This analysis included pregnant women from 2 hospitals in Thailand participating in the Pregnancy and Influenza Multinational Epidemiologic (PRIME) study, a cohort study conducted in 2017 and 2018 across India, Peru, and Thailand to estimate antenatal influenza illness incidence and its association with adverse perinatal outcomes [6]. In Thailand, the influenza season coincides with the circulation period of RSV and hMPV, providing an opportunity to study these viruses among pregnant women in the cohort [7, 8].

Pregnant women were enrolled from 10 weeks before the anticipated start of the influenza season through the first 4 weeks of the season. Eligibility criteria included being ≥18 years old, having an estimated delivery date that allowed for ≥8 weeks of observation during the influenza season, planning to deliver at study hospitals, and consenting to biweekly symptom surveillance throughout pregnancy [6]. At enrollment, participants who did not yet have an ultrasound received one for gestational age dating. All participants completed in-person structured interviews to gather demographic and baseline clinical data.

Study staff contacted participants twice weekly throughout their pregnancies to monitor for ARI, defined as the sudden onset or worsening of ≥1 symptom within the past 7 days (myalgia, cough, runny nose/congestion, sore throat, or difficulty breathing). Participants were encouraged to contact the study staff if they experienced ARI episodes between scheduled contacts. Symptomatic participants were asked to self-collect midturbinate nasal specimens for influenza virus testing. In Thailand, these specimens were also tested for RSV and hMPV. The performance of self-collection for the detection of respiratory viruses was validated [9]. Participants resumed symptom surveillance when illness resolved, defined as 48 hours without symptoms or ≥13 days after illness onset, whichever occurred first. Any new or worsening symptoms after resolution or ≥13 days following a previous illness were considered a new episode.

Neonatal birth weights were recorded by study-approved scales, as close to delivery as possible but no later than 48 hours afterward. Participants completed in-person interviews about pregnancy complications, prenatal care, and delivery/perinatal outcomes within 7 days of delivery. Medical records were reviewed to confirm prenatal care, pregnancy course, and delivery/perinatal outcomes.

Midturbinate nasal specimens were tested for RNase P to assess quality and for RSV and hMPV by real-time reverse transcription polymerase chain reaction per the protocol of the US Centers for Disease Control and Prevention. Participants with respiratory symptoms meeting the study case definition and testing positive for RSV or hMPV were classified as having RSV or hMPV illness, respectively.

Characteristics and illnesses of fully enrolled study participants (those with ≥1 surveillance contact) were summarized as frequencies or medians. Incidences were calculated from enrollment to the end of pregnancy, focusing on episodes of antenatal RSV and hMPV illnesses.

The impact of antenatal RSV illness was assessed for preterm birth and SGA infants. Only fully enrolled participants with singleton pregnancies and live births were included. The effect of hMPV illness was not assessed due to a small sample size. Preterm birth was defined as a live birth at <37 weeks of gestation, determined by the earliest available ultrasound. The INTERGROWTH-21 Project Standards were used to assess SGA status [10]. Participants with influenza or hMPV illness were excluded from the analyses. Additionally, participants were excluded if they had RSV and hMPV coinfection (n = 1) or RSV followed by hMPV illness (n = 1).

Cox proportional hazards regression models were used to estimate hazard ratios for the 2 outcomes, comparing participants with antenatal RSV illness against those without RSV illness (reference group). The time scale began at enrollment and ended at 36 weeks’ gestation for preterm birth analysis or at delivery for the SGA analysis. The exposure of interest was antenatal RSV illness, with participants considered to have had RSV exposure starting from the date of a positive midturbinate nasal swab result through the end of pregnancy. Associated 95% CIs were obtained by robust SEs. Data were analyzed in Stata version 16 (StataCorp). Two-tailed P values <.05 were considered statistically significant.

The study was approved by the institutional review boards (IRBs) of Abt Associates, Walter Reed Army Institute of Research, and the local ethical review committees. The US Centers for Disease Control and Prevention’s IRB relied on the determination of Abt Associates’ IRB. Written informed consent was obtained from all study participants.

Of 2764 fully enrolled participants (Supplementary Figure 1), 1230 were enrolled in 2017 and 1534 in 2018 (Supplementary Table 1). At enrollment, 75% were aged 20 to 35 years, and 24%, 62%, and 13% were in the first, second, and third trimesters, respectively, while 1% had missing trimester data. Overall, 11% self-reported chronic medical conditions, and 8% had preterm deliveries while 7% delivered SGA infants.

During follow-up, 3% had ≥1 episode of RSV illness, and 1% had hMPV illness. Among them, 1 participant had repeat RSV illness, 1 had RSV and hMPV during the same episode, and 1 had RSV followed by hMPV. A total of 71 RSV cases and 29 hMPV cases occurred (Supplementary Figure 2). Among RSV cases, 8% patients reported fever, 27% could not complete daily activities, 42% sought medical care, and 1% were hospitalized (Table 1). The median RSV illness duration was 11 days. For hMPV cases, 21% had fever, 34% could not complete daily activities, 34% sought medical care, and none were hospitalized. The median hMPV illness duration was 10 days.

Participants contributed 12 468 person-months of observation. RSV incidences were 47 per 10 000 person-months in 2017 (95% CI, 31–70) and 64 in 2018 (95% CI, 47–86; Table 2), with no significant difference between years (P = .24). Trimester-specific RSV incidences were 24.3 (first), 53.2 (second), and 61.8 (third) per 10 000 person-months, with no significant differences across trimesters (P > .05). For hMPV, the incidences were 40 per 10 000 person-months in 2017 (95% CI, 25–61) and 10 in 2018 (95% CI, 4–21), significantly higher in 2017 (P < .01). Trimester-specific hMPV incidences were 24.3 (first), 20.9 (second), and 25.0 (third) per 10 000 person-months, with no significant differences across trimesters (P > .05).

In the RSV–preterm birth analysis, 192 (9%) participants had preterm deliveries, and 41 (2%) had RSV illnesses. Adjusting for maternal age, chronic heart conditions, gestational hypertension, and history of SGA infants, RSV illnesses in the third trimester were associated with preterm birth (adjusted hazard ratio [aHR], 2.50; 95% CI, 1.04–6.00; P = .04). No significant effect was observed in the second trimester (aHR, 2.01; 95% CI, .70–5.72; P = .193), and first trimester data were too limited for analysis. In the RSV-SGA analysis, 193 (9%) participants delivered SGA infants, and 51 (2%) had RSV illnesses. No significant difference in the likelihood of delivering an SGA infant was found between those with and without RSV illnesses (aHR, 0.79; 95% CI, .29–2.16; P = .64).