Skip to main content
. Author manuscript; available in PMC: 2025 Mar 24.
Published in final edited form as: JAMA Intern Med. 2013 Jul 22;173(14):1359–1367. doi: 10.1001/jamainternmed.2013.7056

Table 4.

Multivariable Analysis for Sources of Clostridium difficile Acquisition in Patients Having Community-Associated C difficile Infection With No or Low-Level Outpatient Health Care Exposure, 2009 Through 2011a

Odds Ratio (95% CI)
Variable No Exposure Low-Level Exposure
No medical conditions 1.1 (0.8–1.6) 1.7 (1.3–2.3)
Household members
 Infant younger than 1 y 1.8 (0.7–4.6) 2.1 (1.1–4.5)
 With active C difficile infection 6.8 (0.7–65.9) 6.9 (0.9–56.7)
a

Odds ratios were calculated using multinomial logistic regression using high-level outpatient health care exposure as the reference group. Candidate variables included clinical characteristics (age, sex, and no medical conditions), potential sources of C difficile infection (infant younger than 1 year, household member with active C difficile infection, and children younger than 4 years who attended child care settings), and antibiotic use as an interaction term with potential sources of C difficile infection. The final model was selected using a backward logistic regression strategy with a stay criterion of P ≤ .10 for variables.