Letters to the Editor

Military Medicine, May 2005 by Bohnker, Bruce K, Pocock, James, Gray, Gregory C

To the Editor:

I wanted to thank LCDR Riddle et al. for their detailed analysis of DNBI (disease and non-battle injury) rates for ships within the US Fifth Fleet during 2000-2001 [Milit Med 2004; 169:787-794]. Their study supported our findings for differential DBNI rates for various ship types. They also expanded the analysis to include time of year changes, and included SIQ (sick in quarters) and light duty rates that are significant contributions to the literature in understanding shipboard medical needs. Shipboard DNBI rates continue to warrant monitoring and study, with expansion to other areas of responsibility and operational tempo.

The importance of monitoring DNBI rates was demonstrated by Whittaker et al. in their manuscript dealing with shipboard gastroenteritis [Milit Med 2004; 169:747-750]. The USS Theodore Roosevelt (CVN 71) reported an outbreak of 451 cases of viral gastroenteritis attributed to norovirus. This follows previous reports of similar outbreaks on USS Forrestal (CV 59), USS Saratoga (CV 60), USS Constellation (CV 64), USS America (CV 66), USS Pelleliu (LHA 5) and others, as well as from numerous cruise ships.

Their report was notable for an outbreak occurring in CONUS, while many previous Navy shipboard outbreaks have been reported from overseas liberty ports. That outbreak demonstrated the importance of trained shipboard medical personnel and close attention to shipboard medical supplies. The threat of shipboard norovirus continues to justify funding support by US Department of Defense Global Emerging Infections Surveillance and Response System (DoD GEIS).

CAPTBruce K. Bohnker, MC, USN, FS

Chesapeake, Vriginia

To the Editor:

I am a writer looking for information about military nurses during the Korean War. I am starting to outline a novel in which one of the principal characters is a nurse either at a MASH location or at an Army General Hospital in Japan. She cares for a badly wounded Army lieutenant who becomes infatuated with her. I am looking for medical personnel who can answer these questions:

How long was the service obligation for a military nurse (specifically Army nurse) in the period 1950-53?

Was a bachelor's degree a requirement to become a military nurse in 1950-53?

Were there Army nurses in Korea or Japan who did not have college degrees in 1950-53?

If a nurse entered the military (Army) in 1949-1950, how long was her service obligation?

If anyone can answer any of these questions, I would appreciate hearing from you. My email is: jaglpo@cs.com, and my address is: 2349 Sapphire Lane, East Lansing, MI 48823. Thank you.

MG James Pocock, USAR (Ret.)

East Lansing, Michigan

To the Editor:

We read with great interest the recent paper by Levine et al. [Milit Med 2005; 170:149-153] suggesting increased evidence of testicular cancer among veterans of the 1991 Gulf War. We understand that the authors based their conclusions on cancer registry data from New Jersey and the District of Columbia for the period 1991 to 1999. We wonder if the authors may have been misled by ascertainment biases.

During this time period, many Gulf War veterans were referred to Department of Veterans Affairs and Department of Defense referral hospitals in New Jersey and Washington, DC for Gulf War veteran health examinations, and many of these veterans were referred from nearby states. As such comprehensive examinations were not available to the comparison group of nondeployers, the Gulf War veterans' referrals and systematic evaluations could have artificially elevated testicular cancer detection and proportion estimates. There are several ways to assess ascertainment bias. A first step might be to perform the linkages again and reexamine case data.

How many of the 17 Gulf War veteran cases had a temporally-related evaluation in the Gulf War veteran registries at the time of their testicular cancer diagnoses? How many had homes of record from outside of the District of Columbia or New Jersey? We encourage the authors to conduct additional analyses to rule out ascertainment bias.

We also encourage the authors to address the inherent limitations associated with the use of "proportional incidence ratios" (PIR). Analogous to proportional mortality ratios, this technique evaluates only the relative proportion of one cancer among all cancers in the population. An elevated PIR can be misleading if the overall incidence of cancer, or the incidence of some other cancers, is actually lower in the population of interest. If the PIR for testicular cancer was higher among deployers compared to nondeployers, this implies that the PIR for some other cancer(s) was lower among deployers compared to nondeployers. It would be revealing to display data from all cancers and their adjusted incidence rates, rather than PIRs for this very small number of cancer cases.

Gregory C. Gray, MD, MPH

University of Iowa

Iowa City, IA

[References are available from the author.]