Volume 14, Number 5—May 2008
Dispatch
Lakes as Source of Cholera Outbreaks, Democratic Republic of Congo
Didier Bompangue*†, Patrick Giraudoux†, Pascal Handschumacher‡, Renaud Piarroux†
, Bertrand Sudre†, Mosiana Ekwanzala§, Ilunga Kebela*, and Martine Piarroux†
, Bertrand Sudre†, Mosiana Ekwanzala§, Ilunga Kebela*, and Martine Piarroux† Table
Model parameters and odds ratios of the negative binomial model selected for cholera cases in Katanga and Eastern Kasai, Democratic Republic of the Congo, 2000–2005*
| Characteristic | Coefficient estimate | Standard error | t value | Pr(>|t|) | Odds ratio | 95% CI |
|---|---|---|---|---|---|---|
| Intercept | 5.50 | 1.04 | 5.27 | 8.92 × 10−7 | ||
| Ln (area) | −0.28 | 0.11 | −2.44 | 1.64 × 10−2 | ||
| Population | 1.97 × 10−6 | 1.17 × 10−6 | 1.69 | 9.46 × 10−2 | ||
| Railway station | 0.61 | 0.25 | 2.42 | 1.74 × 10−2 | 1.8 | 1.12–3.02 |
| Harbor | 1.39 | 0.33 | 4.16 | 7.06 × 10−5 | 4.0 | 2.09–7.75 |
| Main road | 1.43 | 0.41 | 3.50 | 7.09 × 10−4 | 4.2 | 1.88–9.32 |
| Lake | 2.01 | 0.33 | 6.12 | 2.20 × 10−8 | 7.5 | 3.92–14.23 |
*Coefficient estimate, regression coefficients (for discrete variables, their exponential gives the odds ratio); t value, value of the t distribution; Pr(>|t|), probability of the null hypothesis of a coefficient estimate not statistically different from zero; CI, confidence interval; intercept, average number of cases; Ln, natural logarithm.
