Influenza B Outbreak on a Cruise Ship off the São Paulo Coast, Brazil

Authors


Abstract

Background

In February 2012, crew and passengers of a cruise ship sailing off the coast of São Paulo, Brazil, were hospitalized for acute respiratory illness (ARI). A field investigation was performed to identify the disease involved and factors associated.

Methods

Information on passengers and crew with ARI was obtained from the medical records of hospitalized individuals. Active case finding was performed onboard the ship. ARI was defined as the presence of one nonspecific symptom (fever, chills, myalgia, arthralgia, headache, or malaise) and one respiratory symptom (cough, nasal congestion, sore throat, or dyspnea). A case–control study was conducted among the crew. The cases were crew members with symptoms of influenza-like illness (ILI) (fever and one of the following symptoms: cough, sore throat, and dyspnea) in February 2012. The controls were asymptomatic crew members.

Results

The study identified 104 ARI cases: 54 (51.9%) crew members and 50 (49.1%) passengers. Among 11 ARI hospitalized cases, 6 had influenza B virus isolated in nasopharyngeal swab. One mortality among these patients was caused by postinfluenza Staphylococcus aureus pneumonia. The crew members housed in the two lower decks and those belonging to the 18- to 32-year-old age group were more likely to develop ILI [odds ratio (OR) = 2.39, 95% confidence interval (CI) 1.09–5.25 and OR = 3.72, CI 1.25–11.16, respectively].

Conclusions

In February 2012, an influenza B outbreak occurred onboard a cruise ship. Among crew members, ILI was associated with lower cabin location and younger age group. This was the first influenza outbreak detected by Brazilian public health authorities in a vessel cruising in South American waters.

The number of cruise ship passengers traveling in Brazilian waters increased more than 20 times in the last 10 years. In the 2010 to 2011 cruise season, approximately 760,000 people traveled on cruises along the Brazilian coast.[1] During a cruise, passengers and crew members from several nations intermingle for extended periods in closed environments.[2, 3] Shipboard activities such as dining, games, and movies increase the chance of contact between passengers and crew.[2] The average cruise lasts 6 days, well within the period of incubation and transmission of many infectious diseases.

On February 16, 2012, the São Paulo State Epidemiological Surveillance Center, Brazil, was notified by local health authorities about a crew member who had disembarked and subsequently been hospitalized for acute respiratory illness (ARI) in Santos, a coastal city of the state of São Paulo (Santos/SP). The ship had sailed along the Brazilian, Uruguayan, and Argentinean coasts. This crew member died on the following day of acute respiratory distress syndrome (ARDS). Additionally, among all passengers and crew members, eight patients with respiratory symptoms were hospitalized from February 16 to 18.

Because of continued ARI activity among passengers and crew members until the cruise ship's return to Santos/SP, a public health team from the Agencia Nacional de Vigilância Sanitária (ANVISA—Brazilian Health Surveillance Agency), São Paulo State Epidemiological Surveillance Center, and City of Santos Health Surveillance Service performed a field investigation. The aims of this investigation were to identify the disease involved and associated factors, in addition to implementing control measures.

Methods

Setting

The vessel had a capacity for 1,554 passengers and nearly 700 crew members. During the summer in the southern hemisphere, the cruise ship sailed along the Brazilian, Uruguayan, and Argentinean coasts. On February 15, 2012, new passengers embarked in Santos, while the crew remained relatively unchanged. The ship docked in Santos again on February 18, 2012 for other passengers to embark. The new cruise lasted 8 days and the ship arrived in Santos on February 27.

Epidemiological Investigation

Information on passengers and crew with respiratory illness was obtained from the medical records and laboratorial results of hospitalized individuals. Data were collected using a standard form to obtain relevant information on demographic characteristics, job duties, respiratory illness symptoms, and health care.

On February 27, 2012, when the ship docked at the Santos harbor, active case finding was performed among passengers and crew members. First, before landing, all passengers who had had any symptoms during the cruise were requested through the ship's public address system to be interviewed face-to-face with the standard form and to have medical evaluation performed. Second, all crew members who had had symptoms in February 2012 were summoned for interview and medical evaluation. Third, all asymptomatic crew members were called for interview. They were selected in order of arrival until the number obtained was twice that of symptomatic crew members.

ARI was defined as the presence of illness that occurred from February 1, 2012 to February 27, 2012 and, after being onboard, at least one unspecific symptom and one respiratory symptom. Nonspecific symptoms included fever, chills, myalgia, arthralgia, headache, or malaise, whereas respiratory symptoms were defined as cough, nasal congestion, sore throat, or dyspnea.

For the case–control analysis, a influenza-like illness (ILI) case was defined as a surveyed crew member who reported such illness from February 1, 2012 to February 27, 2012 and, after being onboard, having fever and one of the following symptoms: cough, sore throat, or dyspnea. Controls were asymptomatic surveyed crew members.

Statistical Analysis

Categorical variables were compared between cases and controls. Odds ratio (OR) and 95% confidence interval (CI) were calculated, followed by logistical regression analysis. The logistic regression multivariate model was performed based on variables statistically associated with the outcome, using the stepwise forward strategy. The variables that had p ≤ 0.1 in the univariate analysis were selected for the multivariate model, and those that had p ≤ 0.05 remained in the final multivariate model.[4] epi-info (version 3.5.3) and spss 17 statistical software were used for this analysis.

Ethical Issues

The present investigation was the official response to a public health crisis, thus not requiring ethical council approval.

Results

The study identified 104 ARI cases (Figures 1 and 2): 54 (51.9%) crew members and 50 (49.1%) passengers. There were 11 hospitalized ARI cases in Santos from the cruise that took place between February 1 and 18. Among them, seven were crew members and four were passengers. One crew member died. She was a 31-year-old waitress and smoker, although without a history of comorbidity, whose symptoms began on February 11. She had fever, productive cough, hemoptysis, and developed respiratory failure and shock. Methicillin-resistant Staphylococcus aureus (MRSA) grew in the first blood and endotracheal aspirate cultures. Real-time polymerase chain reaction (RT-PCR) was positive for influenza B virus in the nasopharyngeal swab. Among the remaining 10 hospitalized ARI cases, 6 had RT-PCR positive for influenza B virus in the nasopharyngeal swab, all of whom were crew members. The sequence analysis of influenza B-positive samples identified that the viruses were closely related to B/Brisbane/60/2008 (Victoria lineage).

Figure 1.

ARI outbreak investigation among crew members and passengers of a cruise ship, February 2012, Santos, São Paulo State, Brazil. ARI = acute respiratory illness; ILI = influenza-like illness. *One of the hospitalized crew members went back onboard and was included in the survey.

Figure 2.

Number of cases of acute respiratory illness (ARIs) identified among passengers and crew members on two cruises of a ship by date of onset of symptoms. Santos, São Paulo State, Brazil, February 2012. Surveyed crew = 47; surveyed passengers = 46; hospitalized crew = 7; hospitalized passengers = 4. One hospitalized crew member, whose symptoms began on February 14, went back onboard and was surveyed on February 27.

Among the 1,778 passengers onboard on February 27, 122 (6.8%) were present for interview and medical evaluation before landing, and 46 of them had ARI (Figure 1). The majority of identified passengers with ARI were Brazilians (97.8%), female (62.0%), and aged between 18 and 59 years (66.0%) (Table 1).

Table 1. Descriptive epidemiology of passengers and crew that had ARI on a cruise off the Brazilian coast, Santos, São Paulo State, Brazil, February 2012
VariablesPassengersaCrewb
n%n%
  1. ARI = acute respiratory illness.
  2. aTotal passengers = 50.
  3. bTotal crew = 54.
Sex
Male1938.02953.7
Age (years)
<2
2–<181632.0
18–<603366.054100.0
≥6012.0
Symptoms
Fever2754.04175.9
Sore throat3570.03666.7
Myalgia2754.03666.7
Cough3060.03564.8
Headache2346.03361.1
Coryza3162.03055.6
Dyspnea816.0814.8
Arthralgia48.0814.8
Health care
Onboard medical care1632.03972.2
Oseltamivir prescription510.0814.8
Respiratory isolation12.02138.9
Disembarkation for medical evaluation1018.5

Among the 680 crew members onboard on February 27, 92 (13.5%) reported symptoms throughout the month, and 48 of them had ARI (Figure 1 and Table 1). Of these, one was hospitalized before February 18, and had nasopharyngeal swab positive for influenza B. This crew member was the only hospitalized member who returned onboard after recovery.

Most of the interviewed crew members were Brazilian (87.0%), 39 (72.2%) had medical care onboard, 8 (14.8%) took antiviral medication (oseltamivir), 10 (18.5%) landed for medical evaluation, and 25 (54.3%) kept their regular activities onboard during the symptomatic period.

A total of 33 surveyed crew members had ILI on February 27, and 196 were asymptomatic throughout the month (Table 2). Crew members housed in the lower decks (second and third) and belonging to a younger age group (18–32 years) were more likely than others to have self-reported ILI [adjusted odds ratio (aOR) = 2.39, CI 1.09–5.25 and aOR = 3.72, CI 1.25–11.16, respectively].

Table 2. Analysis of ILI associated factors; outbreak of respiratory illness on a cruise ship, Santos, São Paulo State, Brazil, February 2012
 Casea  (%)Controlb  (%)Odds ratio (95% CI)Adjusted odds ratio (95% CI)
  1. CI = confidence interval; ILI = influenza-like illness.
  2. aTotal cases = 33.
  3. bTotal controls = 196.
  4. cWaiters, bartenders, entertainers, receptionists, tourist guides, sellers, and casino staff.
  5. dOfficers, musicians, dancers, cleaning staff, and galley staff.
Housing deck
Lower decks 2 and 322 (66.6)91 (46.4)2.26 (0.98–5.29)2.39 (1.09–5.25)
Higher decks 4–711 (33.3)103 (52.6)1.00 
Age (years)
18–3229 (87.9)131 (66.8)3.60 (1.14–12.63)3.72 (1.25–11.16)
33–604 (12.1)65 (33.2)1.00
Sex
Male17 (51.5)132 (67.3)1.00
Female16 (48.5)62 (31.6)2.0 (0.89–4.50)
Duties on ship
With passenger contactc27 (81.8)122 (62.2)2.73 (1.01–7.76)
Without passenger contactd6 (18.2)74 (37.8)1.00

Discussion

This investigation documented a summertime influenza outbreak in a cruise ship in South American waters. The influenza B/Brisbane/60/2008-like virus (Victoria lineage) was identified in seven cases. The investigation did not find the index case, who could be a passenger or a crew member that had embarked at one of the stops.

A young crew member died of influenza infection followed by MRSA pneumonia. Neither bacterial typing nor pathogenicity test was performed. However, the findings strongly suggest that it was the community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA), producer of a Panton–Valentine leukocidin (PVL) that caused necrotizing pneumonia. CA-MRSA necrotizing pneumonia is closely related to previous influenza infection.[5, 6]

The influenza virus is commonly involved in outbreaks on cruise ships around the world.[2, 7, 8] However, this report is the first description of an influenza outbreak in a vessel cruising in South American waters documented by Brazilian public health authorities. The growth in the cruise industry in Brazil must be accompanied by improvements in surveillance and prevention against influenza transmission and other public health events.

During the outbreak, large groups of new passengers boarded almost weekly, while the crew remained nearly the same. More than half of symptomatic crew members kept their regular duties onboard, as bartenders, waiters, or receptionists, most of them close to passengers. Frequent group activities (eg, bingo, movies, dining, and tours) were held in semi-closed environments (eg, meeting rooms, ballrooms, and cinemas), increasing contact between passengers and crew members and, consequently, the spread of influenza. A rise in the number of new cases was observed after new passengers boarded. The crew was the reservoir for influenza virus infection and new groups of passengers fed the outbreak with susceptible people, thereby prolonging it. Under these conditions and without control measures, an influenza outbreak in a cruise ship could last several weeks.[9]

ILI was more likely to occur among crew members housed in the lower decks (second and third). The lower-ranking crew members were on such decks, usually sharing cabins that fit two to four people. There were no windows and circulated air came from air conditioners. The influenza virus is spread through droplets and aerosol from infected people when coughing and sneezing. Closed and crowded places, such as the ship's lower decks, facilitate the spread of influenza virus.[10] ILI was also more likely to occur among the younger age group. Younger crew members had more contact with people as they were commonly allocated in passenger services, facilitating the dissemination of respiratory diseases.

Although influenza outbreaks in Brazil occur more often in the winter months (June to September),[11]the period of this outbreak illustrates that influenza can occur at any time of the year. Influenza outbreaks have been reported during the summer on other cruise ships elsewhere,[2, 9] but the unusual period may delay etiological identification and therefore delay the intervention.[12-14]

The Victoria lineage influenza B virus was globally circulating during 2011 and 2012, including Brazil.[15, 16] During 2011, the Influenza Sentinel Surveillance of São Paulo State identified 312 samples positive for respiratory viruses, 10% of them for influenza B virus. The genetic sequencing identified B/Brisbane/60/2008-like and B/Victoria/02/87-like strains.[17] The influenza viruses identified among hospitalized crew members were closely related to the B/Brisbane/60/2008-like lineage, the vaccine virus for the 2011 and 2012 southern hemisphere's season.[15-17]

The influenza vaccine was introduced in the Brazilian immunization schedule in 2000. Until 2008, it was recommended to people older than 60 years. After the influenza A (H1N1) pandemic in 2009, the vaccine recommendation was extended to children younger than 2 years, pregnant women, indigenous people, and those with chronic diseases. The surveyed individuals were asked about their immunization status, but nobody had any documents to confirm this during the survey and most of them did not answer the questions (data not shown). The majority of them were in the age group for which influenza vaccine was not recommended by Brazilian health authorities.

The passengers volunteered to participate in the survey, and some ARI cases could have not been detected. Only passengers who had symptoms were interviewed; therefore, case–control analysis was not performed in this group. ILI incidence between February 18, 2012 and February 27, 2012 alone was calculated among passengers of the cruise, and previous cruise data were obtained from medical reports. The total number of passengers from the previous cruise was unavailable and other symptomatic passengers could have been missed. The investigation team had access to the list of passengers' and crew members' names. However, this list did not specify the sex, nationality, birth date, or age of people onboard. This missing information would enable better understanding of whether those who became ill were representative of the crew and passengers in general. This outbreak investigation faced a common problem found in outbreaks involving vessels and airplanes: a detailed list of people onboard. The absence of specific legislation in Brazil and elsewhere, and companies' privacy policy hinder access to detailed information.

Laboratory confirmation of hospitalized cases was performed by the São Paulo State public health laboratory. The hospitalized crew and passengers were tested following medical request. Influenza screening [influenza A(H1N1)pdm09, influenza A (H3N2), influenza A(H1N1), and influenza B viruses] was performed, but no other respiratory viruses were investigated and no tests were performed during the onboard survey.

The interviewed passengers and crew members received medical care, and those who had ILI on February 27 received antiviral treatment (oseltamivir). After the survey, crew members were instructed to report to the ship's infirmary if any symptoms appeared. Antiviral medications were left on the ship and the health staff were instructed to enhance the surveillance for febrile respiratory illness. Alcohol-based hand sanitizer dispensers were installed throughout the vessel and symptomatic crew members were removed from their duties for 5 days after the onset of ILI symptoms. Despite the limited epidemiological evidence, some studies support that nonpharmaceutical interventions, such as the isolation of symptomatic individuals and hand hygiene, are generally effective in reducing the spread of respiratory viruses.[18-20]

This outbreak was an opportunity to review health conducts on vessels and to recommend some preventive measures. Despite Brazilian health authorities not recommending universal immunization against influenza, companies that operate cruise lines should vaccinate their crew against influenza every year and document it. Large numbers of people in closed and semi-closed settings, such as crew members, prisoners, or institutionalized individuals, can facilitate the transmission of influenza viruses and other respiratory viruses from person to person through droplet spread or potentially through contact with contaminated surfaces.[21, 22] Apart from the influenza vaccine, the crew should be up-to-date on routine vaccines, such as measles–mumps–rubella (MMR), varicella, meningococcal, and pneumococcal vaccines. Tourism companies should recommend that passengers seek medical evaluation before a trip and have the flu vaccine at least 2 weeks before embarking.

Declaration of Interests

The authors state they have no conflicts of interest to declare.

Ancillary