See the Editorial by Robert Steffen, pp. 1–3 of this issue.
Anna Vilella, MD, PhD, Preventive Medicine and Epidemiology Department, Hospital Clínic, Villarroel 170, Barcelona 08036, Spain. E-mail: firstname.lastname@example.org
Background. From the beginning of the influenza pandemic until the time the outbreak described here was detected, 77,201 cases of pandemic influenza A(H1N1) with 332 deaths had been reported worldwide, mostly in the United States and Mexico. All of the cases reported in Spain until then had a recent history of travel to Mexico, the Dominican Republic, or Chile. We describe an outbreak of influenza among medical students who traveled from Spain to the Dominican Republic in June 2009.
Methods. We collected diagnostic samples and clinical histories from consenting medical students who had traveled to the Dominican Republic and from their household contacts after their return to Spain.
Results. Of 113 students on the trip, 62 (55%) developed symptoms; 39 (45%) of 86 students tested had laboratory evidence of influenza A(H1N1) infection. Most students developed symptoms either just before departure from the Dominican Republic or within days of returning to Spain. The estimated secondary attack rate of influenza-like illness among residential contacts of ill students after return to Spain was 2.1%.
Conclusions. The attack rate of influenza A(H1N1) can vary widely depending on the circumstances of exposure. We report a high attack rate among a group of traveling medical students but a much lower secondary attack rate among their contacts after return from the trip. These findings may aid the development of recommendations to prevent influenza.
The first cases of human infection with the 2009 pandemic influenza A(H1N1) virus were detected in two children in Southern California during late April 2009.1 A few weeks earlier, health officials in Mexico had detected an increase in severe pneumonia affecting mainly young, healthy adults that was subsequently determined to be because of infection with a nontypeable influenza A(H1N1) virus genetically similar to that isolated from the children in California.2 From then until late 2009, this pandemic virus caused more than 500,000 cases of influenza worldwide, including over 10,000 deaths.3 We describe an outbreak of H1N1 influenza among medical students who traveled from Spain to the Dominican Republic in June 2009.
Most 2009 pandemic influenza A(H1N1) virus infections resulted in clinically mild disease without complications. However, the virus caused substantial morbidity and mortality, even in young, healthy people.4 Compared to seasonal influenza, the incidence of 2009 pandemic influenza A(H1N1) was higher among people aged 5–65 years.5–7 In Europe, about 80% of reported cases occurred in people aged <30 years.8 From the beginning of the influenza pandemic until the time the outbreak described here was detected, 77,201 cases with 332 deaths had been reported worldwide, mostly in the United States and Mexico. By June 29, 2009, 6,173 cases of influenza A(H1N1) disease had been reported in Europe; 541 of these were in Spain.9 In the Hospital Clinic of Barcelona, 13 cases had been detected, all with a recent history of travel to Mexico, the Dominican Republic, or Chile. Concurrently, 108 cases had been reported in the Dominican Republic.9
On June 27, 2009, a group of 113 sixth-year medical students from the University of Barcelona returned from an 8-day vacation in the Dominican Republic. From 1–3 days before the return trip, six students developed mild influenza-like illness (ILI) manifested primarily by respiratory symptoms and accompanied in some cases by fever and diarrhea. On their return, one student presented to the emergency department of the Hospital Clinic, where a nasal swab was positive for 2009 pandemic influenza A(H1N1) by polymerase chain reaction (PCR). Four students with similar symptoms were seen at the same emergency department in the following hours. This led to suspicion of an outbreak, and an epidemiological investigation was initiated to assess the impact of pandemic influenza A(H1N1) infection in this group of students and their residential contacts.
We attempted to contact all the 113 medical students who traveled to the Dominican Republic between June 19 and June 27. Data on demographics, details of illness, history of influenza vaccination, activities during the trip, location of seating on the return aircraft, and illness among residential contacts after returning from the trip were collected from all consenting students using a standardized questionnaire. Data were collected from June 29 to July 2, 2009.
A probable case of 2009 pandemic A(H1N1) influenza was defined as any medical student who traveled to the Dominican Republic and had onset of ILI between June 19 and July 1, 2009. ILI was defined as recent onset of any of the following: fever, cough, sore throat, rhinorrhea, asthenia, breathing difficulties, myalgia, or malaise.
A confirmed case was defined as any probable case with influenza virus A(H1N1) infection confirmed by the laboratory testing described below. When a probable case was detected, measures to prevent the spread of the virus were recommended to all symptomatic cases, including home isolation, use of a separate bathroom, use of surgical masks when in contact with cohabitants, and regular hand washing.
A secondary case was defined as a household contact who developed an ILI or laboratory-confirmed influenza within 7 days of symptom onset of the corresponding medical student case.
Throat and nasal swabs were collected from all consenting students in the group of travelers, whether symptomatic or not from June 29 to July 2. The samples were transported in 2.5 mL of viral transport medium (VTM) (fluid with 2% fetal bovine serum, penicillin 100 U/mL, streptomycin 100 g/mL, amphotericin B 20 g/mL, neomycin 40 g/mL, and NaHCO3 buffer). Respiratory specimens were placed in a tube containing VTM. Within the first 24 hours they were processed and stored at 2–4°C in several aliquots until use. Total nucleic acids were extracted from 200 µL of fresh specimen and eluted in 25 µL of RNase-free elution buffer using NucliSense easyMAG (bioMérieux, Marcy l’Etoile, France) according to the manufacturer's instructions. Nucleic acids were kept frozen until use.
Two specific one-step multiplex real-time reverse transcription-PCR were used for typing and subtyping the influenza virus, as previously described.10,11 An additional third assay amplified a housekeeping gene (RNase P) of human cells to assess the correct progress of DNA extraction and to underline the absence of PCR inhibitors as an internal control.12 The viral nucleotide sequences obtained from infected students were compared using sequences of viruses in GenBank from the Dominican Republic and Spain. No additional tests were done to assess etiology of gastrointestinal illness.
Differences in student characteristics between pandemic influenza A(H1N1) positive and negative students were evaluated using the chi-square test or Fisher's exact test as necessary. Logistic regression models were used to assess risks factors for having a positive lab test for influenza A(H1N1) adjusting for time between onset of symptoms and collection of swabs. p Values ≤0.05 were considered statistically significant. All analyses were performed using the SPSS v15 (Chicago, IL, USA) for Windows.
The study protocol was approved by the ethics committee of the Hospital Clinic of Barcelona.
On June 19, 2009, the medical students traveled from Barcelona to Santo Domingo, Dominican Republic on a scheduled flight with a stopover in Madrid. A bus took them to their hotel located at a beach resort in Punta Cana. Meals were shared depending on daily activities organized and the number of students participating in each activity. The students slept in rooms for two, three, or four people with members of their own travel group. Activities were organized according to the interests of the students, and included sightseeing excursions. On June 26, all students traveled on the same bus on a 4-hour trip from Punta Cana to Santo Domingo, where they boarded an Airbus aircraft with 284 economy class and 20 business class seats. The flight back to Spain lasted 8 hours. Of the 113 students, 86 (76%) were contacted and agreed to participate in the study. The rest could not be contacted or declined to participate. Of the 86 students, 58 (67%) were female. The median age was 24 years, (range 22–56 y). A total of 62 (72%) students developed ILI, and influenza A(H1N1) was confirmed in 39 (45%) (two confirmed cases were asymptomatic). Thus, assuming that none of the students who did not participate were ill or infected, the minimum attack rate among all 113 students was 55% for probable influenza and 35% for confirmed influenza.
Two of the 37 confirmed cases developed symptoms during the stay in the Dominican Republic. The first confirmed case first developed symptoms on June 24, followed by a second case on June 25 (2 and 1 d before starting the return trip, respectively). Between June 26 (day of departure from Santo Domingo) and 48 hours after arriving in Barcelona, 29/39 (74%) of the students with confirmed A(H1N1) infection developed symptoms; 6 students (15%) developed symptoms more than 72 hours later, and 2 remained asymptomatic (Figure 1).
The predominant symptoms in confirmed cases (Table 1) were cough (87%), malaise (60%), and sore throat (51%). Gastrointestinal symptoms (diarrhea) were reported by 16 (43%) of the confirmed cases. Univariate analyses showed that cough, fever, myalgia, rhinorrhea, and malaise were significantly associated with confirmed infections, and this was supported by the logistic regression analysis (Table 1). Laboratory testing for influenza was more likely to be negative when the time between the onset of illness and the day of diagnostic sampling was longer (Figure 2). The mean time between onset of symptoms and blood sampling was 3.5 days; most (92%) of the positive samples were obtained between 1 and 3 days after onset, whereas most (83%) of the negative samples were obtained 3 or more days after onset.
Table 1. Clinical symptoms of students with positive and negative nasal aspirates for influenza A(H1N1)* and probability of a positive test according to the symptoms presented and adjusted by time between onset of symptoms and swabs (logistic regression)
On arriving home from the trip, the students went to their homes, where they lived with family or other students. We were able to collect information regarding all 188 household contacts of 68 students. No information was available for contacts of the remaining 18 students. Those who were reported by students to have developed influenza-like symptoms a few days after the students arrived home were asked to provide diagnostic samples. Four contacts of three confirmed cases reported ILI and all were tested. Only one of these contacts tested positive for influenza A(H1N1). Thus, the secondary attack rate was 0.5% (1/188) for confirmed influenza and 2.1% (4/188) for probable influenza. Of the 188 contacts, 137 were contacts of students with symptoms, and 4 (2.9%) of these reported illness. The 39 students with confirmed influenza had 98 contacts so, for this group, the secondary attack rate for confirmed influenza was 1% (1/98), and the secondary attack rate for probable influenza was 4% (4/98). All patients seen were advised about preventive measures to avoid further transmission: the use of masks, home isolation, and hand washing.
No student received prophylaxis or antiviral treatment with oseltamivir, as all had mild illness and none had underlying diseases or risk factors for severe illness. The symptoms resolved without specific treatment after a mean of 4–5 days. All confirmed and probable cases recovered satisfactorily and none required hospitalization. Only seven students had been vaccinated with the seasonal influenza vaccine before the trip, of whom four developed laboratory-confirmed A(H1N1) 2009 pandemic influenza.
Figure 3 shows the aircraft seats occupied by the 113 students on the return flight from the Dominican Republic. Students who became ill were seated throughout the aircraft with no apparent clustering. We were not able to obtain information on illness among other passengers who shared the return flight.
Phylogenetic Study of the Virus
The viral nucleotide sequences obtained from infected students were indistinguishable from sequences of viruses in GenBank from the Dominican Republic. However, the sequences were also indistinguishable from other viral sequences identified in Spain.
We describe here a high primary attack rate of pandemic A(H1N1) influenza among a group of medical students who traveled to the Dominican Republic during a period of epidemic influenza transmission, followed by a low secondary attack rate among household contacts after the students' return to Spain.
The relatively mild clinical presentation of pandemic A(H1N1) influenza in this group of students is consistent with previously described outbreaks.6 However, we found a higher frequency of gastrointestinal symptoms than reported in previous studies in cases of influenza,6,13 but the high percentage of gastrointestinal symptoms in students with (64%) and without (43.2%) ILI suggests the possible coexistence of travelers' diarrhea in this group.14
The epidemic curve (Figure 1) suggests that initial transmission of influenza in this group occurred in the Dominican Republic, especially because the pandemic H1N1 virus was circulating widely there at the time of the students' visit. Given that the usual incubation period of pandemic H1N1 influenza is 2–4 days and because all the cases appeared in a short time period, it was not possible to identify the index case. The close contact between students, with many group activities, may have facilitated viral transmission between students once it was encountered.15,16
Transmission was probably more intense just before the return trip, when the group spent even more time in close contact (a 4-h coach trip to the airport, waiting in the airport, boarding).17,18 We considered the possibility that transmission had predominantly occurred during the return flight. Reports show that transmission of an infectious agent in the interior of an aircraft may be influenced by the length of the flight, the stage of the disease, the ventilation system and size of the airplane, and the number of persons onboard.19 It has been reported that the design or malfunction of aircraft ventilation systems could influence viral transmission. In an outbreak of influenza reported in 1979, which also described a high attack rate, a technical failure in the aircraft ventilation system was demonstrated.20 Previous studies have suggested that proximity to the index case (sitting in the same row or in the three anterior rows) increases the probability of infection.15,21,22 We were unable to verify this relationship in the current outbreak. One of the limitations of our study is that we only had information on the group of students and thus do not know whether other passengers were infected.
In our study, the probability of laboratory confirmation of A(H1N1) infection by PCR of nasal aspirates diminished with increasing time from onset of symptoms to testing. This seems consistent with an expected decrease in viral abundance in nasal secretions as the illness resolves. The longer sampling times for some students could result in underestimation of the primary attack rate of confirmed A(H1N1) influenza in this group.
Once the outbreak was recognized, vigorous control and prevention measures were recommended to prevent the spread of the virus. Home isolation, the use of a separate bathroom, the use of surgical masks when in contact with cohabitants, and hand washing precautions were recommended to all cases. These medical students were probably highly motivated to practice preventive measures, and this could have limited secondary transmission to their close contacts. In addition, the majority of household contacts were adults and the infective load of many of the students may have been low once they arrived home. Low rates of secondary transmission, although higher than those in our study, and data showing easier transmission among young children than among adults have been reported in seasonal influenza outbreaks23 and for pandemic influenza in different settings, including on an airline flight.15,24,25
We found no evidence that seasonal influenza vaccination protected against 2009 A(H1N1) pandemic influenza virus infection in this group. However, our data is limited to address this question; only seven students had been previously vaccinated, of whom four were confirmed cases. Previous reports have also failed to demonstrate such protection.26
The difference between the high attack rate among this group of medical students and the much lower secondary attack rate in household contacts after their return home supports the idea that the transmission dynamics of pandemic influenza A(H1N1) virus can vary widely, depending on the level and duration of interpersonal contact and the rigor of preventive measures. The low incidence of secondary cases in our study might signal that the application of preventive measures helped to decrease disease transmission. Similarly intensive preventive interventions in sites such as airports, tourist resorts, and military camps might reduce secondary transmission of influenza. Infection of close-knit groups of travelers, such as students, businessmen, peacekeepers, and tour groups, likely facilitates intense transmission and spread of influenza virus.27 Better understanding the dynamics of diffusion of influenza virus in such groups could help design and support relevant preventive measures, including the recommendation of influenza vaccination before traveling.
The emergence of a new influenza virus may involve changes in the epidemiological pattern of the virus. The investigation of outbreaks such as that described here, especially at the beginning of an epidemic, is important because it may allow early detection of possible changes.
The authors wish to thank all sixth-year medical students of the Clinic Campus, University of Barcelona for their collaboration. We thank S. Polbach for assistance in data collection and M. Domenech for her invaluable cooperation in managing the outbreak.
Declaration of Interests
The authors state that they have no conflicts of interest to declare.