Seroincidence of non-typhoid Salmonella infections: convenience vs. random community-based sampling

TY - JOUR

T1 - Seroincidence of non-typhoid Salmonella infections

T2 - convenience vs. random community-based sampling

AU - Emborg, H-D

AU - Simonsen, J

AU - Jørgensen, C S

AU - Harritshøj, Lene Holm

AU - Krogfelt, K A

AU - Linneberg, A

AU - Mølbak, K

PY - 2016/1/1

Y1 - 2016/1/1

N2 - The incidence of reported infections of non-typhoid Salmonella is affected by biases inherent to passive laboratory surveillance, whereas analysis of blood sera may provide a less biased alternative to estimate the force of Salmonella transmission in humans. We developed a mathematical model that enabled a back-calculation of the annual seroincidence of Salmonella based on measurements of specific antibodies. The aim of the present study was to determine the seroincidence in two convenience samples from 2012 (Danish blood donors, n = 500, and pregnant women, n = 637) and a community-based sample of healthy individuals from 2006 to 2007 (n = 1780). The lowest antibody levels were measured in the samples from the community cohort and the highest in pregnant women. The annual Salmonella seroincidences were 319 infections/1000 pregnant women [90% credibility interval (CrI) 210-441], 182/1000 in blood donors (90% CrI 85-298) and 77/1000 in the community cohort (90% CrI 45-114). Although the differences between study populations decreased when accounting for different age distributions the estimates depend on the study population. It is important to be aware of this issue and define a certain population under surveillance in order to obtain consistent results in an application of serological measures for public health purposes.

AB - The incidence of reported infections of non-typhoid Salmonella is affected by biases inherent to passive laboratory surveillance, whereas analysis of blood sera may provide a less biased alternative to estimate the force of Salmonella transmission in humans. We developed a mathematical model that enabled a back-calculation of the annual seroincidence of Salmonella based on measurements of specific antibodies. The aim of the present study was to determine the seroincidence in two convenience samples from 2012 (Danish blood donors, n = 500, and pregnant women, n = 637) and a community-based sample of healthy individuals from 2006 to 2007 (n = 1780). The lowest antibody levels were measured in the samples from the community cohort and the highest in pregnant women. The annual Salmonella seroincidences were 319 infections/1000 pregnant women [90% credibility interval (CrI) 210-441], 182/1000 in blood donors (90% CrI 85-298) and 77/1000 in the community cohort (90% CrI 45-114). Although the differences between study populations decreased when accounting for different age distributions the estimates depend on the study population. It is important to be aware of this issue and define a certain population under surveillance in order to obtain consistent results in an application of serological measures for public health purposes.

KW - Adolescent

KW - Adult

KW - Aged

KW - Cohort Studies

KW - Denmark

KW - Epidemiologic Methods

KW - Female

KW - Humans

KW - Incidence

KW - Male

KW - Middle Aged

KW - Salmonella

KW - Salmonella Infections

KW - Seroepidemiologic Studies

KW - Serologic Tests

KW - Young Adult

KW - Journal Article

KW - Research Support, Non-U.S. Gov't

U2 - 10.1017/s0950268815001417

DO - 10.1017/s0950268815001417

M3 - Journal article

C2 - 26119415

SN - 0950-2688

VL - 144

SP - 257

EP - 264

JO - Epidemiology and Infection

JF - Epidemiology and Infection

IS - 2

ER -