Preschool-Aged Household Contacts as a Risk Factor for Viral Respiratory Infections in Healthcare Personnel

Abstract

Background: Viral respiratory infections (VRIs) are common and are occupational risks for healthcare personnel (HCP). VRIs can also be acquired at home and other settings among HCPs. We sought to determine if preschool-aged household contacts are a risk factor for VRIs among HCPs working in outpatient settings.

Methods: We conducted a secondary analysis of data from a cluster randomized trial at 7 medical centers in the United States over 4 influenza seasons from 2011-2012 to 2014-2015. Adult HCPs who routinely came within 6 feet of patients with respiratory infections were included. Participants were tested for respiratory viruses whenever symptomatic and at 2 random times each season when asymptomatic. The exposure of interest was the number of household contacts 0-5 years old (preschool-aged) at the beginning of each HCP-season. The primary outcome was the rate of polymerase chain reaction-detected VRIs, regardless of symptoms. The VRI incidence rate ratio (IRR) was calculated using a mixed-effects Poisson regression model that accounted for clustering at the clinic level.

Results: Among the 4476 HCP-seasons, most HCPs were female (85.4%) and between 30 and 49 years of age (54.6%). The overall VRI rate was 2.04 per 100 person-weeks. In the adjusted analysis, HCPs having 1 (IRR, 1.22 [95% confidence interval {CI}, 1.05-1.43]) and ≥2 (IRR, 1.35 [95% CI, 1.09-1.67]) preschool-aged household contacts had higher VRI rates than those with zero preschool-aged household contacts.

Conclusions: Preschool-aged household contacts are a risk factor for developing VRIs among HCPs working in outpatient settings.

Keywords: healthcare personnel; healthcare-associated infections; viral respiratory infections.

Figure 1.

Flowchart showing participation in the Respiratory Protection Effectiveness Clinical Trial (ResPECT) and inclusion in this study over 4 seasons.

Figure 2.

Sensitivity analyses with adjusted rate ratios for polymerase chain reaction (PCR)–detected viral respiratory infections in healthcare personnel (HCP) by number of preschool-aged household contacts (HHCs). For each analysis, “1:0” compares 1 preschool-aged HHC to zero, and “2:0” compares ≥2 preschool-aged HHCs to zero. “Primary” refers to the adjusted primary analysis and is shown for comparison to the sensitivity analyses. “Symptomatic” uses an alternate outcome definition that only included symptomatic individuals, and viral coinfections were counted as 1 event. “ITT cohort” (intention-to-treat) includes all HCP-seasons, but excludes those with missing HHC data (n = 228), missing age data (n = 6), and zero follow-up time (n = 278). “Changing No. HHC” only includes individuals who participated for >1 season and had their number of preschool-aged HHCs change at least once between seasons. “ARI” (acute respiratory illness) uses ARI regardless of PCR testing result as the outcome. Incidence rate ratios were calculated using a mixed-effects Poisson regression model that accounted for clustering at the clinic level.

Figure 3.

Adjusted incidence rate ratios (IRRs) for polymerase chain reaction (PCR)–detected viral respiratory infections with specific viruses. For each analysis, “1:0” compares 1 preschool-aged household contact (HHC) to zero, and “2:0” compares ≥2 preschool-aged HHCs to zero. IRRs were calculated using a mixed-effects Poisson regression model that accounted for clustering at the clinic level and was adjusted for age, sex, and occupational exposure risk. There were too few events for parainfluenza virus to run the model. Adenovirus and human metapneumovirus had very wide confidence intervals and are also excluded. Abbreviations: Flu A, influenza A; Flu B, influenza B; hCOV, endemic human coronaviruses; REV, human rhinovirus or enterovirus; RSV, respiratory syncytial virus.