the corresponding values for the Air Force were 7, 7, and 10 years in the three most recent eras.
When a ±1-year window was used for analysis of “separation” audiograms, the percentages of personnel receiving an audiogram within the 1994–2002 time period were 38, 65, 83, and 86 percent in the Air Force, Army, Marine Corps, and Navy, respectively.3 All these percentages are clearly much higher than the values derived for the same time period using a testing window of ±60 days (Table 6-4), as expected. Even here, however, the percentage of personnel tested is appreciably lower for the Air Force than for the other military services.
It should be noted that this analysis did not evaluate the extent to which the services obtained reference or termination audiograms for personnel entering and leaving hearing conservation programs. The reasons for not focusing exclusively on those personnel in hearing conservation programs were noted previously.
The following findings are based on the data in Tables 6-2 through 6-5, each of which makes use of a ±60-day time window. Although the specific percentages cited in some of the findings are dependent upon the time window used in the analysis, the general features of the data are the same for time windows of ±60 days (Tables 6-3 through 6-5) or ±120 days (Appendix E).
FINDING: Review of a sample of service medical records of military veterans indicates that compliance with requirements for audiometric testing at entrance into service has been limited, even in the most recent eras, and did not exceed 70 percent in any branch or era when using a ±60-day window for analysis.
FINDING: Review of a sample of service medical records of military veterans indicates that audiometric testing at separation from service has
been limited, even in the most recent eras, and did not exceed 54 percent in any branch or era when using a ±60-day window for analysis.
FINDING: Review of a sample of service medical records of military veterans indicates that audiometric testing at both entrance into and separation from service has been extremely limited, even in the most recent eras, and did not exceed 34 percent in any branch or era when using a ±60-day window for analysis.
Department of the Air Force. 1982. AF Regulation 161-35: Hazardous Noise Exposure. Washington, DC: Department of the Air Force.
Department of Veterans Affairs. 2005. Service-Connected Compensation—Veterans with Hearing Impairment as of December 21, 2004. Spreadsheets provided to J. Durch, Institute of Medicine, March 21.
DoD (Department of Defense). 1996. Department of Defense Instruction 6055.12: DoD Hearing Conservation Program. Washington, DC: Department of Defense.
DoD. 1997. Selected Manpower Statistics: Fiscal Year 1997. DTIC/NTIS No. DIOR/M01-9. Washington, DC: Department of Defense.
Gardner MJ, Altman DG. 1989. Statistics with Confidence: Confidence Intervals and Statistical Guidelines. London, United Kingdom: British Medical Journal.
Pluta R (U.S. Air Force). 2005. RE: Seek Help with Technical Review. E-mail to L. Joellenbeck, Institute of Medicine, June 24.