The Burden of Arboviral Infections in the Military Health System 2012–2019

Abstract

Arboviral infections, including dengue (DNV), chikungunya (CHIKV), and Zika (ZIKV), impact both travelers and native populations of endemic regions. We sought to assess the disease burden of arboviral infections in the Military Health System, the validity of arboviral diagnostic codes, and the role of pretravel counseling on insect avoidance precautions. We searched for diagnostic codes consistent with arboviral infection and grouped them into DNV, CHIKV, ZIKV, Japanese encephalitis virus (JEV), and Other. Demographic data were evaluated. A subset of charts in each category were reviewed for diagnostic validity and travel characteristics. In all, 10,547 unique subjects carried 17,135 arboviral diagnostic codes, including 1,606 subjects (15.2%) coded for DNV, 230 (2.2%) for ZIKV, 65 (0.6%) for CHIKV, and 4,317 (40.9%) for JEV. A chart review was performed on 807 outpatient charts, yielding outpatient diagnostic code positive predictive values of 60.5% (DNV), 15.3% (ZIKV), and 64.5% (CHIKV); there were no valid cases of JEV. Dengue represented the greatest burden of arboviral infections with 2.2 cases per 100,000 military healthcare enrollees over the 2012–2019 fiscal years. More than 80% of subjects with arboviral infection did not have documented pretravel counseling. Arboviral infections represent a significant disease burden in young travelers to endemic regions. After adjustment for diagnostic validity, DNV represented the greatest burden. Diagnostic codes for ZIKV and JEV overestimate the burden of these diseases. Low rates of pretravel visits represent an opportunity for increased emphasis on insect exposure precautions.

The risk and risk factors of chikungunya virus infection and rheumatological sequelae in a cohort of U.S. Military Health System beneficiaries: implications for the vaccine era

“…We found a very low number of CHIKV cases identified in the MHS, despite this study period covering the peak CHIKV epidemic in the Americas and leveraging data from about 9.5 million beneficiaries. Our findings correlate with low number of CHIKV infections in other U.S. MHS studies, although such studies were either smaller, earlier in the CHIKV epidemic and/or less comprehensive in their use of electronic medical record (eMR) data sources [28, 31]. Additionally, our findings correlate with a recent post-deployment serology study of 1500 deployments to South America and Puerto Rico with a low post-deployment seropositivity rate of 1.7% (highest in those serving in Puerto Rico) [32].…”

“…The use of the MHS includes a substantive number of beneficiaries who deploy or are stationed in CHIKV endemic regions and allowed for long term follow-up of cases to determine longer term complications. We derived CHIKV cases from a range of eMR data sources, including clinical microbiology laboratory results, to identify more infections than prior MHSbased studies [28][29][30][31]. Our study limitations included the use of ICD codes to diagnose CHIKV and post-CHIKV complications; prior evaluations of CHIKV ICD diagnosis coding have noted imperfect specificity [31] and this study was unable to perform chart adjudications.…”

“…We derived CHIKV cases from a range of eMR data sources, including clinical microbiology laboratory results, to identify more infections than prior MHSbased studies [28][29][30][31]. Our study limitations included the use of ICD codes to diagnose CHIKV and post-CHIKV complications; prior evaluations of CHIKV ICD diagnosis coding have noted imperfect specificity [31] and this study was unable to perform chart adjudications. Conversely, CHIKV diagnoses may have been missed in some clinical presentations, and not all cases may have presented to care (including during deployments).…”