Volume 18, Number 2—February 2012
Dispatch
Plesiomonas shigelloides Infection, Ecuador, 2004–2008
Table
RRs and bootstrapped 95% CIs for single infections and co-infections with Plesiomonas shigelloides, Ecuador, 2004–2008*
| Co-infection | RRSingle P.shig (95% CI) | RRCo-Infection (95% CI) | RRCrude (95% CI) | RRMH-Pooled (95% CI) | Wald test for heterogeneity | p value |
|---|---|---|---|---|---|---|
| Any pathogen | 1.5 (0.9–2.2) | 5.6 (3.5–9.3) | 2.6 (1.9–3.5) | 2.7 (1.9–3.6) | 32.1 | <0.001 |
| Rotavirus | 1.5 (0.9–2.2) | 16.2 (5.5–62.3) | 1.7 (1.1–2.5) | 1.9 (1.2–2.9) | 61.8 | <0.001 |
| Giardia spp. | 1.5 (0.9–2.2) | 2.1 (1.0–3.9) | 1.5 (1.0–2.2) | 1.6 (1.1–2.3) | 1.3 | 0.2 |
| Escherichia coli/ shigellae | 1.5 (0.9–2.2) | 13.8 (3.3–69.3) | 1.6 (1.1–2.4) | 1.7 (1.1–2.6) | 32.8 | <0.001 |
*RR, risk ratio. RRcrude = the unadjusted RR and RRMH-pooled is the pooled Mantel-Haenszel RR ratio estimate. The Wald test assesses whether the strata RRSingle P.shig and RRco-infection differ. Because of the clustered study design and the unequal sampling probabilities of controls, we chose not to use logistic regression models. Instead, we applied a nonparametric approach by using sampling weights to estimate RRs, as one would for a cohort study. We bootstrapped 1,000 samples from the original dataset, and with each new sample, we estimated the RR associated with single infection and co-infection. The lower 0.025 and upper 0.975 percentiles of the bootstrap distribution are reported as 95% CIs. Statistical analyses were conducted by using R version 2.11.1 (www.r-project.org).