Hepatitis E Seroprevalence and Seroconversion. Part 3
Laboratory samples and methods.Aliquots of predeployment and postdeployment frozen serum from the 1500 subjects (3000 samples in total) were shipped on dry ice to the Armed Forces Research Institute of Medical Sciences for anti‐HEV testing. Samples were tested for total anti‐HEV by a noncommercial enzyme immunoassay developed at Walter Reed Army Institute of Research (WRAIR; Silver Spring, MD). To determine whether reactive samples were due to a recent infection (2–3 months prior to sample collection), samples were also tested for immunoglobulin M (IgM) anti‐HEV. Paired predeployment and postdeployment samples were tested simultaneously in duplicate on the same plate. Results were quantified in WRAIR units per milliliter (WRAIR U/mL) by comparing the test sample with the reference serum pool. The reference serum pool was a pool of samples collected from Nepalese donors who were known to have been infected with HEV. A total anti‐HEV level of 20 WRAIR U/mL was considered to be evidence of past infection unless the sample was also reactive for IgM anti‐HEV. Seroconversion was defined as a 4‐fold increase in the anti‐HEV level to 20 WRAIR U/mL. An IgM anti‐HEV level of >100 WRAIR U/mL was considered to be evidence of acute infection.
Statistical analysis.Seroprevalence and 95% confidence intervals (CIs) of total anti‐HEV levels before deployment were calculated. Anti‐HEV seroprevalence was also calculated for specific demographic, location, and service‐related strata. For intrastratum comparisons, because of the small numbers of samples, the Fisher exact test was used to generate P values and exact logistic regression was used to calculate odds ratios. The proportion and 95% CI of service members who seroconverted during deployment was also calculated on the basis of the seroconversion definitions given above. All 95% CIs were calculated using an exact binomial formula. SAS software (version 9.1; SAS Institute) was used for this analysis.
Scientific and human subjects research approvals.This study received scientific review from the WRAIR Scientific Review Committee and was additionally reviewed and approved by the US Army Medical Research and Materiel Command’s Human Subjects Research Review Board.
Study population. The characteristics of the random sample of study subjects with serum samples available were similar to those of the total deployed cohort. A large percentage of the study subjects and a large percentage of the total deployed cohort had missing birth countries (21.2% and 23.4%, respectively). Birth country is reported at the Military Entrance Processing Station and is often missing in the DMSS data.