| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0.05) for infection included age < 20 years, the presence of food or water pans for animals on the property, taro farming, the presence of Ae. aegypti on the property, and presence of Ae. scutellaris group mosquitoes (Ae. Hensilli, Ae. albopictus, and a native species). This was the first outbreak of dengue 4 virus in the Western Pacific, and the first documented epidemic of dengue in Palau since 1988. |
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In most of the tropical and subtropical world, dengue fever is a leading cause of morbidity and mortality, particularly among children.1,2,5–7 In the Pacific, after years of absence due to isolation of the area after World War II, dengue viruses were reintroduced in the 1970s, and epidemic activity in the region has intensified during the 1980s and 1990s.5,8,9 In Palau, epidemic dengue fever had not been reported since 1988, when a large outbreak of dengue fever caused by dengue 2 virus occurred.10
During January and February 1995, 145 patients (an unusually high number) with viral syndrome were reported to the Palau Ministry of Health. On April 3, 1995, a 38-year-old man died at the Palau National Hospital soon after presentation with viral syndrome. He was noted to have had neutropenia and thrombocytopenia. Initially, an outbreak of leptospirosis was suspected (and later this patient was confirmed to have leptospirosis by immunohistochemical analysis). However, the majority of the initial serum samples from patients with febrile illness tested at the Centers for Disease Control and Prevention (CDC) Dengue Laboratory in San Juan, Puerto Rico were consistent with dengue virus infection, suggesting an outbreak of dengue fever.
We conducted an investigation to assess the extent of this outbreak, determine the risk factors associated with infection including the presence and abundance of vector species, and define control strategies. This project was reviewed by the Human Subjects Coordinator in the National Center for Infectious Diseases and determined to be a public health response that did not require Institutional Review Board review.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The medical system of Palau is made up of a central referral hospital in the capital city and two medical clinics. These three facilities serve the majority of the population. In addition, there are primary care clinics in a few outlying islands. Seriously ill patients from the outlying islands are referred to the central hospital in Koror.
Case definition. A case of dengue fever was defined as febrile illness (38.5°C) after January 1, 1995, with body or joint aches and one of the following: headache, rash, nausea, vomiting, or hemorrhagic manifestations in a resident of Palau.
Surveillance review and case finding. To determine the severity of the epidemic, active clinical surveillance was established at the national hospital and two private clinics and was maintained from January 1 to June 30, 1995 (study period.) All patients (outpatients or inpatients) meeting the case definition were included, and their medical records were reviewed using a standard form. Acute-phase serum samples were obtained at the time of initial presentation, and the patient was asked to return in 10 days to obtain a convalescent-phase serum sample.
During the study period, suspected cases of dengue fever were reported to the Public Health Department in Palau. The referral hospital maintained log books that included admission and discharge diagnosis. In addition to establishing active surveillance, we reviewed existing surveillance databases at the Health Department, the referral hospital record, and clinic log books for diagnoses consistent with dengue fever after January 1, 1995. These diagnoses included viral syndrome, dengue fever, and leptospirosis. All records from suspected patients were reviewed. Information was collected on demographic characteristics, date of onset of signs, underlying illness(es), clinical findings, laboratory testing, therapy, and outcome.
Cross-sectional survey. Because infection with dengue virus can result in a broad spectrum of illness with many mild or asymptomatic infections, the attack ratio of dengue infection in the community was estimated by conducting a house-to-house serologic survey in five hamlets in Koror: Meyuns, Idid, Iyebukel, Ngermid, and Ngerbeched, and on the island of Peleliu. Systematic random sampling of every fifth house was performed using 1991 census maps. The head of each household was questioned about household demographics, risk factors, and history of febrile illness within the family since January 1995. Each person with a history of fever, headache, and myalgia since January 1995 was interviewed independently. Blood samples were obtained from all willing family members present at the time of the survey.
Laboratory testing. All laboratory testing for dengue was conducted at the CDC Dengue Laboratory in San Juan, Puerto Rico. Blood samples were allowed to clot at room temperature for 12 hours, centrifuged, and serum was immediately frozen in liquid nitrogen. Serum samples were tested by an IgM capture enzyme-linked immunosorbent assay (ELISA) and an IgG ELISA using methods previously described.11,12 Virus isolation attempts were done by using the mosquito inoculation technique, or in C6/36 mosquito cell cultures, using a 1:5 dilution of serum.13,14 Identification of isolates was done by an indirect immunofluorescence assay using serotype specific monoclonal antibodies.14
Rainfall. We reviewed records from the Palau Meteorological Center to determine precipitation from 1992 to 1995. Records from before 1992 were not available.
Environmental and larval survey. Among a sub-sample of households, environmental and larval habitat surveys were conducted concomitantly with the serologic survey. Household features assessed included the presence and number of animals, type of house construction, presence of screens on doors and windows, and methods of water storage. Potential larval mosquito habitats in which water or mosquito larvae were present were enumerated for each household, and larvae were collected from at least one container of each positive habitat-type. At least 20 larvae were collected for subsequent identification. When possible, larval mosquitoes were reared to the adult stage, frozen, and held in liquid nitrogen for storage and transport. Remaining larvae were killed and transferred to vials filled with fresh 70% ethanol. All mosquito specimens were transported to the Division of Vector-Borne Infectious Diseases of CDC in Fort Collins, Colorado for identification using standard references.15,16
Adult mosquito survey and virus testing. A mosquito aspirator was used to make comparative indoor and outdoor collections of adult Aedes (Stegomyia) mosquitoes for dengue virus isolation.17 From May 2 to May 5, 1995, comparative collections were made at three suspected case houses (residences of persons who presented at the hospital with symptoms compatible with dengue fever), six houses selected randomly from a list of houses participating in the serologic survey, and the dormitory of a private high school. At each of the nine houses, the inside of the house was aspirated as completely as possible with emphasis on likely mosquito resting sites. Comparative outdoor samples were collected by aspirating the grounds of the premise, including the outside walls and foundation of the house and outbuildings and other potential resting sites such as areas of overgrown vegetation, taro and flower gardens, tires, water barrels, and trash piles. Time for each inside and outside sample was brief and varied between 10 and 15 minutes. The freshman and sophomore dormitory rooms of a private school were aspirated for 10 minutes each, and the samples were combined. Time of aspiration for the grounds surrounding the dormitory was 20–25 minutes.
Mosquitoes were placed in cryovials, held in liquid nitrogen, identified, and tested for the presence of virus at the Division of Vector-Borne Diseases at CDC in Fort Collins, Colorado. Attempts to isolate viruses from individual Aedes (Stegomyia) mosquitoes were made using C6/36 cell culture techniques followed by direct fluorescent antibody tests to detect the presence of antigen.14
Statistical analysis. Clinical, epidemiolgic, serologic, and environmental data were entered and analyzed using EpiInfo version 6.0 (CDC, Atlanta, GA).18 Univariate and multivariate analyses were done using generalized estimating equations to control for the clustering effects of families.19
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). For comparison, from the retrospective review of surveillance records at the Palau Ministry of Health, we estimated that the average number of patients with viral syndrome during 1993 and 1994 was 4.5/week (Figure 1).
|
Among patients presenting to health facilities, dengue 4 virus was isolated from 47 (46%) of 102 blood samples tested. No other dengue virus serotype was detected. Antibody testing was done on serum samples from 338 patients, 75% of which were positive for dengue by IgM capture ELISA.
Cross-sectional survey.
Two hundred seventeen households were included in the randomized community survey. Blood samples were collected from 865 household members; results were available for 841, of which 224 (attack ratio = 27%) were positive for IgM antibodies to dengue virus by capture ELISA. Sixty five (31%) of the 217 households surveyed reported having at least one family member who had had an illness consistent with dengue fever in the previous three months. Dengue transmission was widespread, with dengue attack ratios exceeding 20% in all hamlets except Ngermid, which had an attack ratio of only 0.06% (Table 1). The hamlet with the highest attack ratio was Meyuns (44%).
|
0.05).
Clinically, the illness presented as classic dengue with the majority of laboratory-positive patients having fever, headaches, and myalgia (Table 2).
|
0.05). When these rates were extrapolated to the entire relatively homogeneous population of these islands, we estimated the total number of infections to be approximately 3,400.
The results of multivariable clustered analysis of risk factors for infection from the cross-sectional survey (not including environmental data) are summarized in Table 3. Being less than 20 years of age, working in taro fields, reporting increased mosquito biting in the past three months, and reporting an increased mosquito population in comparison with 1994 were all significantly associated with infection (P 0.05).
|
. Based on this comparison with historical rainfall data, we determined that the rainfall for January and February 1995, immediately preceding the outbreak, was approximately three times the expected amount.
|
0.05). The 10 most abundant larvae-positive habitats encountered were tires, cisterns, water barrels, cans and bottles, animal drinking pans, buckets, plastic containers, cooking pots, flower pots. and taro (Table 4). Tires and cisterns were the most common larval habitats on Koror, accounting for 56 (33%) and 28 (16%) of the 172 larvae-positive containers. On Peleliu, larvae were more evenly distributed among the habitat types. Together, water barrels (7.3%), cisterns (3.1%), and tires (3.1%) accounted for 13.5% of the 28 larvae-positive containers discovered in the premise surveys on Peleliu.
|
Larvae of Ae. albopictus, Ae. hensilli, and Aedes (Stegomyia.) sp. 1 are very similar and could not be reliably differentiated. These larvae were identified and treated in the statistical analysis as one group referred to as the Ae. scutellaris group.
The abundance of Stegomyia species, as measured by the percentage of positive houses, was notably different on Koror and Peleliu. Aedes albopictus was present on Koror, but was not collected on Peleliu. On Koror, 24.8% of the houses were Ae. aegypti positive, 19.3% were Ae. albopictus positive, 11% were Ae. hensilli positive, 3.5% were Ae. palauensis positive, 2.1% were Aedes (Stegomyia.) sp. 1 positive, and 6.9% were positive for larvae identified as the Ae. scutellaris group. On Peleliu, 17.9% of houses were Ae. palauensis positive, 14.3% were Ae. hensilli positive, 10.7% were Ae. aegypti positive, and 3.1% were Aedes (Stegomyia.) sp. 1 positive.
Multivariate analysis of risk factors from the environmental and larval investigation revealed that the presence of animal water pans, the presence of Ae. aegypti, and the presence of Ae. scutellaris group mosquitoes (Ae. hensilli, albopictus, and sp. 1), were significantly associated with dengue infection (Table 3
).
Adult mosquito survey and virus testing. Five mosquito taxa were collected during indoor and outdoor adult aspiration studies. All adult Aedes (Stegomyia) mosquitoes (Ae. aegypti, n = 22; Ae. albopictus, n = 13; Ae. hensilli, n = 16; and Ae. (Stg.) sp. 1, n = 2) were individually processed and tested for presence of dengue virus. All specimens were negative. The fifth species, Culex quinquefasciatus, is refractory to infection with dengue viruses, and these specimens were not tested.3,20
Aedes aegypti was the most commonly collected Aedes (Stegomyia) species in and around the nine houses sampled, with a total of 22 adults (mean = 2.4 specimens/house): 8 males and 11 females were collected indoors and 2 males and one female were collected from peridomestic areas. Aedes hensilli was the second most commonly encountered species, with a mean of 1.6 mosquitoes per house: 3 males and 4 females were collected indoors and 3 males and 6 females were collected from peridomestic area. Aedes albopictus, was relatively rare in and around houses sampled during the aspiration survey, with a mean of 0.4 specimens per house. Only one adult specimen of Aedes (Stegomyia.) sp. 1 was collected from the nine sampled houses.
Culex quinquefasciatus was the most frequently encountered species among the nine houses surveyed, with a mean of 4.4 specimens per house. In addition, 127 specimens of Cx. quinquefasciatus were collected inside the freshman and sophomore dormitory rooms of the private school in a 20-minute period.
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Dengue fever is caused by one of four serotypes of the arbovirus. Symptoms include sudden onset of fever, intense headache, and muscle and lower back pain. The duration of illness is usually 1–2 weeks.22 During the 1995 outbreak in Palau, disease was mild and hemorrhagic manifestations were rare. Although the symptoms associated with dengue 4 virus infections have been reported as being milder than symptoms associated with infections with the other subtypes (with outbreaks being characterized by a majority of dengue fever rather than dengue hemorrhagic fever), infections with dengue 4 virus can also be associated with severe and fatal disease.2,20,23
Aedes aegypti, Ae. albopictus, and other known vectors of dengue fever were present on the Palauan islands. In this epidemic, we were able to find a positive correlation between the presence of Ae. aegypti and affected households. We found that dengue fever was associated with young age, food and water pans for animals on the property, and taro farming. Association of dengue fever with younger age groups has been previously reported.20,22,24 The association found between dengue fever and food and water pans for animals may be explained by the fact that the pans provided more breeding sites, or these habitats may have been associated with other breeding sites. While tires, cans, and buckets were some of the most common larval habitats found in the survey, these habitats were not independently associated with case households (the high prevalence in both affected and non-affected households could explain this finding.)
We also found an association with taro cultivation. Taro is a starchy root that is a staple of the Palauan diet. It is grown in patches of moist soil usually located some distance from the house. Some Aedes species of mosquitoes, which may transmit dengue virus, are adapted to living near these plants because water collects in the leaf stalk, forming an ideal breeding site.
All adult Ae. (Stegomyia) collected during aspiration studies were negative for dengue virus. However, the risk factor analysis in conjunction with data on the distribution and abundance of mosquito species suggest that Ae. aegypti and Ae. hensilli were the primary vectors. Specifically, the serologic and entomologic surveys demonstrated a positive association between human infection and the following three variables: presence of Ae. aegypti, participation in taro cultivation, and presence of Ae. scutellaris group mosquitoes (Table 3). Aedes aegypti has been the primary epidemic vector of dengue viruses in urban areas of Asia during the last century and the vector in more recent epidemics in Africa, Central and South America, and the Pacific.1,2 In Palau, Ae. aegypti was the most common Stegomyia collected in aspirator studies and larvae of this species were frequently encountered, being collected at 24.8% of the households on Koror and 10.7% of the households on Peleliu. Although we did collect Ae. hensilli larvae from taro plants, the positive association between human infection and taro cultivation and the presence of Ae. scutellaris group mosquitoes suggest that infection was associated with increased exposure to biting adults of exophylic Aedes (Stegomyia) such as Ae. hensilli and Ae. albopictus. Recently, Ae. hensilli was implicated as the epidemic vector during a dengue 4 virus outbreak in Yap State, Federated States of Micronesia.24 In Palau, Ae. hensilli was the second most commonly collected Stegomyia species in aspirator collections, and this species was also abundant on both Koror and Peleliu, being collected at 11% and 14% of households, respectively. During the 1988 dengue 2 virus outbreak in Palau, 29 (45%) of 65 persons surveyed on the island of Peleliu were positive for IgM antibodies to dengue virus.10 Subsequent identification of all mosquitoes collected in 1988 from 36 houses on Peleliu revealed the presence of only two species of Ae. (Stegomyia): Ae. hensilli and Ae. palauensis.24 The absence of other established vector Stegomyia species on Peleliu during the 1988 outbreak implicated Ae. hensilli as a potential epidemic vector of dengue viruses. In this survey, we have found that Ae. aegypti was apparently introduced into Peleliu sometime between 1988 and 1995, but the relative contribution of Ae. aegypti and Ae. hensilli to transmission during the 1995 outbreak on Peleliu and Koror is unknown. Certainly Ae. aegypti could account for transmission on both islands without the presence of a second vector. However, we believe that Ae. hensilli played an important role in transmission in this epidemic because of the significant positive association between infection and working in taro fields, where one would experience greater exposure to Ae. hensilli, the significantly greater infection rates among persons less than 20 years of age (an age group that may spend a greater proportion of time out-of-doors), the fact that Ae. hensilli was the second most frequently encountered Stegomyia in the aspiration studies, the greater abundance of Ae. hensilli on Peleliu, and the fact that it was responsible for transmission of the dengue 2 virus outbreak on Peleliu in 1988.
We believe that Ae. albopictus was not a primary vector in Palau for the following reasons: Ae. albopictus was absent from the island of Peleliu, infection rates in humans were not significantly different on the islands of Koror and Peleliu, indicating that the absence of Ae. albopictus did not affect transmission rates, and Ae. albopictus was rarely collected in aspirator studies.
As part of this investigation, control measures were begun in the first week. They included a public education campaign (targeting reducing mosquito breeding habitats) and improved solid waste disposal. Pamphlets were produced for house-to-house distribution during our own and other surveys, and these pamphlets were also distributed at public gathering places such as churches and the Koror City baseball park. We recruited the Boy Scouts and Girl Scouts in Koror into our education campaign in a door-to-door survey. Also, we coordinated the production of a video on mosquito control for broadcast on the local television station, we issued several press releases through the local newspaper, and we produced pamphlets for distribution to arriving visitors at the Palau airport.
Continued monitoring of febrile illness, early detection and diagnosis of potential dengue fever outbreaks, and programs of mosquito control should continue in Palau and throughout the Pacific Region. Control measures should include public education with an emphasis on the frequent inversion of any human-made containers that might collect water, including pet food and water pans. Such containers should be inverted at least once each week to interrupt the mosquito breeding cycle. Unless mosquito control and sanitation programs are improved, the continued presence of multiple breeding sites and high mosquito populations increases the likelihood of future outbreaks.
Received December 7, 2002. Accepted for publication April 17, 2003.
Authors’ addresses: David A. Ashford, Rana A. Hajjeh, and Richard A. Spiegel, Meningitis and Special Pathogens Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Mailstop A13, 1600 Clifton Road NE, Atlanta, GA 30333, E-mail: dba4{at}cdc.gov. Harry M. Savage, Vance Vorndam, Gary G. Clark, and Duane G. Gubler, Division of Vector-Borne Infectious Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, PO Box 2087, Fort Collins, CO 80522. Jill McReady, Palau Ministry of Health, Koror, Palau. David M. Bartholomew, Naval Hospital, 100 Brewster Boulevard, Camp Lejeune, NC 28547-2538.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
