Skip directly to search Skip directly to A to Z list Skip directly to page options Skip directly to site content

Volume 17, Number 11—November 2011

CME ACTIVITY - Research

Global Distribution and Epidemiologic Associations of Escherichia coli Clonal Group A, 1998–2007

James R. JohnsonComments to Author , Megan E. Menard, Tsai-Ling Lauderdale, Chris Kosmidis, David Gordon, Peter Collignon, Joel N. Maslow, Arjana Tambić Andrašević, and Michael A. Kuskowski
Author affiliations: Veterans Affairs Medical Center, Minneapolis, Minnesota, USA (J.R. Johnson, M.E. Menard, M.A. Kuskowski); University of Minnesota, Minneapolis (J.R. Johnson, M.A. Kuskowski); National Health Research Institutes, Zhunan, Taiwan (T.-L. Lauderdale); University of Athens Medical School, Athens, Greece (C. Kosmidis); Australian National University, Canberra, Australian Capital Territory, Australia (D. Gordon, P. Collignon); Canberra Hospital, Canberra (P. Collignon); Veterans Affairs Medical Center, Philadelphia, Pennsylvania, USA (J.N. Maslow); University of Pennsylvania, Philadelphia (J.N. Maslow); University Hospital for Infectious Diseases, Zagreb, Croatia (A. Tambić Andrašević)

Main Article

Table 5

Generalized linear modeling and logistic regression analysis of TMP/SMZ phenotype and region as predictors of clonal group A status among extraintestinal Escherichia coli isolates from 32 globally distributed centers, 1998–2007*

Method, type of model, variable† OR (95% CI) p value
GEE‡
Univariable
TMP/SMZ resistance 3.90 (2.04–7.46) <0.001
Africa/Asia 0.39 (0.18–0.89) 0.02
Logistic regression
Univariable
TMP/SMZ resistance 4.14 (2.74–6.26) <0.001
Africa/Asia 0.43 (0.26–0.69) <0.001
Multivariable
TMP/SMZ resistance 3.95 (2.62–5.96) <0.001
Africa/Asia 0.47 (0.30–0.76) 0.002

*TMP/SMZ, trimethoprim/sulfamethoxazole; OR, odds ratio; CI, confidence interval; GEE, generalized estimating equation.
†Univariable models, but not multivariable models, included the following as candidate predictor variables, each of which yielded a p value >0.10: specimen type (urine vs. nonurine), host age group (<18 vs. >18), host hospital status (inpatient vs. outpatient), local prevalence of TMP/SMZ resistance, and year isolate obtained.
‡ Because the multivariable GEE model that used TMP/SMZ phenotype and Africa/Asia as candidate predictor variables could not run to completion, logistic regression analysis was used instead.

*TMP/SMZ, trimethoprim/sulfamethoxazole; OR, odds ratio; CI, confidence interval; GEE, generalized estimating equation.
†Univariable models, but not multivariable models, included the following as candidate predictor variables, each of which yielded a p value >0.10: specimen type (urine vs. nonurine), host age group (<18 vs. >18), host hospital status (inpatient vs. outpatient), local prevalence of TMP/SMZ resistance, and year isolate obtained.
‡ Because the multivariable GEE model that used TMP/SMZ phenotype and Africa/Asia as candidate predictor variables could not run to completion, logistic regression analysis was used instead.

*TMP/SMZ, trimethoprim/sulfamethoxazole; OR, odds ratio; CI, confidence interval; GEE, generalized estimating equation.
†Univariable models, but not multivariable models, included the following as candidate predictor variables, each of which yielded a p value >0.10: specimen type (urine vs. nonurine), host age group (<18 vs. >18), host hospital status (inpatient vs. outpatient), local prevalence of TMP/SMZ resistance, and year isolate obtained.
‡ Because the multivariable GEE model that used TMP/SMZ phenotype and Africa/Asia as candidate predictor variables could not run to completion, logistic regression analysis was used instead.

Main Article

1Investigators who contributed data are listed at the end of this article.

TOP