Volume 26-10
15 May 2000

[Table of Contents]

AN OUTBREAK OF HUMAN BLASTOMYCOSIS: THE EPIDEMIOLOGY OF BLASTOMYCOSIS IN THE KENORA CATCHMENT REGION OF ONTARIO, CANADA*

Introduction

Blastomycosis is a systemic fungal infection caused by the organism Blastomyces dermatitidis. It has a highly variable clinical spectrum and can cause acute pneumonia-like illness or chronic pulmonary disease; in a disseminated form it can have cutaneous manifestations, less commonly involve the genitourinary tract, bone or even the central nervous system, and in rare instances lead to death^(1,2).

The incidence of blastomycosis has not been reliably reported because it is not nationally reportable in Canada or the United States, although it is reportable in Wisconsin^(1,3). The ecological niche of blastomycosis has become better defined, but is not yet fully understood because of the great difficulty in isolating the organism from the environment. Outbreak studies have implicated the building of a hunting lodge, proximity to a construction site, raccoon hunting, exposure to a beaver lodge, and activities by riverbanks as sources of exposure^(4-7). So far, favourable characteristics are thought to be warm moist soil, slightly acidic pH, animal excrement, organic debris and adjacency to waterways such as lakes or rivers ^(3,5,8). However, because of the difficulty in isolating the organism from the soil, the clustering of cases should be used to build upon the growing body of information.

Blastomycosis is referred to as one of the more important endemic mycoses in North America, but its epidemiology is poorly understood⁽⁹⁾. Central Canada has long since been identified as part of the endemic areas, but there have been few studies originating from this region. From 1970 to 1981, the Central Public Health Laboratory in Toronto, Ontario, confirmed 38 cases of blastomycosis⁽⁴⁾. Since then, 16 cases were reported in Ontario from 1982 until 1989, when blastomycosis was removed from the list of diseases reportable under Ontario regulations⁽¹⁰⁾. Recently, there has been a resurgence of interest in blastomycosis in Ontario. In the Kenora, Keewatin, Jaffray Mellick tri-municipal area, and the surrounding reserves in the northwestern region of Ontario, there has been an apparent increase in blastomycosis incidence. This study attempts to describe the epidemiology of blastomycosis in this area, to estimate and compare the incidence with that of other endemic areas, and to contribute to the growing body of information on blastomycosis.

Methods

Design

The study consisted of two parts. In one, a retrospective review was carried out of all incident cases of blastomycosis from 1 January 1997, to 30 April 1999. The second involved a retrospective test for trend of all cases hospitalized with a diagnosis of blastomycosis from 1 January 1990 to 31 December 1998.

Setting

The study involved the Lake of the Woods District Hospital (LWDH) in Kenora, Ontario. The study area consisted of the catchment area of the LWDH: Kenora, Keewatin, Jaffray Mellick tri-municipal area, and the surrounding reserves.

Cases

Incident cases of blastomycosis from 1 January 1997 to 30 April 1999 were identified from the inpatient and outpatient records at LWDH coded with the ICD-9 code 116.0 for blastomycosis, from hospital laboratory records that had a positive microscopic identification of blastomycosis, and from records of individuals with positive results for blastomycosis by culture and microscopy obtained from Thunder Bay Regional Laboratory.

The case definition for a definite case was a positive culture or microscopic visualization of characteristic thick-walled, single budding yeast cells with broad isthmus from a clinical sample, and onset of symptoms characteristic of blastomycosis within the study period. Probable cases were defined as those with onset of symptoms during the study period who had a positive diagnosis by serology or had clinical symptoms typical of blastomycosis that were non-responsive to antibiotics and resolved upon antifungal treatment.

All hospital records at LWDH with the ICD-9 code 116.0 for blastomycosis in any diagnostic field, from 1 January 1990 to 31 December 1998, were analyzed by temporal test for trend. Individuals who were not residents of the study area were excluded.

Questionnaire

One investigator telephoned all individuals classified as incident cases of blastomycosis and administered a standard questionnaire on risk factors.

Analysis

SPSS version 8 was used to analyze incident case data. Population data for the calculations were taken from the 1996 census and, for some reserve communities, from the 1991 census or from Indian and Northern Affairs Canada. In all cases, the more recent population data were used. Chi-square and Fisher's exact test were used for comparison of categorical variables, and t tests for continuous variables. Findings were considered significant if the two-sided p value was < 0.05. Overlapping of 95% confidence intervals (CIs) indicated a non-significant finding. Exposure variables were considered individually and were then grouped into recreational waterway exposure, soil exposure, and wood exposure groups.

The yearly incidence rates of hospitalized cases of blastomycosis from 1990 to 1998 were calculated using the 1996 census population. The hospitalization data were entered into Epi-Info 6.0 and analyzed using the chi-square test for trend. The yearly rates were compared with reference to the 1990 incidence rate of hospitalized cases, and any 95% CI that did not overlap was considered to be significantly different. The overall p value for trend was used to estimate whether there was a significant departure from the horizontal.

Main results

Incident cases

There were 61 individuals with onset of symptoms of blastomycosis during the period from 1 January 1997 to 30 April 1999. Of these, 55 were definite cases identified through culture, microscopic examination, or microscopic visualization; six were probable cases.

Sociodemographic features: The mean age of the 61 individuals was 41.9 years (range: 3 to 80). There were 29 (47.5%) females and 32 (52.5%) males. Sex-specific and age-specific incidence rates are presented in Table 1. The sex-specific incidence rates among males and females were not significantly different. Age-specific rates indicated no significant differences in blastomycosis among age groups.

Table 1 Annual incidence rate of blastomycosis per 100,000

Of the 61 cases identified, 37 (63.8%) were Caucasian and 21 (36.2%) were Aboriginal. Rates specific for ethnicity indicated a significantly higher incidence among Aboriginal individuals. Reliable population data were not available to calculate age-specific rates by ethnic group.

Geographic and temporal distribution: Figure 1 shows the incidence of blastomycosis over the entire study area. The estimated crude incidence rate was 117.2 per 100,000 (95% CI = 90.2 to 152.4). No distinct trends were demonstrated in individual communities; this was likely a result of the wide confidence intervals. The Kenora, Keewatin, Jaffray Mellick tri-municipal area accounted for 40 cases (65.5%), and reserve communities accounted for 14 (22.9%). Comparison of incidence rates showed a significantly higher rate in reserve communities than in the tri-municipal area (Figure 1).

Figure 1 Incidence of blastomycosis in the Kenora catchment area

The distribution of cases by month of onset of initial symptoms showed a tendency to a September to January onset of symptoms; of the 58 (95.1%) cases for which this information was available, onset in 35 (60.3%) occurred during these months.

Morbidity: Table 2 presents the clinical spectrum of blastomycosis in the 61 cases. Overall, two (3.2%) had cutaneous lesions and bone involvement, and another 14 (22.9%) had cutaneous lesions; none had involvement of the central nervous system. There were three deaths; of these, two (3.3%) were directly due to blastomycosis infection.

Both individuals had associated chronic illnesses, one with chronic liver failure, and one who had chronic renal failure and was receiving dialysis. Two individuals were immunocompromised, one as a result of lupus and pyelonephritis, and the other because of testicular cancer. No other individuals had an immunosuppressive illness.

Table 2 Clinical spectrum of cases

There were 26 (43.3%) inpatients compared with 34 (56.7%) outpatients. The mean age of the inpatients was 38.2 years (range: 3 to 80) and was not significantly different from the mean age of the outpatients (mean = 45 years, p = 0.19). The inpatients consisted of 14 (53%) Aboriginal and 12 (46%) non-Aboriginal persons. A significantly higher proportion of Aboriginal than non-Aboriginal cases were of inpatient status (p = 0.015). Aboriginal inpatients had a significantly lower mean age than non-Aboriginal inpatients (34.3 years [range: 3 to 62] compared to 46.4 years [range: 14 to 80], p = 0.2).

Of the 61 individuals, four (6.6%) had had a previous diagnosis of blastomycosis. In one patient, this was 12 months before the study period. In two patients, the diagnosis was made >1 year before; and in another patient, it was made 8 years before the study.

Risk factors: Of the 61 individuals initially identified, 52 (85.2%) were interviewed; 36 (69.2%) were non-Aboriginal and 16 (30.8%) were Aboriginal. This represented a significantly higher response rate among non-Aboriginal than Aboriginal persons (p = 0.01). There was no significant association between gender or inpatient and outpatient status and having been interviewed (p = 0.56, 0.52, respectively).

Of the 52 people interviewed, 32 (61.5%) were employed, of whom 15 (46.9%) had likely exposures in their employment. There were 35 (67.3%) who were exposed to shorelines of streams, rivers, or lakes, which constituted the largest positive response; among these, 26 (74%) had daily or almost daily exposure. The second most common exposures were fishing and swimming, involving 28 (53.8%) individuals, of whom 19 (36.5%) had daily or almost daily exposure. Gardening was also a common activity (22 [42.3%] of cases). The exposure routes were grouped by categories of recreational waterway exposure, wood exposure, and soil exposure (Table 3). Among these exposure groups, 38 (73.1%) individuals participated in activities included in the recreational waterway group, and 34 (65.4%) and 27 (51.9%) participated in activities associated with soil exposure and wood exposure, respectively.

There was no significant association between ethnicity and the grouped exposure variables for recreational waterway exposure, soil exposure, or wood exposure (p = 0.738, 0.529, and 0.768, respectively). Individual exposure variables such as shoreline exposure, fishing or swimming, and gardening showed a similar lack of association with ethnicity (p = 0.75, 0.23, and 0.37, respectively).

Hospitalized cases

The study identified 52 hospitalized cases from 1 January 1990 to 31 December 1998. All were confirmed by laboratory testing, had symptoms characteristic of blastomycosis, and were residents within the study area. The 19 hospitalizations in 1998 were significantly different from the numbers in previous years (relative risk = 4.75, 95% CI = 1.62 to 13.96), which ranged from two to eight. The Chi-square test for trend indicated a significant increasing trend from 1990 to 1998 (p = 0.0003). In 1997, only two cases were hospitalized for blastomycosis.

Table 3 Risk factors

Table 4 Grouped exposure variables Groups*

Table 5 Trend in incidence rate of hospitalized cases from 1990 to 1998 (Chi-square test for trend)

Discussion

The estimated incidence rate of blastomycosis in the Kenora region stands at 117 per 100,000. In comparison to other endemic areas in North America, the Eagle River area in Vilas County, Wisconsin, has the closest reported incidence rate of 100 per 100,000⁽¹¹⁾.

If it can be assumed that the trend seen in hospitalizations reflects the true trend in all cases occurring in the region, the results show that the incidence increased suddenly at a significant level in 1998 (p = 0.0003). The reason for this sudden jump could not be clearly ascertained from this study. Local physicians and public-health professionals have suggested that it was likely due to improved awareness and increased testing. The awareness of the physicians and community may have been heightened by the deaths due to blastomycosis, one of which occurred in June of 1998, preceding the identification of the majority of cases in the same year. On the other hand, this could simply be a result of the higher incidence from September to January.

Specific regions within the study area reported remarkably higher rates of incidence: when the communities were grouped, the reserve communities demonstrated a significantly higher rate of incidence (404.9 per 100,000) than the Kenora, Keewatin, and Jaffray Mellick tri-municipal area (104.9 per 100,000). The differences between the tri-municipal area and the surrounding reserves are mainly that of population density and degree of urbanization. This ties in with the previous literature, which indicates that although blastomycosis can be clustered in urban settings, the ecological niche of the organism is greater in rural environments⁽³⁾. The communities identified by case occurrence, whether in the tri-municipal area or on reserve land, all border the shorelines and waterways along the Lake of the Woods. Klein and Vergermont described similar occurrences in Wisconsin, where cases were distributed along the Eagle and Plover Rivers⁽⁵⁾. The suggested reason for such clustering around waterways is the greater probability of conditions that support the ecological niche of Blastomyces. The waterways provide moist soil, increased humidity, and decaying organic debris. Also, the increased moisture levels enhance the aerosolization of spores.

The second hypothesis for the clustering of cases is the diverse recreational activities that draw individuals to these locations⁽⁵⁾. Therefore, the combination of ideal location for Blastomyces growth and the attractiveness of the location for recreational uses seems to cause this clustering.

Although the exposure due to recreational waterway activity is likely important in this region, it was not able to explain the higher rate of blastomycosis among Aboriginal persons. Since this was a case series study and not a case controlled study, although certain experimental exposures appear to be frequent among cases, without data on the prevalence of these activities, in a controlled population, we can not make definite conclusions about specific risk factors. There was no significant association between ethnicity and the major exposure groups of waterway activities, soil exposure and wood exposure. This supports the hypothesis put forth in a recent study by Baumgardner and Brockman⁽¹¹⁾ that place of residence is of greater influence than specific activities when endemic areas are being considered.

Although cases of blastomycosis occurred throughout the year, there was a tendency for onset to occur from September to January. The incubation period of blastomycosis has been estimated to range from 21 to 106 days^(5,9), which would put the period of infection at June to October. This coincides with the time of the year with the most rainfall and least snow cover on the ground, both of which could contribute to infection from blastomycosis⁽¹²⁾.

The study was able to identify only two (3.8%) individuals who were taking systemic steroids and could be immunocompromised - a finding that suggests that blastomycosis is a primary pathogen and is not opportunistic⁽²⁾. There was no association in the identified cases between age, gender, or the presence of chronic illness and the need for hospitalization (p = 0.19, 0.63, and 0.31, respectively). A significantly higher proportion of Aboriginal than non-Aboriginal individuals were hospitalized, probably because of the lower mean age of Aboriginal inpatients compared with Aboriginal outpatients (p = 0.04). Of the Aboriginal cases who were hospitalized, five (36%) were < 13 years old, whereas there were no non-Aboriginal individuals in that age group. The over-representation of Aboriginal persons in the inpatient category could not be linked to any specific high-risk activity. As previously stated, there was no association between ethnicity and participation in any exposure-related activities. The risk of developing blastomycosis seems to be higher in young children in Aboriginal populations than in non-Aboriginal populations.

Previous studies have identified specific factors that are characteristic of endemic areas of blastomycosis⁽¹³⁾. The soil conditions and the forest type (pine and spruce) that prevail in endemic areas are present in the tri-municipal area and surrounding regions. Although precipitation levels are lower in the study area (mean annual level: 56.3 cm) than in another endemic area, Wisconsin (mean annual level: 79 cm), the lower rainfall is probably compensated for by the Lake of the Woods and Winnipeg River waterways, which supply the surrounding areas with the needed moisture for the ecological niche of blastomycosis.

Conclusions The study identified the Kenora catchment area as having the highest estimated incidence of blastomycosis for a region of this size. The exposure variable of greatest prevalence was waterway recreational activities. In Ontario, blastomycosis was taken off the list of reportable diseases in 1989, which diminishes the ability to track changes in its incidence. The sudden increase in cases in the Kenora catchment region in 1998 is evidence that data gathering should be maintained. A passive reporting system based at the Thunder Bay Laboratory and the LWDH laboratory could be effective in keeping track of blastomycosis in the region. Physicians in the area should be informed of the high rates, particularly in the Aboriginal population, and should be aware of the seasonal and other epidemiologic characteristics of blastomycosis. Further work on the ecological niche of blastomycosis might provide a way of avoiding high-risk locations.

Acknowledgements

The authors thank Dr. R. Summerbell for assessment of risk factors and exposure routes and editorial comments; W. Reynolds for work on the maps; and Dr. P. Pan and the LWDH laboratory and medical records staff, and Dr. F. Ball and staff at the Thunder Bay Public Health Laboratory, for their assistance and support. We also thank the following institutions for their participation: LWDH, 21 Sylvan Street West Kenora, Ontario, P9N 3W7; Northwestern Health Unit, 21 Wolsley Street, Kenora, Ontario, P9N 3W7; Thunder Bay Public Health Laboratory, 336 S. Syndicate Avenue, Thunder Bay, Ontario, P7E 1E3.

References

1. CDC. Blastomycosis - Wisconsin, 1986-1995. MMWR 1996;45:601-03.

2. Kuzo RS, Goodman LR. Blastomycosis. Sem Roentgenol 1991;31:45-51.

3. Manetti CA. Hyperendemic urban blastomycosis. Am J Public Health 1991;81:633-36.

4. Kane J, Righter J, Krajden S et al. Blastomycosis: a new endemic focus in Canada. Can Med Assoc J 1983;129:728-31.

5. Klein BS, Vergermont JM, Disalvo AF et al. Two outbreaks of blastomycosis along rivers in Wisconsin: isolation of Blastomycosis dermatitidis from riverbank soil and evidence for transmission along waterway. Am Rev Respir Dis 1987;136:1333-38.

6. Lowry PW, Keso KY, McFarland LM. Blastomycosis in Washington Parish, Louisiana 1976-1985. Am J Epidemiol 1989;130:151-59.

7. Armstrong CW, Jenkins SR, Kautman L et al. Common-source outbreak of blastomycosis in hunters and their dogs. J Infect Dis 1987;155:568-70.

8. Thompson CA, McEachern R, Norman JR. Blastomycosis as an etiology of acute lung injury. South Med J 1998;91:861-63.

9. Dismukes WE. Blastomycosis: leave it to beaver. N Engl J Med 1986;314:575-76.

10. Public Health Branch. Summary of reportable diseases 1990. Toronto: Communicable Disease Control, Ontario Ministry of Health, 1991.

11. Baumgardner DJ, Brockman K. Epidemiology of human blastomycosis in Vilas County, Wisconsin II: 1991-1996. Wis Med J 1998;97:44-7.

12. Ontario Climate Centre. Kenora A (6034075) 1990-1999. Ottawa: Environment Canada, 1999.

13. Klein BS, Davis JP. Laboratory based surveillance of human blastomycosis in Wisconsin from 1973-1982. Am J Epidemiol 1985;22:897-903.

Source: PJ Dwight, HBSc, MHSc. Department of Community Health, Faculty of Medicine, University of Toronto; M Naus, MD, MHSc, Provincial Epidemiologist and Physician Manager, Disease Control Service, Ontario Ministry of Health and Long-Term Care, Toronto; P Sarsfield, MD, Medical Officer of Health and CEO, Northwestern Health Unit, B Limerick, Environmental Health Team Leader, Northwestern Health Unit, Kenora, Ontario.

* This article is being published in the 26 May 2000 issue of the Public Health & Epidemiology Report Ontario, Volume 11, Number 4.

[Previous] [Table of Contents] [Next]