American Journal of Eptdemiotogy Copyright © 1992 by The Johns Hopkins University School of Hypene and Pubfc Health All rights reserved
Vol 136, No. 4 Printed in U.S.A.
A Multistate Outbreak of Sporotrichosis associated with Sphagnum Moss F. Bruce Cotes,13 Anne Schuchat,2 Jonathan R. Hibbs,14 Stanley F. Kondracki,3 Ira F. Salkin,6 Dennis M. Dixon,5 Hwa Gan Chang,3 Rory A. Duncan,5 Nancy J. Hurd,5 and Dale L. Morse3
In the spring of 1988, the largest documented US outbreak of cutaneous sporotrichosis to date occurred, with 84 cases among persons from 15 states who were exposed to Wisconsin-grown sphagnum moss used in packing evergreen tree seedlings. In New York State, 13 cases occurred among 109 forestry workers. All 13 cases occurred among 76 workers who had handled evergreen seedlings and moss (attack rate = 17%). For those exposed to evergreens and moss, the risk of infection increased as worktime exposure to moss increased (attack rates: 19 hours, 33%). While environmental samples of moss from the Wisconsin supplier were negative, Sporothrix schenckii was cultured from multiple samples of the sphagnum moss obtained from one of six Pennsylvania tree nurseries, representing the nursery that was identified as the source for 79 (94%) of the moss-associated cases. Differences in tree-handling procedures at this nursery—including the use of 1 - to 3year-old moss to pack seedlings, use of a pond water source to wet the moss, use of an organic polymer gel on the seedling root system, and underground storage and longer storage of moss-packed seedlings before shipping—suggested possible explanations for the association. Efforts to prevent sporotrichosis among persons handling evergreen seedlings should include the use of alternate types of packing material (e.g., cedar wood chips or shredded paper) and protective clothing such as gloves and longsleeved shirts. Am J Epidemiol 1992;136:475-87. disease outbreaks; plants; sporotrichosis; trees
Sporotrichosis is caused by the dimorphic fungus Sporothrix schenckii, which exists as a saprophyte on a variety of vegetation, in decomposing wood, and in soil. Lymphocutaneous infection manifesting as skin ulcers and painless lymphadenopathy is the most common presentation and usually follows traumatic inoculation of spores from
barbs or thorns on vegetation harboring the fungus. Cases occur sporadically worldwide and tend to predominate in males. The disease is a recognized occupational hazard among fanners, horticulturists, and forestry workers (1,2). In June 1988, during annual spring reforestation programs in New York State and Illinois, several cases of cutaneous
Received for publication September 11, 1991, and in final form March 6, 1992. Abbreviations. Cl, confidence interval; RR, relative risk. ' Division of Reid Services, Epidemiology Program Office, Centers for Disease Control, Atlanta, GA. 1 Meningitis and Special Pathogens Branch, Division of Bacterial Diseases, Center for Infectious Diseases, Centers for Disease Control, Atlanta, GA 3 Bureau of Communicable Disease Control, New York State Department of Health, Albany, NY 4 Bureau of Epidemiology, Pennsylvania Department of Health, Harristxirg, PA. 6 Laboratories for Mycology, Wadsworth Center for Laboratories and Research, New York State Department of
Health, Albany, NY. Reprint requests to Dr. F. Bruce Coles, Bureau of Communicable Disease Control, New York State Department of Hearth, Room 651, Coming Tower, Empire State Plaza, Albany, NY 12237. The authors thank Dr. David T. Dennis of the Pennsylvania Department of Hearth, Dr. John G. Debbie of the New York State Department of Health, and David H. Pendergast of the New York State Soil and Water Conservation Committee for their invaluable assistance dunng the investigation of this outbreak The authors also thank Dr. Robert A. Gunn of the Division of Reid Services, Centers for Disease Control, for his review of the manuscript.
475
476
Coles et al.
sporotrichosis were reported among state forestry workers who had been exposed to evergreen seedlings packed in Wisconsingrown sphagnum moss. Subsequent investigation identified a total of 84 sphagnum moss-associated cases in 15 states, making this the largest recorded outbreak in US history. This report summarizes the clinical aspects of the cases and the epidemiology of the outbreak. MATERIALS AND METHODS Background
Each year, Soil and Water Conservation departments and private retailers (e.g., Christmas tree farms and garden clubs) in New York, Illinois, and other states purchase seedlings from several nurseries for spring tree sales to the public. The majority of seedlings are shipped packed in Wisconsin-grown sphagnum moss. Because sphagnum moss can absorb up to 20 times its weight in water, it is used as a moistening and protective agent for packing around plants during shipping. Sphagnum moss is a perennial plant which grows in moist areas and thrives in acidic conditions (3). The moss is harvested from several bogs each year from May to November, dried in open fields for approximately 5 days, and then baled. The bales are wrapped in plastic or bound with wire, stored in barns, and shipped to nurseries from January through June. The moss is usually received once per year by the nurseries and stored until the seedling packing season begins in late March or early April of the following year. Depending on when the shipments are received by the nursery, the moss may be 4 months to 2 years old at the time of packing. When a seedling is shipped, the moss is thoroughly soaked in water and packed around the seedling's root system. Buyers usually receive the seedlings within 10 days after they are packed.
tory isolates from these two states and a major private laboratory. Additional cases were reported in response to a fact sheet that was mailed to 12,000 New York State residents who had purchased seedlings from the New York Soil and Water Conservation Department and 3,000 customers in 43 states and Canada who had purchased seedlings directly from a nursery implicated as a source of contaminated material. Cases were defined as persons with physician-diagnosed sporotrichosis who had painless, erythematous cutaneous nodules (with or without ulceration) and lymphadenitis/lymphadenopathy and were exposed to seedlings and/or moss after March 1, 1988. A questionnaire assessing exposure and potential risk factors for acquiring disease was administered to a cohort of New York Soil and Water Conservation Department employees from all counties that had purchased seedlings from nurseries implicated as sources of the fungus. A modified version of the New York questionnaire was administered in Illinois to persons who had attended two meetings at which information about the outbreak was provided. These persons were forestry workers employed at nine seedling distribution sites and members of a garden club. To identify the source of contaminated material, we conducted a follow-up telephone survey in 57 New York counties that had Soil and Water Conservation offices. Information was gathered regarding the number of employees who had worked during the tree distribution, the dates of tree handling, the number of infected employees, the nurseries from which seedlings were purchased, and the number and type of seedlings purchased. A similar survey was mailed to 98 Illinois Soil and Water Conservation offices. An information sheet describing the symptoms of sporotrichosis accompanied the survey to increase recognition and assist with diagnosis.
Clinical and epidemiologic investigation
Cases of sporotrichosis were identified by investigations in New York and Illinois, by physician reports, and by review of labora-
Environmental investigation
On-site evaluations of facility characteristics and operational methods for processing
A Multistate Outbreak of Sporotricriosis
moss and seedlings were conducted at the Wisconsin bogs and at six nurseries located in one Pennsylvania county, including two that were not directly associated with the outbreak. Laboratory investigation
Exudates from lesions of case-patients were collected using sterile, dry cotton swabs and were streaked onto freshly prepared Sabouraud glucose agar supplemented with 10 fig of gentamicin and 25 ^g of chloramphenicol per ml. The specimens were incubated at room temperature. In selected cases, histopathologic identification was performed using our laboratory-generated specific fluorescent antibody procedure (4). Serum samples were collected from four exposure groups, including moss harvesters from Wisconsin, nursery workers from Pennsylvania, and seedling handlers from New York and Illinois. One hundred serum samples from an Illinois blood bank were obtained for comparison. Samples were analyzed for S. schenckii antibodies by slide latex agglutination using reagents prepared in our laboratory (4). Analysis of culture and serologic specimens was performed at the Laboratories for Mycology of the New York State Department of Health's Wadsworth Center for Laboratories and Research. Environmental cultures of sphagnum moss were collected in Wisconsin from three bogs, from piles of both wet and dried moss prior to baling, from plastic-wrapped and wire-bound bales, and from six storage sheds. Multiple samples were also obtained from six Pennsylvania nurseries and from seven New York Soil and Water Conservation district offices. Additional environmental samples were obtained from the six nurseries and included samples from soil, root gel (a potassium propenoate-propenamide copolymer), pond water, faucets, floor drains, pine and spruce seedlings, wood, burlap, and grass. Samples were processed using an aqueous extraction procedure (4). The samples were suspended in sterile water containing antibacterial agents; 0.2-ml volumes and 10-fold dilutions thereof were plated onto selective mycologic media (4).
477
Following 5-7 days of growth, suspicious colonies were subcultured for further study, which included macroscopic and microscopic morphology, growth at 35°C and 37°C, and conversion to the yeast-like phase of growth. Isolates were then identified as 5. schenckii by exoantigen studies as described by Kaufman and Standard (5). Statistical analysis
Relative risks and their 95 percent confidence intervals were calculated for each risk factor using computer software from the Centers for Disease Control (6). For the New York investigation, a multiple logistic regression model from the BMDP software package (7) was used to calculate the adjusted relative risk (as estimated by the odds ratio) for those factors associated with infection in the univariate analysis. To assess potential confounding by exposure to hay or roses in the Illinois investigation, we used stratified analysis to calculate the MantelHaenszel weighted relative risk for worktime exposure to moss (8). RESULTS Clinical characteristics
Eighty-four cases of sporotrichosis associated with sphagnum moss were identified. Follow-up interviews and clinical information were obtained for 74 cases (88 percent). Case-patients reported that they had had first exposure to seedlings and moss from March 1 to June 14, 1988, and that they had handled between 40 and 20,000 seedlings during the exposure period (the median number was 1,800). The median duration of exposure was 12 hours (range, 1-210 hours). Disease onset began on April 15, 1988, and continued through June 30, 1988, with a peak of 16 cases during the week of April 25 (figure 1). The median incubation period from the date of first exposure to seedlings and sphagnum moss was 21 days (range, 6-65 days). Cases were identified in 15 states, including 25 cases in New York, 24 in Illinois, 15 in Pennsylvania, and 20 from 12 other states. Four Pennsylvania nurseries (nurseries A, B, C, and F) were identified as sources
478
Coles et al.
No. Cases 20 15
Week of First Exposure
j-
10 5 0 20 15
mt
11 -
Week of Onset
10 5 0 29
7
14
21
March
28
4
11
18
25
2
9
April
16
May
23
30
6
13
20
27
4
June
1QRR FIGURE 1. Numbers of cases of sporotrichosis, by week of first exposure to seedlings and sphagnum moss and by week of onset, in a multistate sporotnchosis outbreak, 1988.
of seedlings and sphagnum moss for 50 casepatients interviewed during the outbreak investigation. The remaining 34 case-patients were identified through nursery A's customer mailing records. The median age of case-patients was 41 years (range, 16-77 years); 50 (68 percent) were male and 24 (32 percent) were female. Three case-patients (4 percent) reported preexisting skin conditions, including eczema and contact dermatitis; four (5 percent) had diabetes; one (1 percent) had cancer; and none reported liver disease or immune deficiency conditions. Twenty-six case-patients (35 percent) were Soil and Water Conservation Department workers, 12 (16 percent) were tree nursery workers, and 36 (49 percent) were Soil and Water Conservation or nursery customers. No cases were identified among 34 Wisconsin moss harvesters, despite an estimated median of 9,750 lifetime hours of exposure to the moss (range, 650-83,200 hours). All reported cutaneous infections were confined to the hands, arms, or both; no disseminated or extracutaneous infections
occurred (table 1). There was substantial delay in seeking physician consultation and subsequently in obtaining the correct diagnosis once the first contact was made with the physician. Twenty-three persons (31 percent) missed time from work, and 15 (20 percent) were hospitalized. Only 11 case-patients (15 percent) were correctly diagnosed at the time of their initial examination. Fifty-seven persons (77 percent) received oral antibacterial antibiotics, and 10 (13 percent) underwent surgical drainage as initial therapy (table 2). Sixtytwo (84 percent) eventually received an oral saturated solution of potassium iodide containing 1.0 g of potassium iodide per ml. Daily treatment was administered in three divided doses; a 1-g dose is equivalent to 10 drops of the potassium iodide solution. Of the 53 persons who reported dosage, 37 (70 percent) received the recommended initial dose of 1.5-3.0 g per day. Forty persons had completed or discontinued treatment at the time of the survey; 25 (63 percent) reached or exceeded the recommended maximum dose of 9.0-12.0 g
A Multistate Outbreak of Sporotrichosis
479
TABLE 1. Clinical characteristics ol sporotrichosis cases (n = 74) in a multistate sporotrichosis outbreak, 1988 Characteristic Location of lesions Hands Arms Hands and amis Not identified Days from onset of symptoms to date of first physician visit (n = 70) Days from first physician vi3it to date of correct diagnosis (n = 70) No. of physician visits (n = 64) No. of days missed from work Days of hospitalization
TABLE 2. Initial diagnosis and treatment characteristics of sporotrichosis cases (n = 74) in a multistate sporotrichosis outbreak, 1988 No.
%
Initial diagnosis Sporotrichosis Bacterial Infection Spider bite Cancer Unknown
11 48 6 2 7
15 65 8 3 9
Initial treatment Oral antibiotics Surgery Incision and drainage Biopsy Potassium iodide only Unknown
57 23 10 13 11 4
77 31 13 18 15 5
Characteristic
per day. Fifty-seven persons (92 percent) experienced side effects (table 3). Three persons were treated with alternate antibiotic regimens, including intravenous amphotericin B, potassium iodide solution plus flucytosine, and itraconazole. Fifty-two persons reported that all lesions had healed and resolved in a median time of 88 days (range, 28-152 days). New York investigation
The New York State conservation worker cohort study included 109 Soil and Water Conservation Department workers from 22 counties. Seventy-one (65 percent) were males, and the median age was 36 years
No.
%
29 15 26 4
39 20 35 5
23 15
31 20
Median
Range
15
0-47
16 6 6 4
0-174 1-50 1-90 2-24
(range, 12-74 years). Eighty (73 percent) were full-time Soil and Water Conservation employees, and 29 (27 percent) were volunteer workers. Thirteen cases of physician-confirmed sporotrichosis were identified among the Soil and Water Conservation workers from seven counties, for an attack rate of 12 percent. Four exposure groups were identified, including workers exposed to both evergreen seedlings and sphagnum moss (« = 76), workers exposed to both hardwood seedlings and moss (n = 4), persons exposed to evergreen seedlings alone (n = 13), and those reporting no direct contact with either seedlings or moss (n = 16). No worker reported exposure to sphagnum moss alone. The attack rate for workers exposed to evergreen seedlings and sphagnum moss was 17 percent. No cases occurred in the remaining exposure groups. Analysis of the 76 workers who had had both evergreen and moss exposure revealed that the risk for infection was more than two times greater among workers handling seedlings from one Pennsylvania nursery (nursery A) and among volunteer workers. The risk of infection also increased as worktime exposure to moss increased. Workers wearing long-sleeved shirts had a slightly decreased risk of infection (table 4). Differences in age, sex, the use of gloves, and the presence of metabolic or immune deficiency diseases (only two persons, both noninfected, reported a recent history of treatment
480
Coles et al.
TABLE 3. Treatment characteristics of patients who were treated with 1 g/ml of potassium Iodide solution (n = 62) in a murtistate sporotrichosis outbreak, 1988 Characteristic
No
Initial dose (g/day)* (n = 53) Recommended dose (1.5-3.0 g/day) Completed/discontinued therapy Maximum dose attained (g/day) 12.0
37
70
40
64
15 21 4
37 53 10
Median
Range
1.5
0.3-6.9
84
Days of therapy at maximum dose of >9.0 g/day Side effects Acne Upset stomach Lacrimatton/salivation Sore throat Generalized rash Menstrual irregularity Lethargy
57 28 19 8 7 2 1 1
26-171
92 49 33 14 12 4 2 2
• 1 g •=» 10 drops of potassium iodide solution.
TABLE 4. Sporotrichosis attack rates for New York State Soil and Water Conservation Department workers exposed to sphagnum moss (n = 76) in a multistate sporotrichosis outbreak, 1984 No. of cases
Total
Attack rate(%)
RR*
95% C r
Nursery A atone Aand B B alone
8 3 2
29 27 20
28 11 10
2.8 1.1 1.0*
0.6-24.5* 0.1-12.4*
Job status Volunteer Full-time
6 7
20 56
30 13
2.4 1.0*
0.7-7.0*
8 3 2
24 26 26
33 12 8
4.3 1.5
1.0-37.8* 0.2-16.7*
Wore gloves Yes No
2 11
9 67
22 16
1.4 10*
0.2-4.6*
Long sleeves Yes No
6 7
43 33
14 21
0.7 1.0*
0.2-1.8
2 11
11 65
18 17
1.1 1.0*
0.1-3.9*
Risk factor
Hours of exposure to moss
>19 10-19
Vol 136, No. 4 Printed in U.S.A.
A Multistate Outbreak of Sporotrichosis associated with Sphagnum Moss F. Bruce Cotes,13 Anne Schuchat,2 Jonathan R. Hibbs,14 Stanley F. Kondracki,3 Ira F. Salkin,6 Dennis M. Dixon,5 Hwa Gan Chang,3 Rory A. Duncan,5 Nancy J. Hurd,5 and Dale L. Morse3
In the spring of 1988, the largest documented US outbreak of cutaneous sporotrichosis to date occurred, with 84 cases among persons from 15 states who were exposed to Wisconsin-grown sphagnum moss used in packing evergreen tree seedlings. In New York State, 13 cases occurred among 109 forestry workers. All 13 cases occurred among 76 workers who had handled evergreen seedlings and moss (attack rate = 17%). For those exposed to evergreens and moss, the risk of infection increased as worktime exposure to moss increased (attack rates: 19 hours, 33%). While environmental samples of moss from the Wisconsin supplier were negative, Sporothrix schenckii was cultured from multiple samples of the sphagnum moss obtained from one of six Pennsylvania tree nurseries, representing the nursery that was identified as the source for 79 (94%) of the moss-associated cases. Differences in tree-handling procedures at this nursery—including the use of 1 - to 3year-old moss to pack seedlings, use of a pond water source to wet the moss, use of an organic polymer gel on the seedling root system, and underground storage and longer storage of moss-packed seedlings before shipping—suggested possible explanations for the association. Efforts to prevent sporotrichosis among persons handling evergreen seedlings should include the use of alternate types of packing material (e.g., cedar wood chips or shredded paper) and protective clothing such as gloves and longsleeved shirts. Am J Epidemiol 1992;136:475-87. disease outbreaks; plants; sporotrichosis; trees
Sporotrichosis is caused by the dimorphic fungus Sporothrix schenckii, which exists as a saprophyte on a variety of vegetation, in decomposing wood, and in soil. Lymphocutaneous infection manifesting as skin ulcers and painless lymphadenopathy is the most common presentation and usually follows traumatic inoculation of spores from
barbs or thorns on vegetation harboring the fungus. Cases occur sporadically worldwide and tend to predominate in males. The disease is a recognized occupational hazard among fanners, horticulturists, and forestry workers (1,2). In June 1988, during annual spring reforestation programs in New York State and Illinois, several cases of cutaneous
Received for publication September 11, 1991, and in final form March 6, 1992. Abbreviations. Cl, confidence interval; RR, relative risk. ' Division of Reid Services, Epidemiology Program Office, Centers for Disease Control, Atlanta, GA. 1 Meningitis and Special Pathogens Branch, Division of Bacterial Diseases, Center for Infectious Diseases, Centers for Disease Control, Atlanta, GA 3 Bureau of Communicable Disease Control, New York State Department of Health, Albany, NY 4 Bureau of Epidemiology, Pennsylvania Department of Health, Harristxirg, PA. 6 Laboratories for Mycology, Wadsworth Center for Laboratories and Research, New York State Department of
Health, Albany, NY. Reprint requests to Dr. F. Bruce Coles, Bureau of Communicable Disease Control, New York State Department of Hearth, Room 651, Coming Tower, Empire State Plaza, Albany, NY 12237. The authors thank Dr. David T. Dennis of the Pennsylvania Department of Hearth, Dr. John G. Debbie of the New York State Department of Health, and David H. Pendergast of the New York State Soil and Water Conservation Committee for their invaluable assistance dunng the investigation of this outbreak The authors also thank Dr. Robert A. Gunn of the Division of Reid Services, Centers for Disease Control, for his review of the manuscript.
475
476
Coles et al.
sporotrichosis were reported among state forestry workers who had been exposed to evergreen seedlings packed in Wisconsingrown sphagnum moss. Subsequent investigation identified a total of 84 sphagnum moss-associated cases in 15 states, making this the largest recorded outbreak in US history. This report summarizes the clinical aspects of the cases and the epidemiology of the outbreak. MATERIALS AND METHODS Background
Each year, Soil and Water Conservation departments and private retailers (e.g., Christmas tree farms and garden clubs) in New York, Illinois, and other states purchase seedlings from several nurseries for spring tree sales to the public. The majority of seedlings are shipped packed in Wisconsin-grown sphagnum moss. Because sphagnum moss can absorb up to 20 times its weight in water, it is used as a moistening and protective agent for packing around plants during shipping. Sphagnum moss is a perennial plant which grows in moist areas and thrives in acidic conditions (3). The moss is harvested from several bogs each year from May to November, dried in open fields for approximately 5 days, and then baled. The bales are wrapped in plastic or bound with wire, stored in barns, and shipped to nurseries from January through June. The moss is usually received once per year by the nurseries and stored until the seedling packing season begins in late March or early April of the following year. Depending on when the shipments are received by the nursery, the moss may be 4 months to 2 years old at the time of packing. When a seedling is shipped, the moss is thoroughly soaked in water and packed around the seedling's root system. Buyers usually receive the seedlings within 10 days after they are packed.
tory isolates from these two states and a major private laboratory. Additional cases were reported in response to a fact sheet that was mailed to 12,000 New York State residents who had purchased seedlings from the New York Soil and Water Conservation Department and 3,000 customers in 43 states and Canada who had purchased seedlings directly from a nursery implicated as a source of contaminated material. Cases were defined as persons with physician-diagnosed sporotrichosis who had painless, erythematous cutaneous nodules (with or without ulceration) and lymphadenitis/lymphadenopathy and were exposed to seedlings and/or moss after March 1, 1988. A questionnaire assessing exposure and potential risk factors for acquiring disease was administered to a cohort of New York Soil and Water Conservation Department employees from all counties that had purchased seedlings from nurseries implicated as sources of the fungus. A modified version of the New York questionnaire was administered in Illinois to persons who had attended two meetings at which information about the outbreak was provided. These persons were forestry workers employed at nine seedling distribution sites and members of a garden club. To identify the source of contaminated material, we conducted a follow-up telephone survey in 57 New York counties that had Soil and Water Conservation offices. Information was gathered regarding the number of employees who had worked during the tree distribution, the dates of tree handling, the number of infected employees, the nurseries from which seedlings were purchased, and the number and type of seedlings purchased. A similar survey was mailed to 98 Illinois Soil and Water Conservation offices. An information sheet describing the symptoms of sporotrichosis accompanied the survey to increase recognition and assist with diagnosis.
Clinical and epidemiologic investigation
Cases of sporotrichosis were identified by investigations in New York and Illinois, by physician reports, and by review of labora-
Environmental investigation
On-site evaluations of facility characteristics and operational methods for processing
A Multistate Outbreak of Sporotricriosis
moss and seedlings were conducted at the Wisconsin bogs and at six nurseries located in one Pennsylvania county, including two that were not directly associated with the outbreak. Laboratory investigation
Exudates from lesions of case-patients were collected using sterile, dry cotton swabs and were streaked onto freshly prepared Sabouraud glucose agar supplemented with 10 fig of gentamicin and 25 ^g of chloramphenicol per ml. The specimens were incubated at room temperature. In selected cases, histopathologic identification was performed using our laboratory-generated specific fluorescent antibody procedure (4). Serum samples were collected from four exposure groups, including moss harvesters from Wisconsin, nursery workers from Pennsylvania, and seedling handlers from New York and Illinois. One hundred serum samples from an Illinois blood bank were obtained for comparison. Samples were analyzed for S. schenckii antibodies by slide latex agglutination using reagents prepared in our laboratory (4). Analysis of culture and serologic specimens was performed at the Laboratories for Mycology of the New York State Department of Health's Wadsworth Center for Laboratories and Research. Environmental cultures of sphagnum moss were collected in Wisconsin from three bogs, from piles of both wet and dried moss prior to baling, from plastic-wrapped and wire-bound bales, and from six storage sheds. Multiple samples were also obtained from six Pennsylvania nurseries and from seven New York Soil and Water Conservation district offices. Additional environmental samples were obtained from the six nurseries and included samples from soil, root gel (a potassium propenoate-propenamide copolymer), pond water, faucets, floor drains, pine and spruce seedlings, wood, burlap, and grass. Samples were processed using an aqueous extraction procedure (4). The samples were suspended in sterile water containing antibacterial agents; 0.2-ml volumes and 10-fold dilutions thereof were plated onto selective mycologic media (4).
477
Following 5-7 days of growth, suspicious colonies were subcultured for further study, which included macroscopic and microscopic morphology, growth at 35°C and 37°C, and conversion to the yeast-like phase of growth. Isolates were then identified as 5. schenckii by exoantigen studies as described by Kaufman and Standard (5). Statistical analysis
Relative risks and their 95 percent confidence intervals were calculated for each risk factor using computer software from the Centers for Disease Control (6). For the New York investigation, a multiple logistic regression model from the BMDP software package (7) was used to calculate the adjusted relative risk (as estimated by the odds ratio) for those factors associated with infection in the univariate analysis. To assess potential confounding by exposure to hay or roses in the Illinois investigation, we used stratified analysis to calculate the MantelHaenszel weighted relative risk for worktime exposure to moss (8). RESULTS Clinical characteristics
Eighty-four cases of sporotrichosis associated with sphagnum moss were identified. Follow-up interviews and clinical information were obtained for 74 cases (88 percent). Case-patients reported that they had had first exposure to seedlings and moss from March 1 to June 14, 1988, and that they had handled between 40 and 20,000 seedlings during the exposure period (the median number was 1,800). The median duration of exposure was 12 hours (range, 1-210 hours). Disease onset began on April 15, 1988, and continued through June 30, 1988, with a peak of 16 cases during the week of April 25 (figure 1). The median incubation period from the date of first exposure to seedlings and sphagnum moss was 21 days (range, 6-65 days). Cases were identified in 15 states, including 25 cases in New York, 24 in Illinois, 15 in Pennsylvania, and 20 from 12 other states. Four Pennsylvania nurseries (nurseries A, B, C, and F) were identified as sources
478
Coles et al.
No. Cases 20 15
Week of First Exposure
j-
10 5 0 20 15
mt
11 -
Week of Onset
10 5 0 29
7
14
21
March
28
4
11
18
25
2
9
April
16
May
23
30
6
13
20
27
4
June
1QRR FIGURE 1. Numbers of cases of sporotrichosis, by week of first exposure to seedlings and sphagnum moss and by week of onset, in a multistate sporotnchosis outbreak, 1988.
of seedlings and sphagnum moss for 50 casepatients interviewed during the outbreak investigation. The remaining 34 case-patients were identified through nursery A's customer mailing records. The median age of case-patients was 41 years (range, 16-77 years); 50 (68 percent) were male and 24 (32 percent) were female. Three case-patients (4 percent) reported preexisting skin conditions, including eczema and contact dermatitis; four (5 percent) had diabetes; one (1 percent) had cancer; and none reported liver disease or immune deficiency conditions. Twenty-six case-patients (35 percent) were Soil and Water Conservation Department workers, 12 (16 percent) were tree nursery workers, and 36 (49 percent) were Soil and Water Conservation or nursery customers. No cases were identified among 34 Wisconsin moss harvesters, despite an estimated median of 9,750 lifetime hours of exposure to the moss (range, 650-83,200 hours). All reported cutaneous infections were confined to the hands, arms, or both; no disseminated or extracutaneous infections
occurred (table 1). There was substantial delay in seeking physician consultation and subsequently in obtaining the correct diagnosis once the first contact was made with the physician. Twenty-three persons (31 percent) missed time from work, and 15 (20 percent) were hospitalized. Only 11 case-patients (15 percent) were correctly diagnosed at the time of their initial examination. Fifty-seven persons (77 percent) received oral antibacterial antibiotics, and 10 (13 percent) underwent surgical drainage as initial therapy (table 2). Sixtytwo (84 percent) eventually received an oral saturated solution of potassium iodide containing 1.0 g of potassium iodide per ml. Daily treatment was administered in three divided doses; a 1-g dose is equivalent to 10 drops of the potassium iodide solution. Of the 53 persons who reported dosage, 37 (70 percent) received the recommended initial dose of 1.5-3.0 g per day. Forty persons had completed or discontinued treatment at the time of the survey; 25 (63 percent) reached or exceeded the recommended maximum dose of 9.0-12.0 g
A Multistate Outbreak of Sporotrichosis
479
TABLE 1. Clinical characteristics ol sporotrichosis cases (n = 74) in a multistate sporotrichosis outbreak, 1988 Characteristic Location of lesions Hands Arms Hands and amis Not identified Days from onset of symptoms to date of first physician visit (n = 70) Days from first physician vi3it to date of correct diagnosis (n = 70) No. of physician visits (n = 64) No. of days missed from work Days of hospitalization
TABLE 2. Initial diagnosis and treatment characteristics of sporotrichosis cases (n = 74) in a multistate sporotrichosis outbreak, 1988 No.
%
Initial diagnosis Sporotrichosis Bacterial Infection Spider bite Cancer Unknown
11 48 6 2 7
15 65 8 3 9
Initial treatment Oral antibiotics Surgery Incision and drainage Biopsy Potassium iodide only Unknown
57 23 10 13 11 4
77 31 13 18 15 5
Characteristic
per day. Fifty-seven persons (92 percent) experienced side effects (table 3). Three persons were treated with alternate antibiotic regimens, including intravenous amphotericin B, potassium iodide solution plus flucytosine, and itraconazole. Fifty-two persons reported that all lesions had healed and resolved in a median time of 88 days (range, 28-152 days). New York investigation
The New York State conservation worker cohort study included 109 Soil and Water Conservation Department workers from 22 counties. Seventy-one (65 percent) were males, and the median age was 36 years
No.
%
29 15 26 4
39 20 35 5
23 15
31 20
Median
Range
15
0-47
16 6 6 4
0-174 1-50 1-90 2-24
(range, 12-74 years). Eighty (73 percent) were full-time Soil and Water Conservation employees, and 29 (27 percent) were volunteer workers. Thirteen cases of physician-confirmed sporotrichosis were identified among the Soil and Water Conservation workers from seven counties, for an attack rate of 12 percent. Four exposure groups were identified, including workers exposed to both evergreen seedlings and sphagnum moss (« = 76), workers exposed to both hardwood seedlings and moss (n = 4), persons exposed to evergreen seedlings alone (n = 13), and those reporting no direct contact with either seedlings or moss (n = 16). No worker reported exposure to sphagnum moss alone. The attack rate for workers exposed to evergreen seedlings and sphagnum moss was 17 percent. No cases occurred in the remaining exposure groups. Analysis of the 76 workers who had had both evergreen and moss exposure revealed that the risk for infection was more than two times greater among workers handling seedlings from one Pennsylvania nursery (nursery A) and among volunteer workers. The risk of infection also increased as worktime exposure to moss increased. Workers wearing long-sleeved shirts had a slightly decreased risk of infection (table 4). Differences in age, sex, the use of gloves, and the presence of metabolic or immune deficiency diseases (only two persons, both noninfected, reported a recent history of treatment
480
Coles et al.
TABLE 3. Treatment characteristics of patients who were treated with 1 g/ml of potassium Iodide solution (n = 62) in a murtistate sporotrichosis outbreak, 1988 Characteristic
No
Initial dose (g/day)* (n = 53) Recommended dose (1.5-3.0 g/day) Completed/discontinued therapy Maximum dose attained (g/day) 12.0
37
70
40
64
15 21 4
37 53 10
Median
Range
1.5
0.3-6.9
84
Days of therapy at maximum dose of >9.0 g/day Side effects Acne Upset stomach Lacrimatton/salivation Sore throat Generalized rash Menstrual irregularity Lethargy
57 28 19 8 7 2 1 1
26-171
92 49 33 14 12 4 2 2
• 1 g •=» 10 drops of potassium iodide solution.
TABLE 4. Sporotrichosis attack rates for New York State Soil and Water Conservation Department workers exposed to sphagnum moss (n = 76) in a multistate sporotrichosis outbreak, 1984 No. of cases
Total
Attack rate(%)
RR*
95% C r
Nursery A atone Aand B B alone
8 3 2
29 27 20
28 11 10
2.8 1.1 1.0*
0.6-24.5* 0.1-12.4*
Job status Volunteer Full-time
6 7
20 56
30 13
2.4 1.0*
0.7-7.0*
8 3 2
24 26 26
33 12 8
4.3 1.5
1.0-37.8* 0.2-16.7*
Wore gloves Yes No
2 11
9 67
22 16
1.4 10*
0.2-4.6*
Long sleeves Yes No
6 7
43 33
14 21
0.7 1.0*
0.2-1.8
2 11
11 65
18 17
1.1 1.0*
0.1-3.9*
Risk factor
Hours of exposure to moss
>19 10-19
