| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We report an investigation of cutaneous anthrax in Haiti’s Artibonite Valley. Reports of anthrax in Haiti date back to the eighteenth century, when an epidemic consistent with and suspected to be gastrointestinal anthrax killed 15,000 people.4 The World Health Organization (WHO) lists Haiti as 1 of 18 countries worldwide with hyper-endemic anthrax.5–7 Anthrax in Haiti is most commonly reported from the country’s southern peninsula and until recently was relatively rare in the Artibonite Valley, which is located 100 miles north of Port au Prince.2,8,9 In this report we examine the incidence of human anthrax in Haiti’s Artibonite Valley over a 10-year period, describe the clinical presentation of cutaneous anthrax in adults and children, and determine risk factors for anthrax.
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). HAS provides three levels of health care: community health workers in each village, primary care dispensaries in each of 14 rural zones, and a central referral hospital. All patients with acute illness are first seen by a medical officer in their primary care dispensaries, and if necessary are referred to see a physician at the referral hospital. HAS and affiliated dispensaries are the sole health care providers for the population in the catchment area. Therefore, all patients who sought medical care for anthrax in the HAS district would have been seen in a dispensary or the hospital. The Schweitzer Hospital has one of the longest running and most comprehensive population-based health information systems in Haiti.10 The information system includes a household survey conducted every ~10 years, data from community health workers on pregnancies, births, deaths, vaccinations, and children’s weights, and clinical data. The health information system of HAS has been a critical source of data on the health of rural Haiti for 35 years.9,11–17
|
Incidence of reported cases of cutaneous anthrax, 1992–2002. We report the incidence of reported cases of anthrax in each year between 1992 and 2002 as anthrax cases per 1 million persons per year. For each case of anthrax diagnosed at HAS, a national case report form must be completed. The case report form includes a checklist of clinical criteria that must be present for a case to satisfy the WHO clinical definition for suspected cutaneous anthrax.18 Criteria for cutaneous anthrax are: a skin lesion evolving over 2–6 days, starting as a papule or vesicle, ulceration, surrounding edema, and development of a black eschar. The hospital records the number of cases of anthrax reported to the government in monthly morbidity and mortality reports. We reviewed these reports for the past 11 years and recorded the number of cases of anthrax reported per year from 1992–2002.
The population of the hospital district was estimated from periodic household surveys conducted by the HAS Community Health Department since 1968. A survey conducted in 1995 estimated the population of the service district at 250,000. To estimate the population in other years between 1992 and 2002, we assume a population growth of 1.5% per year in intervening years.19,20
Clinical case review of patients diagnosed with anthrax, 1998–2002. We reviewed the medical charts of all cases of suspected anthrax diagnosed between 1998—when an increase in human cases was noted—and 2002 using a structured data collection form. Each patient treated at HAS has a medical chart with a unique identifier number. We generated a list of all anthrax cases and their hospital identifier numbers by review of the hospital logs and the morbidity and mortality reports for the past 5 years. We then retrieved the medical charts and reviewed them. Data summarized from the charts included age, sex, date of diagnosis, place of residence, clinical presentation, treatment, clinical outcome, and clinical laboratory data.
Case control study of cutaneous anthrax diagnosed in 2002. A case control study to identify risk factors for anthrax was prospectively conducted in 2002. The study was conducted by physicians to promptly guide the public health response to the growing number of cases of anthrax seen at the hospital. All patients diagnosed with anthrax at HAS were interviewed by a physician fluent in Haitian Creole using a structured questionnaire on exposure history to animals and animal products. For children, a parent or guardian was also interviewed. Controls were matched for age (+/– 2 years), sex, and dispensary of attendance. Within 1 month of interviewing the case, the physician interviewer visited the primary care dispensary of the rural zone where each case presented and questioned a patient seeking primary care services matched by age and sex to the case. Risk factors for anthrax were compared between the cases and controls.
Analysis methods. Data were analyzed using EPI info 2000 (Centers for Disease Control and Prevention, Atlanta, GA). For categorical variables, frequency and proportions were computed, whereas for continuous variables, median and ranges are presented. The incidence of anthrax is reported as cases per million population per year. Proportions are compared using Fisher’s exact test; risk ratios with 95% confidence intervals (CI) are reported. Means and medians are compared by Student’s t test and the Wilcoxon rank sum test respectively. A P value of less than 0.05 was considered statistically significant.
The study was approved by an institutional review board (IRB) at the HAS and Cornell University.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
details the anthrax incidence in each year between 1992 and 2002.
|
).
|
The most common site of cutaneous anthrax was the face (53 cases, 61%). Of the 53 lesions on the face, 17 were peri-orbital, 11 involved the cheek or jaw, 9 were perioral, 1 was on the forehead, and 15 were unspecified.
The most common associated clinical sign was massive edema (Figure 3). Other common signs included fever, local lymphadenopathy, and abdominal pain. Pain at the site of the lesion was reported in only 12 (14%) cases and was mild to moderate.
|
Residents of the four adjoining rural zones of Perodin, Medor, Petit Riviere, and Plassac were at higher risk for presenting with anthrax than residents of other zones in the district (Figure 1
). Sixty-nine cases (79%) came from these 4 northern zones, which have a population of 78,732 people, for an incidence of 175 cases per million population per year compared with 19 cases (21%) from the remaining 10 zones, which have a population of 171,268 people, for an incidence of 22 cases per million population per year, risk ratio 7.9 (95% CI 4.7–13.1, P < 0.0001).
Between 1998 and 2002, anthrax appeared seasonally in the Artibonite Valley with a peak at the end of the dry season in May and June with a mean of 4 cases per month and a nadir in the rainy months of October and November with a mean of < 1 case per month.
During hospitalization, all patients were treated with intravenous penicillin (1.2 million units per day) for a total course of 14 days per hospital standard procedures. Nine patients (10%) required reconstructive surgery for extensive skin necrosis and 8 (9%) required a prolonged course of antibiotics for secondary infection. Seven patients (8%) died. The essential facts of the deaths are described in Table 2. Four patients died of asphyxiation when facial and neck edema compressed the trachea causing airway obstruction. Two patients died with symptoms consistent with gastrointestinal anthrax. Treatment of the patients who died included tracheotomy for 3 of the patients with airway obstruction and red blood cell transfusions in both patients with gastrointestinal bleeding.
|
Prospective case control study of risk factors for anthrax in 2002.
Twenty patients diagnosed with suspected anthrax at the HAS and twenty controls matched for age, sex, and dispensary of attendance were interviewed with a structured questionnaire to determine risk factors for anthrax. Of the 20 cases there were 9 males and 11 females; the median age was 11 years, and 12 (60%) were from the northern rural zones (Perodin, Medor, Petit Reviere, and Plassac). Risk factors significantly associated with cutaneous anthrax were butchering the corpse of an animal that had died of illness and bringing meat from a diseased animal into the family compound (Table 3).
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We report an increase in cutaneous anthrax cases in the Artibonite Valley during the past 10 years. This increase may be related to several factors. Ecological changes secondary to Haiti’s deforestation have lead to a cyclical pattern of intense flooding followed by long periods of drought.21 Such a flood–drought pattern has been called "anthrax weather" because of its association with anthrax outbreaks in pastures in North America.22 Flooding and subsequent run-off concentrate anthrax spores in low-lying areas; drought evaporates pooled water and further concentrates the spores.23 The seasonality of anthrax cases at HAS supports the relationship with weather patterns. Finally, Haiti has been beset by political, economic, and social turmoil over the past 15 years24; this turmoil has hindered government-sponsored animal health services including anthrax vaccination programs.
The distribution of cases of anthrax within the Artibonite valley displays a clear geographic pattern. The incidence of anthrax in a rural area in the northeast corner of the region was 8-fold higher than the rest of the HAS district with ~80% of cases occurring in this area. Understanding of this geographic distribution enabled HAS to target an animal vaccination program at this area. Educational programs to improve the proper disposal of animal carcasses were also implemented in this region.
Two-thirds of the anthrax cases admitted to HAS between 1998 and 2002 were children. Anthrax has historically been described as a disease mostly affecting adults.25 Our report and previous reports from Haiti demonstrate that children do contract anthrax and that the morbidity and mortality of cutaneous anthrax in this population is significant.26 The high incidence of cutaneous anthrax among children in this series may be related to the integral role that children play in butchering animals in Haitian society (Figure 4).
|
). Asphyxiation from massive edema and tracheal compression has previously been described as a cause of death in cutaneous anthrax in Turkey.27 These cases clearly illustrate the potentially life-threatening nature of cutaneous anthrax of the face and neck and the importance of rapid intervention. Current recommendations for management of cutaneous anthrax in patients with systemic symptoms, extensive edema, or involvement of the head or neck now include the use of at least 2 intravenous antibiotics and consideration of steroids.28 In patients with extensive edema of the head and neck, we also recommend intensive care unit monitoring of the airway and early consultation with a surgical service in case an emergency tracheotomy is needed.
One 7-month-old girl who died had severe anemia with a hematocrit of 20 in the absence of any rectal bleeding. A blood smear showed schistocytes consistent with a microangiopathic hemolytic anemia. A similar syndrome was described in a case of cutaneous anthrax in an infant in the US in 2001.1,29
Two of the patients who died illustrated the classic constellation of fever, abdominal distension, lower GI bleeding, and septic shock that is seen with GI anthrax.30,31 The coexistence of cutaneous and GI anthrax has been reported previously,32 and is not unexpected in a setting where cattle are often butchered and then eaten by the same individuals.
Anthrax was associated with butchering and handling meat from cattle and goats that had died of illness. Reports from other developing countries have found similar associations.33–35 At the time of this anthrax outbreak, local veterinarians had observed an increase in the number of cows dying from a syndrome consistent with enzootic anthrax with sudden onset of fever progressing to hemorrhage from all orifices and death in 3–5 days. A cow in Haiti sells at market for approximately $500 US; the annual per capita income of Haiti is $410 US.36 Understandably, poor families are reluctant to burn or to bury the family livestock. In our experience, families are tempted to explain away the death of the animal by magical or traumatic causes, butcher the animal on site, and quickly sell the meat. Control of anthrax in humans depends on improved animal health and vaccination programs for livestock as well as programs to ensure the appropriate disposal of carcasses.
The current report has several limitations. First, because it was hospital based, we likely missed some subclinical cases of cutaneous anthrax that did not require hospitalization. Indeed, when investigating cases in the field, we encountered several people who described symptoms consistent with cutaneous anthrax for which they did not seek medical care. Several had healing eschars consistent with recent infection. These findings are consistent with prior descriptions of cutaneous anthrax that indicate 80–90% of cases resolve spontaneously.37 Therefore, this case series likely represents only the tip of the iceberg of cutaneous anthrax in the Artibonite Valley.
The second major limitation of this study is that we did not confirm the diagnosis of anthrax by laboratory tests. Therefore, all cases in this report are suspected anthrax by WHO criteria. Current recommendations from the American Academy of Dermatology for laboratory diagnosis of cutaneous anthrax include punch biopsy, serology, and PCR for diagnosis of cutaneous anthrax.12 Punch biopsy is not possible at HAS given the absence of an onsite pathologist. At the time of this report a commercially available serologic test was not available, and PCR is not feasible at this rural hospital. At HAS most physicians are confident in making the diagnosis of cutaneous anthrax on the basis of the WHO clinical criteria. However, unusual cases of cutaneous anthrax and cases of intestinal anthrax or inhalation anthrax without cutaneous manifestations may go undetected without laboratory diagnostics.
Our case control study is limited by the number and matching of controls. This limited the detail and the number of associations that could be made between animal and butchering practices and risk for anthrax. Our case control study was conducted by physicians to promptly guide the public health response to the growing number of cases of anthrax seen at the hospital. Although limited, it provided critical information that confirmed prior suspicions of the etiology of this cutaneous anthrax epidemic.
In conclusion, the number of cases of anthrax in the Artibonite Valley has increased by 20-fold in the past 10 years. Anthrax is a cause of significant morbidity and mortality and is related to butchering and handling meat from cattle or goats that had died of illness. We believe that the evidence presented in this report is sufficient to prompt further public health investigations and animal health interventions including livestock anthrax vaccination programs. We hope that international agencies will work together with the Haitian government in these efforts.
Received March 29, 2007. Accepted for publication August 1, 2007.
Acknowledgments: We thank Warren Johnson and Morton Swartz for their assistance in reviewing this manuscript. We also thank Keith Flanagan and Reuben Petit for advice in the area of veterinary medicine and livestock in Haiti.
* Address correspondence to Daniel W. Fitzgerald, 1300 York Avenue, Room A421, New York, NY 10021. E-mail: dfitzgerald{at}gheskio.org 
Authors’ addresses: Robert Peck, Daniel W. Fitzgerald, 1300 York Avenue, Room A421, New York, NY 10021, E-mail: dfitzgerald{at}gheskio.org; The Hospital Albert Schweitzer, Deschapelles, Haiti; The Haitian Study Group on Kaposi’s Sarcoma and Opportunistic Infection (GHESKIO), Port-au-Prince, Haiti.
We have no conflicts of interest to report.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
