Santa Casa de São Paulo is a tertiary general hospital in downtown São Paulo, with approximately 2000 births annually. The Department of Obstetrics is a referral delivery service for municipal high-risk pregnancy programs, and approximately 15% of newborns require intensive care. The NICU has 28 beds (14 in the intermediate care unit and 14 in the intensive care unit), which are divided among four physically separated wards (Fig. 1). Infants in the main NICU room are kept in either closed incubators or open radiant warmers, and many of them require mechanical ventilation (MV). For those in incubators, most of the routine care is provided through side doors, which are kept closed when not in use. Those in the intermediate care unit stay in open bassinets or closed incubators and do not require MV.

Figure 1.

Schematic and simplified neonatal intensive care unit floor plan and spatial distribution of the cohorts, as part of the control measures used during the outbreaks.

(0.07MB).

In average, more than 65 different healthcare professionals have physical contact with the babies in the NICU daily: six neonatal hospitalists, eight neonatal fellows, four pediatric residents, four medical school graduation students, six nurse practitioners, 34 nurse technicians, and five respiratory therapists. After hand cleaning and replacement of personal clothing by aprons (provided by the hospital), parents are allowed to visit all wards and to participate in the care of their babies. Children under 18 years of age are not allowed to enter the NICU.

Upper respiratory tract infection was defined when cough and rhinorrhea were present. Bronchiolitis diagnosis was made based on the following symptoms: tachypnea, wheezing, cough, crepitus, and/or nasal flaring.

After the identification of a symptomatic infant with confirmed RSV infection, nasopharyngeal aspirate (NPA) samples were collected from all infants hospitalized in the NICU. NPA surveillance was repeated twice a week in all patients. The samples were sent to the Institute of Biomedical Science of the University of São Paulo.

Viral RNA was extracted with the automatic NucliSens easyMAG extractor (Biomerieux Inc., Durham, USA), according to the manufacturer's instructions. Real-time reverse transcription (RRT)-polymerase chain reaction (PCR) was performed with the AgPath-ID one-step RRT-PCR kit and the 7300 Real-Time PCR System (Applied Biosystems, Foster City, USA). To identify the pathogen responsible for the outbreak, the RRT-PCR was performed with a non-influenza virus RRT-PCR Panel Kit to detect 15 respiratory viruses – RSV group A and group B, human metapneumovirus, adenovirus, human coronaviruses (HCoV-HKU, -NL63, -OC43, and 229E), human parainfluenza virus (types 1, 2, 3, and 4), human bocavirus, human enterovirus, and human rhinovirus, which were provided by the Centers for Disease Control and Prevention (CDC – Atlanta, GA, USA),11 and commercial primers and probes for influenza A and B (Invitrogen, Carlsbad, USA).

Positive qPCR samples were amplified by conventional two-step PCR for DNA sequencing: cDNA was synthesized using the SuperScript III kit (Applied Biosystems, Foster City, USA) according to the manufacturer's instructions. Analyses of the second hypervariable region of the G protein gene was carried out using PCR with the primers Gr5_fw12 and FVAB_rev13 in a mixture containing 5μL of cDNA, 5μL of 10× PCR buffer, 25mM of each dNTP, 10pmol of each primer, 1.0U of Platinum Taq DNA Polymerase (Invitrogen, Carlsbad, USA), and nuclease-free water for a final volume of 25μL. The second step of semi-nested PCR was performed for negative samples from the first PCR, using the forward primer GR5_fwd and reverse primer F1AB_rev14 in the same conditions as used for conventional PCR. The amplification was performed using the GeneAmp PCR System 9700 thermocycler (Applied Biosystems, Foster City, USA) using the following program: 95°C for 5min, followed by 35 cycles, each consisting of 30s at 95°C, 30s at 55°C, and 45s at 72°C, and finally 5min of extension at 72°C. The amplified products were analyzed using agarose gel electrophoresis and visualized under ultraviolet light after staining with ethidium bromide. The amplified products of gene G of ∼1052 and 900bp, produced using conventional PCR and semi-nested PCR, consecutively, were purified using ExoSap-IT (Affymetrix Inc., Santa Clara, USA) and submitted to a cycle sequencing reaction (Sanger) using a fluorescent dye terminator kit with all primers, GR5, FV, and F1AB, in the 3100 DNA Sequencer (Applied Biosystems, Foster City, USA). Both strands of each amplification product were sequenced at least twice. Sequence editing, alignments, and phylogenetic analyses were performed with the MegAlign 5.03v software (DNAStar Inc., Madison, USA). Standard published sequences from subgroup A were downloaded from GenBank (https://www.ncbi.nlm.nih.gov/genbank) as a reference of different lineages and genotypes.

The study was approved by the local research ethics committee.

The index case involved a 30-day-old preterm boy (33 weeks of gestational age) with a birth weight of 2290g who had been hospitalized since birth because of prematurity and bronchopulmonary dysplasia. He was being weaned off oxygen supplementation in the intermediate care unit when cough, runny nose, and fever started. Respiratory failure developed rapidly, requiring MV. Five days later, five infants (out of 17 who were in the NICU) developed respiratory symptoms; therefore, an outbreak was declared (Fig. 2). On the same day, NPA samples were collected from all patients in the NICU for RRT-PCR respiratory virus detection. Ten infants were infected with RSV, of whom only three shared the ward with the index case. The attack rate was 58.8%, and only one fatality occurred (CFR=10%). However, the death was not attributed to RSV infection, since the patient did not present with respiratory symptoms and had congenital pulmonary hypoplasia, which was considered the cause of death. Additional information is provided in Table 1.

Figure 2.

Outbreaks timeline.

(0.16MB).

Contact and respiratory precautions were immediately implemented. All infants in the NICU were separated into three different cohorts. RSV-positive infants were placed in ward 1, under contact and respiratory precautions. RSV-negative infants who had previous contact with any of the RSV-positive infants were placed in wards 2 and 3, under contact precautions only. RSV-negative infants who had never had contact with RSV-positive patients were placed in ward 4.

Healthcare staff was also divided across the wards, and their internal rotating schedules were suspended. Fifty of them had NPA and saliva specimens tested for respiratory virus, which were negative. NICU visitation was restricted and its duration was reduced. Infection control measures were reinforced for healthcare staff and infants’ families by the infection control group and NICU members, and they included hand hygiene, environmental cleaning, and increased awareness of respiratory symptoms. On the second day of the outbreak, palivizumab was administered to all infants in the NICU.

NPA surveillance (testing for the 17 respiratory viruses as previously described) was performed twice a week in all infants in the NICU so that viral shedding duration and possible co-infections could be determined. Follow-up continued for four weeks after the last positive sample was identified. RSV-positive infants were kept isolated until NICU discharge and were tested weekly using PCR until a negative result was achieved. The longest viral shedding period was 29 days.

During the following 24 days, no other infants in the NICU developed symptoms suggestive of RSV infection; therefore, the first outbreak was considered controlled. Thirty-four days after the index case, a new outbreak was identified, and control measures were reinstated (Fig. 2). Fourteen infants had RSV identified in respiratory secretions (Table 1). Palivizumab was administered only to RSV-PCR-negative infants with risk factors (eight patients). In this second outbreak, the attack rate was 63.6%, and two deaths occurred (CFR=14.3%). Both infants were premature and were receiving MV when RSV infection was detected. Their deaths were attributed to bacterial sepsis; however, RSV could have contributed. The longest viral shedding time was 14 days.

Among the 24 RSV-positive infants, 15 (62.5%) were born prematurely at less than 37 weeks of gestation (Table 1). The mean gestational age was 33 weeks (range: 28–40 weeks), mean age at diagnosis was 27.5 days (range: 1–80 days), and mean birth weight was 1961g (range: 765–3440g). A total of 11 infants (45.8%) were asymptomatic. Among the symptomatic infants, 12.6% presented with mild upper respiratory symptoms (cough or coryza) or fever, 20.8% presented with bronchiolitis without the need for MV, and an additional 20.8% required invasive ventilatory support. Six infants would be eligible by the Brazilian guidelines to receive palivizumab.15 In both outbreaks, the available data were not sufficient to establish independent risk factors for RSV infection.

Molecular analysis was performed for all 24 RSV-positive samples (Fig. 3). Viruses isolated during the first outbreak were characterized as RSV-A type ON-1, while all infections during the second outbreak were caused by RSV-A type NA-2. Aminoacid analysis evidenced a characteristic genetic signature in the second outbreak samples. No viral coinfection was detected in either outbreak.

Figure 3.

Phylogenetic tree of isolated virus during the outbreaks.

(0.42MB).