Geographic distribution was defined in terms of regional, district, subdistrict, and community distribution of cases. The burden of disease was considered in terms of number of cases affected, age and sex distribution, clinical presentation (preulcerative, ulcer, or deformity), and site of lesion. Preulcerative lesions include nodular, plaque, papular, and nonulcerative edematous forms, as described by the WHO Global Buruli Ulcer Programme (3). Deformities include scars, constriction of limbs, ankylosis of joints, or amputations.

The case search covered every district and known community in Ghana from June to July 1999. A team of 20 national facilitators was trained in the use of the survey instruments and in the clinical presentation of the disease in an endemic focus. Two facilitators were then sent to each region to train regional teams (three from the regional level and two from each district). Seven teams of two persons each from the subdistrict and communities performed the case search.

The permission of the local political and traditional authorities was sought in advance, and the purpose of the search was explained to them and to all participants. The data collectors used a pictorial document designed by the WHO Global Buruli Ulcer Initiative (3). At each village and community, they showed the pictures of Buruli ulcer disease at different stages of development to as many people as possible and asked whether anyone in the village had a similar condition. All persons with lesions that met the WHO standard case definition were interviewed with a simple questionnaire. There was no laboratory confirmation of the cases. A prepackaged dressing was given to each person identified as having ulcers, and the particulars of all the cases were provided to local health authorities for follow-up. The process was repeated in each village until the whole district was covered.

A team from the national level, including a dermatologist familiar with the disease, later validated findings in two randomly selected districts from a region where the disease had not previously been endemic. All cases reported there were found to be consistent with the clinical case definitions used.

Data from all the regions were entered centrally and later cross-checked and edited by EpiInfo 6 software. Data analysis was done both manually and by EpiInfo 6, as appropriate. Yates corrected chi-square and Cornfield 95% confidence interval tests were used as the statistical tests.

Information on regional population distribution was taken from the 2000 national census. About 200 forms from Atwima District of Ashanti Region had to be excluded from the data analysis because information was incomplete on almost all variables. Records with missing data for a particular variable were also excluded from analysis of that variable.

Approximately 49% of all patients for whom data were available (n=5,596) were female. The ages of those with active lesions ranged from 0.8 to 100 years (median 25 years; first quartile 12 years; third quartile 50 years; and mean 32 years.)

Among patients with active lesions, age is significantly associated with sex. There are significantly more females than males among patients ≥20 years than those <20 years old (chi square = 14.9; p=0.0001; odds ratio [OR]=1.3; 95% confidence interval [CI] 1.13-1.48). The age-specific odds ratio for male likelihood of having an active Buruli lesion is 0.58 (CI 0.5-0.68; p=0) while that for females is 0.73 (CI 0.62-0.68; p=0.0002).

Of 5,772 lesions for which the information was available, 25.1% were located on the arms and hands, 65.6% on the legs and feet, 5.4% on the trunk, and 3.8% on the head and neck. The distribution of lesions on the limbs (upper or lower limbs) is significantly associated with the age of the patient. Patients >20 years of age have more lesions on the limbs than the younger ones (chi square=16.4; p=0.0001; OR= 1.43; 95% CI 1.20-1.72).

The distribution of lesions on the trunk and lower limbs did not differ by sex. Males, however, had significantly more lesions on the head and neck than females (chi square=6.71; p=0.006; OR=0.67; 95% CI 0.51-0.89), while females had significantly more lesions on the upper limbs than males (chi square=4.56; p=0.03; OR=1.16; 95% CI 1.01-1.33).

Among females, 64.3% of lesions were on the lower limbs, 27% were on the upper limbs, 5.5% on the truck, and 3.2% on head and neck. Among males, 67% of lesions were on the limbs, 23% were on the upper limbs, 5.6% on the trunk and 4.2% on head and neck.

Cases of Buruli ulcer were identified in all 10 regions. Table 1 shows the prevalence rates per region, based on estimated 1999 population figures from the 2000 census. The Central Region has the highest overall prevalence rate of active cases, followed by the Ashanti Region; the Northern and Upper West Regions had the lowest prevalence rates (Figure 3). Cases of the disease were identified in 90 (81.8%) of Ghana’s 110 districts. Table 2 shows the prevalence rates of the disease in the 10 districts with the highest caseloads. Amansie West had the highest rate (prevalence 150.8 per 100,000), followed by Asante Akim North (prevalence 131.5 per 100,000) and Upper Denkyira (prevalence 114.7 per 100,000).

With regard to access to health care, 38.0% of patients lived in communities >5 miles from the nearest health facility, 37.1% lived 1 to 5 miles away, and 24.9% lived within a mile of the nearest health facility.