Respiratory Viral Infections and Infection Prevention Practices Among Women With Acute Respiratory Illness During Delivery Hospitalizations During the 2019–2020 Influenza Season

Abstract

Background.

We conducted a cross-sectional study of pregnant women with acute respiratory illness during delivery hospitalizations during influenza season to describe clinical testing for respiratory viruses and infection prevention practices.

Methods.

Women had nasal swabs tested for influenza and other respiratory viruses. Among 91 enrolled women, 22 (24%) had clinical testing for influenza.

Results.

Based on clinical and study testing combined, 41 of 91 (45%) women had samples positive for respiratory viruses. The most common virus was influenza (17 of 91, 19%); 53% (9 of 17) of influenza virus infections were identified through study testing alone. Only 16% of women were on droplet precautions.

Conclusions.

Peripartum respiratory infections may be underrecognized.

Childbirth is the most common reason for hospitalizations in the United States and accounts for more than 4 million hospital stays per year [1]. Approximately 1 million delivery hospitalizations occur during the influenza season annually, assuming an influenza season that spans an average of 16 weeks. Influenza and other respiratory illnesses during delivery hospitalizations have unique implications for the health of the mother and for potential transmission to her newborn and others in the hospital. Newborns and young infants that become infected with influenza may be at increased risk for severe complications and death [2]. However, little is known about the burden of respiratory illness among delivery hospitalizations during influenza seasons in the United States or associated influenza diagnostic testing, treatment, and infection prevention practices. Furthermore, peripartum viral respiratory illness during delivery hospitalizations is likely underrecognized because of low clinician-ordered diagnostic testing [3].

Standard infection prevention and control (IPC) for influenza in the healthcare setting includes droplet precautions for patients with suspected and confirmed influenza and may also include placing affected patients in private rooms or cohorting [4]. The Centers for Disease Control and Prevention (CDC) also provides IPC recommendations specifically for the peripartum setting [5]. During the 2019–2020 influenza season, we conducted a cross-sectional study of pregnant women with acute respiratory illness (ARI) during their delivery hospitalizations to describe (1) the clinical testing practices for influenza and other respiratory viruses, (2) the frequency of respiratory virus infections, and (3) influenza antiviral treatment and IPC practices.

METHODS

Setting

This study was conducted at 6 academic medical centers across the United States. Institutional enrollment intervals averaged 10 weeks (range 7–13 weeks). See Supplemental Methods for study sites and a description of influenza virus circulation in the United States during the study season.

Enrollment and Data Collection Procedures

Women were eligible if they were (1) 18–50 years of age, (2) at least 34 weeks gestation at admission, (3) able to speak and read English (or Spanish at some sites), (4) hospitalized for active labor or having delivered a liveborn infant, and (5) had a new cough or worsening of a chronic cough with a new respiratory illness that began within 10 days before enrollment. Women also had to have a nasal swab tested for influenza viruses using a molecular assay during the current hospitalization or be willing to provide a nasal swab for research testing.

Study participants completed a standardized enrollment interview that collected information about demographic characteristics, past medical and prenatal history, respiratory illness symptoms, influenza antiviral treatment before the hospital stay, and receipt of the 2019–2020 influenza vaccine after July 1, 2019. Participants completed a second standardized interview at hospital discharge or within 5 days after hospital discharge. The interview collected information about participant and other family member interactions with the newborn during the hospital stay and the participant’s perceptions of whether healthcare personnel observed standard and droplet precautions when inside the participant’s room. Study staff also collected information from the medical record about parity, insurance status, clinical respiratory viral testing and results, delivery outcomes (live birth, stillbirth, and preterm birth), droplet precaution orders, documentation of temporary separation of mothers and newborns based on maternal respiratory illness, and influenza antiviral treatment.

Respiratory Specimen Collection and Laboratory Methods

Participants who did not have clinical testing for influenza viruses by molecular assays during the delivery hospitalization had a mid-turbinate nasal swab collected by study staff. At some sites, participants who had clinical testing also had a mid-turbinate nasal swab collected as part of the study. Respiratory samples were tested at the Respiratory Virus Diagnostic Laboratory at the Baylor College of Medicine by reverse-transcription polymerase chain reaction (RT-PCR) singleplex assays for influenza A(H3N2), A(H1N1)pdm09, B/Yamagata, B/Victoria, and C viruses, adenovirus, parainfluenza viruses 1–4, rhinovirus, human metapneumovirus, respiratory syncytial viruses (RSVs) A and B, coronaviruses-NL63, HK-U1, 229E, and OC43, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [6–8].

Data Analysis

Frequencies were calculated to characterize participant enrollment characteristics and outcomes of interest. Clinical testing was defined as testing by RT-PCR or rapid antigen test for influenza or RSV and testing by RT-PCR for the other respiratory viruses. Ninety-five percent confidence intervals (CIs) were calculated for survey responses related to participants’ perceptions of IPC practices using standard formulas and assuming a binomial distribution.

To examine the association between clinical testing for influenza and IPC and antiviral treatment practices, the frequency of each practice was compared between participants who had clinical testing and those who did not using χ² or Fisher’s exact test. A P value of <.05 was considered statistically significant. Analyses were conducted in SAS version 9.1.

Ethical Review

The study protocol was reviewed and approved by ethical review committees at participating study sites and the Abt Associates Institutional Review Board (IRB). The US CDC relied on the review of the Abt Associates IRB. Written informed consent was obtained from all study participants.

RESULTS

Enrollment and Baseline Characteristics

During the enrollment period, 6327 women were admitted to Labor and Delivery units, 5176 (82%) of whom were eligible for initial approach for study recruitment. Of these 5176 women, 3815 (74%) were prescreened for ARI and/or completed formal eligibility screening for the study; 119 (3%) were found eligible, 92 (77%) of whom consented to enrollment. One woman who consented to enrollment did not deliver a newborn during her hospitalization and was excluded from analyses.

Among the 91 enrolled women with ARI who delivered newborns during their hospital stay, the median age was 30 years, 58 (64%) women were white, and 59 (65%) women had private insurance (Table 1). Most women had vaginal deliveries (75%). All pregnancies for enrolled women resulted in live births, including 4 (4%) preterm births at 34–36 weeks gestation and 87 (96%) term births. Seventy-six percent of women reported receiving the 2019–2020 influenza vaccine before admission. The most commonly reported respiratory illness symptoms among women were runny nose (86%), sore throat (74%), and difficulty breathing or shortness of breath (45%). Twenty-seven women (30%) had measured or subjective fever. None of the women required intensive care and the median length of stay was 3 days (interquartile range, 3–4 days).

Table 1.

Baseline Characteristics of Pregnant Women With Acute Respiratory Illness During Their Peripartum Hospital Stay, N = 91

Characteristics	N	(Col %)
Age
14–19 years	6	7%
20–34 years	60	66%
35–50 years	25	27%
Race
White	58	64%
Black, African American	17	19%
Other	16	18%
Ethnicity
Not Hispanic/Not Latina	65	71%
Hispanic/Latina	26	29%
Insurance Status
Private	59	65%
Public	28	31%
Self-pay/uninsured	4	4%
Paritya
1	29	32%
>1	62	68%
Delivery Mode
Vaginal	68	75%
Cesarean section, planned	9	10%
Cesarean section, unplanned	14	15%
Live Birth	91	100%
Preterm (34–36 weeks gestation)	4	4%
Term (≥37 weeks gestation)	87	96%
Influenza vaccination in 2019–2020 seasonb	69	76%
Illness Symptoms
Runny nose	78	86%
Sore throat	67	74%
Difficulty breathing/shortness of breath	41	45%
Muscle aches	31	34%
Fever of feverishness	27	30%
Vomiting	13	14%
Diarrhea	15	16%
Tested for Influenzac
Clinical testing	22	24%
Study testing	78	86%
Clinical testing and study testing	9	10%
Tested for RSVc
Clinical testing	12	13%
Study testing	78	86%
Clinical testing and study testing	6	7%

Abbreviations: Col, column; RSV, respiratory syncytial virus.

^aIncludes current pregnancy.

^bVaccination status based on participant self-report.

^cTesting type categories are not mutually exclusive.

Respiratory Virus Testing and Detection

Among the 91 enrolled women, 22 women had clinical testing for influenza (4 by rapid test alone, 16 by RT-PCR, and 2 by both rapid test and RT-PCR) and 12 had clinical testing for RSV (4 by rapid test and 8 by RT-PCR). Sixteen women had clinical testing for various combinations of the other respiratory viruses. Seventy-eight women had study samples taken.

Based on clinical testing and study testing combined, 43 (47%) women had respiratory samples positive for respiratory viruses. The most commonly detected viruses were influenza viruses (17 of 91, 19%), rhinoviruses (11 of 91, 12%), human metapneumoviruses (5 of 91, 5%), and endemic coronaviruses (5 of 91, 5%); there were no codetections of more than 1 respiratory virus. Among the 78 study samples tested, none were positive for SARS-CoV-2 (Supplemental Figure 1, Supplemental Materials). Among the 17 women with influenza, 8 were positive for influenza A(H1N1)pdm09, 5 for influenza B/Victoria, 2 for influenza A(H3N2), and 1 each for influenza A and influenza B not further subtyped or lineage typed. More than half (9 of 17, 53%) of women with influenza did not have clinical testing and had infections that were detected by study testing only; of these 9 women, 7 had measured or subjective fever in addition to cough.

Peripartum Infection Prevention and Control and Influenza Antiviral Treatment Practices

Of these 85 women who completed a discharge survey, 80 (94%) reported that their newborns stayed with them in their hospital room for all or some of the hospital stay. Among 90 women who had chart abstractions completed, 14 (16%) had documentation indicating that they were placed on droplet precautions, 3 (3%) had newborns placed on droplet precautions, and only 1 (1%) had temporary separation of mother and newborn due to maternal laboratory-confirmed influenza (Table 2). Eleven women (12%) received influenza antiviral treatment during their hospitalization. Comparing women who had clinical testing for influenza to those who did not, women who had clinical testing were more likely to be placed on droplet precautions (62% vs 1%, P < .01), have their newborn placed on droplet precautions (10% vs 1%, P < .04), and receive influenza antiviral treatment (41% vs 3%, P < .01).

Table 2.

Delivery Hospitalization Infection Prevention and Control and Antiviral Treatment Practices, N = 90

Infection Prevention and Control Practices or Influenza Antiviral Treatment Practices	All		Clinical Testing for Influenza		No Clinical Testing for Influenza		P Valuea
	(N = 90)		(n = 21)		(n = 69)
	n	(Col %)	n	(Col %)	n	(Col %)
Infection Prevention and Control Practices
Mother placed on droplet precautions	14/90	16%	13/21	62%	1	1%	<.01
Newborn placed on droplet precautions	3/90	3%	2/21	10%	1	1%	.04
Newborn stayed in mother’s room for all or part of stay	80/85	94%	17/20	85%	63/65	97%	.05
Temporary separation of mother from baby	1/90	1%	1/21	5%	0	0%	.06
Influenza Antiviral Treatment Practices
Mother received antiviral treatment before admission	4/90	4%	2/21	10%	2/68	3%	.25
Mother received antiviral treatment during admission	11/90	12%	9/21	43%	2/69	3%	<.0001

Abbreviations: Col, column.

^aχ2 or Fisher’s exact test.

Of the 80 women who completed discharge surveys and had newborns stay with them in their hospital rooms, 41 (51%, 95% CI 40–62%) perceived that their medical team advised them to wash their hands before feeding their newborns, 22 (28%; 95% CI, 18%–37%) reported that they were counseled about the risk of spreading respiratory illness to their newborns, and 12 (15%; 95% CI, 7%–23%) reported that they were advised to wear a facemask before feeding their newborns. Twelve women (15%; 95% CI, 7%–23%) had another caregiver present to care for their newborn while they had respiratory illness symptoms. Only 2 (3%; 95% CI, 0%–6%) women reported that a barrier was used to separate the woman from her newborn while they were in the same room. Overall, 74 (93%; 95% CI, 87%–98%) women breastfed their newborns directly at the breast.

DISCUSSION

In this multicenter study of women with peripartum ARI who are admitted for delivery during the influenza season, almost half of women had respiratory viral infections and 1 in 5 had laboratory-confirmed influenza. Half of the women with influenza were detected only by systematic study testing and would have otherwise been unknown because clinical testing was not done. Droplet precautions and temporary separation of mothers from their newborns because of maternal ARI were rarely implemented and were more common among women who had clinical testing for influenza. Empiric influenza antiviral treatment was also infrequent. Finally, based on maternal report, implementation rates were low for other CDC guidelines to reduce the risk of influenza virus transmission to newborns such as being advised by a medical team to wear a facemask during newborn feeding and using barrier devices to separate women from their newborns while in the same room.

Our findings suggest that ARI among pregnant women admitted for delivery during influenza season is underrecognized even though many of these illnesses are associated with respiratory viral infections that may pose a risk for nosocomial transmission to newborns, other patients, and healthcare personnel. A recent study that included >12 000 delivery hospitalizations among women with acute respiratory illness during influenza season in 4 high-income countries found similar low rates of influenza testing with only 2% of women tested [3]. Although testing for noninfluenza respiratory viruses may not be warranted for most women with uncomplicated peripartum ARI, standard screening for ARI symptoms at hospital admission might improve recognition of women with ARI. Improved recognition of ARI might in turn prompt more consistent implementation of droplet precautions and consideration of empiric influenza antiviral treatment during influenza season.

Strengths of this study include enrollment at 6 medical centers across the United States and systematic testing for influenza among all participants and for other respiratory viruses among more than three quarters of participants. However, several limitations should be considered when interpreting study findings. First, overall enrollment in this study was low, which precluded assessment of differences in virus detection and antiviral treatment and IPC practices between study sites. Second, this study was conducted during a single influenza season in which enrollment was curtailed by the onset of the coronavirus disease 2019 pandemic. Thus, the prevalence of respiratory virus infections may not be representative of all influenza seasons, although overall respiratory virus detection rates in this study are consistent with findings from published studies in pregnant women [6] and outpatient adults [9, 10]. Third, selected measures of adherence to CDC guidelines for the prevention and control of influenza during peripartum hospitalizations were based on participant report and may have been subject to errors in recall.

CONCLUSIONS

In conclusion, in this multicenter study of women with ARI during delivery hospitalizations in the 2019–2020 influenza season, we found that approximately 20% had laboratory-confirmed influenza, although half were not diagnosed by clinical testing. Improved recognition of peripartum ARI through standard screening upon admission might prompt more consistent implementation of droplet precautions and consideration of empiric influenza antiviral treatment during influenza season.