To assess the prevalence of and risk factors for carbapenem-resistant Klebsiella pneumoniae (CRKP) carriage among patients in post-acute-care facilities (PACFs) in Israel.

A cross-sectional prevalence survey was conducted in 12 PACFs. Rectal swab samples were obtained from 1,144 patients in 33 wards. Risk factors for CRKP carriage were assessed among the cohort. Next, a nested, matched case-control study was conducted to define individual risk factors for colonization. Finally, the cohort of patients with a history of CRKP carriage was characterized to determine risk factors for continuous carriage.

The prevalence of rectal carriage of CRKP among 1,004 patients without a history of CRKP carriage was 12.0%. Independent risk factors for CRKP carriage were prolonged length of stay (odds ratio [OR], 1.001; P < .001), sharing a room with a known carrier (OR, 3.09; P = .02), and increased prevalence of known carriers on the ward (OR, 1.02; P = .013). A policy of screening for carriage on admission was protective (OR, 0.41; P = .03). Risk factors identified in the nested case-control study were antibiotic exposure during the prior 3 months (OR, 1.66; P = .03) and colonization with other resistant pathogens (OR, 1.64; P = .03). Among 140 patients with a history of CRKP carriage, 47% were colonized. Independent risk factors for continued CRKP carriage were antibiotic exposure during the prior 3 months (OR, 3.05; P = .04), receipt of amoxicillin-clavulanate (OR, 4.18; P = .007), and screening within 90 days of the first culture growing CRKP (OR, 2.9; P = .012).

We found a large reservoir of CRKP in PACFs. Infection-control polices and antibiotic exposure were associated with patient colonization.

Carbapenem-resistant Enterobacteriaceae (CRE) are important extremely drug-resistant pathogens that have emerged during the past decade. Early identification and isolation of carriers are key components of an effective infection control strategy in healthcare facilities. Very little is known about the natural history of CRE carriage. We aimed to determine the predictors of a positive CRE rectal screen test among patients with known CRE carriage screened at their next hospital encounter.

A case-control study was conducted. Sixty-six patients who tested positive for CRE carriage were surveyed for CRE rectal carriage at the next hospital encounter; screen-positive patients were compared with screen-negative control patients. Data were extracted from the patients' medical records and from the hospital computerized database.

Twenty-three case patients and 43 control patients were identified. Predictors for a positive CRE rectal carriage test were (1) prior fluoroquinolone use (odds ratio [OR], 4.27; 95% confidence interval [CI], 1.10–16.6), (2) admission from an institution or another hospital (OR, 4.04; 95% CI, 1.33–12.37), and (3) time interval less than or equal to 3 months since the first positive CRE test (OR, 3.59; 95% CI, 1.24–10.37). Among patients with no predictor variables, the likelihood of having a positive screen test at the next hospital encounter was 1/7. If they had at least 1 predictor, the likelihood increased to 1/2.

Prior fluoroquinolone use, transfer from another healthcare facility, and admission less than or equal to 3 months since the first CRE isolation are predictors of persistent CRE rectal carriage. These predictors can be used in designing CRE prevention strategies.

To perform a molecular and epidemiologic investigation of multidrug-resistant (MDR) Acinetobacter baumannii in an institution were polyclonal outbreaks have been observed and determine whether these polyclonal outbreaks had an endogenous origin or were caused by in-hospital patient-to-patient transmission.

Retrospective analysis of prospectively collected data.

An epidemiologic and genotypic investigation of incident cases of MDR A. baumannii infection in an Israeli university tertiary care center.

Hospitalized patients with MDR A. baumannii isolated from clinical samples during a 13-week period, from April 15, 2003, through July 15, 2003.

All patients with new MDR A. baumannii infections were recruited, and isolates were typed using pulsed-field gel electrophoresis. Data on in-hospital movements and consultations were extracted from computerized databases. Quantification of transmission opportunities (TOPs), defined as encounters between an established carrier and a future carrier of MDR A. baumannii, and analysis of ward clusters were performed.

We studied 96 MDR A. baumannii isolates, which belonged to 18 different pulsed-field gel electrophoresis clones. In 65% of cases, TOPs involving patients with the same clone were demonstrated, which is significantly greater than the number of TOPs involving patients with different clones (P = .01).

Although outbreaks of MDR A. baumannii infection may be polyclonal, we believe that patient-to-patient transmission explains most cases of transmission. Polyclonal local outbreaks reflect several clonal outbreaks occurring simultaneously. The cause of polyclonal outbreaks of A. baumannii infections clustered by ward and time remains to be explained.